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

Learn More →

Comparing the performance of the public, social security and private health subsystems in Argentina by core dimensions of primary health care

Comparing the performance of the public, social security and private health subsystems in... Background. Most Latin American health systems are comprised of public (PubS), social security (SSS) and private (PrS) subsystems. These subsystems coexist, causing health care fragmentation and population segmentation. Objective. To estimate the extent of subsystem cross-coverage in a geographically bounded population (Rosario city) and to compare the subsystems’ performance on primary health care (PHC) dimensions. Methods. Through a cross-sectional, interviewer-administered survey to a representative sample (n = 822) of the Rosario population, we measured the percentage of cross-coverage (people with usual source of care in one subsystem but also covered by another subsystem) and the health services’ performance by core PHC dimensions, as reported by each subsystem’s usual users. We compared the subsystems’ performance using chi-square analysis and one-way analysis of variance testing. We analyzed whether the observed differences were coherent with the predominant institutional and organizational features of each subsystem. Results. Overall, 39.3% of the population was affiliated with the PubS, 44.8% with the SSS and 15.9% with the PrS. Cross-coverage was reported by 40.6% of respondents. The performance of the PubS was weak on accessibility but strong on person-and-community-oriented care, the opposite of the PrS. The SSS combined the strengths of the other two subsystems. Conclusion. Rosario’s health system has a high percentage of cross-coverage, contributing to issues of fragmentation, segmentation, financial inequity and inefficiency. The overall performance of the SSS was better than that of the PrS and PubS, though each subsystem had a particular performance pattern with areas of strength and weakness that were consistent with their institutional and organizational profiles. Key words. Health Care Quality, Access, and Evaluation, health services research, integrated health care systems, Latin America, primary health care, public health. © The Author 2016. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 250 Family Practice, 2016, Vol. 33, No. 3 to complement or avoid the limitations of the subsystem with which Introduction they were originally affiliated. Most Latin American ‘health systems’ are composed of public (PubS), The objective of this study is to estimate the extent of subsystem social security (SSS) and private (PrS) subsystems. Each has its own cross-coverage in a population in a given geographic area and to philosophical underpinning, funding base, management framework compare the subsystems with respect to their performance on PHC and intended catchment population. Though the performance of core dimensions as reported by the affiliated population. We use the their respective health services is expected to vary depending on the municipality of Rosario, Argentina, as the study case. population served and each subsystem’s organizational and insti- tutional features (1), an oversimplified portrait persists in policy documents and some of the peer-reviewed literature. This portrait Rosario health system characterizes the private subsystem as more efficient, accountable Rosario is the third most populous city in Argentina (948 312 inhab- and quality driven than the public health subsystem, in turn biasing itants). Argentina is a federal democracy, with 23 provinces and the public opinion and policy debate (2). Autonomous City of Buenos Aires. Funding and delivery of public This oversimplified portrait masks the differences within each services are the authority and responsibility of each jurisdiction. The of the subsystems and the fragmented nature of the available data regulatory frameworks share common features across jurisdictions, does not permit a direct comparison between subsystems. In fact, but health services delivery and organization vary. Depending on the so-called private subsystem is actually a collection of different fiscal capacity, funding and delivery has been further decentralized governance structures and delivery systems of varying degrees of to the municipal level. Like many Latin American countries and complexity. The SSS can also include multiple governance organi- jurisdictions, Rosario’s population is covered by segmented public, zations and delivery institutions. Even the PubS can have differ- social security and private systems. It shares most of the institutional ent levels of municipal, state or national government health care and organizational features of the Latin American segmented health schemes. Furthermore, the population can be covered by different systems, synthesized in Table  1. Below, we describe Rosario’s sub- subsystems, with the PubS serving as the ultimate safety net for all. systems, including features that are distinct from the typical Latin Cross-coverage of the population, interdependence between subsys- American case. tems and the changes that have occurred in the Latin American pri- vate health market during recent decades require a more nuanced Rosario’s public subsystem analysis (3). Rosario is one of the country’s richest municipalities, although Across Latin America, countries are making efforts to move ~115 000 of its residents live in slums (8). Since the 1990s, municipal toward the development of integrated, primary health care (PHC)- authorities have invested substantially to develop a publicly funded based health systems to improve access, efficiency, equity and qual - local health system (PubS) that currently provides a wide range of ity for the whole population (4,5). Our purpose is to compare the free health care services, aimed mainly at the population that is not performance of each subsystem using a common set of tools and covered by other systems—approximately a third of the city’s popu- to understand the degree of overlap between subsystems in order lation. Rosario’s PubS is a vertically integrated system of ambulatory to help governments to focus strategies to improve governance and health centres (with multidisciplinary teams), hospitals, an ambula- delivery (6). tory medical specialities centre, a rehabilitation centre and an emer- This study is guided by the World Health Organization (WHO) gency system. Resources are assigned to the system based on public framework for health system performance (1,7). According to this budgets. Per capita health spending in the Rosario PubS was $1907 framework, health systems are characterized by key functions such Argentinean pesos [$479 US dollar (USD), according to the official as stewardship, financing and resource generation, which translate exchange rate at 12 December 2010] in 2010 (9,10). into service provision to the population. High-performing services Furthermore, health professionals have intentionally developed have instrumental goals of providing high quality, responsive and Rosario as a model of participatory management with a multidis- affordable care. This care is the means to achieving the fundamental ciplinary first level of care that has a strong role in coordinating goals of the health system: to improve the average health status of the system. Rosario has been recognized in Latin America as a PHC the population (including reducing health inequalities); to be respon- champion and an example of a PHC-based health system because of sive to the non-medical expectations of the population (respect for the development and achievements of its PubS (11). people and a client orientation); and to ensure fairness in financial contributions. PHC services contribute to these goals through their Rosario’s social security subsystem instrumental goals of timely access, comprehensiveness, continuity of care, person-and-community orientation and technical quality The SSS is an employment-based insurance, with financing based on of care. a percentage of workers’ salaries and not on premiums that can be The premise of this study is that, in a context of multiple health adjusted. The Argentinean Ministry of Health regulatory framework subsystems, differences in the key functions of each system will requires that SSS insurers provide an obligatory broad basket of ser- translate into differences in the extent to which PHC services achieve vices to workers and their families. The Rosario SSS subsystem is their instrumental goals in core PHC dimensions. Our research made up of around 50 of the nearly 300 entities functioning around questions are two. Given the fragmented nature of the Argentinean the country. The entities vary in scope and size and have different health system, what is the extent of cross-coverage between subsys- models of care and management, all of which together attempt to tems? How well is each subsystem performing on core PHC dimen- provide a broad package of services within the limited financing avail - sion, as reported by their regular users? We hypothesize that (i) the able from worker contributions. Unlike other Latin American SSS emerging patterns of strengths and weaknesses in the performance entities, in Argentina and Rosario few entities have their own health of each subsystem will be consistent with its funding, governance, services. Instead, they purchase health services through per capita and organizational particularities and (ii) there will be a high rate of and/or fee-for-services contracts with private providers through selec- cross-coverage largely due to the strategies used by the population tive commissioning. They use different strategies to control, monitor Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 251 Table 1. Typical institutional and organizational features of Latin American health subsystems Latin American subsystems Public Social security Private Funding General taxes Employers’ and employees’ payroll Clients’ voluntary contributions deductions (prepaid insurance and direct payment) Population served Population without contribution Workers in the formal labour market High-, middle- and low-income capacity populations Health services coverage Basket of basic free services. Other Broader basket of services compared For high-income population, a broad services may require out-of-pocket to the PubS, but still requires basket of services is offered at a payments or may not be offered copayments high cost For low- and middle-income popula- tion (uninsured), complementary or supplementary plans are offered at a relatively high cost Authorities/management National, subnational and local Single or multiple social security Multiple for-profit companies and ministries/secretaries organizations non-profit organizations Integration of funding and services Vertical integration: owned health Vertical integration: owned health Contractual integration: purchase provision services services health services through per capita and/or fee-for-services contacts Main professionals’ remuneration Salaries Salaries Fee-for-service system and improve health care quality such as economic incentives, care of the municipality of Rosario. The initial sampling unit was the guidelines and the development of information and auditing systems census tract stratified by the proportion of households below the (12). Per capita health spending in the SSS ($637 USD) is 33% greater poverty threshold. Once the census tract was selected, all residential than in the PubS, but 37% lower than for private insurers (10). dwellings in the city blocks that were located within the perimeter of the census tract were contacted. In active households, all mem- bers were enumerated and one respondent was randomly selected Rosario’s private subsystem until the required sample size in each age strata was obtained The PrS comprises over 30 companies. There are three types of enti- (age < 15 years, 15–49 years and 50+ years). The final sample of 822 ties: voluntary prepaid insurance companies that offer a variety of residents was designed to represent the age–sex demographic strata, premiums; emergency services companies that provide emergency with a 95% margin of error of 3.5%. The survey was conducted care and urgent home care; and social services companies that offer between December 2010 and January 2011. a basic basket of essential ambulatory services through cheap premi- We assigned respondents to a particular subsystem based on their ums. Like the SSS, PrS entities contract service delivery from private affiliation to a specific facility that they reported as their regular or providers and professional associations. Clients choose their own most frequently used place of medical care (from now on, referred to providers from the range of entities and professionals offered by as usual source of care). As some facilities might provide health care their insurer. Services are provided in free-standing health practi- for people affiliated with different subsystems, the respondents were tioner offices, independent clinics and/or clinics owned by insurance asked to indicate the type of coverage used to receive medical care companies. Clinics tend to be organized by discipline, so multi- at that facility. We did not assign respondents based on their health disciplinary work is uncommon. Most health professionals work coverage status because it is likely to find people who are affiliated under an entrepreneur model and with fee-for-service remuneration with one subsystem and also covered by another subsystem (e.g. cov- depending on the number and type of medical services. In contrast to ered by the PrS and the SSS but affiliated with the SSS). the SSS, the PrS has a more passive role in controlling and coordinat- Those without a usual source of care (n = 14) were assigned to ing health services, with minimal mechanisms to regulate clients’ use the subsystem where they had their last visit. We excluded 21 people of services and health professionals’ practices (12). Per capita health who reported affiliation to either non-governmental health organiza - spending in the PrS ($873 UDS) is 82% greater than in the PubS and tions or traditional medicine, since these were beyond the scope of 37% higher than in the SSS (10). this study. In this study, we selected a representative sample of the popula- tion of Rosario to assess the subsystem to which individuals are pre- dominantly affiliated and reported each subsystem’s performance on Survey instrument core dimensions of PHC as reported by each subsystem’s usual users. The questionnaire included 67 questions related to seven sections: We then compared the performance of each of the subsystems to dis- (i) demographic and socio-economic conditions of the respondent cern whether the observed differences are consistent with the pre- and his/her family, with questions on health insurance, employment, dominant institutional and organizational features of the subsystems. literacy and housing; (ii) access to health care within populations with chronic conditions, where respondents were asked to indicate whether they had been diagnosed with any of a number of condi- Methods tions and if they were currently receiving treatment for those condi- Study design, sample and data collection tions; (iii) health care needs, with questions about their experiences We conducted a cross-sectional interviewer-administered survey in the last 12  months; (iv) continuity of care, where respondents using a multistage, stratified sampling strategy designed to obtain were asked to identify their usual source of care and respond to a a sample representative of the age/sex and socio-economic strata number of questions about relational and informational continuity Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 252 Family Practice, 2016, Vol. 33, No. 3 of care at the usual source of care, all based on their experience in Given that there is no consensus on performance standards to the last 12 months; (v) person-and-community-oriented care, again determine whether a specific benchmark was achieved or surpassed, with a focus on the usual source of care during the last 12 months; we described performance achievements in relative terms as per (vi) accommodation, which included two sets of questions related to Murray and Frenk (1). As in other studies, we used a combination of the respondents’ experience with their usual source of care, one set statistical significance and judgment to denote ‘weak’ or ‘strong’ per - based on their experience during the last 12 months and the other formance (2,6). For each variable that was statistically significantly based on their experience during the last consultation; and (vii) different by subsystem, we used our judgment to denote the ‘strong’ Health care use, including questions on health services utilization or ‘weak’ subsystem(s) relative to the other(s). For each PHC dimen- during the last 12 months and the last medical consultation. sion and sub-dimension, we identified as ‘strong’ the subsystem with The questions for respondents younger than 15 years of age were the strongest variables and as ‘weak’ the subsystem with the weakest answered by the adult with the greatest responsibility for their care. variables. The questions were mostly closed, with a mix of reporting and evalu- ative response options, and were taken as much as possible from Results a Canadian validated instrument (13) and other previous surveys conducted in Argentina (11,12). Demographics and socio-economic characteristics by health subsystems Analytic strategy Overall, 39.3% of the population was affiliated with the PubS, Analysis was conducted in two stages: (i) reduction of variables into 44.8% with the SSS and 15.9% with the PrS. We found significant underlying dimensions of PHC performance and (ii) comparison of differences among subsystems by participant sex and age and house- subsystems by variable and dimension. hold socio-economic conditions, including the head of household’s employment status and educational level and the overall household living conditions (Table 2). The respondents affiliated with the PubS Dimensions of primary health care performance subsystem were more likely to be younger, male, unemployed, less We reduced the large number of reported variables into broad educated and to report worse living conditions than respondents in dimensions of PHC using a categorical nonlinear principal compo- the other subsystems. For respondents in both the SSS and PrS sub- nents analysis (CATPCA) (14). CATPCA has the same objectives as systems, the respondent profile was older, more female and educated, traditional principal components analysis, but it allows for variables and among household heads, unemployment was lower and the with mixed measurement levels, transforming them into categorical percentage of retired workers was higher. The higher proportion of variables and creating linear combinations between the transformed older adults in the SSS is expected, given that it includes the national variables. insurance for seniors and pensioners. The relationship between the components and variable loadings Cross-coverage (usual care from one subsystem but also covered can be represented graphically through component loading plots by another subsystem) was reported by 40.6% of respondents. Half of the principal component space, where the axes are the principal of those affiliated with the SSS had complementary or supplemen - components. In these plots, variables with relatively long vectors fit tary PrS coverage. Over a third of PubS affiliates had SSS and/or PrS well into the solution, with the squared length of the loading vec- coverage and a quarter of PrS affiliates had SSS coverage. tor equalling the variance accounted for. When vectors are long, the cosines of the angles between the vectors indicate the Pearson’s cor- relation between the quantified variables. The slope of the vectors Relationship between components by system indicates the relationship of each variable with each component. An In all three subsystems, the first component combines the dimensions angle of 90° indicates no correlation between variables, 180° indi- we hypothesized as continuity of care, person-and-community-ori- cates negative correlation and 0° indicates positive correlation (14). ented care and health utilization while the second maps onto accessi- All analyses were conducted using SPSS version 21. bility. Health care access within populations with chronic conditions We hypothesized a priori which variables best represented each maps onto the first component in both the SSS and PrS and onto the dimension and sub-dimension. We used 19 variables: 12 ordinal, 6 second component in the PubS. Except for question 13 in the SSS, nominal and 1 continuous. Two components accounted for >30% of unmet health care needs maps onto the first component. the variance in each subsystem (PubS 31.8%; SSS 32.2%; PrS 31.0%), The CATPCA component loading plot showed a positive correla- with Cronbach’s alphas of ~0.90 (PubS 0.899; SSS 0.888; PrS 0.912). tion between components in the PubS (Fig. 1); a null correlation in the SSS (Fig. 2); and a negative correlation in the PrS (Fig. 3). Comparison of subsystems Subsystems were compared to look for significant differences in the Primary health care performance by health demographic and socio-economic characteristics of the respondents subsystems in each system, using chi-square analysis for categorical variables Subsystems showed significant differences in performance for all and one-way analysis of variance tests for continuous variables. We PHC dimensions and sub-dimensions, with the exception of unmet conducted a similar analysis for each of the variables for the fol- health care needs and medical consultation rates (Table 3). lowing PHC performance dimensions: access to health care services; The main strengths of the PubS were continuity of care, appro- continuity of care; person-and-community-oriented care and health priateness of place and provider and person-and-community-ori- services utilization. According to the WHO framework for health ented care. The performance of the PubS in continuity of care and system performance (1,7), high performance on these dimensions is a appropriateness of place and provider approximates the SSS, but it means to achieve the health system’s fundamental goals, particularly has a lower percentage of people whose usual source of care is an the goal of being responsive to the non-medical expectations of the emergency or after-hour service, a higher percentage of people whose population. regular provider is a general physician (instead of a specialist) and Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 253 Table 2. Study population demographics, socio-economic conditions and health insurance status by health subsystem, Rosario, 2011 (n = 801) Total Public Social security Private Significance level subsystem subsystem subsystem n % n % n % n % Demographics Population 801 100 315 39.3 359 44.8 127 15.9 – Sex Male 418 52.2 182 57.8 179 49.9 57 44.9 0.024 Age 0–12 years 155 19.4 80 25.4 58 16.2 17 13.4 0.000 13–20 years 120 15.0 68 21.6 37 10.3 15 11.8 21–60 years 358 44.7 149 47.3 145 40.4 64 50.4 61–99 years 168 21.0 18 5.7 119 33.1 31 24.4 Socio-economic conditions Household head, Employed 635 79.4 277 87.9 262 73.2 96 75.6 0.000 employment condition Unemployed 17 2.1 15 4.8 1 0.3 1 0.8 Retired 148 18.5 23 7.3 95 26.5 30 23.6 Household head, Never attended school or incomplete primary 131 16.4 51 16.2 57 15.9 23 18.1 0.002 educational level Complete primary 209 26.1 104 33.0 80 22.3 25 19.7 Incomplete secondary 138 17.3 62 19.7 54 15.1 22 17.3 Complete secondary 207 25.9 73 23.2 99 27.7 35 27.6 Incomplete or complete post-secondary or university 115 14.4 25 7.9 68 19.6 22 17.3 Type of housing Deficient 334 41.7 172 54.6 114 31.8 48 37.8 0.000 Health coverage Cross-coverage between No cross-coverage 426 53.2 201 63.8 164 45.7 61 48.0 – subsystems Cross-coverage 325 40.6 114 36.2 179 49.9 32 25.2 Lost insurance given by the subsystem of affiliation during the last year 50 6.2 – – 16 4.5 34 26.8 A house is considered deficient when it has at least one of the following characteristics: (i) floor—dirt or loose brick, (ii) walls—wood, metal or fibrocement sheets, (iii) walls without external plaster or coating, (iv) roof without internal coating or ceiling only or (v) has a water source outside of the home. The subsystem affiliation refers to the subsystem of the provider reported as the usual or most frequently used place of medical care. The term coverage includes two situations: people with insurance from the SSS or the PrS and people with a factual link to the PubS, as the PubS does not provide insurance to anyone and theoretically provides coverage to everyone. Shifted from SSS to PrS, vice versa or lost insurance. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 254 Family Practice, 2016, Vol. 33, No. 3 a higher percentage of people who are more likely to have very or for the last variable, there were no significant differences between totally complete medical records in their usual source of care. The subsystems. The performance of the SSS in person-and-community- performance of the PubS for person-and-community-oriented care is oriented care was better than that of the PrS but substantially lower far superior to that of the PrS and the SSS. than that of the PubS. Regarding appropriate use of place, the per- Accommodation, access to health care in populations with formance of the SSS was similar but inferior to that of the PubS. chronic conditions and comprehensive use of services are the major With respect to appropriate use of provider, the SSS had the high- performance weaknesses in the PubS. The PubS had the weakest per- est proportion of population whose regular provider is a specialist, formance on the accessibility dimension, with statistically significant probably due to the higher proportion of senior population and the differences on all nine indicators and respondents expressing particu- higher rate of chronic illness detection and treatment. lar difficulty with the appointment process. Compared to the SSS and PrS, the percentage of people diagnosed with chronic conditions in Component loadings the PubS is significantly lower, as is the percentage of people diag - nosed with at least one chronic condition who are currently receiving 1,00 treatment. It should be noted that in the PubS, health care access in the population with chronic conditions and accommodation are com- bined in component 2 (Fig.  1). Although there were not significant 0,75 differences between subsystems, the PubS had the worst performance on unmet health care needs. Finally, affiliates in the PubS were least likely to consult a wide range of services (comprehensive access). 0,50 The PrS had a performance profile that was the opposite of the 50 PubS. It had the weakest performance in appropriateness of place and provider, continuity and person-and-community-oriented care. 0,25 The performance of the PrS for access to health care in populations with chronic conditions was better than that of the PubS, but the rate of chronic illness detection and treatment were 5.0 and 8.8 points 18 0,00 lower, respectively, than those of the SSS. The performance of the 21 PrS on comprehensiveness was almost as good as the SSS, and it had 11 the strongest performance on all but one (question 51) of the nine variables related to accommodation. Despite the strong performance -0,25 of the PrS, respondents in this subsystem were the most likely to have used the emergency room or after-hours service, as these are the -0,20,0 0,2 0,4 0,6 0,8 1,0 usual source of care for part of this population. Component 1 The SSS was similar to the PrS on accessibility conditions and to the PubS on continuity of care, but it outperformed the other two Figure  1. Public subsystem: component loadings from categorical non- subsystems on comprehensiveness, access to health care in popula- linear principal components analysis on 19 primary health care performance tions with chronic conditions and unmet health care needs, though variables. Dimension Sub-dimension Shape Question number Access to health care Unmet health care needs 11 and 13 services Health care access in population with chronic conditions 7 and 8 Accommodation 29, 30, 32, 50 and 51 Continuity of care Continuity relational 17 and 18 Informational continuity 21 Person-and-community- Whole-person care 22, 23 oriented care Population orientation 28 Health services utilization Medical consultations 37 Comprehensiveness 35 Appropriateness of place and provider 16 and 19 Component 2 Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 255 Component loadings accessibility but weak on person-and-community-oriented care. The SSS combines the strengths of the other two subsystems and is par- ticularly strong in continuity of care and detection and treatment 1,00 29 30 of people with chronic conditions. In this section, we interpret the performance and the correlational structures between performance dimensions in light of each subsystem’s institutional and organiza- 0,75 tional arrangements. The pitfall of accessibility to the PubS is due to the lack of incen- tives to attract users and the insufficient supply of services to meet user demands (6,15–17). For instance, the PubS regulates (or rations) 0,50 the demand for services by requiring users to present requests for services in person rather than offering the possibility of making tel- 21 ephone appointments. This situation makes it difficult for certain 0,25 populations to achieve a continuous affiliation and linkage to this subsystem (e.g. working age males, sick or disabled people) and may explain the low levels of detection and treatment of individuals with chronic conditions. The positive correlation between the CAPTCA 0,00 8 components suggests that improvements in overall accessibility and in the identification and ongoing care of people with chronic con - ditions would have a positive effect on the overall direction and -0,2 0,0 0,2 0,4 0,6 0,8 1,0 continuity of care and would lower emergency care usage, in turn resulting in more integrated use of the health system. Component 1 Although timely accessibility is problematic in the PubS, the Figure  2. Social security subsystem: component loadings from categorical annual consultation rate and unmet health care needs are not sta- non-linear principal components analysis on 19 primary health care tistically significantly different from the other subsystems. This con - performance variables. For explanations on symbols, please see Figure 1. trasts with other studies that show higher use rates in the PrS (2). Our findings may be not generalizable to other settings as Rosario’s PubS has been the focus of public policy and has experienced an Component loadings extraordinary level of service development and expansion with free health care delivery. Though this is not an efficiency study, it should 1,00 be noted that this level of heath service utilization is reached by the PubS, with significantly lower expenditure than the SSS or the PrS. In contrast to other studies, we found that the PubS outper- 0,75 forms the PrS in health care orientation and continuity of care (6). Again, this likely reflects Rosario’s investment in multidiscipli - nary teams with a strong orientation to PHC principles and values 0,50 (18). We anticipate that the PubS in other contexts, which have not benefited from orientation or investment similar to those in Rosario, will have even more problematic accessibility and fewer 0,25 achievements in person-and-community-oriented care or health service utilization. 11 37 Our study confirms the findings from several studies that the 0,00 strength of the PrS comes from its comparatively more agile appoint- ment processes (2,6,19). To this, Rosario adds the provision of home-based care services for urgent care, and the wide supply and -0,25 relatively unrestricted access to specialized care. These features are reinforced by the PrS’s market strategy to capture low-income cli- -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 ents who are unsatisfied with the PubS’s appointment conditions and middle- and high-income clients who are reluctant to use regulated Component 1 SSS services. This strategy leads to two types of populations within Figure  3. Private subsystem: component loadings from categorical non- the PrS: one made up of individuals from a middle-to-low socio-eco- linear principal components analysis on 19 primary health care performance nomic level who are covered by a reduced basket of essential services variables. For explanations on symbols, please see Figure 1. and/or emergency services (34.6% in our sample) and another com- posed of individuals of a middle-to-high socio-economic level who Discussion are covered by a comprehensive insurance (65.4% in our sample). In this study of a geographically bounded population, we found This explains why the PrS had a higher percentage of people with cross-coverage by the different subsystems in a high proportion of low-quality living conditions than the SSS in our sample. respondents. When analysing respondents’ experiences with their Our data show that a high proportion of middle- and low-income affiliated subsystem or usual source of care, the performance of populations pays for private coverage in order to have easier access the PubS is weak on accessibility but strong on person-and-com- to urgent and emergency care. Though duplicate or supplementary munity-oriented care, the opposite of the PrS, which is strong on coverage provides additional access options for affiliates of the PubS Component 2 Component 2 Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 256 Family Practice, 2016, Vol. 33, No. 3 Table 3. Differences in variables that measure the primary health care dimensions and sub-dimensions among the user populations of different health subsystems Variables, by primary health care dimensions and sub-dimensions Public subsystem Social security subsystem Private subsystem Significance level n % n % n % Access to health care services Unmet health care Required health care during the last 12 months 263 83.5 311 86.6 106 83.5 0.466 needs Sought out health care each time he/she felt a health care need during the 239 91.2 294 94.5 100 94.3 0.209 last 12 months (Q11) Abandoned health care seeking during the last 12 months (Q13) 19 7.3 14 4.5 5 4.8 0.327 Access to health care Has been diagnosed with a chronic condition (Q7) 64 20.3 117 32.6 35 27.6 0.002 in populations with chronic conditions Currently receiving treatment for chronic conditions (Q8) 49 76.9 111 94.9 30 86.1 0.002 Accommodation/ Level of comfort with appointment Not at all or only somewhat com- 97 32.8 47 13.5 10 8.1 0.000 accessibility system of his/her USC during the fortable last 12 months (Q29) Moderately comfortable 71 24.0 56 16.1 10 8.1 Very or completely comfortable 128 43.2 244 70.3 103 83.7 Possibility of obtaining appoint- Not at all or only somewhat possible 210 77.5 48 14.1 7 5.8 0.000 ment in his/her USC by phone dur- ing the last 12 months (Q30) Moderately possible 18 6.6 36 10.6 12 10.0 Very or completely possible 43 15.9 257 75.4 101 84.2 Ease of obtaining an appointment Not at all or only somewhat easy 54 20.1 33 11.4 18 15.1 0.000 for an acute problem in his/her USC during the last 12 months (Q32) Moderately easy 105 39.0 100 34.6 31 26.1 Very or completely easy 110 40.9 156 54.0 70 58.8 Ease of obtaining an appointment Not at all or only somewhat easy 34 12.8 26 8.5 7 7.3 0.000 for a routine check-up in his/her USC during the last 12 months Moderately easy 103 38.7 91 29.6 26 27.1 Very or completely easy 129 48.5 190 61.9 63 65.6 Number of days between the >week 30 27.3 25 22.3 4 8.9 0.042 request for an acute problem and the visit, considering the last visit in his/her USMC 2–7 days 35 31.8 44 39.3 13 28.9 Next day or same day 45 40.9 43 38.4 28 62.2 Time invested to obtain the last >1 hour 19 10.4 9 3.7 3 3.9 0.000 visit appointment in his/her USC among those who had to request one in person (Q50) 30 minutes to 1 hour 39 21.4 26 10.6 6 7.8 <30 minutes or by phone 124 68.1 211 85.8 68 88.3 Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 257 Table 3. Continued Variables, by primary health care dimensions and sub-dimensions Public subsystem Social security subsystem Private subsystem Significance level n % n % n % Time spent in the waiting room ≥1 hour 55 21.2 40 12.9 14 14.0 0.014 prior to receiving health care during the last visit at his/her USC (Q51) 30 minutes < 1 hour 64 24.6 56 18.0 22 22.0 <30 minutes 141 54.2 215 69.1 64 64.0 Number of variables for which performance is high (over the mean) within those variables with 0/9 8/9 9/9 9 significant differences Continuity of care Relational continuity Currently has a USC, a place that she/he regularly or frequently uses each 303 96.2 352 98.1 118 92.9 0.024 time that she/he needs medical care (affiliation) Has been seeking care at the USC for a year or more (Q17) 271 88.3 321 90.7 109 86.5 0.155 Is currently a regular patient of a particular physician at the USC (Q18) 195 63.5 230 65.0 61 48.4 0.003 Informational How complete is her/his medical Not at all or only somewhat com- 15 6.3 19 6.4 12 13.8 0.004 continuity based on last record at the USC (Q21) plete 12-month experience Moderately complete 33 13.9 54 18.3 10 11.5 Very or totally complete 190 79.8 222 75.3 65 74.7 Number of variables for which performance is high (over the mean) within those variables with 3/3 2/3 0/3 3 significant differences Person-and-community-oriented care Whole-person How well do the staff know her/ Not at all or only somewhat well 142 46.3 180 50.8 72 57.1 0.013 care based on last him at the USC (Q22) 12 months experience Moderately well 59 19.2 81 22.9 26 20.6 Very or completely well 106 34.5 93 26.3 28 22.2 How possible would it be for the Not at all or only somewhat possible 187 60.9 260 73.4 99 78.6 0.000 USC to contact her/him to go to her/his appointments (Q23) Moderately possible 53 17.3 57 16.1 18 14.3 Very or completely possible 67 21.8 37 10.5 9 7.1 Population orienta- Which phrase best describes your Visits or visits and patients 171 55.7 274 77.4 101 80.2 0.000 tion based on last USC: it takes care of… 12 months experience Patients and their families or patients 136 44.3 80 22.6 25 19.8 and their families and communities Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 258 Family Practice, 2016, Vol. 33, No. 3 Table 3. Continued Variables, by primary health care dimensions and sub-dimensions Public subsystem Social security subsystem Private subsystem Significance level n % n % n % Participated in a community activity organized by the USC during the last 25 8.1 4 1.1 2 1.6 0.000 12 moths How active the USC has been in Not at all or only somewhat active 382 69.0 539 77.2 195 79.3 0.000 seeking out solutions to community problems (Q28) Moderately active 27 4.9 28 4.0 15 6.1 Very or completely active 145 26.2 131 18.8 36 14.6 Number of variables for which performance is high (over the mean) within those variables with 5/5 0/5 0/5 5 significant differences Health services utilization Annual consultation rate for population aged 18–64 years (Q37) 3.1 2.9 3.1 0.528 Comprehensiveness Health services used during the last General physician or specialist 146 55.9 115 36.5 47 44.8 0.000 12 months (Q35) General physician + specialist 69 26.4 108 34.3 28 26.7 At least general physician + specialist 46 17.6 92 29.2 30 28.6 + dentist Appropriateness of Population whose USC is an emergency or after-hour service (Q16) 13 4.2 29 8.2 29 23.4 0.000 place and provider Professional profile of her/his regu- General physician 183 93.3 191 83.0 53 86.9 0.006 lar physician at her/his USC (Q19) Specialist 13 6.7 39 17.0 8 13.1 Number of variables for which performance is high (over the mean) within those variables with 2/3 2/3 1/3 3 significant differences USC, usual source of care; USMC, usual source of medical care. Dimension definitions: • Access to health care services: the ease with which the population is able to use appropriate services in proportion to their needs. • Continuity of care: the degree to which a series of discrete health care events are experienced as coherent, connected and consistent with the patient’s medical needs and personal context. • Person-and-community-oriented care: the extent to which primary care clinicians consider and respond to their patients’ physical, emotional and social needs considering their familiar and community context. • Health services utilization: measures include type and degree of services used. The number of questions is to identify only those variables included in the CATCPA. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 259 and SSS, it introduces more fragmentation into their health care pro- Conclusions cess and more segmentation and financial inequity into the health We used a common set of tools in a geographically bounded popula- system. It also increases health system costs and inefficiencies ( 3,4), tion to compare health services performance in different subsystems since most SSS and PubS users covered by the PrS do not use their in the Rosario health system. This allowed us to determine the per- private insurance on a regular basis. centage of cross-coverage between respondents, which is likely to be The emphasis of the PrS system on ease of access that is not sup- similar to other Argentinean settings; measure performance based ported by strong continuity of care or a person-and-community ori- on the experience of the subsystems’ regular users; and make direct entation disrupts the virtuous synergy between access and continuity comparisons regarding the performance of these health subsystems. of care and may compromise quality of care. The lack of mecha- We found a high percentage of cross-coverage between sub- nisms to regulate and monitor service use and to coordinate health systems, which is a marker of system fragmentation and seg- care processes is associated with overprescription and compromised mentation and which may also indicate financial inequity and quality of technical care (6). The absence of health providers who inefficiency. We found different performance patterns in each sub - take responsibility for and orient patients’ care processes fosters the system, consistent with the subsystems’ institutional and organi- observed pattern of consultations oriented toward acute care, with zational profiles. Our findings allowed us to refute the claim that the highest use of emergency and specialist services, which generate the PrS services perform better than those of the PubS or the higher costs and inefficiencies in the system. SSS in any scenario. Our findings also suggest that the PrS is not The SSS’s strong performance in accessibility, continuity of more efficient than the PubS or the SSS, at least in the case being care and detection and treatment of people with chronic condi- studied. tions relates to a combination of market and regulation mecha- We showed that in Rosario, as in many Latin American coun- nisms. SSS entities purchase services using contractual mechanisms tries, the private subsystem does not exclusively serve the high- including economic incentives to promote timely access. Through income population, as often claimed (20), but that it is embedded in auditing and gate-keeping functions, they promote continuity and the middle- and low-income populations. appropriate use of services. The SSS’s even and strong performance This study provides evidence on the performance of the PubS and is consistent with SSS entities’ use of different management strate- the SSS, which is limited and poor quality in low- and middle-income gies to improve access and quality of care. Despite these strengths, countries, and on the performance of the PrS, which is scarce in all the SSS and PrS need to develop care that is more person-and- countries and particularly in middle- and low-income countries. It community oriented. contributes to understanding the dynamics between subsystems, par- Lessons learned from the evidence indicate that the performance ticularly through the determination and analysis of cross-coverage of the PrS could be improved by introducing incentives and/or new between subsystems. Based on our findings, we suggested potential contractual modalities aimed at improving the continuity of care areas of focus for quality improvement. and increasing the efficiency of service use through the enhancement Further research on the relationship between health systems’ of its monitoring and information systems. Both the SSS and PrS functions, performance and health outcomes in middle- and low- would benefit from adopting a PHC focus, especially to improve income countries with fragmented health systems is necessary to bet- person-and-community-oriented care. PubS accessibility condi- ter understand their relationships and effects in different contexts. tions could be improved by the implementation of better appoint- ment systems, including the option to schedule appointments by phone. Monitoring cross-coverage could help to track the flows of Acknowledgements resources and patients between subsystems, justify and inform the We thank Milva Geri and Maria Eugenia Elorza for their contributions in pro- implementation of mechanisms to compensate public expenditures cessing and analysing the survey data. We acknowledge the invaluable inputs in patients covered by the other subsystems, and develop mecha- of the reviewers and Dr Angela Bayer, which led to an improved version of nisms to reduce the fragmentation of the health care process and the the original manuscript. We thank the International Development Research segmentation of the system. Centre (IDRC), the National Scientific and Technical Research Council The major strength of this study is that it compares subsys- (Consejo Nacional de Investigaciones Científicas y Técnicas, CONICET) and tem performance using common indicators based on respondents’ the Program of Transdisciplinary Understanding and Training on Research— predominant subsystem affiliation. The level of cross-coverage Primary Health Care (TUTOR-PHC), whose support was essential to conduct and communicate these research findings. between subsystems demonstrates the importance of basing per- formance on the subsystem affiliation rather than the client or enrolment list from each subsystem. However, there are important Declaration limitations. The indicators provide very little information on the technical quality or appropriateness of care. The judgement of Funding: International Development Research Centre (IDRC, Project No stronger or weaker performance is determined by judgement about 103998) and the National Agency for Scientific and Technologic Promotion (ANPCyT, Project PICT No 1925/07). what is considered ‘better’ based on statistically significant dif - Ethical approval: IDRC and ANPCyT. ferences and not on a widely recognized benchmark. Finally, the Conflict of interest: none. choice of Rosario as a case limits the generalizability of the find - ings to other low- and middle-income contexts in Latin America, including within Argentina. Rosario’s publicly funded system is an References exemplar of a PHC-based and community-oriented system that has 1. Murray CJ, Frenk J. A framework for assessing the performance of health been the focus of public investment and social policy for decades. systems. Bull World Health Organ 2000; 78: 717–31. Although this case is not generalizable to other settings, the find - 2. Basu S, Andrews J, Kishore S, Panjabi R, Stuckler D. Comparative perfor- ings about system weaknesses and how they might be reinforced mance of private and public healthcare systems in low- and middle-income countries: a systematic review. PLoS Med 2012; 9: e1001244. are transferable to other settings. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 260 Family Practice, 2016, Vol. 33, No. 3 3. Santos IS. Evidência sobre o mix público-privado em países com cobertura tema de Cuentas de Salud de la Organización Mundial de la Salud (SHA duplicada: agravamento das iniquidades e da segmentação em sistemas 2011). Rosario, Argentina: Centro de Estudios Interdisciplinarios, Univer- nacionais de saúde. Ciên Saúde Colet 2011; 16: 2743–52. sidad Nacional de Rosario, 2012. 4. Organización Panamericana de la Salud. Redes Integradas de Servicios de 11. Báscolo E, Yavich N. Gobernanza del desarrollo de la APS en Rosario, Salud. Conceptos, opciones de política y hoja de ruta para su implement- Argentina. Rev Salud Pública 2010; 12: 89–104. ación en las Américas. Washington, DC: Organización Panamericana de la 12. Yavich N, Báscolo E, Haggerty J. Construyendo un marco de evaluación Salud, 2010. de la atención primaria de la salud para Latinoamérica. Salud Pública Méx 5. Organización Panamericana de la Salud. Estrategia para el acceso uni- 2010; 52: 39–45. versal a la salud y la cobertura universal de salud; 66ª sesión del Comité 13. Haggerty J, Fortin M, Beaulieu M et al. At the interface of community and Regional de la OMS para las Américas. Washington, DC: OPS, 2014. healthcare systems: a longitudinal cohort study on evolving health and the 6. Berendes S, Heywood P, Oliver S, Garner P. Quality of private and public impact of primary healthcare from the patient’s perspective. BMC Health ambulatory health care in low and middle income countries: systematic Serv Res 2010; 10: 258. review of comparative studies. PLoS Med 2011; 8: e1000433. 14. Linting M, Melman JJ, Groenen PJ, van der Koojj AJ. Nonlinear princi- 7. World Health Organization. The World Health Report—Health Systems: pal components analysis: introduction and application. Psychol Methods Improving Performance. Geneva, Switzerland: World Health Organiza- 2007; 12: 336. tion, 2000. 15. Vergara C. El contexto de las reformas del sector de la salud. Rev Panam 8. Organización Panamericana de la Salud/Organización Mundial de la Salud Salud Pública 2000; 8: 7–12. (OPS/OMS). Perfil de la Ciudad de Rosario en Salud Urbana. Provincia de 16. Báscolo E. Gobernanza y economía política de las políticas de APS en Santa Fe. República Argentina. Reunión de consulta técnica sobre salud América Latina. Ciên Saúde Colet 2011; 16: 2763–72. urbana “Desafíos de la Salud Urbana en las Américas: análisis del impacto 17. Lewis M, Eskeland G, Traa-Valerezo X. Primary health care in practice: is de la urbanización en la salud y en sus determinants,” Unidad de Entornos it effective? Health Policy 2004; 70: 303–25. Saludables del Área de Desarrollo Sostenible y Salud Ambiental. Santiago, 18. Lamarche P, Beaulieu M, Pineault R et al. Choices for Change: The Path Chile: OPS/OMS, 2005. for Restructuring Primary Healthcare Services in Canada. Ottawa, Can- 9. World Health Organization (WHO). System of Health Accounts (SHA 2011). ada: Canadian Health Services Research Foundation, 2003. http://www.who.int/nha/sha_revision/en/ (accessed on 2 October 2015). 19. Colombo F, Tapay N. Private Health Insurance in OECD Countries: 10. Báscolo E; Universidad Nacional de Rosario, Centro de Estudios Interdis- The Benefits and Costs of Individuals and Health Systems . Paris, France: ciplinarios, Maestría en Gestión de Sistemas y Servicios de Salud, Red de Organization for Economic Co-operation and Development, 2004. las Américas de Cuentas de Salud en América Latina y el Caribe. Cálculo 20. Cotlear D, Gómez-Dantés O, Knaul F et al. Overcoming social segregation del gasto en salud de la ciudad de Rosario mediante la aplicación del Sis- in health care in Latin America. Lancet 2015; 385: 1248–59. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Family Practice Oxford University Press

Comparing the performance of the public, social security and private health subsystems in Argentina by core dimensions of primary health care

Download PDF
12 pages

Loading next page...
 
/lp/oxford-university-press/comparing-the-performance-of-the-public-social-security-and-private-NQrz3XvDk5

References (28)

Publisher
Oxford University Press
Copyright
Copyright © 2022 Oxford University Press
ISSN
0263-2136
eISSN
1460-2229
DOI
10.1093/fampra/cmw043
pmid
27377651
Publisher site
See Article on Publisher Site

Abstract

Background. Most Latin American health systems are comprised of public (PubS), social security (SSS) and private (PrS) subsystems. These subsystems coexist, causing health care fragmentation and population segmentation. Objective. To estimate the extent of subsystem cross-coverage in a geographically bounded population (Rosario city) and to compare the subsystems’ performance on primary health care (PHC) dimensions. Methods. Through a cross-sectional, interviewer-administered survey to a representative sample (n = 822) of the Rosario population, we measured the percentage of cross-coverage (people with usual source of care in one subsystem but also covered by another subsystem) and the health services’ performance by core PHC dimensions, as reported by each subsystem’s usual users. We compared the subsystems’ performance using chi-square analysis and one-way analysis of variance testing. We analyzed whether the observed differences were coherent with the predominant institutional and organizational features of each subsystem. Results. Overall, 39.3% of the population was affiliated with the PubS, 44.8% with the SSS and 15.9% with the PrS. Cross-coverage was reported by 40.6% of respondents. The performance of the PubS was weak on accessibility but strong on person-and-community-oriented care, the opposite of the PrS. The SSS combined the strengths of the other two subsystems. Conclusion. Rosario’s health system has a high percentage of cross-coverage, contributing to issues of fragmentation, segmentation, financial inequity and inefficiency. The overall performance of the SSS was better than that of the PrS and PubS, though each subsystem had a particular performance pattern with areas of strength and weakness that were consistent with their institutional and organizational profiles. Key words. Health Care Quality, Access, and Evaluation, health services research, integrated health care systems, Latin America, primary health care, public health. © The Author 2016. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 250 Family Practice, 2016, Vol. 33, No. 3 to complement or avoid the limitations of the subsystem with which Introduction they were originally affiliated. Most Latin American ‘health systems’ are composed of public (PubS), The objective of this study is to estimate the extent of subsystem social security (SSS) and private (PrS) subsystems. Each has its own cross-coverage in a population in a given geographic area and to philosophical underpinning, funding base, management framework compare the subsystems with respect to their performance on PHC and intended catchment population. Though the performance of core dimensions as reported by the affiliated population. We use the their respective health services is expected to vary depending on the municipality of Rosario, Argentina, as the study case. population served and each subsystem’s organizational and insti- tutional features (1), an oversimplified portrait persists in policy documents and some of the peer-reviewed literature. This portrait Rosario health system characterizes the private subsystem as more efficient, accountable Rosario is the third most populous city in Argentina (948 312 inhab- and quality driven than the public health subsystem, in turn biasing itants). Argentina is a federal democracy, with 23 provinces and the public opinion and policy debate (2). Autonomous City of Buenos Aires. Funding and delivery of public This oversimplified portrait masks the differences within each services are the authority and responsibility of each jurisdiction. The of the subsystems and the fragmented nature of the available data regulatory frameworks share common features across jurisdictions, does not permit a direct comparison between subsystems. In fact, but health services delivery and organization vary. Depending on the so-called private subsystem is actually a collection of different fiscal capacity, funding and delivery has been further decentralized governance structures and delivery systems of varying degrees of to the municipal level. Like many Latin American countries and complexity. The SSS can also include multiple governance organi- jurisdictions, Rosario’s population is covered by segmented public, zations and delivery institutions. Even the PubS can have differ- social security and private systems. It shares most of the institutional ent levels of municipal, state or national government health care and organizational features of the Latin American segmented health schemes. Furthermore, the population can be covered by different systems, synthesized in Table  1. Below, we describe Rosario’s sub- subsystems, with the PubS serving as the ultimate safety net for all. systems, including features that are distinct from the typical Latin Cross-coverage of the population, interdependence between subsys- American case. tems and the changes that have occurred in the Latin American pri- vate health market during recent decades require a more nuanced Rosario’s public subsystem analysis (3). Rosario is one of the country’s richest municipalities, although Across Latin America, countries are making efforts to move ~115 000 of its residents live in slums (8). Since the 1990s, municipal toward the development of integrated, primary health care (PHC)- authorities have invested substantially to develop a publicly funded based health systems to improve access, efficiency, equity and qual - local health system (PubS) that currently provides a wide range of ity for the whole population (4,5). Our purpose is to compare the free health care services, aimed mainly at the population that is not performance of each subsystem using a common set of tools and covered by other systems—approximately a third of the city’s popu- to understand the degree of overlap between subsystems in order lation. Rosario’s PubS is a vertically integrated system of ambulatory to help governments to focus strategies to improve governance and health centres (with multidisciplinary teams), hospitals, an ambula- delivery (6). tory medical specialities centre, a rehabilitation centre and an emer- This study is guided by the World Health Organization (WHO) gency system. Resources are assigned to the system based on public framework for health system performance (1,7). According to this budgets. Per capita health spending in the Rosario PubS was $1907 framework, health systems are characterized by key functions such Argentinean pesos [$479 US dollar (USD), according to the official as stewardship, financing and resource generation, which translate exchange rate at 12 December 2010] in 2010 (9,10). into service provision to the population. High-performing services Furthermore, health professionals have intentionally developed have instrumental goals of providing high quality, responsive and Rosario as a model of participatory management with a multidis- affordable care. This care is the means to achieving the fundamental ciplinary first level of care that has a strong role in coordinating goals of the health system: to improve the average health status of the system. Rosario has been recognized in Latin America as a PHC the population (including reducing health inequalities); to be respon- champion and an example of a PHC-based health system because of sive to the non-medical expectations of the population (respect for the development and achievements of its PubS (11). people and a client orientation); and to ensure fairness in financial contributions. PHC services contribute to these goals through their Rosario’s social security subsystem instrumental goals of timely access, comprehensiveness, continuity of care, person-and-community orientation and technical quality The SSS is an employment-based insurance, with financing based on of care. a percentage of workers’ salaries and not on premiums that can be The premise of this study is that, in a context of multiple health adjusted. The Argentinean Ministry of Health regulatory framework subsystems, differences in the key functions of each system will requires that SSS insurers provide an obligatory broad basket of ser- translate into differences in the extent to which PHC services achieve vices to workers and their families. The Rosario SSS subsystem is their instrumental goals in core PHC dimensions. Our research made up of around 50 of the nearly 300 entities functioning around questions are two. Given the fragmented nature of the Argentinean the country. The entities vary in scope and size and have different health system, what is the extent of cross-coverage between subsys- models of care and management, all of which together attempt to tems? How well is each subsystem performing on core PHC dimen- provide a broad package of services within the limited financing avail - sion, as reported by their regular users? We hypothesize that (i) the able from worker contributions. Unlike other Latin American SSS emerging patterns of strengths and weaknesses in the performance entities, in Argentina and Rosario few entities have their own health of each subsystem will be consistent with its funding, governance, services. Instead, they purchase health services through per capita and organizational particularities and (ii) there will be a high rate of and/or fee-for-services contracts with private providers through selec- cross-coverage largely due to the strategies used by the population tive commissioning. They use different strategies to control, monitor Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 251 Table 1. Typical institutional and organizational features of Latin American health subsystems Latin American subsystems Public Social security Private Funding General taxes Employers’ and employees’ payroll Clients’ voluntary contributions deductions (prepaid insurance and direct payment) Population served Population without contribution Workers in the formal labour market High-, middle- and low-income capacity populations Health services coverage Basket of basic free services. Other Broader basket of services compared For high-income population, a broad services may require out-of-pocket to the PubS, but still requires basket of services is offered at a payments or may not be offered copayments high cost For low- and middle-income popula- tion (uninsured), complementary or supplementary plans are offered at a relatively high cost Authorities/management National, subnational and local Single or multiple social security Multiple for-profit companies and ministries/secretaries organizations non-profit organizations Integration of funding and services Vertical integration: owned health Vertical integration: owned health Contractual integration: purchase provision services services health services through per capita and/or fee-for-services contacts Main professionals’ remuneration Salaries Salaries Fee-for-service system and improve health care quality such as economic incentives, care of the municipality of Rosario. The initial sampling unit was the guidelines and the development of information and auditing systems census tract stratified by the proportion of households below the (12). Per capita health spending in the SSS ($637 USD) is 33% greater poverty threshold. Once the census tract was selected, all residential than in the PubS, but 37% lower than for private insurers (10). dwellings in the city blocks that were located within the perimeter of the census tract were contacted. In active households, all mem- bers were enumerated and one respondent was randomly selected Rosario’s private subsystem until the required sample size in each age strata was obtained The PrS comprises over 30 companies. There are three types of enti- (age < 15 years, 15–49 years and 50+ years). The final sample of 822 ties: voluntary prepaid insurance companies that offer a variety of residents was designed to represent the age–sex demographic strata, premiums; emergency services companies that provide emergency with a 95% margin of error of 3.5%. The survey was conducted care and urgent home care; and social services companies that offer between December 2010 and January 2011. a basic basket of essential ambulatory services through cheap premi- We assigned respondents to a particular subsystem based on their ums. Like the SSS, PrS entities contract service delivery from private affiliation to a specific facility that they reported as their regular or providers and professional associations. Clients choose their own most frequently used place of medical care (from now on, referred to providers from the range of entities and professionals offered by as usual source of care). As some facilities might provide health care their insurer. Services are provided in free-standing health practi- for people affiliated with different subsystems, the respondents were tioner offices, independent clinics and/or clinics owned by insurance asked to indicate the type of coverage used to receive medical care companies. Clinics tend to be organized by discipline, so multi- at that facility. We did not assign respondents based on their health disciplinary work is uncommon. Most health professionals work coverage status because it is likely to find people who are affiliated under an entrepreneur model and with fee-for-service remuneration with one subsystem and also covered by another subsystem (e.g. cov- depending on the number and type of medical services. In contrast to ered by the PrS and the SSS but affiliated with the SSS). the SSS, the PrS has a more passive role in controlling and coordinat- Those without a usual source of care (n = 14) were assigned to ing health services, with minimal mechanisms to regulate clients’ use the subsystem where they had their last visit. We excluded 21 people of services and health professionals’ practices (12). Per capita health who reported affiliation to either non-governmental health organiza - spending in the PrS ($873 UDS) is 82% greater than in the PubS and tions or traditional medicine, since these were beyond the scope of 37% higher than in the SSS (10). this study. In this study, we selected a representative sample of the popula- tion of Rosario to assess the subsystem to which individuals are pre- dominantly affiliated and reported each subsystem’s performance on Survey instrument core dimensions of PHC as reported by each subsystem’s usual users. The questionnaire included 67 questions related to seven sections: We then compared the performance of each of the subsystems to dis- (i) demographic and socio-economic conditions of the respondent cern whether the observed differences are consistent with the pre- and his/her family, with questions on health insurance, employment, dominant institutional and organizational features of the subsystems. literacy and housing; (ii) access to health care within populations with chronic conditions, where respondents were asked to indicate whether they had been diagnosed with any of a number of condi- Methods tions and if they were currently receiving treatment for those condi- Study design, sample and data collection tions; (iii) health care needs, with questions about their experiences We conducted a cross-sectional interviewer-administered survey in the last 12  months; (iv) continuity of care, where respondents using a multistage, stratified sampling strategy designed to obtain were asked to identify their usual source of care and respond to a a sample representative of the age/sex and socio-economic strata number of questions about relational and informational continuity Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 252 Family Practice, 2016, Vol. 33, No. 3 of care at the usual source of care, all based on their experience in Given that there is no consensus on performance standards to the last 12 months; (v) person-and-community-oriented care, again determine whether a specific benchmark was achieved or surpassed, with a focus on the usual source of care during the last 12 months; we described performance achievements in relative terms as per (vi) accommodation, which included two sets of questions related to Murray and Frenk (1). As in other studies, we used a combination of the respondents’ experience with their usual source of care, one set statistical significance and judgment to denote ‘weak’ or ‘strong’ per - based on their experience during the last 12 months and the other formance (2,6). For each variable that was statistically significantly based on their experience during the last consultation; and (vii) different by subsystem, we used our judgment to denote the ‘strong’ Health care use, including questions on health services utilization or ‘weak’ subsystem(s) relative to the other(s). For each PHC dimen- during the last 12 months and the last medical consultation. sion and sub-dimension, we identified as ‘strong’ the subsystem with The questions for respondents younger than 15 years of age were the strongest variables and as ‘weak’ the subsystem with the weakest answered by the adult with the greatest responsibility for their care. variables. The questions were mostly closed, with a mix of reporting and evalu- ative response options, and were taken as much as possible from Results a Canadian validated instrument (13) and other previous surveys conducted in Argentina (11,12). Demographics and socio-economic characteristics by health subsystems Analytic strategy Overall, 39.3% of the population was affiliated with the PubS, Analysis was conducted in two stages: (i) reduction of variables into 44.8% with the SSS and 15.9% with the PrS. We found significant underlying dimensions of PHC performance and (ii) comparison of differences among subsystems by participant sex and age and house- subsystems by variable and dimension. hold socio-economic conditions, including the head of household’s employment status and educational level and the overall household living conditions (Table 2). The respondents affiliated with the PubS Dimensions of primary health care performance subsystem were more likely to be younger, male, unemployed, less We reduced the large number of reported variables into broad educated and to report worse living conditions than respondents in dimensions of PHC using a categorical nonlinear principal compo- the other subsystems. For respondents in both the SSS and PrS sub- nents analysis (CATPCA) (14). CATPCA has the same objectives as systems, the respondent profile was older, more female and educated, traditional principal components analysis, but it allows for variables and among household heads, unemployment was lower and the with mixed measurement levels, transforming them into categorical percentage of retired workers was higher. The higher proportion of variables and creating linear combinations between the transformed older adults in the SSS is expected, given that it includes the national variables. insurance for seniors and pensioners. The relationship between the components and variable loadings Cross-coverage (usual care from one subsystem but also covered can be represented graphically through component loading plots by another subsystem) was reported by 40.6% of respondents. Half of the principal component space, where the axes are the principal of those affiliated with the SSS had complementary or supplemen - components. In these plots, variables with relatively long vectors fit tary PrS coverage. Over a third of PubS affiliates had SSS and/or PrS well into the solution, with the squared length of the loading vec- coverage and a quarter of PrS affiliates had SSS coverage. tor equalling the variance accounted for. When vectors are long, the cosines of the angles between the vectors indicate the Pearson’s cor- relation between the quantified variables. The slope of the vectors Relationship between components by system indicates the relationship of each variable with each component. An In all three subsystems, the first component combines the dimensions angle of 90° indicates no correlation between variables, 180° indi- we hypothesized as continuity of care, person-and-community-ori- cates negative correlation and 0° indicates positive correlation (14). ented care and health utilization while the second maps onto accessi- All analyses were conducted using SPSS version 21. bility. Health care access within populations with chronic conditions We hypothesized a priori which variables best represented each maps onto the first component in both the SSS and PrS and onto the dimension and sub-dimension. We used 19 variables: 12 ordinal, 6 second component in the PubS. Except for question 13 in the SSS, nominal and 1 continuous. Two components accounted for >30% of unmet health care needs maps onto the first component. the variance in each subsystem (PubS 31.8%; SSS 32.2%; PrS 31.0%), The CATPCA component loading plot showed a positive correla- with Cronbach’s alphas of ~0.90 (PubS 0.899; SSS 0.888; PrS 0.912). tion between components in the PubS (Fig. 1); a null correlation in the SSS (Fig. 2); and a negative correlation in the PrS (Fig. 3). Comparison of subsystems Subsystems were compared to look for significant differences in the Primary health care performance by health demographic and socio-economic characteristics of the respondents subsystems in each system, using chi-square analysis for categorical variables Subsystems showed significant differences in performance for all and one-way analysis of variance tests for continuous variables. We PHC dimensions and sub-dimensions, with the exception of unmet conducted a similar analysis for each of the variables for the fol- health care needs and medical consultation rates (Table 3). lowing PHC performance dimensions: access to health care services; The main strengths of the PubS were continuity of care, appro- continuity of care; person-and-community-oriented care and health priateness of place and provider and person-and-community-ori- services utilization. According to the WHO framework for health ented care. The performance of the PubS in continuity of care and system performance (1,7), high performance on these dimensions is a appropriateness of place and provider approximates the SSS, but it means to achieve the health system’s fundamental goals, particularly has a lower percentage of people whose usual source of care is an the goal of being responsive to the non-medical expectations of the emergency or after-hour service, a higher percentage of people whose population. regular provider is a general physician (instead of a specialist) and Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 253 Table 2. Study population demographics, socio-economic conditions and health insurance status by health subsystem, Rosario, 2011 (n = 801) Total Public Social security Private Significance level subsystem subsystem subsystem n % n % n % n % Demographics Population 801 100 315 39.3 359 44.8 127 15.9 – Sex Male 418 52.2 182 57.8 179 49.9 57 44.9 0.024 Age 0–12 years 155 19.4 80 25.4 58 16.2 17 13.4 0.000 13–20 years 120 15.0 68 21.6 37 10.3 15 11.8 21–60 years 358 44.7 149 47.3 145 40.4 64 50.4 61–99 years 168 21.0 18 5.7 119 33.1 31 24.4 Socio-economic conditions Household head, Employed 635 79.4 277 87.9 262 73.2 96 75.6 0.000 employment condition Unemployed 17 2.1 15 4.8 1 0.3 1 0.8 Retired 148 18.5 23 7.3 95 26.5 30 23.6 Household head, Never attended school or incomplete primary 131 16.4 51 16.2 57 15.9 23 18.1 0.002 educational level Complete primary 209 26.1 104 33.0 80 22.3 25 19.7 Incomplete secondary 138 17.3 62 19.7 54 15.1 22 17.3 Complete secondary 207 25.9 73 23.2 99 27.7 35 27.6 Incomplete or complete post-secondary or university 115 14.4 25 7.9 68 19.6 22 17.3 Type of housing Deficient 334 41.7 172 54.6 114 31.8 48 37.8 0.000 Health coverage Cross-coverage between No cross-coverage 426 53.2 201 63.8 164 45.7 61 48.0 – subsystems Cross-coverage 325 40.6 114 36.2 179 49.9 32 25.2 Lost insurance given by the subsystem of affiliation during the last year 50 6.2 – – 16 4.5 34 26.8 A house is considered deficient when it has at least one of the following characteristics: (i) floor—dirt or loose brick, (ii) walls—wood, metal or fibrocement sheets, (iii) walls without external plaster or coating, (iv) roof without internal coating or ceiling only or (v) has a water source outside of the home. The subsystem affiliation refers to the subsystem of the provider reported as the usual or most frequently used place of medical care. The term coverage includes two situations: people with insurance from the SSS or the PrS and people with a factual link to the PubS, as the PubS does not provide insurance to anyone and theoretically provides coverage to everyone. Shifted from SSS to PrS, vice versa or lost insurance. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 254 Family Practice, 2016, Vol. 33, No. 3 a higher percentage of people who are more likely to have very or for the last variable, there were no significant differences between totally complete medical records in their usual source of care. The subsystems. The performance of the SSS in person-and-community- performance of the PubS for person-and-community-oriented care is oriented care was better than that of the PrS but substantially lower far superior to that of the PrS and the SSS. than that of the PubS. Regarding appropriate use of place, the per- Accommodation, access to health care in populations with formance of the SSS was similar but inferior to that of the PubS. chronic conditions and comprehensive use of services are the major With respect to appropriate use of provider, the SSS had the high- performance weaknesses in the PubS. The PubS had the weakest per- est proportion of population whose regular provider is a specialist, formance on the accessibility dimension, with statistically significant probably due to the higher proportion of senior population and the differences on all nine indicators and respondents expressing particu- higher rate of chronic illness detection and treatment. lar difficulty with the appointment process. Compared to the SSS and PrS, the percentage of people diagnosed with chronic conditions in Component loadings the PubS is significantly lower, as is the percentage of people diag - nosed with at least one chronic condition who are currently receiving 1,00 treatment. It should be noted that in the PubS, health care access in the population with chronic conditions and accommodation are com- bined in component 2 (Fig.  1). Although there were not significant 0,75 differences between subsystems, the PubS had the worst performance on unmet health care needs. Finally, affiliates in the PubS were least likely to consult a wide range of services (comprehensive access). 0,50 The PrS had a performance profile that was the opposite of the 50 PubS. It had the weakest performance in appropriateness of place and provider, continuity and person-and-community-oriented care. 0,25 The performance of the PrS for access to health care in populations with chronic conditions was better than that of the PubS, but the rate of chronic illness detection and treatment were 5.0 and 8.8 points 18 0,00 lower, respectively, than those of the SSS. The performance of the 21 PrS on comprehensiveness was almost as good as the SSS, and it had 11 the strongest performance on all but one (question 51) of the nine variables related to accommodation. Despite the strong performance -0,25 of the PrS, respondents in this subsystem were the most likely to have used the emergency room or after-hours service, as these are the -0,20,0 0,2 0,4 0,6 0,8 1,0 usual source of care for part of this population. Component 1 The SSS was similar to the PrS on accessibility conditions and to the PubS on continuity of care, but it outperformed the other two Figure  1. Public subsystem: component loadings from categorical non- subsystems on comprehensiveness, access to health care in popula- linear principal components analysis on 19 primary health care performance tions with chronic conditions and unmet health care needs, though variables. Dimension Sub-dimension Shape Question number Access to health care Unmet health care needs 11 and 13 services Health care access in population with chronic conditions 7 and 8 Accommodation 29, 30, 32, 50 and 51 Continuity of care Continuity relational 17 and 18 Informational continuity 21 Person-and-community- Whole-person care 22, 23 oriented care Population orientation 28 Health services utilization Medical consultations 37 Comprehensiveness 35 Appropriateness of place and provider 16 and 19 Component 2 Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 255 Component loadings accessibility but weak on person-and-community-oriented care. The SSS combines the strengths of the other two subsystems and is par- ticularly strong in continuity of care and detection and treatment 1,00 29 30 of people with chronic conditions. In this section, we interpret the performance and the correlational structures between performance dimensions in light of each subsystem’s institutional and organiza- 0,75 tional arrangements. The pitfall of accessibility to the PubS is due to the lack of incen- tives to attract users and the insufficient supply of services to meet user demands (6,15–17). For instance, the PubS regulates (or rations) 0,50 the demand for services by requiring users to present requests for services in person rather than offering the possibility of making tel- 21 ephone appointments. This situation makes it difficult for certain 0,25 populations to achieve a continuous affiliation and linkage to this subsystem (e.g. working age males, sick or disabled people) and may explain the low levels of detection and treatment of individuals with chronic conditions. The positive correlation between the CAPTCA 0,00 8 components suggests that improvements in overall accessibility and in the identification and ongoing care of people with chronic con - ditions would have a positive effect on the overall direction and -0,2 0,0 0,2 0,4 0,6 0,8 1,0 continuity of care and would lower emergency care usage, in turn resulting in more integrated use of the health system. Component 1 Although timely accessibility is problematic in the PubS, the Figure  2. Social security subsystem: component loadings from categorical annual consultation rate and unmet health care needs are not sta- non-linear principal components analysis on 19 primary health care tistically significantly different from the other subsystems. This con - performance variables. For explanations on symbols, please see Figure 1. trasts with other studies that show higher use rates in the PrS (2). Our findings may be not generalizable to other settings as Rosario’s PubS has been the focus of public policy and has experienced an Component loadings extraordinary level of service development and expansion with free health care delivery. Though this is not an efficiency study, it should 1,00 be noted that this level of heath service utilization is reached by the PubS, with significantly lower expenditure than the SSS or the PrS. In contrast to other studies, we found that the PubS outper- 0,75 forms the PrS in health care orientation and continuity of care (6). Again, this likely reflects Rosario’s investment in multidiscipli - nary teams with a strong orientation to PHC principles and values 0,50 (18). We anticipate that the PubS in other contexts, which have not benefited from orientation or investment similar to those in Rosario, will have even more problematic accessibility and fewer 0,25 achievements in person-and-community-oriented care or health service utilization. 11 37 Our study confirms the findings from several studies that the 0,00 strength of the PrS comes from its comparatively more agile appoint- ment processes (2,6,19). To this, Rosario adds the provision of home-based care services for urgent care, and the wide supply and -0,25 relatively unrestricted access to specialized care. These features are reinforced by the PrS’s market strategy to capture low-income cli- -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 ents who are unsatisfied with the PubS’s appointment conditions and middle- and high-income clients who are reluctant to use regulated Component 1 SSS services. This strategy leads to two types of populations within Figure  3. Private subsystem: component loadings from categorical non- the PrS: one made up of individuals from a middle-to-low socio-eco- linear principal components analysis on 19 primary health care performance nomic level who are covered by a reduced basket of essential services variables. For explanations on symbols, please see Figure 1. and/or emergency services (34.6% in our sample) and another com- posed of individuals of a middle-to-high socio-economic level who Discussion are covered by a comprehensive insurance (65.4% in our sample). In this study of a geographically bounded population, we found This explains why the PrS had a higher percentage of people with cross-coverage by the different subsystems in a high proportion of low-quality living conditions than the SSS in our sample. respondents. When analysing respondents’ experiences with their Our data show that a high proportion of middle- and low-income affiliated subsystem or usual source of care, the performance of populations pays for private coverage in order to have easier access the PubS is weak on accessibility but strong on person-and-com- to urgent and emergency care. Though duplicate or supplementary munity-oriented care, the opposite of the PrS, which is strong on coverage provides additional access options for affiliates of the PubS Component 2 Component 2 Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 256 Family Practice, 2016, Vol. 33, No. 3 Table 3. Differences in variables that measure the primary health care dimensions and sub-dimensions among the user populations of different health subsystems Variables, by primary health care dimensions and sub-dimensions Public subsystem Social security subsystem Private subsystem Significance level n % n % n % Access to health care services Unmet health care Required health care during the last 12 months 263 83.5 311 86.6 106 83.5 0.466 needs Sought out health care each time he/she felt a health care need during the 239 91.2 294 94.5 100 94.3 0.209 last 12 months (Q11) Abandoned health care seeking during the last 12 months (Q13) 19 7.3 14 4.5 5 4.8 0.327 Access to health care Has been diagnosed with a chronic condition (Q7) 64 20.3 117 32.6 35 27.6 0.002 in populations with chronic conditions Currently receiving treatment for chronic conditions (Q8) 49 76.9 111 94.9 30 86.1 0.002 Accommodation/ Level of comfort with appointment Not at all or only somewhat com- 97 32.8 47 13.5 10 8.1 0.000 accessibility system of his/her USC during the fortable last 12 months (Q29) Moderately comfortable 71 24.0 56 16.1 10 8.1 Very or completely comfortable 128 43.2 244 70.3 103 83.7 Possibility of obtaining appoint- Not at all or only somewhat possible 210 77.5 48 14.1 7 5.8 0.000 ment in his/her USC by phone dur- ing the last 12 months (Q30) Moderately possible 18 6.6 36 10.6 12 10.0 Very or completely possible 43 15.9 257 75.4 101 84.2 Ease of obtaining an appointment Not at all or only somewhat easy 54 20.1 33 11.4 18 15.1 0.000 for an acute problem in his/her USC during the last 12 months (Q32) Moderately easy 105 39.0 100 34.6 31 26.1 Very or completely easy 110 40.9 156 54.0 70 58.8 Ease of obtaining an appointment Not at all or only somewhat easy 34 12.8 26 8.5 7 7.3 0.000 for a routine check-up in his/her USC during the last 12 months Moderately easy 103 38.7 91 29.6 26 27.1 Very or completely easy 129 48.5 190 61.9 63 65.6 Number of days between the >week 30 27.3 25 22.3 4 8.9 0.042 request for an acute problem and the visit, considering the last visit in his/her USMC 2–7 days 35 31.8 44 39.3 13 28.9 Next day or same day 45 40.9 43 38.4 28 62.2 Time invested to obtain the last >1 hour 19 10.4 9 3.7 3 3.9 0.000 visit appointment in his/her USC among those who had to request one in person (Q50) 30 minutes to 1 hour 39 21.4 26 10.6 6 7.8 <30 minutes or by phone 124 68.1 211 85.8 68 88.3 Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 257 Table 3. Continued Variables, by primary health care dimensions and sub-dimensions Public subsystem Social security subsystem Private subsystem Significance level n % n % n % Time spent in the waiting room ≥1 hour 55 21.2 40 12.9 14 14.0 0.014 prior to receiving health care during the last visit at his/her USC (Q51) 30 minutes < 1 hour 64 24.6 56 18.0 22 22.0 <30 minutes 141 54.2 215 69.1 64 64.0 Number of variables for which performance is high (over the mean) within those variables with 0/9 8/9 9/9 9 significant differences Continuity of care Relational continuity Currently has a USC, a place that she/he regularly or frequently uses each 303 96.2 352 98.1 118 92.9 0.024 time that she/he needs medical care (affiliation) Has been seeking care at the USC for a year or more (Q17) 271 88.3 321 90.7 109 86.5 0.155 Is currently a regular patient of a particular physician at the USC (Q18) 195 63.5 230 65.0 61 48.4 0.003 Informational How complete is her/his medical Not at all or only somewhat com- 15 6.3 19 6.4 12 13.8 0.004 continuity based on last record at the USC (Q21) plete 12-month experience Moderately complete 33 13.9 54 18.3 10 11.5 Very or totally complete 190 79.8 222 75.3 65 74.7 Number of variables for which performance is high (over the mean) within those variables with 3/3 2/3 0/3 3 significant differences Person-and-community-oriented care Whole-person How well do the staff know her/ Not at all or only somewhat well 142 46.3 180 50.8 72 57.1 0.013 care based on last him at the USC (Q22) 12 months experience Moderately well 59 19.2 81 22.9 26 20.6 Very or completely well 106 34.5 93 26.3 28 22.2 How possible would it be for the Not at all or only somewhat possible 187 60.9 260 73.4 99 78.6 0.000 USC to contact her/him to go to her/his appointments (Q23) Moderately possible 53 17.3 57 16.1 18 14.3 Very or completely possible 67 21.8 37 10.5 9 7.1 Population orienta- Which phrase best describes your Visits or visits and patients 171 55.7 274 77.4 101 80.2 0.000 tion based on last USC: it takes care of… 12 months experience Patients and their families or patients 136 44.3 80 22.6 25 19.8 and their families and communities Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 258 Family Practice, 2016, Vol. 33, No. 3 Table 3. Continued Variables, by primary health care dimensions and sub-dimensions Public subsystem Social security subsystem Private subsystem Significance level n % n % n % Participated in a community activity organized by the USC during the last 25 8.1 4 1.1 2 1.6 0.000 12 moths How active the USC has been in Not at all or only somewhat active 382 69.0 539 77.2 195 79.3 0.000 seeking out solutions to community problems (Q28) Moderately active 27 4.9 28 4.0 15 6.1 Very or completely active 145 26.2 131 18.8 36 14.6 Number of variables for which performance is high (over the mean) within those variables with 5/5 0/5 0/5 5 significant differences Health services utilization Annual consultation rate for population aged 18–64 years (Q37) 3.1 2.9 3.1 0.528 Comprehensiveness Health services used during the last General physician or specialist 146 55.9 115 36.5 47 44.8 0.000 12 months (Q35) General physician + specialist 69 26.4 108 34.3 28 26.7 At least general physician + specialist 46 17.6 92 29.2 30 28.6 + dentist Appropriateness of Population whose USC is an emergency or after-hour service (Q16) 13 4.2 29 8.2 29 23.4 0.000 place and provider Professional profile of her/his regu- General physician 183 93.3 191 83.0 53 86.9 0.006 lar physician at her/his USC (Q19) Specialist 13 6.7 39 17.0 8 13.1 Number of variables for which performance is high (over the mean) within those variables with 2/3 2/3 1/3 3 significant differences USC, usual source of care; USMC, usual source of medical care. Dimension definitions: • Access to health care services: the ease with which the population is able to use appropriate services in proportion to their needs. • Continuity of care: the degree to which a series of discrete health care events are experienced as coherent, connected and consistent with the patient’s medical needs and personal context. • Person-and-community-oriented care: the extent to which primary care clinicians consider and respond to their patients’ physical, emotional and social needs considering their familiar and community context. • Health services utilization: measures include type and degree of services used. The number of questions is to identify only those variables included in the CATCPA. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 Comparing primary health care performance across health subsystems 259 and SSS, it introduces more fragmentation into their health care pro- Conclusions cess and more segmentation and financial inequity into the health We used a common set of tools in a geographically bounded popula- system. It also increases health system costs and inefficiencies ( 3,4), tion to compare health services performance in different subsystems since most SSS and PubS users covered by the PrS do not use their in the Rosario health system. This allowed us to determine the per- private insurance on a regular basis. centage of cross-coverage between respondents, which is likely to be The emphasis of the PrS system on ease of access that is not sup- similar to other Argentinean settings; measure performance based ported by strong continuity of care or a person-and-community ori- on the experience of the subsystems’ regular users; and make direct entation disrupts the virtuous synergy between access and continuity comparisons regarding the performance of these health subsystems. of care and may compromise quality of care. The lack of mecha- We found a high percentage of cross-coverage between sub- nisms to regulate and monitor service use and to coordinate health systems, which is a marker of system fragmentation and seg- care processes is associated with overprescription and compromised mentation and which may also indicate financial inequity and quality of technical care (6). The absence of health providers who inefficiency. We found different performance patterns in each sub - take responsibility for and orient patients’ care processes fosters the system, consistent with the subsystems’ institutional and organi- observed pattern of consultations oriented toward acute care, with zational profiles. Our findings allowed us to refute the claim that the highest use of emergency and specialist services, which generate the PrS services perform better than those of the PubS or the higher costs and inefficiencies in the system. SSS in any scenario. Our findings also suggest that the PrS is not The SSS’s strong performance in accessibility, continuity of more efficient than the PubS or the SSS, at least in the case being care and detection and treatment of people with chronic condi- studied. tions relates to a combination of market and regulation mecha- We showed that in Rosario, as in many Latin American coun- nisms. SSS entities purchase services using contractual mechanisms tries, the private subsystem does not exclusively serve the high- including economic incentives to promote timely access. Through income population, as often claimed (20), but that it is embedded in auditing and gate-keeping functions, they promote continuity and the middle- and low-income populations. appropriate use of services. The SSS’s even and strong performance This study provides evidence on the performance of the PubS and is consistent with SSS entities’ use of different management strate- the SSS, which is limited and poor quality in low- and middle-income gies to improve access and quality of care. Despite these strengths, countries, and on the performance of the PrS, which is scarce in all the SSS and PrS need to develop care that is more person-and- countries and particularly in middle- and low-income countries. It community oriented. contributes to understanding the dynamics between subsystems, par- Lessons learned from the evidence indicate that the performance ticularly through the determination and analysis of cross-coverage of the PrS could be improved by introducing incentives and/or new between subsystems. Based on our findings, we suggested potential contractual modalities aimed at improving the continuity of care areas of focus for quality improvement. and increasing the efficiency of service use through the enhancement Further research on the relationship between health systems’ of its monitoring and information systems. Both the SSS and PrS functions, performance and health outcomes in middle- and low- would benefit from adopting a PHC focus, especially to improve income countries with fragmented health systems is necessary to bet- person-and-community-oriented care. PubS accessibility condi- ter understand their relationships and effects in different contexts. tions could be improved by the implementation of better appoint- ment systems, including the option to schedule appointments by phone. Monitoring cross-coverage could help to track the flows of Acknowledgements resources and patients between subsystems, justify and inform the We thank Milva Geri and Maria Eugenia Elorza for their contributions in pro- implementation of mechanisms to compensate public expenditures cessing and analysing the survey data. We acknowledge the invaluable inputs in patients covered by the other subsystems, and develop mecha- of the reviewers and Dr Angela Bayer, which led to an improved version of nisms to reduce the fragmentation of the health care process and the the original manuscript. We thank the International Development Research segmentation of the system. Centre (IDRC), the National Scientific and Technical Research Council The major strength of this study is that it compares subsys- (Consejo Nacional de Investigaciones Científicas y Técnicas, CONICET) and tem performance using common indicators based on respondents’ the Program of Transdisciplinary Understanding and Training on Research— predominant subsystem affiliation. The level of cross-coverage Primary Health Care (TUTOR-PHC), whose support was essential to conduct and communicate these research findings. between subsystems demonstrates the importance of basing per- formance on the subsystem affiliation rather than the client or enrolment list from each subsystem. However, there are important Declaration limitations. The indicators provide very little information on the technical quality or appropriateness of care. The judgement of Funding: International Development Research Centre (IDRC, Project No stronger or weaker performance is determined by judgement about 103998) and the National Agency for Scientific and Technologic Promotion (ANPCyT, Project PICT No 1925/07). what is considered ‘better’ based on statistically significant dif - Ethical approval: IDRC and ANPCyT. ferences and not on a widely recognized benchmark. Finally, the Conflict of interest: none. choice of Rosario as a case limits the generalizability of the find - ings to other low- and middle-income contexts in Latin America, including within Argentina. Rosario’s publicly funded system is an References exemplar of a PHC-based and community-oriented system that has 1. Murray CJ, Frenk J. A framework for assessing the performance of health been the focus of public investment and social policy for decades. systems. Bull World Health Organ 2000; 78: 717–31. Although this case is not generalizable to other settings, the find - 2. Basu S, Andrews J, Kishore S, Panjabi R, Stuckler D. Comparative perfor- ings about system weaknesses and how they might be reinforced mance of private and public healthcare systems in low- and middle-income countries: a systematic review. PLoS Med 2012; 9: e1001244. are transferable to other settings. Downloaded from https://academic.oup.com/fampra/article/33/3/249/1750001 by DeepDyve user on 19 July 2022 260 Family Practice, 2016, Vol. 33, No. 3 3. Santos IS. Evidência sobre o mix público-privado em países com cobertura tema de Cuentas de Salud de la Organización Mundial de la Salud (SHA duplicada: agravamento das iniquidades e da segmentação em sistemas 2011). Rosario, Argentina: Centro de Estudios Interdisciplinarios, Univer- nacionais de saúde. Ciên Saúde Colet 2011; 16: 2743–52. sidad Nacional de Rosario, 2012. 4. Organización Panamericana de la Salud. Redes Integradas de Servicios de 11. Báscolo E, Yavich N. Gobernanza del desarrollo de la APS en Rosario, Salud. Conceptos, opciones de política y hoja de ruta para su implement- Argentina. Rev Salud Pública 2010; 12: 89–104. ación en las Américas. Washington, DC: Organización Panamericana de la 12. Yavich N, Báscolo E, Haggerty J. Construyendo un marco de evaluación Salud, 2010. de la atención primaria de la salud para Latinoamérica. Salud Pública Méx 5. Organización Panamericana de la Salud. Estrategia para el acceso uni- 2010; 52: 39–45. versal a la salud y la cobertura universal de salud; 66ª sesión del Comité 13. Haggerty J, Fortin M, Beaulieu M et al. At the interface of community and Regional de la OMS para las Américas. Washington, DC: OPS, 2014. healthcare systems: a longitudinal cohort study on evolving health and the 6. Berendes S, Heywood P, Oliver S, Garner P. Quality of private and public impact of primary healthcare from the patient’s perspective. BMC Health ambulatory health care in low and middle income countries: systematic Serv Res 2010; 10: 258. review of comparative studies. PLoS Med 2011; 8: e1000433. 14. Linting M, Melman JJ, Groenen PJ, van der Koojj AJ. Nonlinear princi- 7. World Health Organization. The World Health Report—Health Systems: pal components analysis: introduction and application. Psychol Methods Improving Performance. Geneva, Switzerland: World Health Organiza- 2007; 12: 336. tion, 2000. 15. Vergara C. El contexto de las reformas del sector de la salud. Rev Panam 8. Organización Panamericana de la Salud/Organización Mundial de la Salud Salud Pública 2000; 8: 7–12. (OPS/OMS). Perfil de la Ciudad de Rosario en Salud Urbana. Provincia de 16. Báscolo E. Gobernanza y economía política de las políticas de APS en Santa Fe. República Argentina. Reunión de consulta técnica sobre salud América Latina. Ciên Saúde Colet 2011; 16: 2763–72. urbana “Desafíos de la Salud Urbana en las Américas: análisis del impacto 17. Lewis M, Eskeland G, Traa-Valerezo X. Primary health care in practice: is de la urbanización en la salud y en sus determinants,” Unidad de Entornos it effective? Health Policy 2004; 70: 303–25. Saludables del Área de Desarrollo Sostenible y Salud Ambiental. Santiago, 18. Lamarche P, Beaulieu M, Pineault R et al. Choices for Change: The Path Chile: OPS/OMS, 2005. for Restructuring Primary Healthcare Services in Canada. Ottawa, Can- 9. World Health Organization (WHO). System of Health Accounts (SHA 2011). ada: Canadian Health Services Research Foundation, 2003. http://www.who.int/nha/sha_revision/en/ (accessed on 2 October 2015). 19. Colombo F, Tapay N. Private Health Insurance in OECD Countries: 10. Báscolo E; Universidad Nacional de Rosario, Centro de Estudios Interdis- The Benefits and Costs of Individuals and Health Systems . Paris, France: ciplinarios, Maestría en Gestión de Sistemas y Servicios de Salud, Red de Organization for Economic Co-operation and Development, 2004. las Américas de Cuentas de Salud en América Latina y el Caribe. Cálculo 20. Cotlear D, Gómez-Dantés O, Knaul F et al. Overcoming social segregation del gasto en salud de la ciudad de Rosario mediante la aplicación del Sis- in health care in Latin America. Lancet 2015; 385: 1248–59.

Journal

Family Practice – Oxford University Press

Published: Jun 1, 2016

Keywords: primary health care; social security

There are no references for this article.