Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Interventions aimed at reducing problems in adult patients discharged from hospital to home: a systematic meta-review

Interventions aimed at reducing problems in adult patients discharged from hospital to home: a... Background: Many patients encounter a variety of problems after discharge from hospital and many discharge (planning and support) interventions have been developed and studied. These primary studies have already been synthesized in several literature reviews with conflicting conclusions. We therefore set out a systematic review of the reviews examining discharge interventions. The objective was to synthesize the evidence presented in literature on the effectiveness of interventions aimed to reduce post-discharge problems in adults discharged home from an acute general care hospital. Methods: A comprehensive search of seventeen literature databases and twenty-five websites was performed for the period 1994–2004 to find relevant reviews. A three-stage inclusion process consisting of initial sifting, checking full-text papers on inclusion criteria, and methodological assessment, was performed independently by two reviewers. Data on effects were synthesized by use of narrative and tabular methods. Results: Fifteen systematic reviews met our inclusion criteria. All reviews had to deal with considerable heterogeneity in interventions, populations and outcomes, making synthesizing and pooling difficult. Although a statistical significant effect was occasionally found, most review authors reached no firm conclusions that the discharge interventions they studied were effective. We found limited evidence that some interventions may improve knowledge of patients, may help in keeping patients at home or may reduce readmissions to hospital. Interventions that combine discharge planning and discharge support tend to lead to the greatest effects. There is little evidence that discharge interventions have an impact on length of stay, discharge destination or dependency at discharge. We found no evidence that discharge interventions have a positive impact on the physical status of patients after discharge, on health care use after discharge, or on costs. Conclusion: Based on fifteen high quality systematic reviews, there is some evidence that some interventions may have a positive impact, particularly those with educational components and those that combine pre-discharge and post-discharge interventions. However, on the whole there is only limited summarized evidence that discharge planning and discharge support interventions have a positive impact on patient status at hospital discharge, on patient functioning after discharge, on health care use after discharge, or on costs. Page 1 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 and primary care providers [74,75], and many others Background Going back home from hospital is not always a smooth [76,77]. process. Many studies from all over the world have repeat- edly reported that many people who have been dis- These 'discharge interventions' mostly aim to smoothen charged from hospital to home, especially the elderly, the discharge itself (generally measured by length of stay encounter a variety of problems in the first weeks after and discharge destination) or to prevent, ease or solve their return home. Problems after discharge include problems in patient's functioning after discharge (gener- dependence on others with regard to household activities ally measured by function-measures) or to prevent [1-6], lower levels of independence in activities of daily readmissions to the hospital (which are generally seen as living and self-care deficits [2,3,5-12], difficulty with read- a proxy for patient problems after discharge) or to lower ing medication labels or instilling eyedrops [13,14], not health care costs, related to hospital readmissions and getting the help they needed [4,5,13,15-23], not being treatment of post discharge problems. aware of available services [24-26], informational needs [4,13,26-30], symptom distress [28,31-33], social prob- Reviews of these studies come to different conclusions on lems [34] and emotional problems as anxiety and uncer- the effectiveness of these interventions, varying from "Dis- tainty [7,29,35]. The post-discharge problems seem to be charge planning and support teams are cost effective and should more common with increased age and in women [36] and be in place universally" [78] to "The impact of discharge plan- may lead to further complications and unplanned hospi- ning on readmission rates, hospital length of stay, health out- tal readmissions. comes and cost is uncertain" [79] to 'In general, the evidence is a mixture of benefit, deficit and uncertainty, due to the com- In addition, lengths of hospital stay have dropped steeply plexity and variability of the interventions and methodological in the last few decades, e.g. from 6.5 days in 1985 to 4.8 problems with the evaluations' [80] and "Evidence from RCT's days in 2003 in the USA (with the greatest decline for peo- is not available to support the general adoption of discharge ple aged 65 years and older [37]), from 10.5 days in 1985 planning protocols, geriatric assessment processes or discharge to 6.9 days in 2003 in the European Union [38], and from support schemes as means of improving discharge outcomes" 12.5 days in 1985 to 7.3 days in 2003 in the Netherlands [81]. [39]. Consequently, the time available to a healthcare team to adequately prepare patients for discharge has vir- The mixed results of the reviews may, however, be caused tually evaporated [40]. by different study populations, heterogeneity of interven- tions, or a variety of outcomes that have been chosen. A Discharge planning and aftercare initiatives have received lot of questions with regard to the optimal content and much and increased attention over the past few years as a the organization of discharge planning and support result. Rorden & Taft defined discharge planning as 'a remain unanswered. We therefore set out a systematic process made up of several steps or phases whose imme- review of reviews dealing with discharge interventions. diate goal is to anticipate changes in patient care needs and whose long-term goal is to ensure continuity of As mentioned earlier, we defined discharge interventions health care' [41]. We defined discharge interventions as as in-hospital interventions or interventions after dis- in-hospital interventions or interventions after discharge charge performed (partly) by hospital-based profession- performed (partly) by hospital-based professionals, als, explicitly targeted to smooth the transition from explicitly targeted to smooth the transition from hospital hospital to home or to prevent or diminish problems after to home or to prevent or diminish problems after hospital hospital discharge. These can roughly be classified in two discharge. groups: Many studies were performed with various forms of dis- - Discharge preparation: interventions that mainly take charge planning and aftercare, e.g. screening patients with place during admission in the hospital, with the objective a high risk of post discharge problems [42,43], intensive of organizing care and preparing patients in such a way in-hospital discharge preparation [44], discharge rounds that the length of hospital stay is as short as possible for [45,46], transitional and intermediate care units [32,47- most patients, that the condition of most patients is such 50], written information leaflets [51], liaison nurses and that they can be discharged home and not into institu- discharge coordinators [52-55], clinical nurse specialists tional care, that they will need as little care as possible [56-58], home visits prior to discharge [59,60], preventive post discharge, and that care (organizations) needed after home visits of district nurses after discharge [61-63], post- discharge are informed and organized as well as possible, hospital support programs [7,64-68], telephone follow- so that patients will not have unmet needs, will not have up after discharge [69-72], discharge planning protocols to be readmitted and will not die due to complications or [18,73], ameliorated communication between hospital deterioration after discharge. Page 2 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 - Discharge support/aftercare: interventions that mainly take sionals providing services in the patient's home after dis- place after discharge from hospital and that are targeted to charge from hospital. prevent, ease or solve problems after discharge in order to prevent readmissions to hospital or admissions to institu- - 'Educational interventions' are interventions targeted at tional care and to maximize recovery and improve func- patients undergoing discharge from hospital that are tional, emotional, social and health status in the post- intended to improve their ability to manage aspects of discharge period. their care after discharge through the provision of infor- mation or more active education. The interventions may Besides this rough two categories classification system, we be limited to education, or supplemented by other activi- considered the categorization of discharge interventions ties such as home visits or telephone calls after discharge. put forward by Parker et al. [81] as a useful additional framework for ordering the results of the included The objective of this meta-review was to identify, appraise reviews. Parker et al. have four broad classes of 'discharge and synthesize the evidence presented in reviews of the lit- arrangements': comprehensive discharge planning proto- erature for the effectiveness of discharge interventions in cols, comprehensive geriatric assessment programmes, reducing post-discharge problems in adults discharged discharge support arrangements and educational inter- home from an acute general care hospital. In addition to ventions, all of which can be either generic or disease spe- problems in patient's functioning after discharge we cific. They define these as follows: sought for evidence about the effects of discharge inter- ventions on discharge status and on health care services - 'Comprehensive discharge planning protocols' are inter- use and costs after discharge. ventions involving standardised actions or interventions carried out by an individual, including assessment, coor- The following questions were addressed: dination and implementation of the discharge plan, which project post-discharge needs with the aim of pre- - What are the effects of 'discharge interventions' on the venting unnecessary readmission, maintaining the health discharge status of patients? status of patients or lessening carers' burdens. (length of hospital stay, discharge destination, depend- - 'Comprehensive geriatric assessment (CGA) pro- ency at discharge) grammes' are programmes based either in hospital or sup- porting older people recently discharged from hospital. In - What are the effects of 'discharge interventions' on the CGA programmes the multidisciplinary, multidimen- functioning of patients in the first 3 months after dis- sional nature of the assessment of health, rehabilitation charge? and social care needs is formalized, often using standard- ized assessment instruments. The results of these formal (physical status, emotional status, social status, health sta- assessments are then used either to inform or prompt tus) treatment and management recommendations, which may be carried out in dedicated inpatient units, provided - What are the effects of these interventions on health care as recommendations to the referring physician or team, or services use and costs in the first 3 months after discharge? delivered in the patient's home or other ambulatory care setting such as the day hospital or outpatient clinic. Dis- (readmissions, use of health care services post discharge, charge planning is usually regarded as an important com- costs) ponent of inpatient CGA programmes, although most are not focused on discharge itself, but on improving func- Outcomes in carers or relatives were not considered. tional health status, and thereby independent living, through medical intervention and rehabilitation. Methods Data sources - 'Discharge support arrangements' are schemes that are We searched for reviews of the literature and reviews that designed to provide support for (older) people after expe- are part of evidence-based guidelines containing synthe- riencing discharge from inpatient hospital care. These are sized evidence relating to discharge planning and support interventions in which hospital or community staff are in interventions aimed at preventing or diminishing prob- contact with the patient around the time of hospital dis- lems in adult patients following hospital discharge. charge, with the specific intention of providing support during the post-discharge period. The interventions may Searches were performed in seventeen literature databases be limited to a post-discharge telephone contact at one and on twenty-five websites, which are listed in Appendix extreme, or, at the other extreme, involve teams of profes- 1 (see Additional file 1). All databases were searched from Page 3 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 1994 (or from their inception if this was later than 1994) - The review has sufficient methodological quality (= until December 2004. Overview Quality Assessment Questionnaire score ≥ 5 [83-85]) A search strategy for PUBMED was developed; which was partly based on the search filters of the Dutch Cochrane Publications were excluded when: Centre for searching systematic reviews and for searching guidelines in PUBMED [82]. Suitable search strategies - They were primary research studies were developed for the other databases, as adaptations of the PUBMED search. No limits were applied where lan- - The outcomes in the review were only reported for carers guages were concerned. All detailed search strategies can or professionals be found in Appendix 2 (see Additional file 2). - The review involved only paediatric or psychiatric The words "discharge planning", "aftercare", "hospital patients discharge" and "continuity of care" (or equivalents in Dutch, French or German for the non-English sites) were - The review involved only emergency department (ED) sequentially entered in the search frame of the sites, for patients or one-day stay procedures the purpose of searching the websites to find systematic reviews as part of a guideline. - The review concerned interventions that are primarily intended to address the problems of caregivers rather than The hits of all searches were entered into Reference Man- of patients ager , duplicates were sifted out in this program, and the inclusion process were executed thereafter. - The experimental interventions discussed in the review are performed after discharge solely by primary care pro- Study selection viders The manuscripts had to fulfil all of the following criteria in order to be included: Since there is no generally accepted definition of what a postdischarge period means, and the duration of postdis- - The manuscript is a systematic review of the literature, charge problems may vary for different illnesses and treat- either as an independent manuscript or as a part of a ment procedures, the choice of a time period of 3 months guideline (we considered a review as a systematic review if as inclusion criterion had to be arbitrary. There is evi- at least two out of three of the following criteria were met: dence, however, that most postdischarge problems occur a search strategy was reported, a search was performed in in the period immediately after discharge: Naylor states in Pubmed at least, and the included studies were subjected her review [86] that '4 to 6 weeks post discharge represents a to some kind of methodological assessment) critical period when many elders are at highest risk for poor dis- charge outcomes' and empirical research in a mixed popu- - The review concerns 'discharge interventions' (= in-hos- lation has shown that postdischarge problems are greater pital interventions or interventions after discharge per- at 7 days post discharge than at 30 days post discharge formed (partly) by hospital-based professionals, explicitly [43]. Moreover, three months is a period for which it is targeted to smooth the transition from hospital to home reasonable to assume that outcomes can be related to the or to prevent or diminish problems after hospital dis- intervention around or in the first month after discharge. charge) A three-stage inclusion process was applied. Titles and - The interventions discussed in the review relate to adult abstracts of articles identified from the search strategies patients discharged home from an acute general care hos- were screened in the first stage of initial sifting, in order to pital, who were admitted for a primarily physical problem determine their relevance and whether they fulfilled the inclusion criteria. For each study the criteria were judged - The outcomes studied in the review concern patient sta- from top to bottom of the inclusion criteria referred to; no tus at discharge, patient functioning after discharge, or further analysis was done on the subsequent criteria as health care service use and costs after discharge soon as one criterion was not met. In this first stage (which is more focused on excluding than on including), - The outcomes studied in the review are measured within one reviewer screened all references and the second 3 months after discharge from hospital reviewer independently checked a 10% random sample of the references. If agreement between the two reviewers on - None of the exclusion criteria listed below are met whether to exclude studies was lower than 95% for the 10% sample, the second reviewer would proceed to check Page 4 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 the other 90% of the sample. In addition, 10% of the ref- comes, and selected patient populations, effects on erences that were excluded by the first reviewer were patients, effects on health care use and costs. checked by a second reviewer. When the title and/or abstract provided insufficient information to determine As stated earlier two categorizations for the interventions relevance, full paper copies of the articles were ordered were used to organize the data. Firstly, the rough two cat- and they proceeded to the second stage. In case articles egories system of discharge interventions, divided in dis- were published in a language in which the reviewers were charge preparation and discharge support interventions; not fluent, assistance was sought from other colleagues secondly the categorization of Parker et al[81], who dis- who mastered that language. tinguish four broad classes of 'discharge arrangements': comprehensive discharge planning protocols, compre- In the second stage, two reviewers independently exam- hensive geriatric assessment programmes, discharge sup- ined all full paper copies of the articles selected in the first port arrangements and educational interventions, all of stage, in order to determine whether they fulfilled the which can be either generic or disease specific. The defini- inclusion criteria. tions of each category are already given in the Background of this article. The criteria were again judged from top to bottom for each study; no further assessment was done on the subsequent The outcomes were classified according to the research criteria as soon as a criterion was not met. Any disagree- questions: ments were resolved by discussion between the two reviewers; if no agreement could be reached, a third - The discharge status of patients: length of hospital stay, reviewer decided. discharge destination, dependency at discharge The third stage of inclusion related to the methodological - The functioning of patients in the first 3 months after assessment of the reviews. All reviews remaining after the discharge: physical status, emotional status, social status, second stage were assessed with the Overview Quality health status Assessment Questionnaire [83-85]. This instrument is one of the most frequently used appraisal instruments for sys- - Health care services use and costs: readmissions, use of tematic reviews in the biomedical literature [87], besides health care services post discharge, costs being one of the few found for which psychometric prop- erties had been documented [88] and which had been Physical status concerns all measures about level of activ- found to meet several important criteria, such as construct ities of daily living, self-care abilities, self efficacy or inde- validity, inter-observer reliability and coverage of the pendence. Emotional status concerns all measures about items in the QUORUM statement for reporting systematic the level of well-being of patients such as uncertainty, anx- reviews [89]. Scores on this instrument can vary from 1 iety, depression, informational needs, mood or coping. (extensive flaws) to 7 (minimal flaws). Two reviewers per- Social status refers to the extent a patient is able to partic- formed this assessment independently. The mean of the ipate in normal social activities and relationships. Health scores of the two reviewers was computed and classified as status concerns symptom prevalence and burden, organ the final quality judgment; in case the scores of the review- dysfunction, mortality, morbidity and physical complica- ers differed more than 2 points, reviewers discussed their tions. However, these categories are not always mutual assessments and came to a new joint score (this was only exclusive, e.g. in the case where multi-dimensional quality needed once, mean difference score was 0.91). of life measures were used. Only high quality reviews (= with mean scores of 5 Whether an outcome was regarded as a positive or a neg- (minor flaws) and above) were used for the data-extrac- ative effect, was primarily based on the perspective and tion, as is proposed by Jadad et al. [90] and Peach [91], definitions used by the review authors. However, in gen- since it is known that low quality reviews may reach dif- eral a shorter length of hospital stay, home as discharge ferent conclusions than high quality reviews [92-94], and destination, better physical, emotional and social func- also to avoid false conclusions that are based on low qual- tioning, better health status, less readmissions, less use of ity evidence. health care services and less costs were regarded as posi- tive outcomes by the review authors, and consequently by Data-analysis and synthesis us. Data were extracted about the applied in- and exclusion criteria for the primary studies, search strategies, studied Data-analysis was done primarily by description of the interventions, time frame of the searches, selected out- interventions and by making cross-tables for the different Page 5 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 interventions, populations and effects. No quantitative Type and number of studies included in the reviews pooling was performed across the reviews. Since all included reviews were focused on effectiveness, all reviews limited their inclusion criteria to comparative Conclusions for the meta-review were based on the con- research designs. Seven reviews [81,98,102,104-107] were clusions and results of meta-analyses presented in the limited to randomized controlled trials only, while the reviews studied. other eight also included other comparative designs, such as quasi-randomized trials, non-randomized comparative studies and before-after designs. Two review authors Results Search and inclusion results [97,103] additionally searched for other reviews and After duplicates had been removed, the searches in the dif- guidelines and used these to reach their conclusions. ferent databases resulted in an initial set of 7442 refer- ences of potential interest. Initial sifting based on title and The fifteen reviews included a total of 265 different pri- abstract reduced this set to 117 references. As said, the first mary studies, the number of primary studies included in reviewer carried out this process and a 10% random sam- an individual review varying from 8 [98] to 71 [81]. Most ple was also done independently by a second reviewer (200 of the 265) of the primary studies were included (crude agreement between reviewers was 99% with a only once in a review, with the exception of a few papers kappa coefficient of 0.33). In addition, when a second that were included in more than one review, extending to reviewer checked a 10% random sample of the excluded four inclusions for ten primary studies and with a maxi- references, discussion was only needed for two references mum of five inclusions for two primary studies. A list of and resulted in an exclusion-decision. The set of the 117 all primary studies included in one of the reviews can be references, representing 108 reviews, was ordered full text found in Appendix 4 (see Additional file 4). for the second stage of the inclusion process. Two review- Aims of the reviews ers performed this second phase independently; agree- ment between reviewers in this phase was 79% with a The aims of the reviews included are all related to the kappa coefficient of 0.56. Discussion was needed for 23 effectiveness of discharge interventions, but there is a wide references and agreement was subsequently reached. A set variation in what review authors describe as their objec- of 49 references, representing 41 reviews, finally proved to tives, as can be seen in Table 2. fulfil the inclusion criteria for type and content of study. Patients of interest in the reviews In the following stage, two reviewers independently Some of the reviews included studies in which interven- assessed the remaining 41 reviews on their methodologi- tions targeted several or mixed patient populations, while cal quality, using the Overview Quality Assessment Ques- others were restricted to studies with a specified patient tionnaire [83-85] proposed by Oxman. A mean of the two group only (e.g. stroke patients, hip fracture patients, eld- scores was computed and classified as the final quality erly or patients with heart failure). A combination was judgment. Twenty-six reviews had a mean quality score sometimes made of elderly patients and a specific medical lower than 5 and were excluded, while the remaining fif- condition. An overview is presented in Table 3. teen high quality reviews [79,81,95-107] advanced to the next stage of the review, for data-extraction and analysis. Interventions studied in the reviews As said, we used two categorization systems for the dis- The flow diagram of the inclusion process is shown in Fig- charge interventions. For this paragraph only the results ure 1. References of the studies excluded and the reason for the rough two categories system is presented. The for exclusion can be found in Appendix 3 (see Additional grouping of the results by the second categorization sys- file 3). tem of Parker et al. [81] is presented in the more detailed section about the effectiveness of interventions later on. Characteristics of the final 15 reviews Publication date of the reviews and the journals in which they were According to the first system discharge interventions are published classified into two groups, discharge preparation and dis- All reviews included date from 2000 or later and five were charge support interventions. published in 2004. The oldest reference included in a review dates from 1964 and the most recent one from Some of the reviews included only studies that used inter- 2004. Search periods for each review are shown in Table 1. ventions from the first group, others only included studies that used interventions from the second group, and a The reviews were published in eight different journals; six third category comprised reviews that included studies in reviews [79,95,99,101,102,106] were published as a which interventions from both groups had to be applied. review in the Cochrane Database of Systematic Reviews. The focus of the reviews is shown in Table 4. Page 6 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Database searches, number of hits BIOMED C2-RIPE CDSR CINAHL DARE ERIC EMBASE GUIDELINE HTA 124 0 7 75 2 1 2202 20 2 INVERT LILACS NEED PICARTA PSYCH PUBMED SOCIO SCI-E VHL 2 6 26 41 66 4823 0 339 23 references ↓ After eliminating duplicates TITLE & ABSTRACT 7742 references ↓ -7625: one or more inclusioncriteria not met references 117 references, representing 108 studies studies FULL TEXT ↓ -35: Not a systematic review 73 studies -26: Review does not contain manuscripts concerning discharge interventions in adults explicitly targeted to smoothen transition from hospital to home 47 studies -5: Does not concern discharge home from an acute general hospital 42 studies ↓ -1: No outcomes measured within 3 months 41 studies ↓ -26: Mean methodological score <5 FINAL SET 15 studies F Figure 1 low diagram of the inclusion process Flow diagram of the inclusion process. Page 7 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Table 1: Search periods in included reviews Review Search period Cameron 2002 inception-2002 Cole 2001 1975–2000 Day 2004 1980–2003 Gwadry 2004 inception-2000 Handoll 2004 inception-2004 Hyde 2000 inception-1997 Kwan 2002 1975–2003 Outpatient Service Trialists (OST) 2003 inception-2001 Parker G 2000 1988–1999 Parker S 2002 inception-2001 Phillips 2004 inception-2003 Richards 2003 inception-2000 Shepperd 2001 inception-2001 Shepperd 2004 inception-2002 Teasell 2003 1995–2002 Interventions included in discharge preparation reviews (hospital at home), educational interventions and inten- were care pathways, patient management schemes, spe- sified rehabilitation/(physio)therapy schemes. cialized units (for stroke, hip fracture or geriatric patients for example), geriatric assessment and/or consultation, The interventions included in a particular review showed discharge coordinators, nurse specialists, educational considerable heterogeneity in terms of what exactly was interventions, intensified rehabilitation/(physio)therapy done, by whom it was done, the way it was done, the fre- schemes, adjusting skill-mix of hospital professionals, quency with which it was done, and the duration of the and discharge plans. intervention. Interventions included in the discharge support reviews Control conditions in the reviews were telephone follow-up, home visits, geriatric assess- Most reviews included studies in which patients in the ment and/or consultation, intensified post-discharge care control condition received usual care (according to the Table 2: Aim of review, as worded by review-authors Review Aim Cameron 2002 to examine the effectiveness and cost effectiveness of specialised multidisciplinary inpatient rehabilitation supervised by a geriatrician or rehabilitation physician compared with usual (orthopedic) care, for older patients with proximal femoral fracture Cole 2001 to determine the impact of geriatric post-discharge services on mental state Day 2004 to provide the evidence base on the effectiveness of specialist geriatric services for developing a sound practice framework Gwadry 2004 to evaluate the effectiveness of multidisciplinary heart failure management programs on hospital admission rates Handoll 2004 to evaluate the effects of different mobilisation strategies and programmes after hip fracture surgery Hyde 2000 to investigate the effects of supported discharge after an acute admission in older people with undifferentiated clinical problems Kwan 2002 to assess the effects of care pathways, compared with standard medical care, among patients with acute stroke who had been admitted to hospital. In particular we aimed to assess the effects on functional outcome, process of care, quality of life and the hospitalisation costs OST 2003 to assess the effects of therapy-based rehabilitation services targeted towards stroke patients resident in the community within one year of stroke onset or discharge from hospital following stroke Parker G 2000 to establish both the volume and strength of existing evaluative research on the costs, quality and effectiveness of different locations of acute, post- and subacute and rehabilitation care for older people Parker S 2002 to test the following hypotheses: 1. There is an inadequate number of comparable rct's to allow a definitive analysis; 2. Hospital discharge process, outcome and cost-effectiveness can be improved through the use of a variety of interventions; 3. Some interventions are more effective than others; 4. there are priority areas for future research Phillips 2004 to evaluate the effect of comprehensive discharge planning plus post-discharge support in patients with chronic heart failure on the rate of readmission, all cause mortality, length of stay, quality of life and medical costs Richards 2003 to determine the effectiveness and costs of interventions intended to improve access to health and social care for older patients following discharge from acute hospitals Shepperd 2001 to assess the effects of hospital at home compared with in-patient hospital care Shepperd 2004 to determine the effectiveness of planning the discharge of patients moving from hospital Teasell 2003 to assess the effectiveness of early supported discharge programs in the context of stroke rehabilitation Page 8 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Table 3: Patients of interest in the reviews Review Several/mixed Elderly Stroke patients Patients with hip or femur fractures Patients with heart failure Cameron 2002 X Cole 2001 X Day 2004 X Gwadry 2004 X Handoll 2004 X Hyde 2000 X Kwan 2002 X OST 2003 X Parker G 2000 X Parker S 2002 X Phillips 2004 X Richards 2003 X Shepperd 2001 X Shepperd 2004 X Teasell 2003 X Total 2 6 3 2 2 trial authors); other reviews included studies in which the relevant intervention. Many of the outcomes, in both the different interventions were compared against each other primary studies and the reviews, lacked a clear definition, (e.g. different rehabilitation/therapy schemes). The prob- however, e.g. functional status or quality of life or mental lem with the first category for all review authors was that state. In addition, different terms were used across pri- the trial authors were not clear on what constituted 'usual mary studies and reviews for outcomes that are related or care'. that are probably the same (e.g. physical status or func- tional status or ability in activities in daily living). Above Outcomes studied in the reviews this, even similar outcomes were measured with different Some of the included reviews had well described primary (frequently not validated) instruments at different times outcomes that to had be described in the trials before they post discharge, posing problems for the review authors in could be included, while others had no criteria at all with combining the effects across trials, but also in combining regard to outcomes as long as the studies dealt with the the results from reviews for this meta-review. Table 4: Focus of interventions in reviews Review Focus on discharge preparation Focus on discharge support/aftercare Cameron 2002 X Cole 2001 X Day 2004 X X Gwadry 2004 X Handoll 2004 X X Hyde 2000 X Kwan 2002 X OST 2003 X Parker G 2000 X Parker S 2002 X X Phillips 2004 X X Richards 2003 X X Shepperd 2001 X Shepperd 2004 X Teasell 2003 X Total 8 13 Page 9 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Effectiveness of the discharge interventions Effect of discharge interventions on physical status after discharge General picture The effect of interventions from the discharge planning Although a statistically significant effect was occasionally category on physical status in the first 3 months after dis- found for a particular intervention on a particular out- charge was studied in three reviews [79,81,105]. Parker et come, most review authors reached no firm conclusions al. [81] included RCT's only and found eight articles rep- that the discharge interventions they studied were effec- resenting seven studies in which discharge planning was tive. Only two review authors [104,105] were firm in their studied. All studies involved patients who had experi- conclusions. The conclusions as formulated by the enced discharge from an acute inpatient hospital stay and authors are shown in Table 5, with formulations indicat- evaluated a comprehensive discharge protocol imple- ing no effects or inconclusive ones are shown in italics and mented by an individual who was either a specialist nurse, formulations indicating firm conclusions are shown in a social worker or an admitting clerk. The comprehensive bold typeface. discharge protocols were similar in design and were com- pared with usual discharge care. The protocols all had Effect of discharge interventions on discharge status similar elements, including the assessment of patients, Length of stay was studied in nine reviews. The findings liaising with the patient's carer and other professionals to were inconclusive in four reviews [95,97,101,107], no sig- coordinate discharge and providing follow-up visits or tel- nificant differences were found in another four reviews ephone calls. Only two of the seven studies included in [79,81,99,104] and one review [106] concludes that hos- this part of the review considered outcomes related to pital length of stay was significantly shorter for 'hospital- physical function. No differences were found between at-home' interventions. experimental and control groups within 3 months after discharge. Richards and Coast [105] included five RCT's Discharge destination was studied in six reviews. Findings dealing with comprehensive discharge planning and came were inconclusive in one review [97] and no significant to the same conclusion as Parker et al. that no differences differences were found in four reviews [79,81,101,106], had been shown with regard to physical status. Shepperd while one review [103] found a significant difference in et al. [79] included 11 RCT's, six of which presented data the number of patients being discharged home when they concerning physical status. Here too, no effects of dis- were cared for at a stroke unit (based on three trials) but charge planning on physical status were found. not when they were treated in hip units or geriatric units. So, these three reviews discussing the impact of discharge Dependency at discharge was studied in one review [101] planning on physical status after discharge are mutually and it was found, on the basis of two studies (one rand- consistent and all conclude that no effect of discharge omized and one non-randomized) that patients from the planning has been demonstrated on physical status. care pathway group were more dependent at discharge than the control group. The effect of interventions from the comprehensive geriatric assessment category on physical status in the first 3 months There is no evidence on the whole that discharge interven- after discharge was studied in three reviews on generic tions have a positive impact at length of stay, discharge patient populations [81,97,105] and in one review on destination, or dependency at discharge. patients with femoral fractures [95]. Day and Rasmussen [97] conclude that measures of functional status were sim- Effect of discharge interventions on patient functioning after ilar and showed no significant difference between the discharge intervention and control groups. Parker et al. [81] point to As was specified in the second research question, patient the great variety of measures used to report physical func- functioning after discharge was divided into four types: tion outcomes, making comparisons and pooling diffi- physical, emotional, social and health status. The effects cult. They say that the majority of studies appeared to have of the discharge interventions are given for each of these, found no significant differences in the physical function and subdivided according to the intervention classifica- outcomes of study patients and control patients over time. tion scheme put forward by Parker et al. [81], in which With regard to improvement in physical function over there are four broad classes of discharge interventions: time, Parker et al. were able to calculate an odds-ratio over comprehensive discharge planning protocols, compre- six studies and found a significant effect suggesting that hensive geriatric assessment programmes, discharge sup- the intervention was beneficial for physical functioning. port arrangements and educational interventions, all of These outcomes, however, were not measured within our which can be either generic or disease specific. stated timeframe of 3 months post discharge. Richards and Coast [105] included two studies in which functional status outcomes were measured within the 3 months after discharge and both found no differences. Finally, Cam- Page 10 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Table 5: Conclusions in included reviews Review Conclusions Cameron 2002 The available trials had different aims, interventions and outcomes. Combined outcome measures (e.g. death or institutional care) tended to be better for patients receiving coordinated inpatient rehabilitation, but the results were heterogeneous and not statistically significant. Cole 2001 There is little evidence that geriatric post-discharge services have an impact on the mental state of aged subjects. Day 2004 This review generally supports the efficacy of specialist geriatric team services trained in geriatrics with a multidisciplinary collaborative focus undertaking assessment, rehabilitation and coordinated case management in community settings; both preventive care and supportive discharge in these settings appear to provide greater benefit over usual care; however these benefits are not consistent across all outcomes and although improvement in outcomes was often apparent, these were not always significant when compared with the comparison group. Efficacy of specialist geriatric services for inpatient settings was more diverse; this was due to the diversity of studies across the continuum of subacute, acute, postacute care in unit or ward settings with resulting heterogeneous outcomes and only some of these outcomes showing significance over usual care. With regard to day hospital and outpatient care, evidence for the efficacy of specialist geriatric services was lacking, with no conclusive evidence that the services are of greater benefit than usual care. Gwadry 2004 This review suggests that specific heart failure targeted interventions significantly decrease hospital readmissions but do not affect mortality rates. Handoll 2004 There is insufficient evidence from randomised trials to determine the effectiveness of the various mobilisation strategies that start either in the early post-operative period or during the later rehabilitation period Hyde 2000 We believe that the results of this review provide reassurance that supporting discharge from hospital to home is of value. However, important sources of uncertainty remain, suggesting the need for further research. There was relative certainty that the proportion of those at home 6–12 months after admission is greater with supported discharge; this was associated with a consistent pattern of reduction in admission to long-stay care over the same period, without apparent increases in mortality. There was uncertainty about the effect of supported discharge on hospitalization. There were no rigorous data on functional status, patient and carer satisfaction and in consequence uncertainty about the overall effectiveness of supported discharge. Kwan 2002 Use of stroke care pathways may be associated with positive and negative effects. Since most of the results have been derived from non-randomised studies, they are likely to be influenced by potential biases and confounding factors. There is currently insufficient supporting evidence to justify the routine implementation of care pathways for acute stroke management or stroke rehabilitation. OST 2003 Therapy-based rehabilitation services targeted towards stroke patients living at home reduces the odds of a poor outcome and has a beneficial effect on a patient's ability to perform activities of daily living. However, the evidence is derived from a review of heterogeneous interventions and therefore further exploration of the interventions is justifiable. Parker G 2000 Despite considerable recent development of different forms of care for older patients, evidence about effectiveness and costs is weak. However, evidence is also weak for longer-standing care models. Parker S 2002 The evidence from these trials does not suggest that discharge arrangements have effects on mortality or length of hospital stay. This review supports the concept that arrangements for discharging older people from hospital can have beneficial effects on subsequent readmission rates. Interventions provided across the hospital-community interface, both in hospital and in the patient's home, showed the largest effects. Evidence from RCT's is not available to support the general adoption of discharge planning protocols, geriatric assessment processes or discharge support schemes as means of improving discharge outcomes. Phillips 2004 Comprehensive discharge planning plus postdischarge support for older people with chronic heart failure significantly reduced readmission rates and may improve health outcomes such as survival and quality of life without increasing costs. Richards 2003 The interventions provided and patient groups targeted by these services were heterogeneous. There was, however, some evidence that services combining needs assessment, discharge planning and a method for facilitating the implementation of these plans were more effective than services that do not include the latter action. The assessment of need may be insufficient in itself for the adequate provision of post-discharge care; needs assessment should be combined with a service that facilitates the implementation of care plans. Shepperd 2001 This review does not support the development of hospital at home services as a cheaper alternative to in-patient care. Early discharge schemes for patients recovering from elective surgery and elderly patients with a medical condition may have a place in reducing the pressure on acute hospital beds, providing the views of the carers are taken into account. The evidence supporting hospital at home for patients recovering from stroke is conflicting. There is some evidence that admission avoidance schemes may provide a less costly alternative to hospital care. Shepperd 2004 The impact of discharge planning on readmission rates, hospital length of stay, health outcomes and cost is uncertain. Teasell 2003 Although the majority of studies reported no statistically significant differences in functional outcomes between the two groups, there was a reduction in hospital stays for patients receiving home-based therapy. These results suggest that patients with milder strokes who receive home-based therapies have similar functional outcomes to patients who receive traditional inpatient rehabilitation. There is strong evidence that high-level stroke patients discharged from an acute hospital unit can be rehabilitated in the community by an interdisciplinary stroke rehabilitation team without negative consequences. These patients attain similar functional outcomes compared to patients with equivalent stroke severity who receive inpatient rehabilitation. Community based programs also appear to reduce hospital length of stay, although we do not have evidence of an overall cost reduction. Although the effectiveness of early supported discharge programs for patients with moderate-to-severe deficits has not been well studied, limited evidence suggests that these patients are unsuitable candidates and should receive inpatient rehabilitation instead. Page 11 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 eron et al. [95] examined the effects of coordinated multi- outcome measures. It was found on the basis of twelve tri- disciplinary inpatient rehabilitation by a geriatrician or als that patients who received therapy-based rehabilita- rehabilitation physician compared with usual care for tion services after stroke were significantly more older patients with hip fracture, and they state that the independent in personal activities of daily living than available trials reviewed had a variety of aims, interven- those patients who received no care or usual care. Most of tions and outcomes, making them difficult to combine. the studies measured this outcome at 6 or 12 months after They conclude on the basis of nine trials that functional starting the therapy, however, and it is not clear how long status did not improve consistently. this was after hospital discharge; no (pooled) data at 3 months post discharge are given in this review. Teasell et On the basis of these four reviews, therefore, it appears al. [107] studied the effectiveness of early supported dis- that comprehensive geriatric assessment has not been charge programs in stroke patients. Ten studies were shown to have a positive impact on functional status included, eight of which reported some kind of functional within 3 months after discharge, in comparison with the outcome. None of these studies reported statistically sig- control groups. nificant differences between the treatment groups, indi- cating that functional outcome was not affected negatively The effect of interventions from the discharge support cate- or positively by the intervention. Pooling was not per- gory on physical status after discharge was studied in four formed in this review. generic [81,100,105,106] and two disease specific reviews [102,107], both in stroke patients. Hyde et al. [100] inves- On the basis of these six reviews, therefore, there are no tigated the effects of supported discharge after an acute indications that patients who receive supported discharge admission in older people with undifferentiated clinical have a better physical status at 3 months after discharge problems, in which supported discharge was defined as than patients from the control groups. actual additional support from any source provided to patients or their carers and commencing within one week The effect of educational interventions on physical status of discharge following an acute admission. They included after discharge was covered by two reviews [81,105]. nine studies of which six provided data on functional sta- Parker et al. [81] studied if education interventions tus; however, there were no rigorous data on functional improved the outcome of discharge of elderly people status that made pooled conclusions possible. Parker et al. from hospital; the interventions studied were described as [81] point to the wide range of types of intervention, var- mainly educational and could be limited to education or ying from a single phone call after discharge to complex supplemented by other activities, such as home visits or multidisciplinary interventions. They included twenty- telephone calls after discharge. Eleven studies were eight controlled trials, nineteen of which reported on included, two of which contained data on physical status; some aspect of physical functioning and eight of which one study found better results in the intervention group, were comparable enough to pool, but showed no signifi- but the other study found no effects. Richards et al. [105] cant effect on physical functioning. Richards and Coast studied discharge co-ordinator roles, which may incorpo- [105] evaluated the effectiveness of organizational inter- rate educational interventions. Five studies were included, ventions that influence access to health and social care four of which contained data on physical status after dis- after discharge. They found considerable heterogeneity in charge; none of these found significant differences the content of interventions and the selection of patient between experimental and control groups. groups. They identified two trials that reported on func- tional status within 3 months of discharge, but both of On the basis of these two reviews, therefore, there are no these were inconclusive and did not suggest improve- clear indications that educational interventions have an ment. Shepperd et al. [106] assessed the effects of hospi- effect on physical status after discharge. tal-at-home compared with in-patient hospital care. Sixteen studies were included, eight of which measured Finally, Handol et al. [99] studied mobilisation strategies functional status in elderly medical patients and two trials in hip fracture surgery patients. They conclude that there in patients following elective surgery. Although pooling is insufficient evidence from randomized trials to deter- was not possible, there were no indications that the func- mine the effectiveness of the various mobilization strate- tional status in the intervention groups was better at 3 gies. months post discharge. The review of the Outpatient Serv- ice Trialists [102] considered interventions targeting In summary, we found no evidence base that discharge stroke patients resident in the community setting. Four- interventions have a positive impact on the physical status teen trials were included, twelve of which involved of patients after discharge. All the reviews included, how- patients who had experienced discharge from hospital; ever, had to contend with extensive heterogeneity in inter- the trials included used a large number of heterogeneous ventions, patient populations, and outcomes scales and Page 12 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 times and with inadequate descriptions of control condi- ety of ways and in multiple domains, making tions, all of which made pooling difficult. interpretation or synthesis across studies problematic, and that in general, these measures remained unchanged Effect of discharge interventions on emotional status after discharge between intervention and control groups. In addition, The effect of interventions from the discharge planning cat- Parker et al. refer to sixteen trials measuring dimensions of egory on emotional status after discharge was studied in quality of life, which may incorporate emotional status. three reviews [79,81,105]. Parker et al. [81] found one dis- Here too, they found many different instruments and that charge planning study that included emotional status out- the data on the whole did not suggest that discharge sup- comes, which stated that mean satisfaction scores changed port arrangements had a major impact on the quality of little over time. Richards and Coast [105] included two life of subjects when compared to controls. Finally, Parker studies that reported emotional function outcome within et al. refer to six trials in which satisfaction was recorded. 3 months and both found no differences. Shepperd et al. Four of the trials suggested some increased satisfaction [79] found two studies containing some kind of emo- with the service provided, but the data were neither con- tional function; one found some improvement on one sistently nor reliably reported. Richards and Coast [105] parameter but not on two other emotional outcomes, included two trials in this category; neither of which while the second study failed to detect a difference. found differences in emotional status outcomes. To the extent that early discharge can be regarded as 'discharge On the basis of these three reviews, therefore, there are no support', Shepperd et al. [106] found eight trials involving indications that discharge planning affects emotional medical patients in which some dimensions of psycho- functioning after discharge. social well-being or quality of life were measured. Six failed to detect a difference between intervention groups The effect of interventions from the comprehensive geriatric and control groups, while two studies reported more psy- assessment category on emotional status after discharge was cho-social dysfunction for the intervention group. Two tri- covered by two reviews [81,105]. Parker et al. [81] found als involving surgery patients were included and failed to eight studies reporting on aspects of emotional status, detect differences in this dimension. With regard to only one of which reported a significantly greater patient satisfaction, there was a mixed and ambivalent improvement in cognitive scores in the intervention picture, but satisfaction tended to be higher in the hospi- group than found in the controls. On the whole, however, tal-at-home groups. No pooling was possible on these var- the outcomes of intervention and control group patients iables. The Outpatient Service Trialists [102] pooled were broadly similar, with no obvious benefit observable results from five studies of quality of life in stroke patients for patients undergoing comprehensive geriatric assess- and found no significant difference between experimental ment. Richards and Coast [105] included three studies in groups and control groups, which also applied to the find- which some emotional outcome was reported within 3 ings of six studies in which mood/distress was measured. months after discharge, but none of the three found dif- ferences between intervention and control groups. On the basis of these five reviews, therefore, there are no indications that discharge support interventions enhance On the basis of these two reviews, therefore, there are no emotional functioning after discharge. indications that comprehensive geriatric assessment has a positive impact on emotional status after discharge. The effect of educational interventions on emotional status after discharge was covered by two reviews [81,105]. The effect of interventions from the discharge support cate- Parker et al. [81] found three studies of educational inter- gory on emotional status after discharge was studied in ventions that investigated the effect on emotional func- four generic reviews [81,96,105,106] and in one disease tion; pooling was impossible and the effects were mixed: specific review [102]. Cole [96] found eleven trials report- one study found no differences except for increased self- ing emotional status outcomes after geriatric post-dis- efficacy for walking; the second study had no measure- charge services, with the type of intervention and the type ments after discharge, and the third study, in which an of emotional status outcomes varying from one study to education intervention in hospital was supported with the next. Emotional status outcomes included depression, extensive telephone follow-up after discharge, showed sig- morale, life satisfaction, contentment, emotional func- nificantly lower levels of anxiety and a higher level of tion, self perceived health or cognition. Three trials knowledge at 6 weeks after discharge. They also found reported small effects and eight reported no effect. Parker four studies that considered the effect of educational inter- et al. [81] found nine trials reporting on emotional func- ventions on adherence to medication advice, in which dif- tioning, including cognitive function (five trials) and ferent measures were used to assess adherence, including measures of anxiety (three trials) or depression (two tri- tablet counts, self-reports of compliance and knowledge als). They state that emotional status is measured in a vari- of medication regimens. All but one of these studies Page 13 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 showed some improvements in adherence to medication control patients. The other four reviews, however, found or knowledge and it is concluded that more intensive no differences with regard to social status after discharge. interventions appear to be relatively effective, but that brief counselling or education is of little effect. Richards Effect of discharge interventions on health status after discharge and Coast [105] studied discharge coordinator roles, Mortality is certainly the outcome that has been looked at which may incorporate educational interventions. Five most frequently in the reviews, regardless the focus of the studies were included, three of which contained data on interventions. Most of the reviews (and the underlying tri- emotional status after discharge, and none of these found als), however, looked at mortality over more extended significant differences between experimental groups and periods of time than the 3 months that are of interest in control groups. this meta review; mortality was mostly measured at 6 or 12 months. Twelve reviews [79,81,95,98-106] found no On the basis of these two reviews, therefore, it appears significant differences in mortality and only Day and Ras- that educational interventions might have some effect on mussen [97] conclude that stroke units showed significant aspects of emotional status after discharge, on knowledge benefits in terms of mortality reduction, but do not spec- and medication adherence, but the results of the reviews ify the trials on which this conclusion is based. are not straightforward and the effects seem to depend on the dose and format of the educational interventions. The four reviews [95,97,99,106] in which morbidity or com- plications after discharge was studied and that were able to In summary, discharge interventions appear to have no include trials, found no significant differences. effect, or only a very limited one, on the emotional status after discharge. In summary, we found no firm evidence that discharge interventions have a positive impact on health status of Effect of discharge interventions on social status after discharge patients after discharge. Data on the effect of interventions from the comprehensive discharge planning category on social status in first 3 Effect of discharge interventions on health care use after discharge months after discharge were found in one review [105]. and costs Readmissions were measured in eleven reviews, but the Richard and Coast [105] included five studies with com- prehensive discharge planning coordinators, and none measurement period was frequently 6 or 12 months and found differences in social support experienced. not the 3 months that is of interest for this meta-review. The effect of interventions from the postdischarge support Seven reviewers [79,95,97,100,102,103,105] are incon- category on social status was found in three reviews clusive about the effect of discharge interventions on [81,103,106]. Parker G et al. [103] report about three tri- readmission rates. One reviewer [106] found no statisti- als in which there was no difference in patients at home at cally significant difference for patients in a hospital-at- 3 months. Parker S et al. [81] found no statistical differ- home intervention. Three reviews [81,98,101] found a ence between experimental groups and control groups in positive effect on readmissions. Parker et al. [81] reviewed number of patients being at home, based on six trials that four types of discharge interventions and conclude that measured this within first six months. On the basis of when all interventions groups are taken together, the three trials, Shepperd et al. [106] found a significantly patients in the intervention groups have a significant larger number of patients from the hospital-at-home lower risk of being readmitted and this was more marked group being at home at 6 weeks. among interventions provided both at hospital and at home. In the subgroups they did not find a significant dif- No reviews discussed effects of interventions from geriatric ference for discharge planning activities, discharge sup- assessment category or of educational interventions on port or geriatric assessment but they did find a significant social status. difference in favour of patients receiving some kind of educational intervention. This is congruent with the posi- Finally, Handoll et al. [99] mention one small trial, in tive finding of Gwadry et al. [98], that patients receiving a which no difference was found in loss of social independ- heart failure management program are less frequently ence between intensive physical training and placebo readmitted. Finally, Kwan and Sandercock [101] found activities started post discharge. fewer readmissions for patients that were cared for in a stroke care pathway. In summary, there is a little bit of evidence, based on one review [106], that patients treated in hospital-at-home Three reviews [81,105,106] had included and discussed interventions more frequently remain at home than the trials relating to the use of services after discharge and all were inconclusive on this subject. Page 14 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 All reviewers comment on the variety of ways that costs, ciplinary heart failure management programs on hospital cost-benefit and cost-effectiveness were measured in the readmission rates and found a significant decrease in trials, making synthesis difficult. Costs are also largely these rates. Phillips et al. [104] also conclude that compre- dependent on the organization of health care in an indi- hensive discharge planning plus postdischarge support for vidual country, making cross-country synthesis difficult. older people with chronic heart failure significantly reduced readmission rates, and may improve health out- With this in mind, all reviewers comes such as survival and quality of life without increas- [79,81,95,97,99,101,103,105-107] who report on costs ing costs. Based on above two reviews, it appears that are inconclusive about the impact of discharge interven- readmissions in heart failure patients can be reduced by tions on costs. some kind of intervention. In summary, there is little evidence that discharge inter- Discussion ventions have an impact on health care use after dis- We found more than forty systematic reviews of discharge charge, or on costs, except that educational interventions interventions, fifteen of which scored highly on method- may reduce readmissions in heart failure patients. ological quality. Our conclusions on the basis of these fif- teen reviews, is that there is only limited evidence for the Effects of discharge interventions in specific patient groups positive impact of discharge interventions. We found a Three reviews [101,102,107] focused on stroke patients and few indications that discharge interventions may be effec- compared several care delivery models and rehabilitation tive. Three reviews [81,104,105] state that effects are services. The main aim of this group of studies was more mainly observed when interventions from the discharge on the post-discharge period than on the discharge itself. planning and discharge support side were combined Kwan and Sandercock [101] conclude that stroke care across the hospital-home interface. In addition, two pathways may be associated with positive and negative reviews [81,105], appear to show that educational inter- effects and that there is currently insufficient evidence to ventions might have some effect on aspects of the emo- justify the implementation of care pathways for acute tional status after discharge, on knowledge and stroke management or stroke rehabilitation. The Outpa- medication adherence. tient Service Trialists [102] conclude that therapy-based rehabilitation services targeting stroke patients living at The limited evidence about effectiveness of discharge home reduce the odds of a poor outcome and have a ben- interventions may be due to the heterogeneity of several eficial effect on a patient's ability to perform activities of aspects which review authors had to deal with. All review daily living. They warn, however, that the evidence is authors were confronted with heterogeneity in interven- derived from heterogeneous interventions and further tions, control conditions, patient populations, outcome exploration of the interventions is justifiable as a result. definition, methods of outcome measurement, outcomes Teasell et al. [107] conclude that there is strong evidence assessment times, and in other aspects. This heterogeneity that high-level stroke patients discharged from an acute made it difficult for the review authors to synthesize the hospital unit can be rehabilitated in the community by an results of the underlying trials and this mostly led to interdisciplinary stroke rehabilitation team without nega- inconclusive conclusions. tive consequences, and that community based programs also appear to reduce hospital length of stay. It may be that discharge interventions do have an impact, but that measurements of outcomes are not reliable or not Two reviews [95,99] concentrated on patients with frac- sensitive enough. There is also a possibility that discharge tures. Cameron et al. [95] state that the available RCT's interventions do have an effect, but that this is not long- had different aims, interventions and outcomes and were standing and can no longer be measured at the time of the of poor to moderate quality, thus allowing only tentative outcome assessments. On the other hand, there is a possi- conclusions. Combined outcome measures (e.g. death or bility that effects of discharge interventions only show up institutional care) tended to be better for patients receiv- after the three months after discharge to which we had ing coordinated inpatient rehabilitation, but the results limited the meta-review. There is no good theoretical base were heterogeneous and not statistically significant. for either option, however, whether very short-term or Handoll et al. [99] conclude that there is insufficient evi- very long-term. It may also be that patients in control con- dence to determine the effectiveness of the various mobi- ditions received more care than is suggested by the term lization strategies that start either in the early post- 'usual care', which was mostly ill-defined. Another possi- operative period or during the later rehabilitation period'. bility is that discharge interventions are only working in specific subgroups of patients, or that discharge interven- Two reviews [98,104] concentrated on cardiac patients. tions are only effective in higher intensities. Gwadry et al. [98] evaluated the effectiveness of multidis- Page 15 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 On the other hand, we did find a few indications that dis- ing a telemanagement program directed by an advanced charge interventions may be effective. Three reviews practice nurse after hospital discharge decreases the costs [81,104,105] state that effects are seen in particular when and frequent rehospitalizations associated with heart fail- interventions from the discharge planning and discharge ure and improves the patient's quality of life [108], but support side were combined across the hospital-home also a review that states that the evidence, as it stands at interface. If discharge planning interventions are to be present, raises a number of issues about current hospital effective, they should have to be combined with discharge discharge policy [109], one that concludes that hospital- support interventions and vice versa. In addition, two based case management did not reduce length of hospital reviews [81,105] appear to show that educational inter- stay or readmissions in adult inpatients [110], and ventions might have some effect on parts of the emotional another review that states that there was inconclusive evi- status after discharge, on knowledge and medication dence about the effects of telephone follow-up after dis- adherence, but the results of the reviews are not straight- charge [111]. forward and effects appear to be dependent on the quan- tity and format of the educational interventions. From a research point of view, many challenges remain in proving the (in)effectiveness of discharge interventions: We also had one review [101], however, in which it was better designs, better instruments, better descriptions of concluded that the effect of a discharge intervention was interventions and control conditions, and many more. in the opposite direction to what had been expected, since they found that patients from the care pathway group Challenges also remain for reviewers in applying strate- were more dependent at discharge then the control group. gies to find all available research data, but also in finding methods of synthesizing results containing a high degree An interesting finding in this meta-review is that only a of heterogeneity. Questions remain when reviews are few trials were included in more than one review, comparable enough to allow synthesizing the results in although all included reviews had a related topic of the way it was done in this meta-review; maybe the research and all applied sensitive methods to find the pri- umbrella concept of 'discharge interventions' is too broad mary research. It is possible that the final inclusion sets of to endeavour synthesizing by means of a review of system- each review differ due to different focuses of each review, atic reviews already dealing with vast heterogeneity. what causes differences in search strategies and inclusion criteria. However, the question remains that, if a meta- Finally, challenges remain for meta-reviewers in develop- review were to be done on the data from all of the 265 pri- ing methods for synthesizing results of the relevant mary studies included in one of the reviews, whether this reviews available. The methodology for doing systematic would lead to conclusions similar to those we have now reviews is well developed nowadays and well described obtained. for instance in the Cochrane handbook for reviewers, but a well founded methodology and rationale for performing It could be argued that this meta-review does not give a a systematic review of reviews is currently lacking, espe- complete picture of the state of art, because there are cially with regard to the ways of synthesizing data. Such many more reviews on discharge interventions than were methodology is hardly needed due to rapidly growing included in this review. Inclusion of reviews of a lower amount of published reviews on a same or related topic. methodological quality would certainly have added some In this respect, we advise to follow closely the ongoing information, but these findings are less reliable in our work of the recently started Cochrane Umbrella Reviews opinion and would have led to more uncertainty. Moreo- Working Group. ver, we believe it gives cause for concern that we excluded more than half of the reviews found, solely on the basis of From a practical point of view, this meta-review is rather the suboptimal methodological quality of the systematic disappointing, since there is only limited evidence to give review. directions to how health care professionals and organisa- tions can adopt discharge planning or discharge support It could also be argued that this meta-review is not up to interventions. Usual care seems to be equally as effective date, since it was limited to reviews dated pre-2005. There or ineffective as discharge interventions. Post-discharge may be more recent systematic reviews with conclusions problems continue to be an important issue, however, different to those presented here. When a quick search for which means that professionals and organisations must recent reviews was made in PUBMED and CINAHL in consider ways of preventing, easing or solving post-dis- November 2006, however, and without a formal inclu- charge problems. sion process applied, we found no indications that this would have altered our conclusions. There is a review, for example, that reaches firm conclusions that implement- Page 16 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Conclusion Additional file 4 Based on fifteen high quality systematic reviews, there is Appendix 4: List of references of primary studies included in one of the some evidence that some interventions, particularly those reviews. The data show the references of the primary studies included in with educational components and those which combine one of the reviews pre-discharge and post-discharge interventions, may have Click here for file a positive impact but there is, on the whole, limited sum- [http://www.biomedcentral.com/content/supplementary/1472- marized evidence that discharge planning and discharge 6963-7-47-S4.doc] support interventions have a positive impact on patient status at hospital discharge, on patient functioning after discharge, or on health care use after discharge and costs. Acknowledgements The library of NIVEL is acknowledged for the suggestions in the develop- Competing interests ment of the search strategies and for their efforts in obtaining the docu- The author(s) declare that they have no competing inter- ments for the review. ests. References Authors' contributions 1. Hawe P, Gebski V, Andrews G: Elderly patients after they leave All authors made a substantive contribution to all parts of hospital. Med J Aust 1986, 145:251-254. 2. Lin PC, Chang SY: Functional recovery among elderly people this study. PM initiated and designed the study, developed one year after hip fracture surgery. J Nurs Res 2004, 12:72-82. and performed literature searches, was the primary 3. Michels N: The transition from hospital to home: an explora- responsible person for study inclusion and quality assess- tory study. Home Health Care Serv Q 1988, 9:29-44. 4. Mistiaen P, Duijnhouwer E, Wijkel D, de Bont M, Veeger A: The ment, did the data-extraction and synthesis and made the problems of elderly people at home one week after dis- first draft of this manuscript. AF en EP were involved in charge from an acute care setting. J Adv Nurs 1997, 25:1233-1240. commenting the study design, participated in study inclu- 5. Tierney A, Worth A, Closs SJ, King C, Macmillan M: Older patients' sion and quality assessment, and contributed substan- experiences of discharge from hospital. Nurs Times 1994, tively in drafting and revising the manuscript. 90:36-39. 6. Waters K, Allsopp D, Davidson I, Dennis A: Sources of support for older people after discharge from hospital: 10 years on. J Adv All authors read and approved the final manuscript. Nurs 2001, 33:575-582. 7. Boter H, Rinkel GJ, de Haan RJ: Outreach nurse support after stroke: a descriptive study on patients' and carers' needs, Additional material and applied nursing interventions. Clin Rehabil 2004, 18:156-163. 8. Jopp M, Carroll MC, Waters L: Using self-care theory to guide Additional file 1 nursing management of the older adult after hospitalization. Rehabil Nurs 1993, 18:91-94. Appendix 1: Data sources. The table shows all the data sources that were 9. Mahoney JE, Eisner J, Havighurst T, Gray S, Palta M: Problems of searched older adults living alone after hospitalization. J Gen Intern Med Click here for file 2000, 15:611-619. [http://www.biomedcentral.com/content/supplementary/1472- 10. Rosswurm MA, Lanham DM: Discharge planning for elderly 6963-7-47-S1.doc] patients. J Gerontol Nurs 1998, 24:14-21. 11. Stone SP, Herbert P, Chrisostomou J, Vessey C, Horwood C: The assessment of disability in patients on an acute medical ward Additional file 2 for elderly people. Disabil Rehabil 1993, 15:35-37. Appendix 2: Search strategies. The table shows the applied search strate- 12. Waters KR: Outcomes of discharge from hospital for elderly gies for each data source people. J Adv Nurs 1987, 12:347-355. 13. Boter H, Mistiaen P, Duijnhouwer E, Groenewegen I: The problems Click here for file of elderly patients at home after ophthalmic treatment. J [http://www.biomedcentral.com/content/supplementary/1472- Ophthalmic Nurs Technol 1998, 17:59-65. 6963-7-47-S2.doc] 14. Smith SJ, Drance SM: Difficulties patients have at home after cataract surgery. Can J Ophthalmol 1984, 19:6-9. 15. Bowles KH, Naylor MD, Foust JB: Patient characteristics at hos- Additional file 3 pital discharge and a comparison of home care referral deci- Appendix 3: Excluded studies and reason for exclusion. The data show the sions. J Am Geriatr Soc 2002, 50:336-342. references that were excluded and the reason why 16. Harding J, Modell M: Elderly peoples' experiences of discharge Click here for file from hospital. J R Coll Gen Pract 1989, 39:17-20. [http://www.biomedcentral.com/content/supplementary/1472- 17. Iturria Sierra JA, Marquez CS: [Health and social services: needs, preferences and their use by the elderly following hospitali- 6963-7-47-S3.doc] zation]. Rev Esp Salud Publica 1997, 71:281-291. 18. Mamon J, Steinwachs DM, Fahey M, Bone LR, Oktay J, Klein L: Impact of hospital discharge planning on meeting patient needs after returning home. Health Serv Res 1992, 27:155-175. 19. Schultz AA, Geary PA, Casey FS, Fournier MA: Joining education and service in exploring discharge needs. J Community Health Nurs 1997, 14:141-153. 20. Simon EP, Showers N, Blumenfield S, Holden G, Wu X: Delivery of home care services after discharge: what really happens. Health Soc Work 1995, 20:5-14. Page 17 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 21. Williams EI, Fitton F: Use of nursing and social services by eld- patients: lessons from an acute care for elders unit. Jt Comm erly patients discharged from hospital. Br J Gen Pract 1991, J Qual Improv 1998, 24:63-76. 41:72-75. 48. Griffin KM: Evolution of transitional care settings: past, 22. Wilson K, Pateman B, Beaver K, Luker KA: Patient and carer present, future. AACN Clin Issues 1998, 9:398-408. needs following a cancer-related hospital admission: the 49. Griffiths P, Edwards M, Forbes A, Harris R, Ritchie G: Effectiveness importance of referral to the district nursing service. J Adv of intermediate care in nursing-led in-patient units. Cochrane Nurs 2002, 38:245-253. Database Syst Rev 2004:CD002214. 23. Yost LS: Cancer patients and home care: extent to which 50. Keenan SP, Massel D, Inman KJ, Sibbald WJ: A systematic review required services are not received. Cancer Pract 1995, 3:83-87. of the cost-effectiveness of noncardiac transitional care 24. Jones EW, Densen PM, Brown SD: Posthospital needs of elderly units. Chest 1998, 113:172-177. people at home: findings from an eight-month follow-up 51. Johnson A, Sandford J, Tyndall J: Written and verbal information study. Health Serv Res 1989, 24:643-664. versus verbal information only for patients being discharged 25. Schultz AA: Identification of needs of and utilization of from acute hospital settings to home. Cochrane Database Syst resources by rural and urban elders after hospital discharge Rev 2003:CD003716. to the home. Public Health Nurs 1997, 14:28-36. 52. Arts SE, Francke AL, Hutten JB: Liaison nursing for stroke 26. Tyson S, Turner G: Discharge and follow-up for people with patients: results of a Dutch evaluation study. J Adv Nurs 2000, stroke: what happens and why. Clin Rehabil 2000, 14:381-392. 32:292-300. 27. Driscoll A: Managing post-discharge care at home: an analysis 53. Dukkers van Emden DM, Ros WJ, Berns MP: Transition of care: an of patients' and their carers' perceptions of information evaluation of the role of the discharge liaison nurse in The received during their stay in hospital. J Adv Nurs 2000, Netherlands. J Adv Nurs 1999, 30:1186-1194. 31:1165-1173. 54. Houghton A, Bowling A, Clarke KD, Hopkins AP, Jones I: Does a 28. Galloway S, Bubela N, McKibbon A, Rebeyka D, Saxe-Braithwaite M: dedicated discharge coordinator improve the quality of hos- Symptom distress, anxiety, depression, and discharge infor- pital discharge? Qual Health Care 1996, 5:89-96. mation needs after peripheral arterial bypass. J Vasc Nurs 55. Peters P, Fleuren M, Wijkel D: The quality of the discharge plan- 1995, 13:35-40. ning process: the effect of a liaison nurse. Int J Qual Health Care 29. Galloway SC, Graydon JE: Uncertainty, symptom distress, and 1997, 9:283-287. information needs after surgery for cancer of the colon. Can- 56. Conger MM: Integration of the clinical nurse specialist into the cer Nurs 1996, 19:112-117. nurse case manager role. Nurs Case Manag 1996, 1:230-234. 30. Jackson MF: Use of community support services by elderly 57. Dahl J, Penque S: APN spells success for a heart failure pro- patients discharged from general medical and geriatric med- gram. Nurs Manage 2002, 33:46-48. ical wards. J Adv Nurs 1990, 15:167-175. 58. de Loor S, Jaarsma T: Nurse-managed heart failure pro- 31. Grimmer K, Moss J, Falco J: Experiences of elderly patients grammes in the Netherlands. Eur J Cardiovasc Nurs 2002, regarding independent community living after discharge 1:123-129. from hospital: a longitudinal study. Int J Qual Health Care 2004, 59. Mountain G, Pighills A: Pre-discharge home visits with older 16:465-472. people: time to review practice. Health Soc Care Community 32. O'Hare PA, Malone D, Lusk E, McCorkle R: Unmet needs of black 2003, 11:146-154. patients with cancer posthospitalization: a descriptive study. 60. Nygard L, Grahn U, Rudenhammar A, Hydling S: Reflecting on Oncol Nurs Forum 1993, 20:659-664. practice: are home visits prior to discharge worthwhile in 33. Styrborn K: Elderly patients discharged from acute hospital - geriatric inpatient care? Scand J Caring Sci 2004, 18:193-203. a follow-up study. Arch Gerontol Geriatr 1994, 19:53-73. 61. Hansen FR, Spedtsberg K, Schroll M: [Follow-up home visits to 34. Bowman GS, Howden J, Allen S, Webster RA, Thompson DR: A tel- elderly patients after hospitalization. A randomized control- ephone survey of medical patients 1 week after discharge led study]. Ugeskr Laeger 1994, 156:3305-1. from hospital. J Clin Nurs 1994, 3:369-373. 62. van Harteveld JT, Mistiaen PJ, Dukkers van Emden DM: Home visits 35. LeClerc CM, Wells DL, Craig D, Wilson JL: Falling short of the by community nurses for cancer patients after discharge mark: tales of life after hospital discharge. Clin Nurs Res 2002, from hospital: an evaluation study of the continuity visit. Can- 11:242-263. cer Nurs 1997, 20:105-114. 36. Mistiaen P, Duijnhouwer E, Ettema T: The construction of a 63. Williams EI, Greenwell J, Groom LM: The care of people over 75 research model on post-discharge problems based on a years old after discharge from hospital: an evaluation of review of the literature 1990-1995. Soc Work Health Care 1999, timetabled visiting by Health Visitor Assistants. J Public Health 29:33-68. Med 1992, 14:138-144. 37. Center for Disease Control and Prevention 2006 [http:// 64. Boter H: Multicenter randomized controlled trial of an out- www.cdc.gov/nchs/]. reach nursing support program for recently discharged 38. WHO, Regional Office for Europe 2006 [http:// stroke patients. Stroke 2004, 35:2867-2872. data.euro.who.int/hfadb/]. 65. Nikolaus T, Specht-Leible N, Bach M, Oster P, Schlierf G: A rand- 39. Statistiek CB: Vademecum gezondheidsstatistiek Nederland 2003 Den omized trial of comprehensive geriatric assessment and Haag, Centraal Bureau voor de Statistiek; 2003. home intervention in the care of hospitalized patients. Age 40. Maramba PJ, Richards S, Myers AL, Larrabee JH: Discharge plan- Ageing 1999, 28:543-550. ning process: applying a model for evidence-based practice. 66. Oktay JS, Volland PJ: Post-hospital support program for the frail J Nurs Care Qual 2004, 19:123-129. elderly and their caregivers: a quasi-experimental evalua- 41. Rorden JW, Taft E: Discharge planning guide for nurses Philadelphia, tion. Am J Public Health 1990, 80:39-46. W.B. Suanders Company; 1990. 67. Rodgers H, Soutter J, Kaiser W, Pearson P, Dobson R, Skilbeck C, 42. Blaylock A, Cason CL: Discharge planning predicting patients' Bond J: Early supported hospital discharge following acute needs. J Gerontol Nurs 1992, 18:5-10. stroke: pilot study results. Clin Rehabil 1997, 11:280-287. 43. Mistiaen P, Duijnhouwer E, Prins-Hoekstra A, Ros W, Blaylock A: 68. Stewart S, Pearson S, Horowitz JD: Effects of a home-based inter- Predictive validity of the BRASS index in screening patients vention among patients with congestive heart failure dis- with post-discharge problems. Blaylock Risk Assessment charged from acute hospital care. Arch Intern Med 1998, Screening Score. J Adv Nurs 1999, 30:1050-1056. 158:1067-1072. 44. Victor CR, Vetter NJ: Preparing the elderly for discharge from 69. Boter H, Mistiaen P, Groenewegen I: A randomized trial of a Tel- hospital: a neglected aspect of patient care? Age Ageing 1988, ephone Reassurance Programme for patients recently dis- 17:155-163. charged from an ophthalmic unit. J Clin Nurs 2000, 9:199-206. 45. O'Hare PA: Comparing two models of discharge planning 70. Dunn RB, Guy PM, Hardman CS, Lewis PA, Vetter NJ: Can a house rounds in acute care. Appl Nurs Res 1992, 5:66-73. call by a public health nurse improve the quality of the dis- 46. Shardien K: Multidisciplinary discharge rounds ensure conti- charge process for geriatric patients? Clin Perform Qual Health nuity of care. Oncol Nurs Forum 1997, 24:967. Care 1995, 3:151-155. 47. Covinsky KE, Palmer RM, Kresevic DM, Kahana E, Counsell SR, Fort- insky RH, Landefeld CS: Improving functional outcomes in older Page 18 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 71. Shu E, Mirmina Z, Nystrom K: A telephone reassurance pro- 95. Cameron ID, Handoll HH, Finnegan TP, Madhok R, Langhorne P: Co- gram for elderly home care clients after discharge. Home ordinated multidisciplinary approaches for inpatient rehabil- Healthc Nurse 1996, 14:154-161. itation of older patients with proximal femoral fractures. 72. Turner D: Can telephone follow-up improve post-discharge Cochrane Database Syst Rev 2002:CD000106. outcomes? Br J Nurs 1996, 5:1361-1365. 96. Cole MG: The impact of geriatric post-discharge services on 73. Watanabe A, Watanabe M, Yamagishi K, Hamaguchi K, Ishigaki Y, mental state. Age Ageing 2001, 30:415-418. Kondo A, Akiyama M, Tamura S, Sugano H, Setoyama O: [The effec- 97. Day P, Rasmussen P: What is the evidence for the effectiveness tiveness of discharge planning (DSP) in home care--compar- of specialist geriatric services in acute, post-acute and sub- ison of conditions of cancer patient shifted to home acute settings? A critical appraisal of the literature. In NZHTA treatment before and after DSP introduction]. Gan To Kagaku REPORT Volume 7. New Zealand Health Technology Assessment Ryoho 1995, 22 Suppl 4:331-338. (NZHTA); 2004. 74. Payne S, Kerr C, Hawker S, Hardey M, Powell J: The communica- 98. Gwadry-Sridhar FH, Flintoft V, Lee DS, Lee H, Guyatt GH: A sys- tion of information about older people between health and tematic review and meta-analysis of studies comparing social care practitioners. Age Ageing 2002, 31:107-117. readmission rates and mortality rates in patients with heart 75. Robinson A, Street A: Improving networks between acute care failure. Arch Intern Med 2004, 164:2315-2320. nurses and an aged care assessment team. J Clin Nurs 2004, 99. Handoll HH, Parker MJ, Sherrington C: Mobilisation strategies 13:486-496. after hip fracture surgery in adults. Cochrane Database Syst Rev 76. Anderson C, Ni MC, Brown PM, Carter K: Stroke rehabilitation 2004:CD001704. services to accelerate hospital discharge and provide home- 100. Hyde CJ, Robert IE, Sinclair AJ: The effects of supporting dis- based care: an overview and cost analysis. Pharmacoeconomics charge from hospital to home in older people. Age Ageing 2002, 20:537-552. 2000, 29:271-279. 77. Lim WK, Lambert SF, Gray LC: Effectiveness of case manage- 101. Kwan J, Sandercock P: In-hospital care pathways for stroke. ment and post-acute services in older people after hospital Cochrane Database Syst Rev 2002:CD002924. discharge. Med J Aust 2003, 178:262-266. 102. Outpatient Service Trialists: Therapy-based rehabilitation serv- 78. O'Neill DF, Meade F: What evidence is ther to demonstrate the effective- ices for stroke patients at home. Cochrane Database Syst Rev ness of health and social care interventions and services in reducing pres- 2003:CD002925. sures on the acute hospital system? Turnbridge Wells, Centre for Health 103. Parker G, Bhakta P, Katbamna S, Lovett C, Paisley S, Parker S, Phelps Services Studies, University of Kent; 2001. K, Baker R, Jagger C, Lindesay J, Shepperdson B, Wilson A: Best 79. Shepperd S, Parkes J, McClaren J, Phillips C: Discharge planning place of care for older people after acute and during suba- from hospital to home. Cochrane Database Syst Rev cute illness: a systematic review. J Health Serv Res Policy 2000, 2004:CD000313. 5:176-189. 80. Ali W, Rasmussen P: What is the evidence for the effectiveness 104. Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S, Rubin HR: of managing the hospital/community interface for older peo- Comprehensive discharge planning with postdischarge sup- ple? A critical appraisal of the literature. In NZHTA REPORT Vol- port for older patients with congestive heart failure: a meta- ume 7. New Zealand Health Technology Assessment (NZHTA); analysis. JAMA 2004, 291:1358-1367. 2004:1-146. 105. Richards S, Coast J: Interventions to improve access to health 81. Parker SG, Peet SM, McPherson A, Cannaby AM, Abrams K, Baker R, and social care after discharge from hospital: a systematic Wilson A, Lindesay J, Parker G, Jones DR: A systematic review of review. J Health Serv Res Policy 2003, 8:171-179. discharge arrangements for older people. Health Technol Assess 106. Shepperd S, Iliffe S: Hospital at home versus in-patient hospital 2002, 6:1-183. care. Cochrane Database Syst Rev 2001:CD000356. 82. Cochrane.nl/downloads/searchfilters [http://www.cochrane.nl/ 107. Teasell RW, Foley NC, Bhogal SK, Speechley MR: Early supported index.html] discharge in stroke rehabilitation. Top Stroke Rehabil 2003, 83. Oxman AD, Guyatt GH: Validation of an index of the quality of 10:19-33. review articles. J Clin Epidemiol 1991, 44:1271-1278. 108. Delgado - Passler P, McCaffrey R: The influences of postdis- 84. Oxman AD, Guyatt GH, Singer J, Goldsmith CH, Hutchison BG, Mil- charge management by nurse practitioners on hospital ner RA, Streiner DL: Agreement among reviewers of review readmission for heart failure. J Am Acad Nurse Pract 2006, articles. J Clin Epidemiol 1991, 44:91-98. 18:154-160. 85. Oxman AD: Systematic reviews: Checklists for review arti- 109. Glasby J, Littlechild R, Pryce K: All dressed up but nowhere to go? cles. BMJ 1994, 309:648-651. Delayed hospital discharges and older people. J Health Serv Res 86. Naylor MD: Transitional care of older adults. Annu Rev Nurs Res Policy 2006, 11:52-58. 2002, 20:127-147. 110. Harris R: Review: hospital based case management does not 87. Ospina MB, Kelly K, Klassen TP, Rowe BH: Overview of reports on reduce length of hospital stay or readmissions in adults. Evi- the quality of systematic reviews and meta-analysis in the dence-Based Nursing 2006, 9:54. biomedical literature. Canadian Journal of Emergency Medicine 111. Mistiaen P, Poot E: Telephone follow-up, initiated by a hospital- 2005, 7:174-175. based health professional, for postdischarge problems in 88. Katrak P, Bialocerkowski A, Massy-Westropp N, Kumar V, Grimmer patients discharged from hospital to home. Cochrane Database K: A systematic review of the content of critical appraisal Syst Rev 2006:CD004510. tools. BMC Medical Research Methodology 2004, 4:22. 89. Shea B, Dubé C, Moher D: Assessing the quality of reports of Pre-publication history systematic reviews: the Quorom statement compared to The pre-publication history for this paper can be accessed other tools. In Systematic reviews in health care: meta-analysis in con- text Volume 7. second edition. Edited by: Egger M, Smith GD and Alt- here: man DG. London, BMJ Publishing Group; 2001:122-139. 90. Jadad AR, Cook DJ, Browman GP: A guide to interpreting dis- cordant systematic reviews. CMAJ 1997, 156:1411-1416. http://www.biomedcentral.com/1472-6963/7/47/prepub 91. Peach H: Reading systematic reviews. Aust Fam Physician 2002, 31:736-740. 92. Jadad AR, McQuay HJ: Meta-analyses to evaluate analgesic interventions: a systematic qualitative review of their meth- odology. J Clin Epidemiol 1996, 49:235-243. 93. Assendelft WJ, Koes BW, Knipschild PG, Bouter LM: The relation- ship between methodological quality and conclusions in reviews of spinal manipulation. JAMA 1995, 274:1942-1948. 94. Canter P, Ernst E: Sources of bias in reviews of spinal manipu- lation for back pain. Wien Klin Wochenschr 2005, 117:333-341. Page 19 of 19 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Health Services Research Springer Journals

Interventions aimed at reducing problems in adult patients discharged from hospital to home: a systematic meta-review

Loading next page...
 
/lp/springer-journals/interventions-aimed-at-reducing-problems-in-adult-patients-discharged-pMiTnKyaSa

References (124)

Publisher
Springer Journals
Copyright
Copyright © 2007 by Mistiaen et al; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Public Health; Health Administration; Health Informatics; Nursing Management/Nursing Research
eISSN
1472-6963
DOI
10.1186/1472-6963-7-47
pmid
17408472
Publisher site
See Article on Publisher Site

Abstract

Background: Many patients encounter a variety of problems after discharge from hospital and many discharge (planning and support) interventions have been developed and studied. These primary studies have already been synthesized in several literature reviews with conflicting conclusions. We therefore set out a systematic review of the reviews examining discharge interventions. The objective was to synthesize the evidence presented in literature on the effectiveness of interventions aimed to reduce post-discharge problems in adults discharged home from an acute general care hospital. Methods: A comprehensive search of seventeen literature databases and twenty-five websites was performed for the period 1994–2004 to find relevant reviews. A three-stage inclusion process consisting of initial sifting, checking full-text papers on inclusion criteria, and methodological assessment, was performed independently by two reviewers. Data on effects were synthesized by use of narrative and tabular methods. Results: Fifteen systematic reviews met our inclusion criteria. All reviews had to deal with considerable heterogeneity in interventions, populations and outcomes, making synthesizing and pooling difficult. Although a statistical significant effect was occasionally found, most review authors reached no firm conclusions that the discharge interventions they studied were effective. We found limited evidence that some interventions may improve knowledge of patients, may help in keeping patients at home or may reduce readmissions to hospital. Interventions that combine discharge planning and discharge support tend to lead to the greatest effects. There is little evidence that discharge interventions have an impact on length of stay, discharge destination or dependency at discharge. We found no evidence that discharge interventions have a positive impact on the physical status of patients after discharge, on health care use after discharge, or on costs. Conclusion: Based on fifteen high quality systematic reviews, there is some evidence that some interventions may have a positive impact, particularly those with educational components and those that combine pre-discharge and post-discharge interventions. However, on the whole there is only limited summarized evidence that discharge planning and discharge support interventions have a positive impact on patient status at hospital discharge, on patient functioning after discharge, on health care use after discharge, or on costs. Page 1 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 and primary care providers [74,75], and many others Background Going back home from hospital is not always a smooth [76,77]. process. Many studies from all over the world have repeat- edly reported that many people who have been dis- These 'discharge interventions' mostly aim to smoothen charged from hospital to home, especially the elderly, the discharge itself (generally measured by length of stay encounter a variety of problems in the first weeks after and discharge destination) or to prevent, ease or solve their return home. Problems after discharge include problems in patient's functioning after discharge (gener- dependence on others with regard to household activities ally measured by function-measures) or to prevent [1-6], lower levels of independence in activities of daily readmissions to the hospital (which are generally seen as living and self-care deficits [2,3,5-12], difficulty with read- a proxy for patient problems after discharge) or to lower ing medication labels or instilling eyedrops [13,14], not health care costs, related to hospital readmissions and getting the help they needed [4,5,13,15-23], not being treatment of post discharge problems. aware of available services [24-26], informational needs [4,13,26-30], symptom distress [28,31-33], social prob- Reviews of these studies come to different conclusions on lems [34] and emotional problems as anxiety and uncer- the effectiveness of these interventions, varying from "Dis- tainty [7,29,35]. The post-discharge problems seem to be charge planning and support teams are cost effective and should more common with increased age and in women [36] and be in place universally" [78] to "The impact of discharge plan- may lead to further complications and unplanned hospi- ning on readmission rates, hospital length of stay, health out- tal readmissions. comes and cost is uncertain" [79] to 'In general, the evidence is a mixture of benefit, deficit and uncertainty, due to the com- In addition, lengths of hospital stay have dropped steeply plexity and variability of the interventions and methodological in the last few decades, e.g. from 6.5 days in 1985 to 4.8 problems with the evaluations' [80] and "Evidence from RCT's days in 2003 in the USA (with the greatest decline for peo- is not available to support the general adoption of discharge ple aged 65 years and older [37]), from 10.5 days in 1985 planning protocols, geriatric assessment processes or discharge to 6.9 days in 2003 in the European Union [38], and from support schemes as means of improving discharge outcomes" 12.5 days in 1985 to 7.3 days in 2003 in the Netherlands [81]. [39]. Consequently, the time available to a healthcare team to adequately prepare patients for discharge has vir- The mixed results of the reviews may, however, be caused tually evaporated [40]. by different study populations, heterogeneity of interven- tions, or a variety of outcomes that have been chosen. A Discharge planning and aftercare initiatives have received lot of questions with regard to the optimal content and much and increased attention over the past few years as a the organization of discharge planning and support result. Rorden & Taft defined discharge planning as 'a remain unanswered. We therefore set out a systematic process made up of several steps or phases whose imme- review of reviews dealing with discharge interventions. diate goal is to anticipate changes in patient care needs and whose long-term goal is to ensure continuity of As mentioned earlier, we defined discharge interventions health care' [41]. We defined discharge interventions as as in-hospital interventions or interventions after dis- in-hospital interventions or interventions after discharge charge performed (partly) by hospital-based profession- performed (partly) by hospital-based professionals, als, explicitly targeted to smooth the transition from explicitly targeted to smooth the transition from hospital hospital to home or to prevent or diminish problems after to home or to prevent or diminish problems after hospital hospital discharge. These can roughly be classified in two discharge. groups: Many studies were performed with various forms of dis- - Discharge preparation: interventions that mainly take charge planning and aftercare, e.g. screening patients with place during admission in the hospital, with the objective a high risk of post discharge problems [42,43], intensive of organizing care and preparing patients in such a way in-hospital discharge preparation [44], discharge rounds that the length of hospital stay is as short as possible for [45,46], transitional and intermediate care units [32,47- most patients, that the condition of most patients is such 50], written information leaflets [51], liaison nurses and that they can be discharged home and not into institu- discharge coordinators [52-55], clinical nurse specialists tional care, that they will need as little care as possible [56-58], home visits prior to discharge [59,60], preventive post discharge, and that care (organizations) needed after home visits of district nurses after discharge [61-63], post- discharge are informed and organized as well as possible, hospital support programs [7,64-68], telephone follow- so that patients will not have unmet needs, will not have up after discharge [69-72], discharge planning protocols to be readmitted and will not die due to complications or [18,73], ameliorated communication between hospital deterioration after discharge. Page 2 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 - Discharge support/aftercare: interventions that mainly take sionals providing services in the patient's home after dis- place after discharge from hospital and that are targeted to charge from hospital. prevent, ease or solve problems after discharge in order to prevent readmissions to hospital or admissions to institu- - 'Educational interventions' are interventions targeted at tional care and to maximize recovery and improve func- patients undergoing discharge from hospital that are tional, emotional, social and health status in the post- intended to improve their ability to manage aspects of discharge period. their care after discharge through the provision of infor- mation or more active education. The interventions may Besides this rough two categories classification system, we be limited to education, or supplemented by other activi- considered the categorization of discharge interventions ties such as home visits or telephone calls after discharge. put forward by Parker et al. [81] as a useful additional framework for ordering the results of the included The objective of this meta-review was to identify, appraise reviews. Parker et al. have four broad classes of 'discharge and synthesize the evidence presented in reviews of the lit- arrangements': comprehensive discharge planning proto- erature for the effectiveness of discharge interventions in cols, comprehensive geriatric assessment programmes, reducing post-discharge problems in adults discharged discharge support arrangements and educational inter- home from an acute general care hospital. In addition to ventions, all of which can be either generic or disease spe- problems in patient's functioning after discharge we cific. They define these as follows: sought for evidence about the effects of discharge inter- ventions on discharge status and on health care services - 'Comprehensive discharge planning protocols' are inter- use and costs after discharge. ventions involving standardised actions or interventions carried out by an individual, including assessment, coor- The following questions were addressed: dination and implementation of the discharge plan, which project post-discharge needs with the aim of pre- - What are the effects of 'discharge interventions' on the venting unnecessary readmission, maintaining the health discharge status of patients? status of patients or lessening carers' burdens. (length of hospital stay, discharge destination, depend- - 'Comprehensive geriatric assessment (CGA) pro- ency at discharge) grammes' are programmes based either in hospital or sup- porting older people recently discharged from hospital. In - What are the effects of 'discharge interventions' on the CGA programmes the multidisciplinary, multidimen- functioning of patients in the first 3 months after dis- sional nature of the assessment of health, rehabilitation charge? and social care needs is formalized, often using standard- ized assessment instruments. The results of these formal (physical status, emotional status, social status, health sta- assessments are then used either to inform or prompt tus) treatment and management recommendations, which may be carried out in dedicated inpatient units, provided - What are the effects of these interventions on health care as recommendations to the referring physician or team, or services use and costs in the first 3 months after discharge? delivered in the patient's home or other ambulatory care setting such as the day hospital or outpatient clinic. Dis- (readmissions, use of health care services post discharge, charge planning is usually regarded as an important com- costs) ponent of inpatient CGA programmes, although most are not focused on discharge itself, but on improving func- Outcomes in carers or relatives were not considered. tional health status, and thereby independent living, through medical intervention and rehabilitation. Methods Data sources - 'Discharge support arrangements' are schemes that are We searched for reviews of the literature and reviews that designed to provide support for (older) people after expe- are part of evidence-based guidelines containing synthe- riencing discharge from inpatient hospital care. These are sized evidence relating to discharge planning and support interventions in which hospital or community staff are in interventions aimed at preventing or diminishing prob- contact with the patient around the time of hospital dis- lems in adult patients following hospital discharge. charge, with the specific intention of providing support during the post-discharge period. The interventions may Searches were performed in seventeen literature databases be limited to a post-discharge telephone contact at one and on twenty-five websites, which are listed in Appendix extreme, or, at the other extreme, involve teams of profes- 1 (see Additional file 1). All databases were searched from Page 3 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 1994 (or from their inception if this was later than 1994) - The review has sufficient methodological quality (= until December 2004. Overview Quality Assessment Questionnaire score ≥ 5 [83-85]) A search strategy for PUBMED was developed; which was partly based on the search filters of the Dutch Cochrane Publications were excluded when: Centre for searching systematic reviews and for searching guidelines in PUBMED [82]. Suitable search strategies - They were primary research studies were developed for the other databases, as adaptations of the PUBMED search. No limits were applied where lan- - The outcomes in the review were only reported for carers guages were concerned. All detailed search strategies can or professionals be found in Appendix 2 (see Additional file 2). - The review involved only paediatric or psychiatric The words "discharge planning", "aftercare", "hospital patients discharge" and "continuity of care" (or equivalents in Dutch, French or German for the non-English sites) were - The review involved only emergency department (ED) sequentially entered in the search frame of the sites, for patients or one-day stay procedures the purpose of searching the websites to find systematic reviews as part of a guideline. - The review concerned interventions that are primarily intended to address the problems of caregivers rather than The hits of all searches were entered into Reference Man- of patients ager , duplicates were sifted out in this program, and the inclusion process were executed thereafter. - The experimental interventions discussed in the review are performed after discharge solely by primary care pro- Study selection viders The manuscripts had to fulfil all of the following criteria in order to be included: Since there is no generally accepted definition of what a postdischarge period means, and the duration of postdis- - The manuscript is a systematic review of the literature, charge problems may vary for different illnesses and treat- either as an independent manuscript or as a part of a ment procedures, the choice of a time period of 3 months guideline (we considered a review as a systematic review if as inclusion criterion had to be arbitrary. There is evi- at least two out of three of the following criteria were met: dence, however, that most postdischarge problems occur a search strategy was reported, a search was performed in in the period immediately after discharge: Naylor states in Pubmed at least, and the included studies were subjected her review [86] that '4 to 6 weeks post discharge represents a to some kind of methodological assessment) critical period when many elders are at highest risk for poor dis- charge outcomes' and empirical research in a mixed popu- - The review concerns 'discharge interventions' (= in-hos- lation has shown that postdischarge problems are greater pital interventions or interventions after discharge per- at 7 days post discharge than at 30 days post discharge formed (partly) by hospital-based professionals, explicitly [43]. Moreover, three months is a period for which it is targeted to smooth the transition from hospital to home reasonable to assume that outcomes can be related to the or to prevent or diminish problems after hospital dis- intervention around or in the first month after discharge. charge) A three-stage inclusion process was applied. Titles and - The interventions discussed in the review relate to adult abstracts of articles identified from the search strategies patients discharged home from an acute general care hos- were screened in the first stage of initial sifting, in order to pital, who were admitted for a primarily physical problem determine their relevance and whether they fulfilled the inclusion criteria. For each study the criteria were judged - The outcomes studied in the review concern patient sta- from top to bottom of the inclusion criteria referred to; no tus at discharge, patient functioning after discharge, or further analysis was done on the subsequent criteria as health care service use and costs after discharge soon as one criterion was not met. In this first stage (which is more focused on excluding than on including), - The outcomes studied in the review are measured within one reviewer screened all references and the second 3 months after discharge from hospital reviewer independently checked a 10% random sample of the references. If agreement between the two reviewers on - None of the exclusion criteria listed below are met whether to exclude studies was lower than 95% for the 10% sample, the second reviewer would proceed to check Page 4 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 the other 90% of the sample. In addition, 10% of the ref- comes, and selected patient populations, effects on erences that were excluded by the first reviewer were patients, effects on health care use and costs. checked by a second reviewer. When the title and/or abstract provided insufficient information to determine As stated earlier two categorizations for the interventions relevance, full paper copies of the articles were ordered were used to organize the data. Firstly, the rough two cat- and they proceeded to the second stage. In case articles egories system of discharge interventions, divided in dis- were published in a language in which the reviewers were charge preparation and discharge support interventions; not fluent, assistance was sought from other colleagues secondly the categorization of Parker et al[81], who dis- who mastered that language. tinguish four broad classes of 'discharge arrangements': comprehensive discharge planning protocols, compre- In the second stage, two reviewers independently exam- hensive geriatric assessment programmes, discharge sup- ined all full paper copies of the articles selected in the first port arrangements and educational interventions, all of stage, in order to determine whether they fulfilled the which can be either generic or disease specific. The defini- inclusion criteria. tions of each category are already given in the Background of this article. The criteria were again judged from top to bottom for each study; no further assessment was done on the subsequent The outcomes were classified according to the research criteria as soon as a criterion was not met. Any disagree- questions: ments were resolved by discussion between the two reviewers; if no agreement could be reached, a third - The discharge status of patients: length of hospital stay, reviewer decided. discharge destination, dependency at discharge The third stage of inclusion related to the methodological - The functioning of patients in the first 3 months after assessment of the reviews. All reviews remaining after the discharge: physical status, emotional status, social status, second stage were assessed with the Overview Quality health status Assessment Questionnaire [83-85]. This instrument is one of the most frequently used appraisal instruments for sys- - Health care services use and costs: readmissions, use of tematic reviews in the biomedical literature [87], besides health care services post discharge, costs being one of the few found for which psychometric prop- erties had been documented [88] and which had been Physical status concerns all measures about level of activ- found to meet several important criteria, such as construct ities of daily living, self-care abilities, self efficacy or inde- validity, inter-observer reliability and coverage of the pendence. Emotional status concerns all measures about items in the QUORUM statement for reporting systematic the level of well-being of patients such as uncertainty, anx- reviews [89]. Scores on this instrument can vary from 1 iety, depression, informational needs, mood or coping. (extensive flaws) to 7 (minimal flaws). Two reviewers per- Social status refers to the extent a patient is able to partic- formed this assessment independently. The mean of the ipate in normal social activities and relationships. Health scores of the two reviewers was computed and classified as status concerns symptom prevalence and burden, organ the final quality judgment; in case the scores of the review- dysfunction, mortality, morbidity and physical complica- ers differed more than 2 points, reviewers discussed their tions. However, these categories are not always mutual assessments and came to a new joint score (this was only exclusive, e.g. in the case where multi-dimensional quality needed once, mean difference score was 0.91). of life measures were used. Only high quality reviews (= with mean scores of 5 Whether an outcome was regarded as a positive or a neg- (minor flaws) and above) were used for the data-extrac- ative effect, was primarily based on the perspective and tion, as is proposed by Jadad et al. [90] and Peach [91], definitions used by the review authors. However, in gen- since it is known that low quality reviews may reach dif- eral a shorter length of hospital stay, home as discharge ferent conclusions than high quality reviews [92-94], and destination, better physical, emotional and social func- also to avoid false conclusions that are based on low qual- tioning, better health status, less readmissions, less use of ity evidence. health care services and less costs were regarded as posi- tive outcomes by the review authors, and consequently by Data-analysis and synthesis us. Data were extracted about the applied in- and exclusion criteria for the primary studies, search strategies, studied Data-analysis was done primarily by description of the interventions, time frame of the searches, selected out- interventions and by making cross-tables for the different Page 5 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 interventions, populations and effects. No quantitative Type and number of studies included in the reviews pooling was performed across the reviews. Since all included reviews were focused on effectiveness, all reviews limited their inclusion criteria to comparative Conclusions for the meta-review were based on the con- research designs. Seven reviews [81,98,102,104-107] were clusions and results of meta-analyses presented in the limited to randomized controlled trials only, while the reviews studied. other eight also included other comparative designs, such as quasi-randomized trials, non-randomized comparative studies and before-after designs. Two review authors Results Search and inclusion results [97,103] additionally searched for other reviews and After duplicates had been removed, the searches in the dif- guidelines and used these to reach their conclusions. ferent databases resulted in an initial set of 7442 refer- ences of potential interest. Initial sifting based on title and The fifteen reviews included a total of 265 different pri- abstract reduced this set to 117 references. As said, the first mary studies, the number of primary studies included in reviewer carried out this process and a 10% random sam- an individual review varying from 8 [98] to 71 [81]. Most ple was also done independently by a second reviewer (200 of the 265) of the primary studies were included (crude agreement between reviewers was 99% with a only once in a review, with the exception of a few papers kappa coefficient of 0.33). In addition, when a second that were included in more than one review, extending to reviewer checked a 10% random sample of the excluded four inclusions for ten primary studies and with a maxi- references, discussion was only needed for two references mum of five inclusions for two primary studies. A list of and resulted in an exclusion-decision. The set of the 117 all primary studies included in one of the reviews can be references, representing 108 reviews, was ordered full text found in Appendix 4 (see Additional file 4). for the second stage of the inclusion process. Two review- Aims of the reviews ers performed this second phase independently; agree- ment between reviewers in this phase was 79% with a The aims of the reviews included are all related to the kappa coefficient of 0.56. Discussion was needed for 23 effectiveness of discharge interventions, but there is a wide references and agreement was subsequently reached. A set variation in what review authors describe as their objec- of 49 references, representing 41 reviews, finally proved to tives, as can be seen in Table 2. fulfil the inclusion criteria for type and content of study. Patients of interest in the reviews In the following stage, two reviewers independently Some of the reviews included studies in which interven- assessed the remaining 41 reviews on their methodologi- tions targeted several or mixed patient populations, while cal quality, using the Overview Quality Assessment Ques- others were restricted to studies with a specified patient tionnaire [83-85] proposed by Oxman. A mean of the two group only (e.g. stroke patients, hip fracture patients, eld- scores was computed and classified as the final quality erly or patients with heart failure). A combination was judgment. Twenty-six reviews had a mean quality score sometimes made of elderly patients and a specific medical lower than 5 and were excluded, while the remaining fif- condition. An overview is presented in Table 3. teen high quality reviews [79,81,95-107] advanced to the next stage of the review, for data-extraction and analysis. Interventions studied in the reviews As said, we used two categorization systems for the dis- The flow diagram of the inclusion process is shown in Fig- charge interventions. For this paragraph only the results ure 1. References of the studies excluded and the reason for the rough two categories system is presented. The for exclusion can be found in Appendix 3 (see Additional grouping of the results by the second categorization sys- file 3). tem of Parker et al. [81] is presented in the more detailed section about the effectiveness of interventions later on. Characteristics of the final 15 reviews Publication date of the reviews and the journals in which they were According to the first system discharge interventions are published classified into two groups, discharge preparation and dis- All reviews included date from 2000 or later and five were charge support interventions. published in 2004. The oldest reference included in a review dates from 1964 and the most recent one from Some of the reviews included only studies that used inter- 2004. Search periods for each review are shown in Table 1. ventions from the first group, others only included studies that used interventions from the second group, and a The reviews were published in eight different journals; six third category comprised reviews that included studies in reviews [79,95,99,101,102,106] were published as a which interventions from both groups had to be applied. review in the Cochrane Database of Systematic Reviews. The focus of the reviews is shown in Table 4. Page 6 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Database searches, number of hits BIOMED C2-RIPE CDSR CINAHL DARE ERIC EMBASE GUIDELINE HTA 124 0 7 75 2 1 2202 20 2 INVERT LILACS NEED PICARTA PSYCH PUBMED SOCIO SCI-E VHL 2 6 26 41 66 4823 0 339 23 references ↓ After eliminating duplicates TITLE & ABSTRACT 7742 references ↓ -7625: one or more inclusioncriteria not met references 117 references, representing 108 studies studies FULL TEXT ↓ -35: Not a systematic review 73 studies -26: Review does not contain manuscripts concerning discharge interventions in adults explicitly targeted to smoothen transition from hospital to home 47 studies -5: Does not concern discharge home from an acute general hospital 42 studies ↓ -1: No outcomes measured within 3 months 41 studies ↓ -26: Mean methodological score <5 FINAL SET 15 studies F Figure 1 low diagram of the inclusion process Flow diagram of the inclusion process. Page 7 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Table 1: Search periods in included reviews Review Search period Cameron 2002 inception-2002 Cole 2001 1975–2000 Day 2004 1980–2003 Gwadry 2004 inception-2000 Handoll 2004 inception-2004 Hyde 2000 inception-1997 Kwan 2002 1975–2003 Outpatient Service Trialists (OST) 2003 inception-2001 Parker G 2000 1988–1999 Parker S 2002 inception-2001 Phillips 2004 inception-2003 Richards 2003 inception-2000 Shepperd 2001 inception-2001 Shepperd 2004 inception-2002 Teasell 2003 1995–2002 Interventions included in discharge preparation reviews (hospital at home), educational interventions and inten- were care pathways, patient management schemes, spe- sified rehabilitation/(physio)therapy schemes. cialized units (for stroke, hip fracture or geriatric patients for example), geriatric assessment and/or consultation, The interventions included in a particular review showed discharge coordinators, nurse specialists, educational considerable heterogeneity in terms of what exactly was interventions, intensified rehabilitation/(physio)therapy done, by whom it was done, the way it was done, the fre- schemes, adjusting skill-mix of hospital professionals, quency with which it was done, and the duration of the and discharge plans. intervention. Interventions included in the discharge support reviews Control conditions in the reviews were telephone follow-up, home visits, geriatric assess- Most reviews included studies in which patients in the ment and/or consultation, intensified post-discharge care control condition received usual care (according to the Table 2: Aim of review, as worded by review-authors Review Aim Cameron 2002 to examine the effectiveness and cost effectiveness of specialised multidisciplinary inpatient rehabilitation supervised by a geriatrician or rehabilitation physician compared with usual (orthopedic) care, for older patients with proximal femoral fracture Cole 2001 to determine the impact of geriatric post-discharge services on mental state Day 2004 to provide the evidence base on the effectiveness of specialist geriatric services for developing a sound practice framework Gwadry 2004 to evaluate the effectiveness of multidisciplinary heart failure management programs on hospital admission rates Handoll 2004 to evaluate the effects of different mobilisation strategies and programmes after hip fracture surgery Hyde 2000 to investigate the effects of supported discharge after an acute admission in older people with undifferentiated clinical problems Kwan 2002 to assess the effects of care pathways, compared with standard medical care, among patients with acute stroke who had been admitted to hospital. In particular we aimed to assess the effects on functional outcome, process of care, quality of life and the hospitalisation costs OST 2003 to assess the effects of therapy-based rehabilitation services targeted towards stroke patients resident in the community within one year of stroke onset or discharge from hospital following stroke Parker G 2000 to establish both the volume and strength of existing evaluative research on the costs, quality and effectiveness of different locations of acute, post- and subacute and rehabilitation care for older people Parker S 2002 to test the following hypotheses: 1. There is an inadequate number of comparable rct's to allow a definitive analysis; 2. Hospital discharge process, outcome and cost-effectiveness can be improved through the use of a variety of interventions; 3. Some interventions are more effective than others; 4. there are priority areas for future research Phillips 2004 to evaluate the effect of comprehensive discharge planning plus post-discharge support in patients with chronic heart failure on the rate of readmission, all cause mortality, length of stay, quality of life and medical costs Richards 2003 to determine the effectiveness and costs of interventions intended to improve access to health and social care for older patients following discharge from acute hospitals Shepperd 2001 to assess the effects of hospital at home compared with in-patient hospital care Shepperd 2004 to determine the effectiveness of planning the discharge of patients moving from hospital Teasell 2003 to assess the effectiveness of early supported discharge programs in the context of stroke rehabilitation Page 8 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Table 3: Patients of interest in the reviews Review Several/mixed Elderly Stroke patients Patients with hip or femur fractures Patients with heart failure Cameron 2002 X Cole 2001 X Day 2004 X Gwadry 2004 X Handoll 2004 X Hyde 2000 X Kwan 2002 X OST 2003 X Parker G 2000 X Parker S 2002 X Phillips 2004 X Richards 2003 X Shepperd 2001 X Shepperd 2004 X Teasell 2003 X Total 2 6 3 2 2 trial authors); other reviews included studies in which the relevant intervention. Many of the outcomes, in both the different interventions were compared against each other primary studies and the reviews, lacked a clear definition, (e.g. different rehabilitation/therapy schemes). The prob- however, e.g. functional status or quality of life or mental lem with the first category for all review authors was that state. In addition, different terms were used across pri- the trial authors were not clear on what constituted 'usual mary studies and reviews for outcomes that are related or care'. that are probably the same (e.g. physical status or func- tional status or ability in activities in daily living). Above Outcomes studied in the reviews this, even similar outcomes were measured with different Some of the included reviews had well described primary (frequently not validated) instruments at different times outcomes that to had be described in the trials before they post discharge, posing problems for the review authors in could be included, while others had no criteria at all with combining the effects across trials, but also in combining regard to outcomes as long as the studies dealt with the the results from reviews for this meta-review. Table 4: Focus of interventions in reviews Review Focus on discharge preparation Focus on discharge support/aftercare Cameron 2002 X Cole 2001 X Day 2004 X X Gwadry 2004 X Handoll 2004 X X Hyde 2000 X Kwan 2002 X OST 2003 X Parker G 2000 X Parker S 2002 X X Phillips 2004 X X Richards 2003 X X Shepperd 2001 X Shepperd 2004 X Teasell 2003 X Total 8 13 Page 9 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Effectiveness of the discharge interventions Effect of discharge interventions on physical status after discharge General picture The effect of interventions from the discharge planning Although a statistically significant effect was occasionally category on physical status in the first 3 months after dis- found for a particular intervention on a particular out- charge was studied in three reviews [79,81,105]. Parker et come, most review authors reached no firm conclusions al. [81] included RCT's only and found eight articles rep- that the discharge interventions they studied were effec- resenting seven studies in which discharge planning was tive. Only two review authors [104,105] were firm in their studied. All studies involved patients who had experi- conclusions. The conclusions as formulated by the enced discharge from an acute inpatient hospital stay and authors are shown in Table 5, with formulations indicat- evaluated a comprehensive discharge protocol imple- ing no effects or inconclusive ones are shown in italics and mented by an individual who was either a specialist nurse, formulations indicating firm conclusions are shown in a social worker or an admitting clerk. The comprehensive bold typeface. discharge protocols were similar in design and were com- pared with usual discharge care. The protocols all had Effect of discharge interventions on discharge status similar elements, including the assessment of patients, Length of stay was studied in nine reviews. The findings liaising with the patient's carer and other professionals to were inconclusive in four reviews [95,97,101,107], no sig- coordinate discharge and providing follow-up visits or tel- nificant differences were found in another four reviews ephone calls. Only two of the seven studies included in [79,81,99,104] and one review [106] concludes that hos- this part of the review considered outcomes related to pital length of stay was significantly shorter for 'hospital- physical function. No differences were found between at-home' interventions. experimental and control groups within 3 months after discharge. Richards and Coast [105] included five RCT's Discharge destination was studied in six reviews. Findings dealing with comprehensive discharge planning and came were inconclusive in one review [97] and no significant to the same conclusion as Parker et al. that no differences differences were found in four reviews [79,81,101,106], had been shown with regard to physical status. Shepperd while one review [103] found a significant difference in et al. [79] included 11 RCT's, six of which presented data the number of patients being discharged home when they concerning physical status. Here too, no effects of dis- were cared for at a stroke unit (based on three trials) but charge planning on physical status were found. not when they were treated in hip units or geriatric units. So, these three reviews discussing the impact of discharge Dependency at discharge was studied in one review [101] planning on physical status after discharge are mutually and it was found, on the basis of two studies (one rand- consistent and all conclude that no effect of discharge omized and one non-randomized) that patients from the planning has been demonstrated on physical status. care pathway group were more dependent at discharge than the control group. The effect of interventions from the comprehensive geriatric assessment category on physical status in the first 3 months There is no evidence on the whole that discharge interven- after discharge was studied in three reviews on generic tions have a positive impact at length of stay, discharge patient populations [81,97,105] and in one review on destination, or dependency at discharge. patients with femoral fractures [95]. Day and Rasmussen [97] conclude that measures of functional status were sim- Effect of discharge interventions on patient functioning after ilar and showed no significant difference between the discharge intervention and control groups. Parker et al. [81] point to As was specified in the second research question, patient the great variety of measures used to report physical func- functioning after discharge was divided into four types: tion outcomes, making comparisons and pooling diffi- physical, emotional, social and health status. The effects cult. They say that the majority of studies appeared to have of the discharge interventions are given for each of these, found no significant differences in the physical function and subdivided according to the intervention classifica- outcomes of study patients and control patients over time. tion scheme put forward by Parker et al. [81], in which With regard to improvement in physical function over there are four broad classes of discharge interventions: time, Parker et al. were able to calculate an odds-ratio over comprehensive discharge planning protocols, compre- six studies and found a significant effect suggesting that hensive geriatric assessment programmes, discharge sup- the intervention was beneficial for physical functioning. port arrangements and educational interventions, all of These outcomes, however, were not measured within our which can be either generic or disease specific. stated timeframe of 3 months post discharge. Richards and Coast [105] included two studies in which functional status outcomes were measured within the 3 months after discharge and both found no differences. Finally, Cam- Page 10 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Table 5: Conclusions in included reviews Review Conclusions Cameron 2002 The available trials had different aims, interventions and outcomes. Combined outcome measures (e.g. death or institutional care) tended to be better for patients receiving coordinated inpatient rehabilitation, but the results were heterogeneous and not statistically significant. Cole 2001 There is little evidence that geriatric post-discharge services have an impact on the mental state of aged subjects. Day 2004 This review generally supports the efficacy of specialist geriatric team services trained in geriatrics with a multidisciplinary collaborative focus undertaking assessment, rehabilitation and coordinated case management in community settings; both preventive care and supportive discharge in these settings appear to provide greater benefit over usual care; however these benefits are not consistent across all outcomes and although improvement in outcomes was often apparent, these were not always significant when compared with the comparison group. Efficacy of specialist geriatric services for inpatient settings was more diverse; this was due to the diversity of studies across the continuum of subacute, acute, postacute care in unit or ward settings with resulting heterogeneous outcomes and only some of these outcomes showing significance over usual care. With regard to day hospital and outpatient care, evidence for the efficacy of specialist geriatric services was lacking, with no conclusive evidence that the services are of greater benefit than usual care. Gwadry 2004 This review suggests that specific heart failure targeted interventions significantly decrease hospital readmissions but do not affect mortality rates. Handoll 2004 There is insufficient evidence from randomised trials to determine the effectiveness of the various mobilisation strategies that start either in the early post-operative period or during the later rehabilitation period Hyde 2000 We believe that the results of this review provide reassurance that supporting discharge from hospital to home is of value. However, important sources of uncertainty remain, suggesting the need for further research. There was relative certainty that the proportion of those at home 6–12 months after admission is greater with supported discharge; this was associated with a consistent pattern of reduction in admission to long-stay care over the same period, without apparent increases in mortality. There was uncertainty about the effect of supported discharge on hospitalization. There were no rigorous data on functional status, patient and carer satisfaction and in consequence uncertainty about the overall effectiveness of supported discharge. Kwan 2002 Use of stroke care pathways may be associated with positive and negative effects. Since most of the results have been derived from non-randomised studies, they are likely to be influenced by potential biases and confounding factors. There is currently insufficient supporting evidence to justify the routine implementation of care pathways for acute stroke management or stroke rehabilitation. OST 2003 Therapy-based rehabilitation services targeted towards stroke patients living at home reduces the odds of a poor outcome and has a beneficial effect on a patient's ability to perform activities of daily living. However, the evidence is derived from a review of heterogeneous interventions and therefore further exploration of the interventions is justifiable. Parker G 2000 Despite considerable recent development of different forms of care for older patients, evidence about effectiveness and costs is weak. However, evidence is also weak for longer-standing care models. Parker S 2002 The evidence from these trials does not suggest that discharge arrangements have effects on mortality or length of hospital stay. This review supports the concept that arrangements for discharging older people from hospital can have beneficial effects on subsequent readmission rates. Interventions provided across the hospital-community interface, both in hospital and in the patient's home, showed the largest effects. Evidence from RCT's is not available to support the general adoption of discharge planning protocols, geriatric assessment processes or discharge support schemes as means of improving discharge outcomes. Phillips 2004 Comprehensive discharge planning plus postdischarge support for older people with chronic heart failure significantly reduced readmission rates and may improve health outcomes such as survival and quality of life without increasing costs. Richards 2003 The interventions provided and patient groups targeted by these services were heterogeneous. There was, however, some evidence that services combining needs assessment, discharge planning and a method for facilitating the implementation of these plans were more effective than services that do not include the latter action. The assessment of need may be insufficient in itself for the adequate provision of post-discharge care; needs assessment should be combined with a service that facilitates the implementation of care plans. Shepperd 2001 This review does not support the development of hospital at home services as a cheaper alternative to in-patient care. Early discharge schemes for patients recovering from elective surgery and elderly patients with a medical condition may have a place in reducing the pressure on acute hospital beds, providing the views of the carers are taken into account. The evidence supporting hospital at home for patients recovering from stroke is conflicting. There is some evidence that admission avoidance schemes may provide a less costly alternative to hospital care. Shepperd 2004 The impact of discharge planning on readmission rates, hospital length of stay, health outcomes and cost is uncertain. Teasell 2003 Although the majority of studies reported no statistically significant differences in functional outcomes between the two groups, there was a reduction in hospital stays for patients receiving home-based therapy. These results suggest that patients with milder strokes who receive home-based therapies have similar functional outcomes to patients who receive traditional inpatient rehabilitation. There is strong evidence that high-level stroke patients discharged from an acute hospital unit can be rehabilitated in the community by an interdisciplinary stroke rehabilitation team without negative consequences. These patients attain similar functional outcomes compared to patients with equivalent stroke severity who receive inpatient rehabilitation. Community based programs also appear to reduce hospital length of stay, although we do not have evidence of an overall cost reduction. Although the effectiveness of early supported discharge programs for patients with moderate-to-severe deficits has not been well studied, limited evidence suggests that these patients are unsuitable candidates and should receive inpatient rehabilitation instead. Page 11 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 eron et al. [95] examined the effects of coordinated multi- outcome measures. It was found on the basis of twelve tri- disciplinary inpatient rehabilitation by a geriatrician or als that patients who received therapy-based rehabilita- rehabilitation physician compared with usual care for tion services after stroke were significantly more older patients with hip fracture, and they state that the independent in personal activities of daily living than available trials reviewed had a variety of aims, interven- those patients who received no care or usual care. Most of tions and outcomes, making them difficult to combine. the studies measured this outcome at 6 or 12 months after They conclude on the basis of nine trials that functional starting the therapy, however, and it is not clear how long status did not improve consistently. this was after hospital discharge; no (pooled) data at 3 months post discharge are given in this review. Teasell et On the basis of these four reviews, therefore, it appears al. [107] studied the effectiveness of early supported dis- that comprehensive geriatric assessment has not been charge programs in stroke patients. Ten studies were shown to have a positive impact on functional status included, eight of which reported some kind of functional within 3 months after discharge, in comparison with the outcome. None of these studies reported statistically sig- control groups. nificant differences between the treatment groups, indi- cating that functional outcome was not affected negatively The effect of interventions from the discharge support cate- or positively by the intervention. Pooling was not per- gory on physical status after discharge was studied in four formed in this review. generic [81,100,105,106] and two disease specific reviews [102,107], both in stroke patients. Hyde et al. [100] inves- On the basis of these six reviews, therefore, there are no tigated the effects of supported discharge after an acute indications that patients who receive supported discharge admission in older people with undifferentiated clinical have a better physical status at 3 months after discharge problems, in which supported discharge was defined as than patients from the control groups. actual additional support from any source provided to patients or their carers and commencing within one week The effect of educational interventions on physical status of discharge following an acute admission. They included after discharge was covered by two reviews [81,105]. nine studies of which six provided data on functional sta- Parker et al. [81] studied if education interventions tus; however, there were no rigorous data on functional improved the outcome of discharge of elderly people status that made pooled conclusions possible. Parker et al. from hospital; the interventions studied were described as [81] point to the wide range of types of intervention, var- mainly educational and could be limited to education or ying from a single phone call after discharge to complex supplemented by other activities, such as home visits or multidisciplinary interventions. They included twenty- telephone calls after discharge. Eleven studies were eight controlled trials, nineteen of which reported on included, two of which contained data on physical status; some aspect of physical functioning and eight of which one study found better results in the intervention group, were comparable enough to pool, but showed no signifi- but the other study found no effects. Richards et al. [105] cant effect on physical functioning. Richards and Coast studied discharge co-ordinator roles, which may incorpo- [105] evaluated the effectiveness of organizational inter- rate educational interventions. Five studies were included, ventions that influence access to health and social care four of which contained data on physical status after dis- after discharge. They found considerable heterogeneity in charge; none of these found significant differences the content of interventions and the selection of patient between experimental and control groups. groups. They identified two trials that reported on func- tional status within 3 months of discharge, but both of On the basis of these two reviews, therefore, there are no these were inconclusive and did not suggest improve- clear indications that educational interventions have an ment. Shepperd et al. [106] assessed the effects of hospi- effect on physical status after discharge. tal-at-home compared with in-patient hospital care. Sixteen studies were included, eight of which measured Finally, Handol et al. [99] studied mobilisation strategies functional status in elderly medical patients and two trials in hip fracture surgery patients. They conclude that there in patients following elective surgery. Although pooling is insufficient evidence from randomized trials to deter- was not possible, there were no indications that the func- mine the effectiveness of the various mobilization strate- tional status in the intervention groups was better at 3 gies. months post discharge. The review of the Outpatient Serv- ice Trialists [102] considered interventions targeting In summary, we found no evidence base that discharge stroke patients resident in the community setting. Four- interventions have a positive impact on the physical status teen trials were included, twelve of which involved of patients after discharge. All the reviews included, how- patients who had experienced discharge from hospital; ever, had to contend with extensive heterogeneity in inter- the trials included used a large number of heterogeneous ventions, patient populations, and outcomes scales and Page 12 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 times and with inadequate descriptions of control condi- ety of ways and in multiple domains, making tions, all of which made pooling difficult. interpretation or synthesis across studies problematic, and that in general, these measures remained unchanged Effect of discharge interventions on emotional status after discharge between intervention and control groups. In addition, The effect of interventions from the discharge planning cat- Parker et al. refer to sixteen trials measuring dimensions of egory on emotional status after discharge was studied in quality of life, which may incorporate emotional status. three reviews [79,81,105]. Parker et al. [81] found one dis- Here too, they found many different instruments and that charge planning study that included emotional status out- the data on the whole did not suggest that discharge sup- comes, which stated that mean satisfaction scores changed port arrangements had a major impact on the quality of little over time. Richards and Coast [105] included two life of subjects when compared to controls. Finally, Parker studies that reported emotional function outcome within et al. refer to six trials in which satisfaction was recorded. 3 months and both found no differences. Shepperd et al. Four of the trials suggested some increased satisfaction [79] found two studies containing some kind of emo- with the service provided, but the data were neither con- tional function; one found some improvement on one sistently nor reliably reported. Richards and Coast [105] parameter but not on two other emotional outcomes, included two trials in this category; neither of which while the second study failed to detect a difference. found differences in emotional status outcomes. To the extent that early discharge can be regarded as 'discharge On the basis of these three reviews, therefore, there are no support', Shepperd et al. [106] found eight trials involving indications that discharge planning affects emotional medical patients in which some dimensions of psycho- functioning after discharge. social well-being or quality of life were measured. Six failed to detect a difference between intervention groups The effect of interventions from the comprehensive geriatric and control groups, while two studies reported more psy- assessment category on emotional status after discharge was cho-social dysfunction for the intervention group. Two tri- covered by two reviews [81,105]. Parker et al. [81] found als involving surgery patients were included and failed to eight studies reporting on aspects of emotional status, detect differences in this dimension. With regard to only one of which reported a significantly greater patient satisfaction, there was a mixed and ambivalent improvement in cognitive scores in the intervention picture, but satisfaction tended to be higher in the hospi- group than found in the controls. On the whole, however, tal-at-home groups. No pooling was possible on these var- the outcomes of intervention and control group patients iables. The Outpatient Service Trialists [102] pooled were broadly similar, with no obvious benefit observable results from five studies of quality of life in stroke patients for patients undergoing comprehensive geriatric assess- and found no significant difference between experimental ment. Richards and Coast [105] included three studies in groups and control groups, which also applied to the find- which some emotional outcome was reported within 3 ings of six studies in which mood/distress was measured. months after discharge, but none of the three found dif- ferences between intervention and control groups. On the basis of these five reviews, therefore, there are no indications that discharge support interventions enhance On the basis of these two reviews, therefore, there are no emotional functioning after discharge. indications that comprehensive geriatric assessment has a positive impact on emotional status after discharge. The effect of educational interventions on emotional status after discharge was covered by two reviews [81,105]. The effect of interventions from the discharge support cate- Parker et al. [81] found three studies of educational inter- gory on emotional status after discharge was studied in ventions that investigated the effect on emotional func- four generic reviews [81,96,105,106] and in one disease tion; pooling was impossible and the effects were mixed: specific review [102]. Cole [96] found eleven trials report- one study found no differences except for increased self- ing emotional status outcomes after geriatric post-dis- efficacy for walking; the second study had no measure- charge services, with the type of intervention and the type ments after discharge, and the third study, in which an of emotional status outcomes varying from one study to education intervention in hospital was supported with the next. Emotional status outcomes included depression, extensive telephone follow-up after discharge, showed sig- morale, life satisfaction, contentment, emotional func- nificantly lower levels of anxiety and a higher level of tion, self perceived health or cognition. Three trials knowledge at 6 weeks after discharge. They also found reported small effects and eight reported no effect. Parker four studies that considered the effect of educational inter- et al. [81] found nine trials reporting on emotional func- ventions on adherence to medication advice, in which dif- tioning, including cognitive function (five trials) and ferent measures were used to assess adherence, including measures of anxiety (three trials) or depression (two tri- tablet counts, self-reports of compliance and knowledge als). They state that emotional status is measured in a vari- of medication regimens. All but one of these studies Page 13 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 showed some improvements in adherence to medication control patients. The other four reviews, however, found or knowledge and it is concluded that more intensive no differences with regard to social status after discharge. interventions appear to be relatively effective, but that brief counselling or education is of little effect. Richards Effect of discharge interventions on health status after discharge and Coast [105] studied discharge coordinator roles, Mortality is certainly the outcome that has been looked at which may incorporate educational interventions. Five most frequently in the reviews, regardless the focus of the studies were included, three of which contained data on interventions. Most of the reviews (and the underlying tri- emotional status after discharge, and none of these found als), however, looked at mortality over more extended significant differences between experimental groups and periods of time than the 3 months that are of interest in control groups. this meta review; mortality was mostly measured at 6 or 12 months. Twelve reviews [79,81,95,98-106] found no On the basis of these two reviews, therefore, it appears significant differences in mortality and only Day and Ras- that educational interventions might have some effect on mussen [97] conclude that stroke units showed significant aspects of emotional status after discharge, on knowledge benefits in terms of mortality reduction, but do not spec- and medication adherence, but the results of the reviews ify the trials on which this conclusion is based. are not straightforward and the effects seem to depend on the dose and format of the educational interventions. The four reviews [95,97,99,106] in which morbidity or com- plications after discharge was studied and that were able to In summary, discharge interventions appear to have no include trials, found no significant differences. effect, or only a very limited one, on the emotional status after discharge. In summary, we found no firm evidence that discharge interventions have a positive impact on health status of Effect of discharge interventions on social status after discharge patients after discharge. Data on the effect of interventions from the comprehensive discharge planning category on social status in first 3 Effect of discharge interventions on health care use after discharge months after discharge were found in one review [105]. and costs Readmissions were measured in eleven reviews, but the Richard and Coast [105] included five studies with com- prehensive discharge planning coordinators, and none measurement period was frequently 6 or 12 months and found differences in social support experienced. not the 3 months that is of interest for this meta-review. The effect of interventions from the postdischarge support Seven reviewers [79,95,97,100,102,103,105] are incon- category on social status was found in three reviews clusive about the effect of discharge interventions on [81,103,106]. Parker G et al. [103] report about three tri- readmission rates. One reviewer [106] found no statisti- als in which there was no difference in patients at home at cally significant difference for patients in a hospital-at- 3 months. Parker S et al. [81] found no statistical differ- home intervention. Three reviews [81,98,101] found a ence between experimental groups and control groups in positive effect on readmissions. Parker et al. [81] reviewed number of patients being at home, based on six trials that four types of discharge interventions and conclude that measured this within first six months. On the basis of when all interventions groups are taken together, the three trials, Shepperd et al. [106] found a significantly patients in the intervention groups have a significant larger number of patients from the hospital-at-home lower risk of being readmitted and this was more marked group being at home at 6 weeks. among interventions provided both at hospital and at home. In the subgroups they did not find a significant dif- No reviews discussed effects of interventions from geriatric ference for discharge planning activities, discharge sup- assessment category or of educational interventions on port or geriatric assessment but they did find a significant social status. difference in favour of patients receiving some kind of educational intervention. This is congruent with the posi- Finally, Handoll et al. [99] mention one small trial, in tive finding of Gwadry et al. [98], that patients receiving a which no difference was found in loss of social independ- heart failure management program are less frequently ence between intensive physical training and placebo readmitted. Finally, Kwan and Sandercock [101] found activities started post discharge. fewer readmissions for patients that were cared for in a stroke care pathway. In summary, there is a little bit of evidence, based on one review [106], that patients treated in hospital-at-home Three reviews [81,105,106] had included and discussed interventions more frequently remain at home than the trials relating to the use of services after discharge and all were inconclusive on this subject. Page 14 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 All reviewers comment on the variety of ways that costs, ciplinary heart failure management programs on hospital cost-benefit and cost-effectiveness were measured in the readmission rates and found a significant decrease in trials, making synthesis difficult. Costs are also largely these rates. Phillips et al. [104] also conclude that compre- dependent on the organization of health care in an indi- hensive discharge planning plus postdischarge support for vidual country, making cross-country synthesis difficult. older people with chronic heart failure significantly reduced readmission rates, and may improve health out- With this in mind, all reviewers comes such as survival and quality of life without increas- [79,81,95,97,99,101,103,105-107] who report on costs ing costs. Based on above two reviews, it appears that are inconclusive about the impact of discharge interven- readmissions in heart failure patients can be reduced by tions on costs. some kind of intervention. In summary, there is little evidence that discharge inter- Discussion ventions have an impact on health care use after dis- We found more than forty systematic reviews of discharge charge, or on costs, except that educational interventions interventions, fifteen of which scored highly on method- may reduce readmissions in heart failure patients. ological quality. Our conclusions on the basis of these fif- teen reviews, is that there is only limited evidence for the Effects of discharge interventions in specific patient groups positive impact of discharge interventions. We found a Three reviews [101,102,107] focused on stroke patients and few indications that discharge interventions may be effec- compared several care delivery models and rehabilitation tive. Three reviews [81,104,105] state that effects are services. The main aim of this group of studies was more mainly observed when interventions from the discharge on the post-discharge period than on the discharge itself. planning and discharge support side were combined Kwan and Sandercock [101] conclude that stroke care across the hospital-home interface. In addition, two pathways may be associated with positive and negative reviews [81,105], appear to show that educational inter- effects and that there is currently insufficient evidence to ventions might have some effect on aspects of the emo- justify the implementation of care pathways for acute tional status after discharge, on knowledge and stroke management or stroke rehabilitation. The Outpa- medication adherence. tient Service Trialists [102] conclude that therapy-based rehabilitation services targeting stroke patients living at The limited evidence about effectiveness of discharge home reduce the odds of a poor outcome and have a ben- interventions may be due to the heterogeneity of several eficial effect on a patient's ability to perform activities of aspects which review authors had to deal with. All review daily living. They warn, however, that the evidence is authors were confronted with heterogeneity in interven- derived from heterogeneous interventions and further tions, control conditions, patient populations, outcome exploration of the interventions is justifiable as a result. definition, methods of outcome measurement, outcomes Teasell et al. [107] conclude that there is strong evidence assessment times, and in other aspects. This heterogeneity that high-level stroke patients discharged from an acute made it difficult for the review authors to synthesize the hospital unit can be rehabilitated in the community by an results of the underlying trials and this mostly led to interdisciplinary stroke rehabilitation team without nega- inconclusive conclusions. tive consequences, and that community based programs also appear to reduce hospital length of stay. It may be that discharge interventions do have an impact, but that measurements of outcomes are not reliable or not Two reviews [95,99] concentrated on patients with frac- sensitive enough. There is also a possibility that discharge tures. Cameron et al. [95] state that the available RCT's interventions do have an effect, but that this is not long- had different aims, interventions and outcomes and were standing and can no longer be measured at the time of the of poor to moderate quality, thus allowing only tentative outcome assessments. On the other hand, there is a possi- conclusions. Combined outcome measures (e.g. death or bility that effects of discharge interventions only show up institutional care) tended to be better for patients receiv- after the three months after discharge to which we had ing coordinated inpatient rehabilitation, but the results limited the meta-review. There is no good theoretical base were heterogeneous and not statistically significant. for either option, however, whether very short-term or Handoll et al. [99] conclude that there is insufficient evi- very long-term. It may also be that patients in control con- dence to determine the effectiveness of the various mobi- ditions received more care than is suggested by the term lization strategies that start either in the early post- 'usual care', which was mostly ill-defined. Another possi- operative period or during the later rehabilitation period'. bility is that discharge interventions are only working in specific subgroups of patients, or that discharge interven- Two reviews [98,104] concentrated on cardiac patients. tions are only effective in higher intensities. Gwadry et al. [98] evaluated the effectiveness of multidis- Page 15 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 On the other hand, we did find a few indications that dis- ing a telemanagement program directed by an advanced charge interventions may be effective. Three reviews practice nurse after hospital discharge decreases the costs [81,104,105] state that effects are seen in particular when and frequent rehospitalizations associated with heart fail- interventions from the discharge planning and discharge ure and improves the patient's quality of life [108], but support side were combined across the hospital-home also a review that states that the evidence, as it stands at interface. If discharge planning interventions are to be present, raises a number of issues about current hospital effective, they should have to be combined with discharge discharge policy [109], one that concludes that hospital- support interventions and vice versa. In addition, two based case management did not reduce length of hospital reviews [81,105] appear to show that educational inter- stay or readmissions in adult inpatients [110], and ventions might have some effect on parts of the emotional another review that states that there was inconclusive evi- status after discharge, on knowledge and medication dence about the effects of telephone follow-up after dis- adherence, but the results of the reviews are not straight- charge [111]. forward and effects appear to be dependent on the quan- tity and format of the educational interventions. From a research point of view, many challenges remain in proving the (in)effectiveness of discharge interventions: We also had one review [101], however, in which it was better designs, better instruments, better descriptions of concluded that the effect of a discharge intervention was interventions and control conditions, and many more. in the opposite direction to what had been expected, since they found that patients from the care pathway group Challenges also remain for reviewers in applying strate- were more dependent at discharge then the control group. gies to find all available research data, but also in finding methods of synthesizing results containing a high degree An interesting finding in this meta-review is that only a of heterogeneity. Questions remain when reviews are few trials were included in more than one review, comparable enough to allow synthesizing the results in although all included reviews had a related topic of the way it was done in this meta-review; maybe the research and all applied sensitive methods to find the pri- umbrella concept of 'discharge interventions' is too broad mary research. It is possible that the final inclusion sets of to endeavour synthesizing by means of a review of system- each review differ due to different focuses of each review, atic reviews already dealing with vast heterogeneity. what causes differences in search strategies and inclusion criteria. However, the question remains that, if a meta- Finally, challenges remain for meta-reviewers in develop- review were to be done on the data from all of the 265 pri- ing methods for synthesizing results of the relevant mary studies included in one of the reviews, whether this reviews available. The methodology for doing systematic would lead to conclusions similar to those we have now reviews is well developed nowadays and well described obtained. for instance in the Cochrane handbook for reviewers, but a well founded methodology and rationale for performing It could be argued that this meta-review does not give a a systematic review of reviews is currently lacking, espe- complete picture of the state of art, because there are cially with regard to the ways of synthesizing data. Such many more reviews on discharge interventions than were methodology is hardly needed due to rapidly growing included in this review. Inclusion of reviews of a lower amount of published reviews on a same or related topic. methodological quality would certainly have added some In this respect, we advise to follow closely the ongoing information, but these findings are less reliable in our work of the recently started Cochrane Umbrella Reviews opinion and would have led to more uncertainty. Moreo- Working Group. ver, we believe it gives cause for concern that we excluded more than half of the reviews found, solely on the basis of From a practical point of view, this meta-review is rather the suboptimal methodological quality of the systematic disappointing, since there is only limited evidence to give review. directions to how health care professionals and organisa- tions can adopt discharge planning or discharge support It could also be argued that this meta-review is not up to interventions. Usual care seems to be equally as effective date, since it was limited to reviews dated pre-2005. There or ineffective as discharge interventions. Post-discharge may be more recent systematic reviews with conclusions problems continue to be an important issue, however, different to those presented here. When a quick search for which means that professionals and organisations must recent reviews was made in PUBMED and CINAHL in consider ways of preventing, easing or solving post-dis- November 2006, however, and without a formal inclu- charge problems. sion process applied, we found no indications that this would have altered our conclusions. There is a review, for example, that reaches firm conclusions that implement- Page 16 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 Conclusion Additional file 4 Based on fifteen high quality systematic reviews, there is Appendix 4: List of references of primary studies included in one of the some evidence that some interventions, particularly those reviews. The data show the references of the primary studies included in with educational components and those which combine one of the reviews pre-discharge and post-discharge interventions, may have Click here for file a positive impact but there is, on the whole, limited sum- [http://www.biomedcentral.com/content/supplementary/1472- marized evidence that discharge planning and discharge 6963-7-47-S4.doc] support interventions have a positive impact on patient status at hospital discharge, on patient functioning after discharge, or on health care use after discharge and costs. Acknowledgements The library of NIVEL is acknowledged for the suggestions in the develop- Competing interests ment of the search strategies and for their efforts in obtaining the docu- The author(s) declare that they have no competing inter- ments for the review. ests. References Authors' contributions 1. Hawe P, Gebski V, Andrews G: Elderly patients after they leave All authors made a substantive contribution to all parts of hospital. Med J Aust 1986, 145:251-254. 2. Lin PC, Chang SY: Functional recovery among elderly people this study. PM initiated and designed the study, developed one year after hip fracture surgery. J Nurs Res 2004, 12:72-82. and performed literature searches, was the primary 3. Michels N: The transition from hospital to home: an explora- responsible person for study inclusion and quality assess- tory study. Home Health Care Serv Q 1988, 9:29-44. 4. Mistiaen P, Duijnhouwer E, Wijkel D, de Bont M, Veeger A: The ment, did the data-extraction and synthesis and made the problems of elderly people at home one week after dis- first draft of this manuscript. AF en EP were involved in charge from an acute care setting. J Adv Nurs 1997, 25:1233-1240. commenting the study design, participated in study inclu- 5. Tierney A, Worth A, Closs SJ, King C, Macmillan M: Older patients' sion and quality assessment, and contributed substan- experiences of discharge from hospital. Nurs Times 1994, tively in drafting and revising the manuscript. 90:36-39. 6. Waters K, Allsopp D, Davidson I, Dennis A: Sources of support for older people after discharge from hospital: 10 years on. J Adv All authors read and approved the final manuscript. Nurs 2001, 33:575-582. 7. Boter H, Rinkel GJ, de Haan RJ: Outreach nurse support after stroke: a descriptive study on patients' and carers' needs, Additional material and applied nursing interventions. Clin Rehabil 2004, 18:156-163. 8. Jopp M, Carroll MC, Waters L: Using self-care theory to guide Additional file 1 nursing management of the older adult after hospitalization. Rehabil Nurs 1993, 18:91-94. Appendix 1: Data sources. The table shows all the data sources that were 9. Mahoney JE, Eisner J, Havighurst T, Gray S, Palta M: Problems of searched older adults living alone after hospitalization. J Gen Intern Med Click here for file 2000, 15:611-619. [http://www.biomedcentral.com/content/supplementary/1472- 10. Rosswurm MA, Lanham DM: Discharge planning for elderly 6963-7-47-S1.doc] patients. J Gerontol Nurs 1998, 24:14-21. 11. Stone SP, Herbert P, Chrisostomou J, Vessey C, Horwood C: The assessment of disability in patients on an acute medical ward Additional file 2 for elderly people. Disabil Rehabil 1993, 15:35-37. Appendix 2: Search strategies. The table shows the applied search strate- 12. Waters KR: Outcomes of discharge from hospital for elderly gies for each data source people. J Adv Nurs 1987, 12:347-355. 13. Boter H, Mistiaen P, Duijnhouwer E, Groenewegen I: The problems Click here for file of elderly patients at home after ophthalmic treatment. J [http://www.biomedcentral.com/content/supplementary/1472- Ophthalmic Nurs Technol 1998, 17:59-65. 6963-7-47-S2.doc] 14. Smith SJ, Drance SM: Difficulties patients have at home after cataract surgery. Can J Ophthalmol 1984, 19:6-9. 15. Bowles KH, Naylor MD, Foust JB: Patient characteristics at hos- Additional file 3 pital discharge and a comparison of home care referral deci- Appendix 3: Excluded studies and reason for exclusion. The data show the sions. J Am Geriatr Soc 2002, 50:336-342. references that were excluded and the reason why 16. Harding J, Modell M: Elderly peoples' experiences of discharge Click here for file from hospital. J R Coll Gen Pract 1989, 39:17-20. [http://www.biomedcentral.com/content/supplementary/1472- 17. Iturria Sierra JA, Marquez CS: [Health and social services: needs, preferences and their use by the elderly following hospitali- 6963-7-47-S3.doc] zation]. Rev Esp Salud Publica 1997, 71:281-291. 18. Mamon J, Steinwachs DM, Fahey M, Bone LR, Oktay J, Klein L: Impact of hospital discharge planning on meeting patient needs after returning home. Health Serv Res 1992, 27:155-175. 19. Schultz AA, Geary PA, Casey FS, Fournier MA: Joining education and service in exploring discharge needs. J Community Health Nurs 1997, 14:141-153. 20. Simon EP, Showers N, Blumenfield S, Holden G, Wu X: Delivery of home care services after discharge: what really happens. Health Soc Work 1995, 20:5-14. Page 17 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 21. Williams EI, Fitton F: Use of nursing and social services by eld- patients: lessons from an acute care for elders unit. Jt Comm erly patients discharged from hospital. Br J Gen Pract 1991, J Qual Improv 1998, 24:63-76. 41:72-75. 48. Griffin KM: Evolution of transitional care settings: past, 22. Wilson K, Pateman B, Beaver K, Luker KA: Patient and carer present, future. AACN Clin Issues 1998, 9:398-408. needs following a cancer-related hospital admission: the 49. Griffiths P, Edwards M, Forbes A, Harris R, Ritchie G: Effectiveness importance of referral to the district nursing service. J Adv of intermediate care in nursing-led in-patient units. Cochrane Nurs 2002, 38:245-253. Database Syst Rev 2004:CD002214. 23. Yost LS: Cancer patients and home care: extent to which 50. Keenan SP, Massel D, Inman KJ, Sibbald WJ: A systematic review required services are not received. Cancer Pract 1995, 3:83-87. of the cost-effectiveness of noncardiac transitional care 24. Jones EW, Densen PM, Brown SD: Posthospital needs of elderly units. Chest 1998, 113:172-177. people at home: findings from an eight-month follow-up 51. Johnson A, Sandford J, Tyndall J: Written and verbal information study. Health Serv Res 1989, 24:643-664. versus verbal information only for patients being discharged 25. Schultz AA: Identification of needs of and utilization of from acute hospital settings to home. Cochrane Database Syst resources by rural and urban elders after hospital discharge Rev 2003:CD003716. to the home. Public Health Nurs 1997, 14:28-36. 52. Arts SE, Francke AL, Hutten JB: Liaison nursing for stroke 26. Tyson S, Turner G: Discharge and follow-up for people with patients: results of a Dutch evaluation study. J Adv Nurs 2000, stroke: what happens and why. Clin Rehabil 2000, 14:381-392. 32:292-300. 27. Driscoll A: Managing post-discharge care at home: an analysis 53. Dukkers van Emden DM, Ros WJ, Berns MP: Transition of care: an of patients' and their carers' perceptions of information evaluation of the role of the discharge liaison nurse in The received during their stay in hospital. J Adv Nurs 2000, Netherlands. J Adv Nurs 1999, 30:1186-1194. 31:1165-1173. 54. Houghton A, Bowling A, Clarke KD, Hopkins AP, Jones I: Does a 28. Galloway S, Bubela N, McKibbon A, Rebeyka D, Saxe-Braithwaite M: dedicated discharge coordinator improve the quality of hos- Symptom distress, anxiety, depression, and discharge infor- pital discharge? Qual Health Care 1996, 5:89-96. mation needs after peripheral arterial bypass. J Vasc Nurs 55. Peters P, Fleuren M, Wijkel D: The quality of the discharge plan- 1995, 13:35-40. ning process: the effect of a liaison nurse. Int J Qual Health Care 29. Galloway SC, Graydon JE: Uncertainty, symptom distress, and 1997, 9:283-287. information needs after surgery for cancer of the colon. Can- 56. Conger MM: Integration of the clinical nurse specialist into the cer Nurs 1996, 19:112-117. nurse case manager role. Nurs Case Manag 1996, 1:230-234. 30. Jackson MF: Use of community support services by elderly 57. Dahl J, Penque S: APN spells success for a heart failure pro- patients discharged from general medical and geriatric med- gram. Nurs Manage 2002, 33:46-48. ical wards. J Adv Nurs 1990, 15:167-175. 58. de Loor S, Jaarsma T: Nurse-managed heart failure pro- 31. Grimmer K, Moss J, Falco J: Experiences of elderly patients grammes in the Netherlands. Eur J Cardiovasc Nurs 2002, regarding independent community living after discharge 1:123-129. from hospital: a longitudinal study. Int J Qual Health Care 2004, 59. Mountain G, Pighills A: Pre-discharge home visits with older 16:465-472. people: time to review practice. Health Soc Care Community 32. O'Hare PA, Malone D, Lusk E, McCorkle R: Unmet needs of black 2003, 11:146-154. patients with cancer posthospitalization: a descriptive study. 60. Nygard L, Grahn U, Rudenhammar A, Hydling S: Reflecting on Oncol Nurs Forum 1993, 20:659-664. practice: are home visits prior to discharge worthwhile in 33. Styrborn K: Elderly patients discharged from acute hospital - geriatric inpatient care? Scand J Caring Sci 2004, 18:193-203. a follow-up study. Arch Gerontol Geriatr 1994, 19:53-73. 61. Hansen FR, Spedtsberg K, Schroll M: [Follow-up home visits to 34. Bowman GS, Howden J, Allen S, Webster RA, Thompson DR: A tel- elderly patients after hospitalization. A randomized control- ephone survey of medical patients 1 week after discharge led study]. Ugeskr Laeger 1994, 156:3305-1. from hospital. J Clin Nurs 1994, 3:369-373. 62. van Harteveld JT, Mistiaen PJ, Dukkers van Emden DM: Home visits 35. LeClerc CM, Wells DL, Craig D, Wilson JL: Falling short of the by community nurses for cancer patients after discharge mark: tales of life after hospital discharge. Clin Nurs Res 2002, from hospital: an evaluation study of the continuity visit. Can- 11:242-263. cer Nurs 1997, 20:105-114. 36. Mistiaen P, Duijnhouwer E, Ettema T: The construction of a 63. Williams EI, Greenwell J, Groom LM: The care of people over 75 research model on post-discharge problems based on a years old after discharge from hospital: an evaluation of review of the literature 1990-1995. Soc Work Health Care 1999, timetabled visiting by Health Visitor Assistants. J Public Health 29:33-68. Med 1992, 14:138-144. 37. Center for Disease Control and Prevention 2006 [http:// 64. Boter H: Multicenter randomized controlled trial of an out- www.cdc.gov/nchs/]. reach nursing support program for recently discharged 38. WHO, Regional Office for Europe 2006 [http:// stroke patients. Stroke 2004, 35:2867-2872. data.euro.who.int/hfadb/]. 65. Nikolaus T, Specht-Leible N, Bach M, Oster P, Schlierf G: A rand- 39. Statistiek CB: Vademecum gezondheidsstatistiek Nederland 2003 Den omized trial of comprehensive geriatric assessment and Haag, Centraal Bureau voor de Statistiek; 2003. home intervention in the care of hospitalized patients. Age 40. Maramba PJ, Richards S, Myers AL, Larrabee JH: Discharge plan- Ageing 1999, 28:543-550. ning process: applying a model for evidence-based practice. 66. Oktay JS, Volland PJ: Post-hospital support program for the frail J Nurs Care Qual 2004, 19:123-129. elderly and their caregivers: a quasi-experimental evalua- 41. Rorden JW, Taft E: Discharge planning guide for nurses Philadelphia, tion. Am J Public Health 1990, 80:39-46. W.B. Suanders Company; 1990. 67. Rodgers H, Soutter J, Kaiser W, Pearson P, Dobson R, Skilbeck C, 42. Blaylock A, Cason CL: Discharge planning predicting patients' Bond J: Early supported hospital discharge following acute needs. J Gerontol Nurs 1992, 18:5-10. stroke: pilot study results. Clin Rehabil 1997, 11:280-287. 43. Mistiaen P, Duijnhouwer E, Prins-Hoekstra A, Ros W, Blaylock A: 68. Stewart S, Pearson S, Horowitz JD: Effects of a home-based inter- Predictive validity of the BRASS index in screening patients vention among patients with congestive heart failure dis- with post-discharge problems. Blaylock Risk Assessment charged from acute hospital care. Arch Intern Med 1998, Screening Score. J Adv Nurs 1999, 30:1050-1056. 158:1067-1072. 44. Victor CR, Vetter NJ: Preparing the elderly for discharge from 69. Boter H, Mistiaen P, Groenewegen I: A randomized trial of a Tel- hospital: a neglected aspect of patient care? Age Ageing 1988, ephone Reassurance Programme for patients recently dis- 17:155-163. charged from an ophthalmic unit. J Clin Nurs 2000, 9:199-206. 45. O'Hare PA: Comparing two models of discharge planning 70. Dunn RB, Guy PM, Hardman CS, Lewis PA, Vetter NJ: Can a house rounds in acute care. Appl Nurs Res 1992, 5:66-73. call by a public health nurse improve the quality of the dis- 46. Shardien K: Multidisciplinary discharge rounds ensure conti- charge process for geriatric patients? Clin Perform Qual Health nuity of care. Oncol Nurs Forum 1997, 24:967. Care 1995, 3:151-155. 47. Covinsky KE, Palmer RM, Kresevic DM, Kahana E, Counsell SR, Fort- insky RH, Landefeld CS: Improving functional outcomes in older Page 18 of 19 (page number not for citation purposes) BMC Health Services Research 2007, 7:47 http://www.biomedcentral.com/1472-6963/7/47 71. Shu E, Mirmina Z, Nystrom K: A telephone reassurance pro- 95. Cameron ID, Handoll HH, Finnegan TP, Madhok R, Langhorne P: Co- gram for elderly home care clients after discharge. Home ordinated multidisciplinary approaches for inpatient rehabil- Healthc Nurse 1996, 14:154-161. itation of older patients with proximal femoral fractures. 72. Turner D: Can telephone follow-up improve post-discharge Cochrane Database Syst Rev 2002:CD000106. outcomes? Br J Nurs 1996, 5:1361-1365. 96. Cole MG: The impact of geriatric post-discharge services on 73. Watanabe A, Watanabe M, Yamagishi K, Hamaguchi K, Ishigaki Y, mental state. Age Ageing 2001, 30:415-418. Kondo A, Akiyama M, Tamura S, Sugano H, Setoyama O: [The effec- 97. Day P, Rasmussen P: What is the evidence for the effectiveness tiveness of discharge planning (DSP) in home care--compar- of specialist geriatric services in acute, post-acute and sub- ison of conditions of cancer patient shifted to home acute settings? A critical appraisal of the literature. In NZHTA treatment before and after DSP introduction]. Gan To Kagaku REPORT Volume 7. New Zealand Health Technology Assessment Ryoho 1995, 22 Suppl 4:331-338. (NZHTA); 2004. 74. Payne S, Kerr C, Hawker S, Hardey M, Powell J: The communica- 98. Gwadry-Sridhar FH, Flintoft V, Lee DS, Lee H, Guyatt GH: A sys- tion of information about older people between health and tematic review and meta-analysis of studies comparing social care practitioners. Age Ageing 2002, 31:107-117. readmission rates and mortality rates in patients with heart 75. Robinson A, Street A: Improving networks between acute care failure. Arch Intern Med 2004, 164:2315-2320. nurses and an aged care assessment team. J Clin Nurs 2004, 99. Handoll HH, Parker MJ, Sherrington C: Mobilisation strategies 13:486-496. after hip fracture surgery in adults. Cochrane Database Syst Rev 76. Anderson C, Ni MC, Brown PM, Carter K: Stroke rehabilitation 2004:CD001704. services to accelerate hospital discharge and provide home- 100. Hyde CJ, Robert IE, Sinclair AJ: The effects of supporting dis- based care: an overview and cost analysis. Pharmacoeconomics charge from hospital to home in older people. Age Ageing 2002, 20:537-552. 2000, 29:271-279. 77. Lim WK, Lambert SF, Gray LC: Effectiveness of case manage- 101. Kwan J, Sandercock P: In-hospital care pathways for stroke. ment and post-acute services in older people after hospital Cochrane Database Syst Rev 2002:CD002924. discharge. Med J Aust 2003, 178:262-266. 102. Outpatient Service Trialists: Therapy-based rehabilitation serv- 78. O'Neill DF, Meade F: What evidence is ther to demonstrate the effective- ices for stroke patients at home. Cochrane Database Syst Rev ness of health and social care interventions and services in reducing pres- 2003:CD002925. sures on the acute hospital system? Turnbridge Wells, Centre for Health 103. Parker G, Bhakta P, Katbamna S, Lovett C, Paisley S, Parker S, Phelps Services Studies, University of Kent; 2001. K, Baker R, Jagger C, Lindesay J, Shepperdson B, Wilson A: Best 79. Shepperd S, Parkes J, McClaren J, Phillips C: Discharge planning place of care for older people after acute and during suba- from hospital to home. Cochrane Database Syst Rev cute illness: a systematic review. J Health Serv Res Policy 2000, 2004:CD000313. 5:176-189. 80. Ali W, Rasmussen P: What is the evidence for the effectiveness 104. Phillips CO, Wright SM, Kern DE, Singa RM, Shepperd S, Rubin HR: of managing the hospital/community interface for older peo- Comprehensive discharge planning with postdischarge sup- ple? A critical appraisal of the literature. In NZHTA REPORT Vol- port for older patients with congestive heart failure: a meta- ume 7. New Zealand Health Technology Assessment (NZHTA); analysis. JAMA 2004, 291:1358-1367. 2004:1-146. 105. Richards S, Coast J: Interventions to improve access to health 81. Parker SG, Peet SM, McPherson A, Cannaby AM, Abrams K, Baker R, and social care after discharge from hospital: a systematic Wilson A, Lindesay J, Parker G, Jones DR: A systematic review of review. J Health Serv Res Policy 2003, 8:171-179. discharge arrangements for older people. Health Technol Assess 106. Shepperd S, Iliffe S: Hospital at home versus in-patient hospital 2002, 6:1-183. care. Cochrane Database Syst Rev 2001:CD000356. 82. Cochrane.nl/downloads/searchfilters [http://www.cochrane.nl/ 107. Teasell RW, Foley NC, Bhogal SK, Speechley MR: Early supported index.html] discharge in stroke rehabilitation. Top Stroke Rehabil 2003, 83. Oxman AD, Guyatt GH: Validation of an index of the quality of 10:19-33. review articles. J Clin Epidemiol 1991, 44:1271-1278. 108. Delgado - Passler P, McCaffrey R: The influences of postdis- 84. Oxman AD, Guyatt GH, Singer J, Goldsmith CH, Hutchison BG, Mil- charge management by nurse practitioners on hospital ner RA, Streiner DL: Agreement among reviewers of review readmission for heart failure. J Am Acad Nurse Pract 2006, articles. J Clin Epidemiol 1991, 44:91-98. 18:154-160. 85. Oxman AD: Systematic reviews: Checklists for review arti- 109. Glasby J, Littlechild R, Pryce K: All dressed up but nowhere to go? cles. BMJ 1994, 309:648-651. Delayed hospital discharges and older people. J Health Serv Res 86. Naylor MD: Transitional care of older adults. Annu Rev Nurs Res Policy 2006, 11:52-58. 2002, 20:127-147. 110. Harris R: Review: hospital based case management does not 87. Ospina MB, Kelly K, Klassen TP, Rowe BH: Overview of reports on reduce length of hospital stay or readmissions in adults. Evi- the quality of systematic reviews and meta-analysis in the dence-Based Nursing 2006, 9:54. biomedical literature. Canadian Journal of Emergency Medicine 111. Mistiaen P, Poot E: Telephone follow-up, initiated by a hospital- 2005, 7:174-175. based health professional, for postdischarge problems in 88. Katrak P, Bialocerkowski A, Massy-Westropp N, Kumar V, Grimmer patients discharged from hospital to home. Cochrane Database K: A systematic review of the content of critical appraisal Syst Rev 2006:CD004510. tools. BMC Medical Research Methodology 2004, 4:22. 89. Shea B, Dubé C, Moher D: Assessing the quality of reports of Pre-publication history systematic reviews: the Quorom statement compared to The pre-publication history for this paper can be accessed other tools. In Systematic reviews in health care: meta-analysis in con- text Volume 7. second edition. Edited by: Egger M, Smith GD and Alt- here: man DG. London, BMJ Publishing Group; 2001:122-139. 90. Jadad AR, Cook DJ, Browman GP: A guide to interpreting dis- cordant systematic reviews. CMAJ 1997, 156:1411-1416. http://www.biomedcentral.com/1472-6963/7/47/prepub 91. Peach H: Reading systematic reviews. Aust Fam Physician 2002, 31:736-740. 92. Jadad AR, McQuay HJ: Meta-analyses to evaluate analgesic interventions: a systematic qualitative review of their meth- odology. J Clin Epidemiol 1996, 49:235-243. 93. Assendelft WJ, Koes BW, Knipschild PG, Bouter LM: The relation- ship between methodological quality and conclusions in reviews of spinal manipulation. JAMA 1995, 274:1942-1948. 94. Canter P, Ernst E: Sources of bias in reviews of spinal manipu- lation for back pain. Wien Klin Wochenschr 2005, 117:333-341. Page 19 of 19 (page number not for citation purposes)

Journal

BMC Health Services ResearchSpringer Journals

Published: Apr 4, 2007

There are no references for this article.