Introduction

The use of health services can indirectly measure access and equity of the health system. The Brazilian National Health System (SUS) establishes national regulations for its decentralized, comprehensive and universal services. However, regional characteristics generate special features on the system’s operation and access. In geographical and economically peripheral regions, such distortions impact the system’s performance and the population’s health.

The use of health services can be assessed on the whole - use of any health service - or specifically - use of the most common services. In Brazil, a systematic review of population-based studies, conducted between 1998 and 2013 observed an increase in medical and dental visits and a decrease in hospitalizations.¹ In the North region of Brazil, however, medical and dental visits decreased in the same period, and the decrease in hospitalizations was even higher than in the other regions of the country.¹ The Brazilian National Health Survey (PNS), held in 2013, presented a prevalence of 17.5% in the use of any health service in the two weeks prior to the interview in Brazil.²

Low supply and use of health services and low level of self-awareness on chronic diseases can impact on a lower expectancy of life and of healthy life of the population.³ The local burden of diseases also pushes the demand for health services, especially those related to non-communicable diseases,⁴ which require continued health care for control and secondary prevention. Environment and social context also affect the burden of diseases and associated costs,^5-7 whose effects fall upon the health system. Thus, the organization of these services impacts on a population’s health status, by inducing the use of such services.

Therefore, investigating health situation and use of health services is essential, in order to highlight aspects which demand attention to improve the region’s situation. In areas where the situational diagnosis of health services and population’s health status is less known, such as Amazonas, this type of research is even more necessary. The objective of this study was to describe the use of health services among adults in Manaus Metropolitan Region, Amazonas State, Brazil.

Methods

Results

A total of 8,587 households were visited, from which 3,177 (36.9%) were closed or the resident did not suit the predefined quotas by sex and age. A total of 5,410 individuals were invited to participate in the study, of whom 1,314 refused (24.3%) and 95 (1.8%) were ineligible due to cognitive impairment. In the end, 4,001 adults were included in the study (Figure 1).

Figure 1 - Recruitment procedure for the population-based survey in Manaus Metropolitan Region, 2015 

The participants’ characteristics are presented in Table 1. There was equivalence of sex representation, with slightly more women (52.8%) than men. Half of the participants were 18-34 years old, most were brown-skinned (72.2%) and 1% self-declared indigenous. More than half of the participants (54.3%) were single. Participants with at least high school represented 47.5%; 62.7% belonged to economic classes C2/D/E, and 86.7% lived in Manaus. About 80% of the participants rated their health status as good or regular.

Table 2 shows the results for the participants who had used a health service in the previous two weeks (n=838; 20.9% - 95%CI 19.7; 22.2%). The main reason for using a service was health problem (64.5%) and diagnosis or prevention of diseases (22.0%). Most individuals got assistance at the first attempt (84.4%); 22.7% searched for assistance in two or more health services; 44.4% were assisted in primary health care units; and 41.1% were assisted in urgency or emergency services.

Most of the participants who had not used a health service in the two weeks prior to the interview (n=3,163) reported that they did not need it (89.3%). Among the reasons for not getting health care, individuals reported limitations of the service (available timetables, waiting time, distance, specialist availability: 6.9%), and personal limitations (constraints regarding money, transportation or hospital companion: 3.2%).

Medical visit was the most frequently used service in the previous 15 days (76.5%), whereas 2.9% reported they had never visited a doctor (95%CI 2.4;3.4) and 20.6% last visited a doctor over one year before the survey (Table 3).

Most of the 3,886 individuals that had visited a doctor reported the last medical visit as ‘good’/‘very good’ regarding the following aspects: equipment (61.8%); room for consultation (61.1%); attendants (60.9%); cleanliness (55.9%); and assessment of physician’s skills (64.3%), respect (66.0%), transparency (63.3%), privacy (68.0%) and availability for questions (65.4%). The waiting time was considered ‘bad’ or ‘very bad’ by 27.3% of the participants and 25.7% gave negative feedback regarding the freedom to choose a physician. Other negative aspects included travel time (18.9%) and physician’s availability for patients’ questions (12.3%; Table 4).

Discussion

A fifth of the population of Manaus Metropolitan Region used a health service in the fortnight previous to the survey and almost 10% searched for a service, but did not use it due to personal or system’s constraints. Among those who used the service, the primary health care unit was the main place of assistance, whereas medical visit was the most used service. On the overall, the last visit was positively assessed. Waiting time, opportunity to choose the physician and communication with the professional are points that would benefit from improvements.

Adults of Greater Manaus used a health service in the last fortnight with higher frequency than that observed nationwide by the National Health Survey (17.5%; 95%CI 17.0; 18.1%).² However, the proportion of individuals in Greater Manaus who had never visited a doctor was three times higher than the national estimate, highlighting the existing inequalities in Brazil.¹⁷

The positive feedback about the last medical visit was similar to that observed nationwide by the National Health Survey, which also reported less satisfaction regarding waiting time and freedom to choose a physician.¹⁸

It was not possible to estimate the impact of losses caused by refusals and closed households in this study. Inclusion of individuals was random and based on the predefined quotas by sex and age, from official estimates, in order to obtain demographic representativeness of the population as a whole. Another effort that could have been employed to increase the sample’s representativeness would be the random selection of residents and appointment of an interview with those who were not present at the moment of the household visit. The proportion of refusals in epidemiologic studies has been increasing over the last decades, however the refusals’ proportion does not necessarily indicate a selection bias.¹⁹ Thus, it is important to assess possible differences between the characteristics of the participants and of those who declined participating in the study.¹⁹

The participants’ characteristics support the representativeness of the sample. The proportions of women and young adults are similar to those of the last official census.⁹ Most of the individuals live in Manaus, a proportion similar to the official estimates for this metropolitan area, and self-reported brown-skinned, an ethnical mix of indigenous, white and black individuals, which is typical in that region.

Other characteristics emphasize the differences of this area in comparison with Brazil as a whole. Almost 90% of the population belongs to C, D and E economic classes. In Brazil, this stratum represents 75% of the population.¹⁴

Besides the limitations inherent to the applied cross-sectional design, there is a possibility of selection bias. Individuals with more severe diseases - and consequently more demand for health care - might not have been included in the survey because they were under treatment, thus, not in the household. This would lead to an underestimation in the prevalences of health services use.²⁰

A third of the sample assessed their health status as regular, bad or very bad, which shows the need for healthcare assistance in the region. Considering the low economic power of the population of Manaus Metropolitan Region, their need for public health assistance, which depends exclusively on SUS, is even higher.²¹ The strengthening and widening of SUS is essential to the region’s social development.

A second stage of the project was approved by the Ethics Research Committee (Report No. 1,541,710 dated May 10^th 2016; CAAE 53215816.6.0000.5020) and began in June 2016. In this second phase, participants are contacted by telephone to ascertain their interest in performing laboratory tests, such as blood count. Once the individual accepts, a new term of consent is presented and, after signed, the collection of biological material is scheduled to be performed in the household or work place, wherever is more convenient for the participant. This stage depends on reaching the participants by phone, which is a concern for researchers: in Brazil, it is very common for people to change cell phone numbers due to changes in their plans and promotions from other telephone companies. Other factors that may affect adherence are the lack of interest, fear of needles/syringes and misreporting of telephone numbers during the interviews performed in 2015.

The results of this survey point to differences in the use of health services in Manaus Metropolitan Region compared to the national profile. The investigation of associated factors, in future analyses, may help elucidate the health status and use of health services in Manaus Metropolitan Region, drawing attention to the problem and to health policies aimed at its resolution.