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

Learn More →

PRISM: Mounting a community‐randomised trial. Establishing partnerships with local government

PRISM: Mounting a community‐randomised trial. Establishing partnerships with local government Rhonda Small, Stephanie Brown, Wendy Dawson, Lyndsey Watson and Judith Lumley Mother and Child Health Research, School of Public Health, La Trobe University, Victoria Abstract This paper describes initial partnership development in PRISM (Program of Resources, Information and Support for Mothers), a community-randomised trial to improve maternal health in the first year after birth conducted in Victoria in the period 1998-2003. First, we discuss the principles underpinning community recruitment methods in PRISM that guided both our initial approaches to, and our continuing relationships with, communities. Second, we outline the strategies used to recr uit communities and to establish the groundwork for ongoing partnerships over the projected six years of the study. ( Aust N Z J Public Health 2004; 28: 471-5) he importance of collaborati ve partnerships and of community participation and ‘ownership’ for the success and sustainability of population-based health promotion activities are now considered axiomatic, although the benefits have been noted as largely ‘reputed’1,2 rather than firmly established. In a discussion of the implications of the results of population-level health interventions from the 1980s and 1990s, Sorensen et al. called for greater community involvement in program planning and attention to community competence or capacity building in implementation in future community intervention trials.3 They also emphasised the importance of more creativity in dealing with the tension inherent in the need for protocol standardisation in trial design and appropriate flexibility in the intervention to meet differing community needs and priorities. Similarly, Shediac-Rizkallah and Bone discussed the importance of ‘project negotiation’ in design and implementation phases to strengthen project sustainability, stressing consensusbuilding to reach a compromise for addressing the needs of all project partners – researchers, funders and community participants.4 Despite broad-based ag reement about the importance of community involvement for the success of population-based health interventions, at the time we approached local gover nment authorities about par ticipation in PRISM in early 1998 surprisingly little had been written about strategies for initiating community collaborations in the context of communityrandomised trials. This paper is one step in filling that gap, describing how we went about the process of community recr uitment to PRISM, a community-randomised trial that aimed to impro ve the ph ysical and emotional health of recent mothers via an integ rated program of primary care and community-based strategies. The paper f irst outlines the principles underlying our approach to communities and, second , the strategies employed to establish local gover nment par tnerships in order to build the foundations for subsequent broader community collaboration in project implementation and evaluation. Principles underlying the approach to communities First, we provided comprehensiv e infor mation about the rationale for conducting the trial, the aims of the trial and a description of our pre vious research, par ticularly the studies that had developed our understanding of Victorian women’s experiences of early motherhood, their own health and their use of health services in the first years after childbirth. It w as this research that infor med the proposed inter vention program.5 We felt that such infor mation was essential for communities to assess the potential rele vance and value of the proposed intervention program and to judge our credentials as researchers. 1,6,7 Second, we believed it was impor tant for the trial to provide opportunities for people in communities to learn more about research, starting with our discussion of the proposed Correspondence to: Dr Rhonda Small, Mother and Child Health Research, School of Public Health, La T robe Univ ersity, 251 Faraday Street, Carlton, Victor ia 3053. Fax: (03) 8341 8555; e-mail: r.small@latrobe.edu.au Submitted: October 2003 Revision requested: April 2004 Accepted: July 2004 2004 VOL. 28 NO . 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Small et al. Practice Note trial design and key elements of the intervention, and the role of large community trials with rigorous evaluation in developing knowledge about ways to improve health. It was equally important for us to lear n more about the political and social realities for each community and local gover nment area taking part and to communicate our support for local input in tailoring the proposed inter vention program locally7 – a commitment to reciprocity 8-10 . The third aspect was transparency of information and process. We wanted to be explicit from the outset about what participation in the trial would involve, what would be required of communities taking part, and what contributions we as the research team would be able to make. This meant identification upfront of the likely costs, contributions, timeframes and benefits of participation, as clearly as was possible prior to commencement. 1,8,10 Fourth, PRISM was conceived as building on resources within communities,11-13 with the aim of strengthening existing service systems of suppor t to recent mothers. The main providers of primar y care and community services to women in the year after birth in Victoria include the neighbourhood-based universal maternal and child health service (staffed by nurses with additional qualif ications in midwifery and community and child health, and funded by State and local gover nment), primary medical care by general practitioners, and a range of famil y and community services (e.g. child care, family suppor t, etc.) provided by local government and other community agencies. Recognition of the maternal health role played by these local services and acting to strengthen the existing network of service provision were central to PRISM, and impor tant aspects to communicate from our earliest contacts with communities. Fifth, we aimed to develop mutually respectful and collaborative par tnerships8,11,13,14 with par ticipating communities. We recognised that previous interactions between research teams and communities had not always been satisfactor y because of factors such as lack of infor mation, greater burden imposed than communities were first made aware of, lack of detailed feedback, insufficiently clear def inition and agreement about roles and responsibilities, and lack of awareness by researchers of local issues and concer ns.10,11,15 Our strategies included: travelling to communities across the State, having principal investigators involved in briefings, inviting dialogue and feedback, listening to problems and responding with support and action where possible, and providing full and frank infor mation about the trial.1,8,11,14 Finally, our research with recent mothers across Victoria and review of existing evidence about intervention strategies infor med the inclusion of common essential program elements in all intervention communities, but we also appreciated the importance of communities tailoring the proposed intervention strategies to suit local contexts,4,10,12 and the likelihood that local communities might wish to add to the program their own strategies to support mothers. This was an important rationale for co-ordination of the PRISM program in each inter vention community by a local steering committee suppor ted by a full-time community development officer (CDO). Strategies for establishing community partnerships to implement PRISM Figure 1 is a flow char t of the process utilised in PRISM for identifying, recruiting and randomising communities. It is worth noting at this point that eligible communities were considered to be those municipalities with between 300 and 1,500 bir ths per year. In consultation with community development experts, we judged that this upper and lower boundary def ined feasible-sized communities for the work of one full-time CDO. Figure 1: Recruitment of Victorian rural and metropolitan local government authorities (LGAs) to PRISM. 78 Victor ian LGAs 47 rural/31 metro 33 LGAs eligible for PRISM (No. of births per ann um between 300 & 1,500) 17 rural/16 metro 26 LGAs submitted expressions of interest 14 rural/12 metro 1 rural LGA withdrew expression of interest prior to briefing 25 LGAs briefed about PRISM 13 rural/12 metro 21 LGAs signed Memorandum of Understanding 11 rural/ 10 metro 16 LGAs selected from all possible sets of matched pairs âžœ âžœ âžœ âžœ âžœ âžœ 8 rural LGAs (4 matched pairs) âžœ 4 Intervention LGAs 8 metro LGAs (4 matched pairs) âžœ AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH âžœ 4 Comparison LGAs âžœ 4 Intervention LGAs âžœ 4 Comparison LGAs âžœ VOL. 28 NO. 5 Mothers, Children and Adolescents Establishing partnerships for community trials Initial assessment of the acceptability and feasibility of PRISM In the v ery early stages of trial protocol development, discussions were held with community services managers in two metropolitan municipalities about the likely acceptability of PRISM to local government and local communities. These discussions indicated potential interest in the project and a belief in its feasibility, and confirmed our view that local gover nment was the most appropriate auspice for the project. This was because of the statutory responsibility of local government for ser vice provision to all mothers and babies post hospital dischar ge, following receipt of birth notif ications sent by hospitals. Local gover nment in Victoria is also involved in the provision of a range of other family and community services (child care, immunisation services, libraries, family suppor t services). The responsibility for maintaining the data systems (based on birth notifications) that would facilitate contact with mothers for the planned health outcome assessment in the trial also made local government the most appropriate public authority to approach to provide informed consent for community participation in PRISM.15,16 There are ver y real challenges in identifying who can legitimately ‘represent’ a community in decision making about participation in a trial. It could be argued that consent to participation ought to involve the targeted participants, in this case recent mothers. Obtaining such consent was not possible, however, as women w ho would be giving bir th during the trial were not identif iable prior to trial commencement. infor mation about eligible communities prior to our approaching local councils. The compilation included publicly available infor mation about each local gover nment area (such as council annual reports, municipal pub lic health plans, community service directories and organisational char ts identifying key decision makers), documentation of rele vant health service activity (Maternal and Child Health Services, special outreach programs for mothers and babies and General Practice Division projects focused on pre gnancy, birth and the postnatal period). Australian Bureau of Statistics data on each local gover nment area provided a socio-demographic profile of each community, and the most recent data on numbers of bir ths in each municipality were requested from the Victorian Perinatal Data Collection Unit, within the Depar tment of Human Services. Local gover nment had experienced unprecedented change in the years immediately prior to the commencement of PRISM, including: forced amalg amations that reduced the number of municipalities from 210 to 78, elected local councils replaced by State Government appointed commissioners to manage the amalgamation process, and a State Gover nment requirement that a minimum of 50% of all municipal services be put out to open tender. In most areas elected councils had been retur ned by the time we invited participation in PRISM, yet the repercussions of these years of change were still being keenly felt in local communities and acknowledgement of this was important in our early discussions with local gover nment. Inviting local government participation PRISM Reference Group One of the f irst steps in PRISM was establishment of a project reference group, which comprised key stakeholders with relevant skills and expertise in local gover nment, community psychiatry, primary care, r ural health, general practice, maternal and child health nursing, consumer advocacy, psychology, public health, community development and psychiatric epidemiology . Once funding was obtained to commence the trial, we consulted the Municipal Association of Victoria (MA via its representative V), on the reference group, about our planned approach to local municipalities and the project information package to be provided. These discussions also allowed some of the logistical aspects of implementing PRISM to be discussed, such as whether CDOs should be employees of the university or of local government. The MA supported the view that university appointments would V provide uniformity of remuneration and access to research team support and staff development for CDOs and was likely to be preferable. Our rationale in approaching all the eligible communities across the State of Victoria was to be able to comment on the generalisability of the PRISM f indings after implementation, although we were unab le to involve all those interested in taking par t. Constraints on funding and the limited availability of noncontiguous, matched community pairs meant that we were able to include only the minimum number of communities necessary to detect an impor tant effect of the program. We estimated this to be seven pairs. 17 The eligible local government authorities of 17 rural and 16 metropolitan areas 18 were sent a letter of invitation accompanied by an information package about the trial and a form for expression of interest in a formal project briefing, to be returned by a specif ied date. This invitation package was sent to the ma yor, the chief executive officer and the community services manager in all 33 municipalities. A copy of the information package can be vie wed on the PRISM website at www.latrobe.edu.au/csmch/prism/ infopack_bd.pdf. Where no response had been received by the specified date, phone calls w ere made to the community services manager to check whether the package had been received and to ask w hether the non-response reflected lack of interest in par ticipation or whether further time or discussion was required. Twenty-six local government authorities submitted written expressions of interest in par ticipating and in receiving a more detailed briefing about PRISM with the research team. Reasons for not wishing to take part included: the potential costs to Council Collection of information about eligible municipalities Developing an understanding of the local government context and building a prof ile of each of the communities we intended to approach formed the next stage in the process. One of the authors (WD), who had previously worked in local gover nment, spent six months part time compiling background 2004 VOL. 28 NO . 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Small et al. Practice Note (2); lack of agreement with the contracted maternal and child health service about par ticipation (2); council restructuring (1); a combination of factors, including a perceived conflict of interest due to participation on the PRISM reference group (1); and unknown, due to repeated non-return of phone calls (1). One municipality subsequently decided not to have a brief ing session, as they felt over-committed to other projects. Formal brief ing sessions took place with the remaining 25 areas – 12 in the Melbourne metropolitan area and 13 in rural parts of Victoria. authorities returned signed MoUs to the university. These were countersigned by La Trobe University’s dean of health sciences and a copy returned to each local area signatory. The process of selecting the f inal set of par ticipating communities started by identifying all possible pair matches and from these randomly selecting a set of eight pairs.18 All areas not included in the f inal set of 16 were notif ied of this outcome by phone and letter. Trial launch and randomisation Briefing local government about PRISM All briefings were held in local communities, mostly at municipal offices, involving more than 4,500 kilometres of travel by the research team. Local government par ticipants numbered 103 in total (range of par ticipants from 2-11,with a median of three). Two members of the PRISM research team (two of RS, SB, WD, JL) attended all brief ings. The brief ings were interactive sessions that enabled not only the presentation of more detailed information about the trial, but also the opportunity for local government par ticipants to provide comment on the trial, seek clarification, or raise concerns. For the research team the briefings provided an opportunity to develop fur ther understanding about service provision and local issues in each municipality. The PRISM project website includes a detailed paper covering the range of issues raised and discussed at briefing sessions (see www.latrobe.edu.au/csmch/prism/briefings_bd.pdf ). A public launch of the trial at which the randomisation of the 16 communities was undertaken took place in Melbourne in June 1998. Representatives of all participating areas were invited to take par t in a half-day workshop, preceding the randomisation. This provided a for um for public discussion of the evaluation of social interventions and, in par ticular, the rationale for conducting PRISM as a community randomised trial. Invitations to attend the launch and workshop w ere also extended to the f ive municipalities not included in the final group of participating communities and these were taken up by two of the five, with one participant travelling six hours by car to attend. As well as the invited local government representati ves, participants came from Divisions of General Practice, community health services, our own and other universities, research funding bodies, the Victorian Depar tment of Human Services, and hospitals. They included mater nal and child health nurses, community services managers, community health workers, general practitioners, academics, researchers, parliamentarians and midwives. In all, par ticipants numbered 139. Our intention in holding the workshop was to achieve as diverse a for um as possible for presenting and discussing the issues in evaluating community interventions, in the context of the launch of PRISM. The workshop involved presentations and discussion on the theme ‘Evaluating social and community interventions: integrating theory and practice’. Our aims also included making research understandable, fur ther elaborating the background to the trial, putting our previous research and our proposed methods for the trial up for scrutiny, and introducing the economic and ecological evaluation components of the trial in more detail. Prior to randomisation, all the possible randomisation outcomes for the rural and metropolitan community pairs were on display for all workshop par ticipants to peruse, with the aim of making the process as transparent as possible. Each of these combinations was placed in a sealed opaque envelope with 16 yellow envelopes containing the possible metropolitan outcomes and four blue envelopes containing the possible rural outcomes. One envelope was then randomly selected from each colour g roup and the name and allocated status of each metropolitan and rural community was read to all present. The day after the public randomisation, all areas allocated comparison community status were contacted and an opportunity was provided to discuss their response to the outcome of randomisation, to reinforce how important their role as comparison areas would be in the trial and to mention the trial newsletter that VOL. Agreement to participate: a joint memorandum of understanding Each briefing session ended with a discussion of the proposed joint memorandum of understanding (MoU) between La Trobe University and each local gover nment area, with the option left open for amendments to be negotiated. The MoU contained a project overview, roles and responsibilities of project participants, a description of the inter vention program, the process for randomisation and implications of allocation, implementation costs available from project funds, the evaluation plan, timelines and mechanisms for communication between the research team and local government including a process for dispute resolution. A copy of the generic MoU utilised in PRISM can be viewed at www.latrobe.edu/csmch/prism/MOU_bd.pdf. The information provided about the trial included information about the process for selecting a f inal set of participating communities (likely to be fewer than those wanting to participate) and details of the subsequent randomisation process. Effor ts were made to ensure that each area had a g rasp of what was going to happen if they signed up and that doing so meant taking their chance, f irst of being randomly selected from among all areas that agreed to par ticipate and were able to be matched with a pair and, second, if selected, a 50/50 chance of being allocated intervention or comparison community status. Those areas wishing to proceed were asked to retur n a signed MoU to the university by a pre-deter mined date (allowing 2-4 weeks for further discussion at the local level). Twenty-one local go vernment AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 28 NO. 5 Mothers, Children and Adolescents Establishing partnerships for community trials we would be sending out at regular intervals to keep them in touch with trial progress. Reactions to comparison status allocation varied from disappointment: “We’re still in mourning here … Still we did know the process” to mild relief, and acceptance. All eight comparison areas were keen to be kept infor med via the newsletter and confirmed their agreements about continued par ticipation in the trial. We also received positive feedback about the prerandomisation workshop in these conversations: “It’s so rare to get to know so much about a project before it star ts”, and appreciation of efforts made to involv e people. Intervention communities were also contacted by phone shortly after the launch and all 16 communities were then advised in writing of the outcome of randomisation. University and Victorian Depar tment of Human Services Collaborative Industry Grant 1997; NHMRC project grants 199799 and 1999-2001, 2002-04; Victorian Depar tment of Human Services program implementation and evaluation grants 19982000; Victorian Health Promotion Foundation grants for program resources and implementation 1998-2000; Felton Bequest grant 1998; Sidney Myer Fund grant 1999; beyondblue the National Depression Initiative grant 2002; and participating municipalities’ contributions to program implementation and data collection from 1998. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

PRISM: Mounting a community‐randomised trial. Establishing partnerships with local government

Loading next page...
 
/lp/wiley/prism-mounting-a-community-randomised-trial-establishing-partnerships-NBNPcF2H0x

References (23)

Publisher
Wiley
Copyright
Copyright © 2004 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/j.1467-842X.2004.tb00030.x
Publisher site
See Article on Publisher Site

Abstract

Rhonda Small, Stephanie Brown, Wendy Dawson, Lyndsey Watson and Judith Lumley Mother and Child Health Research, School of Public Health, La Trobe University, Victoria Abstract This paper describes initial partnership development in PRISM (Program of Resources, Information and Support for Mothers), a community-randomised trial to improve maternal health in the first year after birth conducted in Victoria in the period 1998-2003. First, we discuss the principles underpinning community recruitment methods in PRISM that guided both our initial approaches to, and our continuing relationships with, communities. Second, we outline the strategies used to recr uit communities and to establish the groundwork for ongoing partnerships over the projected six years of the study. ( Aust N Z J Public Health 2004; 28: 471-5) he importance of collaborati ve partnerships and of community participation and ‘ownership’ for the success and sustainability of population-based health promotion activities are now considered axiomatic, although the benefits have been noted as largely ‘reputed’1,2 rather than firmly established. In a discussion of the implications of the results of population-level health interventions from the 1980s and 1990s, Sorensen et al. called for greater community involvement in program planning and attention to community competence or capacity building in implementation in future community intervention trials.3 They also emphasised the importance of more creativity in dealing with the tension inherent in the need for protocol standardisation in trial design and appropriate flexibility in the intervention to meet differing community needs and priorities. Similarly, Shediac-Rizkallah and Bone discussed the importance of ‘project negotiation’ in design and implementation phases to strengthen project sustainability, stressing consensusbuilding to reach a compromise for addressing the needs of all project partners – researchers, funders and community participants.4 Despite broad-based ag reement about the importance of community involvement for the success of population-based health interventions, at the time we approached local gover nment authorities about par ticipation in PRISM in early 1998 surprisingly little had been written about strategies for initiating community collaborations in the context of communityrandomised trials. This paper is one step in filling that gap, describing how we went about the process of community recr uitment to PRISM, a community-randomised trial that aimed to impro ve the ph ysical and emotional health of recent mothers via an integ rated program of primary care and community-based strategies. The paper f irst outlines the principles underlying our approach to communities and, second , the strategies employed to establish local gover nment par tnerships in order to build the foundations for subsequent broader community collaboration in project implementation and evaluation. Principles underlying the approach to communities First, we provided comprehensiv e infor mation about the rationale for conducting the trial, the aims of the trial and a description of our pre vious research, par ticularly the studies that had developed our understanding of Victorian women’s experiences of early motherhood, their own health and their use of health services in the first years after childbirth. It w as this research that infor med the proposed inter vention program.5 We felt that such infor mation was essential for communities to assess the potential rele vance and value of the proposed intervention program and to judge our credentials as researchers. 1,6,7 Second, we believed it was impor tant for the trial to provide opportunities for people in communities to learn more about research, starting with our discussion of the proposed Correspondence to: Dr Rhonda Small, Mother and Child Health Research, School of Public Health, La T robe Univ ersity, 251 Faraday Street, Carlton, Victor ia 3053. Fax: (03) 8341 8555; e-mail: r.small@latrobe.edu.au Submitted: October 2003 Revision requested: April 2004 Accepted: July 2004 2004 VOL. 28 NO . 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Small et al. Practice Note trial design and key elements of the intervention, and the role of large community trials with rigorous evaluation in developing knowledge about ways to improve health. It was equally important for us to lear n more about the political and social realities for each community and local gover nment area taking part and to communicate our support for local input in tailoring the proposed inter vention program locally7 – a commitment to reciprocity 8-10 . The third aspect was transparency of information and process. We wanted to be explicit from the outset about what participation in the trial would involve, what would be required of communities taking part, and what contributions we as the research team would be able to make. This meant identification upfront of the likely costs, contributions, timeframes and benefits of participation, as clearly as was possible prior to commencement. 1,8,10 Fourth, PRISM was conceived as building on resources within communities,11-13 with the aim of strengthening existing service systems of suppor t to recent mothers. The main providers of primar y care and community services to women in the year after birth in Victoria include the neighbourhood-based universal maternal and child health service (staffed by nurses with additional qualif ications in midwifery and community and child health, and funded by State and local gover nment), primary medical care by general practitioners, and a range of famil y and community services (e.g. child care, family suppor t, etc.) provided by local government and other community agencies. Recognition of the maternal health role played by these local services and acting to strengthen the existing network of service provision were central to PRISM, and impor tant aspects to communicate from our earliest contacts with communities. Fifth, we aimed to develop mutually respectful and collaborative par tnerships8,11,13,14 with par ticipating communities. We recognised that previous interactions between research teams and communities had not always been satisfactor y because of factors such as lack of infor mation, greater burden imposed than communities were first made aware of, lack of detailed feedback, insufficiently clear def inition and agreement about roles and responsibilities, and lack of awareness by researchers of local issues and concer ns.10,11,15 Our strategies included: travelling to communities across the State, having principal investigators involved in briefings, inviting dialogue and feedback, listening to problems and responding with support and action where possible, and providing full and frank infor mation about the trial.1,8,11,14 Finally, our research with recent mothers across Victoria and review of existing evidence about intervention strategies infor med the inclusion of common essential program elements in all intervention communities, but we also appreciated the importance of communities tailoring the proposed intervention strategies to suit local contexts,4,10,12 and the likelihood that local communities might wish to add to the program their own strategies to support mothers. This was an important rationale for co-ordination of the PRISM program in each inter vention community by a local steering committee suppor ted by a full-time community development officer (CDO). Strategies for establishing community partnerships to implement PRISM Figure 1 is a flow char t of the process utilised in PRISM for identifying, recruiting and randomising communities. It is worth noting at this point that eligible communities were considered to be those municipalities with between 300 and 1,500 bir ths per year. In consultation with community development experts, we judged that this upper and lower boundary def ined feasible-sized communities for the work of one full-time CDO. Figure 1: Recruitment of Victorian rural and metropolitan local government authorities (LGAs) to PRISM. 78 Victor ian LGAs 47 rural/31 metro 33 LGAs eligible for PRISM (No. of births per ann um between 300 & 1,500) 17 rural/16 metro 26 LGAs submitted expressions of interest 14 rural/12 metro 1 rural LGA withdrew expression of interest prior to briefing 25 LGAs briefed about PRISM 13 rural/12 metro 21 LGAs signed Memorandum of Understanding 11 rural/ 10 metro 16 LGAs selected from all possible sets of matched pairs ➜ ➜ ➜ ➜ ➜ ➜ 8 rural LGAs (4 matched pairs) ➜ 4 Intervention LGAs 8 metro LGAs (4 matched pairs) ➜ AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH ➜ 4 Comparison LGAs ➜ 4 Intervention LGAs ➜ 4 Comparison LGAs ➜ VOL. 28 NO. 5 Mothers, Children and Adolescents Establishing partnerships for community trials Initial assessment of the acceptability and feasibility of PRISM In the v ery early stages of trial protocol development, discussions were held with community services managers in two metropolitan municipalities about the likely acceptability of PRISM to local government and local communities. These discussions indicated potential interest in the project and a belief in its feasibility, and confirmed our view that local gover nment was the most appropriate auspice for the project. This was because of the statutory responsibility of local government for ser vice provision to all mothers and babies post hospital dischar ge, following receipt of birth notif ications sent by hospitals. Local gover nment in Victoria is also involved in the provision of a range of other family and community services (child care, immunisation services, libraries, family suppor t services). The responsibility for maintaining the data systems (based on birth notifications) that would facilitate contact with mothers for the planned health outcome assessment in the trial also made local government the most appropriate public authority to approach to provide informed consent for community participation in PRISM.15,16 There are ver y real challenges in identifying who can legitimately ‘represent’ a community in decision making about participation in a trial. It could be argued that consent to participation ought to involve the targeted participants, in this case recent mothers. Obtaining such consent was not possible, however, as women w ho would be giving bir th during the trial were not identif iable prior to trial commencement. infor mation about eligible communities prior to our approaching local councils. The compilation included publicly available infor mation about each local gover nment area (such as council annual reports, municipal pub lic health plans, community service directories and organisational char ts identifying key decision makers), documentation of rele vant health service activity (Maternal and Child Health Services, special outreach programs for mothers and babies and General Practice Division projects focused on pre gnancy, birth and the postnatal period). Australian Bureau of Statistics data on each local gover nment area provided a socio-demographic profile of each community, and the most recent data on numbers of bir ths in each municipality were requested from the Victorian Perinatal Data Collection Unit, within the Depar tment of Human Services. Local gover nment had experienced unprecedented change in the years immediately prior to the commencement of PRISM, including: forced amalg amations that reduced the number of municipalities from 210 to 78, elected local councils replaced by State Government appointed commissioners to manage the amalgamation process, and a State Gover nment requirement that a minimum of 50% of all municipal services be put out to open tender. In most areas elected councils had been retur ned by the time we invited participation in PRISM, yet the repercussions of these years of change were still being keenly felt in local communities and acknowledgement of this was important in our early discussions with local gover nment. Inviting local government participation PRISM Reference Group One of the f irst steps in PRISM was establishment of a project reference group, which comprised key stakeholders with relevant skills and expertise in local gover nment, community psychiatry, primary care, r ural health, general practice, maternal and child health nursing, consumer advocacy, psychology, public health, community development and psychiatric epidemiology . Once funding was obtained to commence the trial, we consulted the Municipal Association of Victoria (MA via its representative V), on the reference group, about our planned approach to local municipalities and the project information package to be provided. These discussions also allowed some of the logistical aspects of implementing PRISM to be discussed, such as whether CDOs should be employees of the university or of local government. The MA supported the view that university appointments would V provide uniformity of remuneration and access to research team support and staff development for CDOs and was likely to be preferable. Our rationale in approaching all the eligible communities across the State of Victoria was to be able to comment on the generalisability of the PRISM f indings after implementation, although we were unab le to involve all those interested in taking par t. Constraints on funding and the limited availability of noncontiguous, matched community pairs meant that we were able to include only the minimum number of communities necessary to detect an impor tant effect of the program. We estimated this to be seven pairs. 17 The eligible local government authorities of 17 rural and 16 metropolitan areas 18 were sent a letter of invitation accompanied by an information package about the trial and a form for expression of interest in a formal project briefing, to be returned by a specif ied date. This invitation package was sent to the ma yor, the chief executive officer and the community services manager in all 33 municipalities. A copy of the information package can be vie wed on the PRISM website at www.latrobe.edu.au/csmch/prism/ infopack_bd.pdf. Where no response had been received by the specified date, phone calls w ere made to the community services manager to check whether the package had been received and to ask w hether the non-response reflected lack of interest in par ticipation or whether further time or discussion was required. Twenty-six local government authorities submitted written expressions of interest in par ticipating and in receiving a more detailed briefing about PRISM with the research team. Reasons for not wishing to take part included: the potential costs to Council Collection of information about eligible municipalities Developing an understanding of the local government context and building a prof ile of each of the communities we intended to approach formed the next stage in the process. One of the authors (WD), who had previously worked in local gover nment, spent six months part time compiling background 2004 VOL. 28 NO . 5 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH Small et al. Practice Note (2); lack of agreement with the contracted maternal and child health service about par ticipation (2); council restructuring (1); a combination of factors, including a perceived conflict of interest due to participation on the PRISM reference group (1); and unknown, due to repeated non-return of phone calls (1). One municipality subsequently decided not to have a brief ing session, as they felt over-committed to other projects. Formal brief ing sessions took place with the remaining 25 areas – 12 in the Melbourne metropolitan area and 13 in rural parts of Victoria. authorities returned signed MoUs to the university. These were countersigned by La Trobe University’s dean of health sciences and a copy returned to each local area signatory. The process of selecting the f inal set of par ticipating communities started by identifying all possible pair matches and from these randomly selecting a set of eight pairs.18 All areas not included in the f inal set of 16 were notif ied of this outcome by phone and letter. Trial launch and randomisation Briefing local government about PRISM All briefings were held in local communities, mostly at municipal offices, involving more than 4,500 kilometres of travel by the research team. Local government par ticipants numbered 103 in total (range of par ticipants from 2-11,with a median of three). Two members of the PRISM research team (two of RS, SB, WD, JL) attended all brief ings. The brief ings were interactive sessions that enabled not only the presentation of more detailed information about the trial, but also the opportunity for local government par ticipants to provide comment on the trial, seek clarification, or raise concerns. For the research team the briefings provided an opportunity to develop fur ther understanding about service provision and local issues in each municipality. The PRISM project website includes a detailed paper covering the range of issues raised and discussed at briefing sessions (see www.latrobe.edu.au/csmch/prism/briefings_bd.pdf ). A public launch of the trial at which the randomisation of the 16 communities was undertaken took place in Melbourne in June 1998. Representatives of all participating areas were invited to take par t in a half-day workshop, preceding the randomisation. This provided a for um for public discussion of the evaluation of social interventions and, in par ticular, the rationale for conducting PRISM as a community randomised trial. Invitations to attend the launch and workshop w ere also extended to the f ive municipalities not included in the final group of participating communities and these were taken up by two of the five, with one participant travelling six hours by car to attend. As well as the invited local government representati ves, participants came from Divisions of General Practice, community health services, our own and other universities, research funding bodies, the Victorian Depar tment of Human Services, and hospitals. They included mater nal and child health nurses, community services managers, community health workers, general practitioners, academics, researchers, parliamentarians and midwives. In all, par ticipants numbered 139. Our intention in holding the workshop was to achieve as diverse a for um as possible for presenting and discussing the issues in evaluating community interventions, in the context of the launch of PRISM. The workshop involved presentations and discussion on the theme ‘Evaluating social and community interventions: integrating theory and practice’. Our aims also included making research understandable, fur ther elaborating the background to the trial, putting our previous research and our proposed methods for the trial up for scrutiny, and introducing the economic and ecological evaluation components of the trial in more detail. Prior to randomisation, all the possible randomisation outcomes for the rural and metropolitan community pairs were on display for all workshop par ticipants to peruse, with the aim of making the process as transparent as possible. Each of these combinations was placed in a sealed opaque envelope with 16 yellow envelopes containing the possible metropolitan outcomes and four blue envelopes containing the possible rural outcomes. One envelope was then randomly selected from each colour g roup and the name and allocated status of each metropolitan and rural community was read to all present. The day after the public randomisation, all areas allocated comparison community status were contacted and an opportunity was provided to discuss their response to the outcome of randomisation, to reinforce how important their role as comparison areas would be in the trial and to mention the trial newsletter that VOL. Agreement to participate: a joint memorandum of understanding Each briefing session ended with a discussion of the proposed joint memorandum of understanding (MoU) between La Trobe University and each local gover nment area, with the option left open for amendments to be negotiated. The MoU contained a project overview, roles and responsibilities of project participants, a description of the inter vention program, the process for randomisation and implications of allocation, implementation costs available from project funds, the evaluation plan, timelines and mechanisms for communication between the research team and local government including a process for dispute resolution. A copy of the generic MoU utilised in PRISM can be viewed at www.latrobe.edu/csmch/prism/MOU_bd.pdf. The information provided about the trial included information about the process for selecting a f inal set of participating communities (likely to be fewer than those wanting to participate) and details of the subsequent randomisation process. Effor ts were made to ensure that each area had a g rasp of what was going to happen if they signed up and that doing so meant taking their chance, f irst of being randomly selected from among all areas that agreed to par ticipate and were able to be matched with a pair and, second, if selected, a 50/50 chance of being allocated intervention or comparison community status. Those areas wishing to proceed were asked to retur n a signed MoU to the university by a pre-deter mined date (allowing 2-4 weeks for further discussion at the local level). Twenty-one local go vernment AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH 28 NO. 5 Mothers, Children and Adolescents Establishing partnerships for community trials we would be sending out at regular intervals to keep them in touch with trial progress. Reactions to comparison status allocation varied from disappointment: “We’re still in mourning here … Still we did know the process” to mild relief, and acceptance. All eight comparison areas were keen to be kept infor med via the newsletter and confirmed their agreements about continued par ticipation in the trial. We also received positive feedback about the prerandomisation workshop in these conversations: “It’s so rare to get to know so much about a project before it star ts”, and appreciation of efforts made to involv e people. Intervention communities were also contacted by phone shortly after the launch and all 16 communities were then advised in writing of the outcome of randomisation. University and Victorian Depar tment of Human Services Collaborative Industry Grant 1997; NHMRC project grants 199799 and 1999-2001, 2002-04; Victorian Depar tment of Human Services program implementation and evaluation grants 19982000; Victorian Health Promotion Foundation grants for program resources and implementation 1998-2000; Felton Bequest grant 1998; Sidney Myer Fund grant 1999; beyondblue the National Depression Initiative grant 2002; and participating municipalities’ contributions to program implementation and data collection from 1998.

Journal

Australian and New Zealand Journal of Public HealthWiley

Published: Oct 1, 2004

There are no references for this article.