The assessment of PHC attributes attainment using PCATool among elderly individuals identified that those who consulted in units with FHS had higher PHC scores than those attending to BHU. Such association between model of care and PHC score was independent of socioeconomic status or major health problems. Work status, a surrogated marker of income, played a role in the adherence to PHC orientation and those elderly patients not working were more likely to have low score. Low PHC score was associated with two morbid conditions - hypertension (inversely) and cardiovascular disease (positively). Regardless of the health problem, the PHC scores were directly associated with the mental component of quality of life and negatively with the physical component. These results suggest that the model of care in PHC was associated with quality of life.
Evaluation of the PHC attributes attainment using PCATool [17
] had already been done, but not for elderly patients. The higher PHC scores observed among elderly who consulted at FHS could be anticipated, since this model of care was planned to fulfill these attributes while BHUs aimed to a different target. Even so, is reassuring to detect that the FHS reached its purpose in elderly from Ilheus, Bahia, in Northeast Brazil. However, there are other dimensions to be assessed in the PHC. For instance, patient-physician relationship and teaching activities were positively associated with quality of care in a study conducted in the Catalan Primary Health Care - with universal coverage used by 75% of the population, which is quite similar to the Brazilian coverage by the national health system. In the Spanish study items such as accessibility and doctor-patient relationship were higher in rural areas, less privileged populations, and among teams involved in the care of elderly [29
]. Although PCATooL and other instruments used to evaluate the attributes of PHC do not incorporate the doctor-patient relationship and adoption of evidence-based practices, they included the core attributes [30
FHS units are located in more deprived areas than BHU units. This condition did not account for overall differences between elderly patients who consulted in both models of care. While mean age, work status, and morbid conditions - hypertension, diabetes mellitus, obesity, cardiovascular disease, and mental disorder – were similar among patients from both models of care, those who consulted in the FHS units were more likely to live alone or with a spouse, had low education level, higher prevalence of central obesity and chronic pain. These conditions were directly associated to the use of health care facilities [31
]. In a multivariate analysis, the low PHC score was also associated with work status.
The report of morbid conditions as major health problems is not equivalent to the prevalence of these conditions. Some chronic conditions such as hypertension [33
] and chronic pain [34
], were underestimated. For instance, hypertension prevalence affected 68.9% (95% CI 64.1%–73.3%) of the Brazilian elderly population, according to the JOINT definition and using blood pressure measurement, 49.0% (95% CI 46.8%–51.2%) by self-reported in household surveys, or 53.8% (95% CI 44.8%–62.6%) by telephone surveys [35
]. These numbers are far higher than the 30% of elderly reporting hypertension as a major health problem. Besides the label of being a hypertensive patient and taking lowering blood pressure medicine daily, hypertension does not cause symptoms and is easily underestimated as a burden of disease. Elderly with hypertension seems to benefit of home visiting and reinforcement to take blood pressure-lowering medication [23
], resources offered in units with FHS. Its inverse association with low PHC score, independently of other confounding factors, might be attributed to characteristics of hypertension that demands frequent appointments for dispensing medicine and checking hypertension control.
Quality of life for the elderly population was shown to be associated with the PHC score, independent of confounding factors and even of hypertension, which is one of the determinants of reduced quality of life [33
]. Morbid conditions can reduce health and, in turn, the physical component of quality of life [36
]. Studies conducted in China [31
] and Germany [32
] have shown that low quality of life markedly increased the use of health services, but this relationship was characterized for PHC without comparison with traditional Chinese medicine or other type of healthcare. In the elderly, the deterioration of quality of life mainly due to the physical functioning rather than mental [38
] was shown through the opposite association with PHC score. The inverse association between PHC score and physical component of quality of life in the elderly suggests the difficulty to benefit from health care system due to loss of physical functioning. Findings in elderly living in Australia showed a negative association of PCS-12 with characteristics of the practice, which highlighted the role of chronic diseases [39
]. The presence of chronic pain considerably limits the autonomy of the elderly to perform daily living activities [40
], which could indirectly lead to increased use of units with full care, increasing the proportion of patients with lower quality of life [41
Potential limitations of our study should be addressed. For example, the cross-sectional design does not allow to characterizing a causal association, and reverse causality may play a role for some associations. Nonetheless, our findings reinforce the need for integral action in the care of elderly. Practices of health promotion are relevant to the control of chronic diseases, regardless of which one is the health problem. Increased access to PHC, offered in units with FHS, can be part of the strategy to provide assistance to people with disability and elderly above 80 years, which should an increase in the total burden of disease in the coming years.
In this study, socioeconomic status was determined by formal education and work status, but residual confounding is still possible. In fact, the socioeconomic status was positively associated with quality of life in some, but not in all countries [43
]. In Canada, for example, patients have universal access to health care and socioeconomic status does not prevent to seeing a doctor. In addition, a consultation with a specialist in Canada is determined by need and not by household income, the opposite of the United States [44