Evaluation of comprehensive care for older adults in primary care services

ABSTRACT OBJECTIVE To evaluate the performance of comprehensive care for older adults in primary care services in the Brazilian Unified Health System in the state of São Paulo, Brazil. METHODS A total of 157 primary care services from five health regions in midwestern São Paulo responded, from October to December 2014, the pre-validated 2014 questionnaire for primary care services assessment and monitoring. We selected 155 questions, based on national policies and guidelines on this theme. The responses indicate the service performance in older adults’ care, clustered into three areas of analysis: health care for active and healthy aging (45 indicators, d1), chronic noncommunicable diseases care (89 indicators, d2), and support network in aging care (21 indicators, d3). Performance was measured by the sum of positive (value 1) or negative (value 0) responses for each indicator. Services were clustered according to k-means of the performance scores of each domain. After weighting the domains (Z tests), we estimated the associations between the scores of each domain and independent management variables (typology, planning and evaluation of services), with simple and multiple linear regression. RESULTS Chronic noncommunicable diseases care (d2) showed, for all clusters, better average performance (55.7) than domains d1 (35.4) and d3 (39.2). Service performance in the general area of planning and evaluation associates with the performance of older adults’ care. CONCLUSIONS The evaluated services had incipient implementation of comprehensive care for older adults. The evaluation framework can contribute to processes to improve the quality of primary health care.


INTRODUCTION
Primary health care (PHC) is pointed out as a priority to assist and monitor older adults' health status, besides acting in the prevention of health problems and in the health promotion to a healthy aging. Accordingly, the World Health Organization (WHO) proposed in 2004 that primary care services should adapt to serve older adults properly 1 . In Brazil, in view of the population aging, the Ministry of Health published in 2007 the Caderno de Atenção Básica (Primary Care Guide) no. 19, guiding primary care teams for a greater resolution of the older adults' demands 2 .
The expansion of the PHC public network and the need to strengthen management structures and to redefine roles in the three spheres of government have driven the institutionalization of agreement and evaluation mechanisms to assist the decision-making process and to improve the public health system 3,4 . Especially since the last decade, evaluation and monitoring strategies are essential tools to measure the effectiveness of health systems at all levels of care 4 .
In this context, evaluating services for the older adults' health care, especially in PHC, is increasingly important to know the advances in implementing the recommendations of the Primary Care Guide no. 19 2 . They cover promotion and prevention actions to achieve active and healthy aging, management of chronic noncommunicable diseases (CNCD) and networking.
National and international studies conducted to evaluate the older adults' care in PHC indicate the need to expand preventive practices and health promotion, infrequent and low diverse in actions, in addition to the importance to build networks of comprehensive care to older adults' health, ordained by PHC services [5][6][7][8][9][10] . The QualiAB 2014 instrument (Questionnaire for Primary Care Services Assessment and Monitoring) proposes, among its indicators for the comprehensive evaluation of primary care services, those directed to older adults' care 11 .
Despite the political and technological propositions available in the country for PHC performance in this segment, literature on the actions effectively implemented evaluating their organization and supply is still scarce 5 , as well as on the challenges generated by population aging for these services [5][6][7] . Most studies evaluate the actions developed in PHC according to the older adults' perception of the care they received, and those who take the organization of actions in a comprehensive way and directed to the work carried out by the teams as an evaluative focus are rare.
This study aims to evaluate the performance of comprehensive care for older adults in primary health care services according to their managers and professionals, as well as to analyze the relationship between performance and indicators of health planning and evaluation.

METHODS
Cross-sectional evaluative research based on the analysis of the results of QualiAB application in 2014. The evaluation occurred in PHC services located in five health regions in midwestern São Paulo: Bauru, Jaú, Lins, Polo Cuesta and Vale do Jurumirim, totaling 68 municipalities and 303 services, according to CNES 12 (eliminating duplicate and non-corresponding registrations to PHC services).
This is a convenience sample, consisting of five health regions chosen for a census application of QualiAB 2014. The regions are in areas close to the higher education institution responsible for the research, with which previous partnerships were made in projects to support the management and evaluation of PHC services using the same instrument 13 . A total of 157 services distributed in 41 municipalities adhered to the project, representing 63% of the existing in the period 12 . Adhesion was voluntary since no financial incentive or prize was offered in the process.
Teams of each health service answered QualiAB 2014 online, after the municipal manager and the managers or responsible for each service adhered. This application integrated the instrument upgrade and revalidation process, originally validated in 2007 [13][14][15] .
This QualiAB version consists of 126 multiple-choice questions, which generate composite indicators for the overall evaluation of PHC services, to cover the diverse set of health actions attributed to primary care, as provided for the National Primary Care Policy (PNAB-Política Nacional de Atenção Básica) 16 . It is a questionnaire developed to evaluate and monitor services regarding the structure and organization of the work process by indicators of care and management, regardless of the type of unit, that is, organized both according to the Family Health Strategy and by another type of arrangement between the various existing compositions 13 .
To this study, we chose and categorized the variables of QualiAB 2014 related to older adults' health and aging care, thus defining 155 performance indicators. First, we identified and analyzed documents on this evaluation object, to enable its delimitation and build a logical-theoretical model that will guide the selection of indicators and their organization in domains. Although a program specifically designed for the older adults' health is not instituted, as it is with other segments, guidelines and technical standards are available, which allow delimiting the actions that should be developed at this level of care 17 .
We used the following documents: the Brazilian Health Policy for Older Adults (PNSPI-Política Nacional de Saúde da Pessoa Idosa) 18 , and the guidelines of the Ministry of Health published in the Primary Care Guide no. 19 2 , in the Strategic Action Plan to Tackle Noncommunicable Diseases (NCD) in Brazil 2011-2022 19 , and in the technical manual Guidelines for the care of people with chronic diseases in Health Care Nets and in priority lines of care 20 .
The QualiAB 2014 selected indicators were gathered according to three large sets of attributes considered essential to comprehensive care for the older adults' health and aging in three domains: health care for active and healthy aging (d1, composed of 45 indicators), chronic noncommunicable diseases (CNCD) care (d2, composed of 89 indicators), and structure and support network in the aging care (d3, composed of 21 indicators). Statistical analysis was conducted from different strategies. To each domain, a final score equivalent to the sum of dichotomous responses was assigned to the indicators that compose it. Value 1 refers to achievement and value 0 to non-achievement. We also defined three clusters (performance groups) using k-means related to the scores, to achieve maximum heterogeneity between the different groups and maximum internal homogeneity in each group 21 . The three groups were organized by decreasing gradation of k-means obtained by the services, naming group 1 (G1) the services with the highest k-means, and group 3 (G3) those with the lowest k-means.
The second stage tested the association of each group composition with the responses to 30 QualiAB 2014 indicators that characterize the activities related to the service planning and general evaluation, shown in Chart 1. The associations between the selected indicators and each of the three domains were estimated using chi-square tests followed by Z tests. As each domain consisted of different amounts of indicators (with maximum scores of 45, 89 and 21, respectively), the scores were standardized to keep them within the same magnitude scale.
With the standardized scores, simple and multiple linear regression models were adjusted. As independent variables we had the service selected characteristics (Chart 1), and as response variables, the scores of each domain. The variables that obtained r 2 values above 50% and significance levels of the model lower than 0.25 with normal distribution of residues in simple adjustments were carried to multiple adjustment, in which those with r 2 above 80%, p < 0.05 values, and normal residues distribution were maintained. All steps of statistical analysis were developed with SPSS software version 20.0.

RESULTS
Among the 157 PHC services evaluated, according to self-classification, Family Health Units (FHU; 42%) prevailed, followed by traditional Basic Health Units (BHU; 36.9%). The Family Health Strategy (FHS) was present in 8.2% of the BHU, and the Community Health Agent Program (CHAP), in 8.9%; 1.2% were traditional BHU or FHU integrated to emergency care, and 2.5% of the services chose to be classified with other modalities.
All 45 QualiAB 2014 indicators elected to domain 1 (d1) cover actions of promotion, prevention and assistance related to older adults; aging care in the prevention of health problems and in health promotion; strategies and actions in situations of violence against older adults; and attention to older adults' caregivers. Domain 2 (d2) clustered 89 indicators corresponding to routine actions for people with CNCD and strategies to approach nonadherence to treatment; programmatic care, routine examinations and medications available to people with arterial hypertension and to people with type 2 diabetes mellitus; and actions for bedridden people care. In domain 3 (d3), we clustered 21 indicators related to infrastructure, supplies and professional qualifications that enable prevention and health promotion for an active and healthy aging, in addition to the network of services for older adults' health care in collaborative work with primary care. The services distribution in the different groups (G1, G2 and G3) and the average performance according to each analysis domain are shown in Table 1. In all three areas most services are concentrated in group 2 of performance, although with variations between the means achieved. The highest values were obtained in the domain of CNCD care (d2), in which the highest means are also visible.
In domain 1, it is noteworthy the services presented the lowest means among the groups with worst performances. In domain 3, group 1 obtained its lowest mean (51.0) among all domains. Notably, most services perform less than half of the set of indicators evaluated, composing groups 2 (39.2; d2) and 3 (30.2; d3). The slightest difference between the means of the groups in this domain suggests the services show less differentiation in relation to the indicators. Tables 2 and 3 show the frequencies of indicators by domain and by performance group (G1, G2 and G3), allowing to deepen the understanding of central tendency measures. The differences between groups are significant in 72.9% (113) out of the 155 indicators, shown in Table 2. Services belonging to G1 have high percentages of achievement of most of the actions evaluated; however, they too lacks totality for a comprehensive care for older adults' health and aging.
In domain 2, we observe the smallest percentage differences of achievement between the quality groups. The smaller response percentages are among the indicators of domain 3, focused on equipment with which services can share actions and care directed to older adults and promote the team's continued training on older adults' health-for example, the Specialized Social Assistance Reference Center (SSARC). Table 3 features the indicators with no significance according to the chi-square test. This includes 27% (42) out of the 155 indicators. Notably, most of them are in domain 3, in which non-differentiation between groups is due to the availability of infrastructure and, on the other hand, to the lack of support from the services network to older adults' care in most services. Table 4 provides adjustments to the simple and multiple linear regression (SLR and MLR) models for the domain scores for the health planning and evaluation variables that remained in the final multiple model, although others may have been adjusted according to the criterion defined in the method. Analyzing the independent variables related to the differentiation of quality groups for domain 1 (d1), we found that the service typology, the consequences of evaluations, and changes induced by evaluations are distinctive to compose groups with different performance.
For domain 2 (d2), the independent variables that reflect relation with the differentiation of quality groups relate to the services typology, the use of epidemiological data, and the changes induced by evaluations, to form groups with better or worse performances. In domain 3 (d3), independent variables that influence on the formation of the different performance quality groups for the set of indicators evaluated relate to the changes induced by evaluations.    It is important to note the adjustment model (r 2 ) for each domain proved to be great, a fact corroborated by its results: 0.82%, 0.96% and 0.93%, respectively.

DISCUSSION
The results point to a better performance of PHC services in actions related to chronic noncommunicable diseases, matching with results from other studies 22,23 and reflecting the undifferentiation with which services fulfill the scope of actions in older adults' health and aging care. Prevention and promotion actions were incompletely incorporated into few services, and most of them lack a support network to collaborative work, resulting on the lack of comprehensive care for older adults' health in the sample. This is corroborated by a descriptive study previously conducted by the authors with analysis of the response frequencies of these primary care services on the indicators. This study identified: the presence of a support group for older adults' caregivers in few services (8.2%), the use of a protocol to attend cases of violence against older adults (12.1%), a record with differentiated risk for people with chronic conditions (31.2%), and inclusion of older adults' health as the subject of continuing education for the professional team (33.1%) 11 .
Aging and longevity are strongly associated with developing chronic diseases, because aging without any chronic disease is rare 24 . However, old age cannot be reduced to a set of illnesses in the actions developed in PHC.
The relevance of preventive actions for comprehensive care to the adult's-and especially to the older adult's-health apparently is a consensus; however, notably, practices tend to be restricted to the care of more prevalent CNCD 5,24 . It is obviously not a question of minimizing  the importance of the quality of care for people with chronic diseases, but of emphasizing that older adults are often inserted in this set of problems, which does not automatically offer the necessary for them and for their aging.
Naturally, the actions performed by the HiperDia program, instituted since 2002 by the Brazilian Ministry of Health for people with systemic arterial hypertension and diabetes mellitus 25 , were implemented because chronic diseases represent one of the main demands for continuous follow-up in PHC services. However, to invest in improving the quality of this care is necessary, because the services fail to fully perform the recommended, as in relation to periodic funduscopy or foot care for people with diabetes, which occur in few services evaluated 11 .
In this study, it is possible to verify that the guidelines for health prevention and promotion are not being implemented in older adults and aging care, which also occurs with other stages of the life cycle and health demands 26,27 . In this sense, monitoring older adults' functional capacity becomes a strategic indicator for health services 23 , especially for primary care services, which are potentially instrumentalized by the use of scales for functional and cognitive evaluation, among others, available in the Primary Care Guide no. 19 2 .
Studies on older adults' care by primary care services conducted in Rio de Janeiro (RJ) 28 and in Santos (SP) 29 found a great disarticulation of the health services network in general and, especially, an absence of an older adults' health care network, which hampers the supplying of this population health needs. Associations between the performance of the services in older adults' care and the performance of activities indicative of local planning and evaluation show the importance of service management in determining the technical performance of programmatic activities. In particular, the low incorporation and use of data resulting from evaluation processes highlights the need to strengthen and improve monitoring and evaluation practices as subsidies for reprogramming technical actions of care, such as those directed to older adults' health and to aging.
The limits of a cross-sectional study carried out in certain health regions and with a convenience sample should inhibit generalization of results. However, it brings significant elements in the face of the evaluation gap and the absence of instruments directed to assess older adults' and aging care. In this sense, the study proved feasible to use an instrument that does not focus on this assessment but uses it as part of the diverse set of actions under the responsibility of PHC services.
It is necessary to advance in the provision of practices aimed at older adults by PHC services that seek to face the health demands of this population, aiming to understand the needs arising from aging and to collaborate so that individuals achieve old age with independence, autonomy and productivity 29 , because most chronic diseases that affect older adults have their main risk factor in their own age 24 . A contemporary model focused on aging, but especially in the older adult, needs to gather a sequence of education, health promotion, prevention of evitable diseases and postponement of injuries. To achieve positive results, an articulated, referenced network with information system built is essential 24,30 .

CONCLUSIONS
The performance of comprehensive care for older adults in primary care services in the Brazilian Unified Health System in the state of São Paulo is incipient. Incipience is stronger in healthy aging care activities, while those for treatment of some chronic diseases show relatively better performance.
To distinguish three groups of decreasing performance services was possible according to the three domains evaluated. The intermediate group, group 2, concentrates most services (67%). The best group, group 1, is minority (12%). The worst performing group, group 3, concentrates 31 services (20%) and shows worst performance in all domains, especially in the healthy aging care.
Results indicate the need to improve older adults' care in all domains, emphasizing, however, the urgency of improving local performance and health network in the effective implementation of actions to aging and older adults' care according to the Ministry of Health guidelines. The precariousness of performance of various services also draws attention.
The evaluation framework was able to characterize and differentiate the current performance of older adults' care in PHC services and can be used to establish and disseminate new normative standards for PHC.