Implementation research: towards universal health coverage with more doctors in Brazil.

OBJECTIVE: To evaluate the implementation of a programme to provide primary care physicians for remote and deprived populations in Brazil.
METHODS: The Mais Médicos (More Doctors) programme was launched in July 2013 with public calls to recruit physicians for priority areas. Other strategies were to increase primary care infrastructure investments and to provide more places at medical schools. We conducted a quasi-experimental, before-and-after evaluation of the implementation of the programme in 1708 municipalities with populations living in extreme poverty and in remote border areas. We compared physician density, primary care coverage and avoidable hospitalizations in municipalities enrolled (n = 1450) and not enrolled (n = 258) in the programme. Data extracted from health information systems and Ministry of Health publications were analysed.
FINDINGS: By September 2015, 4917 physicians had been added to the 16 524 physicians already in place in municipalities with remote and deprived populations. The number of municipalities with ≥ 1.0 physician per 1000 inhabitants doubled from 163 in 2013 to 348 in 2015. Primary care coverage in enrolled municipalities (based on 3000 inhabitants per primary care team) increased from 77.9% in 2012 to 86.3% in 2015. Avoidable hospitalizations in enrolled municipalities decreased from 44.9% in 2012 to 41.2% in 2015, but remained unchanged in control municipalities. We also documented higher infrastructure investments in enrolled municipalities and an increase in the number of medical school places over the study period.
CONCLUSION: Other countries having shortages of physicians could benefit from the lessons of Brazil's programme towards achieving universal right to health.

Introduction

Brazil’s large territory, partially covered by dense forests, is a challenge for efforts to achieve universal health access and coverage for the population. After the re-establishment of democracy and the new Constitution in 1988, access to health care was established as a constitutional right and an obligation of the state. A unified national health system – the Sistema Único de Saúde – was founded in 1988, based on the principles of universal, integrated health care with equality of access to all. Financed with public funds, the national health system provides preventive and curative health care free of charge.

Although the health-care system has made major advances in the past 25 years, such as enhancing the coverage of antenatal care and measles vaccination, transforming the constitutional right into a reality still remains a challenge. For example, a study by the federal medical council in October 2012 showed that there were 388 015 physicians for an estimated population of 191 million people. This average density of 2.00 physicians per 1000 inhabitants nationwide varied from 0.71 physicians per 1000 people in the state of Maranhão (the poorest state) to 4.09 per 1000 people in the Federal District (the wealthiest state). Distribution within the states was also unequal; some municipalities in the north and north-east semi-arid states had no physicians. The causes of these inequalities of distribution are related to problems of recruitment and retention of physicians at basic health units in the national health system.– These include job insecurity and low career prospects in the public health system; poor working conditions and infrastructure in primary care facilities; and better job opportunities elsewhere.

Brazil needs an equitable distribution of the health workforce, particularly physicians, if it is to achieve the sustainable development goals (SDGs) on health and equality (SDGs 3 and 10)., The tendency for physicians to concentrate in urban centres is a problem that affects many countries. Strategies for retaining health-care professionals in areas of vulnerability range from providing incentives and support, to enforcement measures such as mandatory civil service.– To increase the number of primary health-care physicians for priority areas the Brazilian government set up the Mais Médicos (More Doctors) programme in July 2013. This employed strategies aligned with those proposed by the World Health Organization to improve the retention of health workers in remote and rural areas.,

The aim of the present study was to describe the implementation of the programme in remote and deprived areas in Brazil and to evaluate its contribution to progress towards achieving universal health access and coverage for these vulnerable populations.

Methods

Intervention

The Mais Médicos programme aimed to actively recruit primary care physicians and provide incentives to retain them through better working conditions and increased investment in the primary care infrastructure. The programme also addressed the shortage of physicians in the mid- and long-term through an expansion of medical training.

All 5570 municipalities of Brazil were eligible to enrol in the first strand of the programme: the emergency provision of physicians for primary health care. Priority was given to municipalities with 20% or more of the population in extreme poverty; in deprived areas of state capitals and metropolitan regions; in rural and remote regions, e.g. the north-east semi-arid region and the north (Amazon) region; and with traditional populations, such as Maroon communities. All the 34 indigenous health districts that provide primary health care to 517 383 Brazilian indigenous people in forest and rural areas (2010 Census data) were prioritized. The programme offers physicians incentives to enrol, with three-year work contracts, competitive salaries (compatible to market values), lodging expenses and airline tickets to home once a year.

The second strand of the programme was to enhance the quality of primary health care and guarantee better working conditions for health professionals by investing in the infrastructure of basic health units. Municipalities could apply to the health ministry for funds to construct or refurbish health units or to purchase basic equipment such as refrigerators, autoclaves, otoscopes and sphygmomanometers.

The third strand was to increase the overall number of physicians in Brazil by providing more undergraduate medical school places in states where there were ≤ 1.5 places per 10 000 population or ≤ 2.7 doctors per 1000 population. The federal government planned to authorize 11 447 new undergraduate medical course places by 2017. Priority was initially for towns with a population of 70 000 inhabitants or more and without a medical school in their territory; later this changed to areas 75 km away from any medical school. The programme also made changes in the national curriculum to tailor undergraduate medical training to the national unified health system.

Implementation

The programme was launched nationwide via a provisional presidential decree on 8 July 2013 and was enacted into Law No. 12.871 of 22 October 2013. Public calls for enrolment of municipalities and recruitment of physicians were opened on the health ministry website (http://www.maismedicos.gov.br). From 8 July 2013 to 2 October 2015 there were six public calls for municipalities to join and eight public calls for physicians to enrol.

At the first public call, in July 2013, 3511 of the 5570 municipalities in Brazil enrolled in the programme and requested 15 460 physicians. Although the programme prioritized the recruitment of Brazilian physicians, only 1096 nationals spontaneously registered and were hired, while another 522 positions were occupied by physicians from other countries where the ratio of physicians to inhabitants was higher than in Brazil. Both Brazilians and the individual exchange foreign professionals could indicate up to six locations of choice. With the assistance of the Pan American Health Organization, an agreement was signed with the Cuban government to supply physicians with primary health-care experience. Cuban physicians were sent to priority municipalities, usually remote places, which had not been chosen by Brazilian doctors. The health ministry licensed physicians who obtained their diplomas abroad to practise medicine for three years only in the assigned municipality.

In November 2013, according to the health minister, all the regions including rural communities and the Brazilian semi-arid region had at least one doctor. By September 2015 the programme had increased to 4004 participating municipalities and provided 17 625 physicians: 11 329 (64.3%) Cubans, 5600 (31.8%) Brazilians and 696 (3.9%) professionals from other countries.

Study design

We conducted a quasi-experimental, before-and-after evaluation of the implementation of the programme in 1708 priority municipalities with remote and deprived populations. These were municipalities with 20% or more of the population living in extreme poverty and those located in the country border areas. The proportion of the population in extreme poverty in each municipality was obtained from the Ministry of Social Development website. The cut-off for defining social vulnerability due to extreme poverty in July 2013 was family monthly income per capita below 70 Brazilian reais (equivalent to 19.4 United States dollars, US$).

We compared the 1450 remote and deprived municipalities that enrolled in 2013 or 2104 and received physicians, with the 258 municipalities that were eligible and prioritized but did not enrol in the programme. Of the original 1450 enrolled municipalities, 39 (2.7%) remained in the programme for two years but dropped out in 2015 and so were lost to follow-up.

The study was approved by the research ethics board of the University of Brasilia (protocol no. 399.461, CAAE 21688313.9.0000.0030).

Outcomes and data sources

We extracted data on outcomes of the programme in the years before and after implementation of the programme in the studied municipalities.

The number of physicians in each municipality in April 2013, just before the programme started, was obtained from the Brazilian National Register of Health Institutions Establishments. The number of physicians allocated by the programme in September 2015 was requested by the health ministry. For this study we included physicians from Mais Médicos and from PROVAB (Programa de Valorização da Atenção Básica), a similar programme,which attracted fewer physicians and only Brazilians. In January 2015 the health ministry merged the two programmes.

We estimated the density of physicians, i.e. the number of physicians per 1000 inhabitants, using official population estimates based on Brazilian national census data. The density of physicians across Brazil was then mapped in five bands: ≥ 1.0, ≥ 0.7 to < 1.0, ≥ 0.4 to < 0.7, ≥ 0.1 to < 0.4, and < 0.1 physicians per 1000. The health ministry defined adequate density as at least 1.0 physician per 1000 inhabitants. Vector maps were created in Quantum geographical information system software, version 2.12.0 (Open Source Geospatial Foundation; http://www.osgeo.org/).

The health ministry stipulates that each family-health team covers 3000 people. To estimate primary health-care coverage, we obtained data on the number of primary health-care teams and the population in each studied municipality over the period 2011–2015. The number of teams was multiplied by 3000, divided by the number of inhabitants and multiplied by 100. Data were extracted from the health ministry’s DATASUS database.

We estimated the rate of avoidable hospitalizations in each studied municipality over the period 2011–2015 from the number of hospitalizations due to ambulatory care-sensitive conditions as a percentage of the total number of clinical admissions. In Brazil 17 ambulatory care-sensitive conditions were defined in the Ministry of Health Directive MS/SAS No. 221 of 17 April 2008. Ambulatory care-sensitive conditions are those which are potentially preventable through early diagnosis, treatment of acute conditions or control and monitoring of chronic diseases. Data were obtained from the health ministry website.

It was not possible to calculate confidence limits because we employed aggregated data.

We estimated the annual expenditure in Brazilian real on construction of new basic health units and renovation of existing units in the different regions of the country over the period 2012–2015. Data were obtained from the health ministry’s Support Unit for Strategic Management database. On 31 May 2016, the conversion rate was 1 Brazilian real to US$ 0.28.

To describe trends in medical school education we obtained data on the number of medical schools and the number of undergraduate medical places per 10 000 inhabitants in the different regions over the period 1994–2015. These data were collected from the medical schools of Brazil website and from the health ministry.

Results

Distribution of municipalities

Fig. 1 shows the geographical distribution of the 1708 studied municipalities with remote or deprived populations. These were mainly concentrated in the north-east and north of the country (Amazon region) or in the mid-west and south regions at the borders. Fig. 1 also maps the distribution of the 1450 (84.9%) municipalities that enrolled in 2013–14 and were allocated physicians by the programme and the 258 municipalities eligible and prioritized, but not enrolled in the programme in the period 2013–2015.

Fig. 1

Municipalities with remote and deprived populations according to enrolment in the

Notes: The Mais Médicos programme was implemented in July 2013. Municipalities with remote and deprived populations were those with 20% or more of the population living in extreme poverty (per capita family income below 70 Brazilian reais, equivalent to 19.4 United States dollars) in 2013 or those in the border areas. Enrolled municipalities were those that joined the programme and received physicians allocated by the programme. Of the 1450 municipalities enrolled in the programme in 2013–14, 39 (2.7%) had dropped out by September 2015. The 258 municipalities not enrolled were eligible for the programme but never enrolled at the time of the study. Other municipalities were not included in the study.

Sources: Secretariat of Labour Management and Health Education and the Department of Planning and Regulation of the Provision of Health Care Professionals, Brazilian Ministry of Health, on request, 2015.

Density of physicians

By September 2015, 4917 physicians had been allocated by the programme to the municipalities with remote and deprived populations. They accounted for 27.9% of the total of 17 625 physicians enrolled in the whole programme. Their nationalities were 3566 from Cuba (72.5%), 1270 from Brazil (25.8%) and 81 (1.6%) from other countries. Table 1 shows the numbers of municipalities with remote and deprived populations in each region of Brazil and the number and percentage of physicians allocated by the programme to these municipalities. The highest number of physicians was allocated to the north-east (3547, 72.2%), followed by the north (952, 19.4%).

Table 1

Enrolment of municipalities with remote and deprived populations in the Mais Médicos programme, and physicians allocated to these municipalities, by region, Brazil, 2013–2015

Region	Total no. of municipalitiesa	No. of municipalities with remote and deprived populationsb		No. (%) of physicians allocatedd
		Total	Enrolledc
North	450	253	228	952 (19.4)
North-east	1795	1287	1077	3547 (72.2)
South-east	1659	73	61	110 (2.2)
South	1200	60	50	204 (4.1)
Mid-west	466	35	34	104 (2.1)
All regions	5570	1708	1450	4917 (100.0)

a Total number of municipalities in each region in 2013.

b Municipalities with remote and deprived populations were those with 20% or more of the population living in extreme poverty (per capita family monthly income below 70 Brazilian reais, equivalent to 19.4 United States dollars) in 2013, and those located in the border areas.

c Of the 1450 municipalities initially enrolled in the programme 39 (2.7%) had dropped out by September 2015.

d Physicians allocated by the programme to municipalities with remote and deprived populations by September 2015.

Sources: Total number of municipalities were obtained from national census data from the Brazilian Institute of Geography and Statistics. Numbers of municipalities with remote and deprived populations and numbers of physicians allocated were obtained from the Secretariat of Labour Management and Health Education and the Department of Planning and Regulation of the Provision of Health Care Professionals at the Brazilian Ministry of Health, on request, 2015.

The map for 2013 shows the distribution of the 16 524 physicians already in place and incorporated into primary care teams before the programme started in the 1450 studied municipalities (Fig. 2). After implementation of the programme, 4917 new doctors joined, increasing by 29.8% the availability of medical doctors. There was an increase in the number of physicians per 1000 inhabitants in municipalities with remote and deprived populations. In 2013 there were 292 municipalities with < 0.4 physicians per 1000 inhabitants, and this declined to 81 municipalities in 2015 (decrease of 72.3%). The number of municipalities with ≥ 1.0 physicians per 1000 inhabitants rose from 163 to 348 during the same period, an increase of 113.5% (Fig. 2).

Fig. 2

Density of physicians in municipalities with remote and deprived populations, before and after implementation of the

Notes: The Mais Médicos programme was implemented in July 2013. Municipalities with remote and deprived populations were those with 20% or more of the population living in extreme poverty (per capita family income below 70 Brazilian reais, equivalent to 19.4 United States dollars) in 2013 or those in the border areas. Of the 1450 municipalities initially enrolled in the programme 39 (2.7%) had dropped out by September 2015. Density of physicians was the number of physicians per 1000 inhabitants in each municipality.

Sources: National Register of Health Institutions, Ministry of Health. Secretariat of Labour Management and Health Education and the Department of Planning and Regulation of the Provision of Health Care Professionals, Ministry of Health, on request, 2015.

Primary care coverage

In 2012, there were a total of 6948 primary care teams for a population of 26 742 974 in the enrolled municipalities, a coverage level of 77.9% (100% coverage was 3000 inhabitants per team). By 2015 there were 8038 teams serving 27 929 381 inhabitants, a coverage level of 86.3% (Table 2). Throughout the period 2011–2015, municipalities that did not enrol in the programme had high coverage of primary health care (around 95%), which may explain why they were not interested in enrolling in the programme.

Table 2

Primary health-care coverage in municipalities with remote and deprived populations and enrolled or not enrolled in the Mais Médicos programme, before and after implementation of the programme in Brazil, 2011–2015

Year	Municipalities enrolled (n = 1450)a					Municipalities not enrolled (n = 258)a
	Totalpopulation	No. of PHC teams	No. ofinhabitants per PHC team	Coverage,b (%)		Totalpopulation	No. of PHC teams	No. ofinhabitants per PHC team	Coverage,b (%)
2011	26 538 610	6901	3846	78.0		2 364 218	753	3140	95.5
2012	26 742 974	6948	3849	77.9		2 375 778	754	3151	95.2
2013	27 583 700	7354	3751	80.0		2 437 540	784	3109	96.5
2014	27 762 204	7933	3500	85.7		2 446 769	789	3101	96.7
2015	27 929 381	8038	3475	86.3		2 458 039	777	3163	94.8

PHC: primary health care.

a Of the 1450 municipalities initially enrolled in the programme 39 (2.7%) had dropped out by September 2015. The 258 municipalities not enrolled were eligible for the programme but never enrolled at the time of the study.

b The total population was divided by the number of PHC teams to obtain the number of inhabitants per primary health-care team. This figure was then converted to a percentage, in which 100% coverage was 3000 inhabitants per PHC team, as recommended by the Brazilian Ministry of Health.

Notes: The Mais Médicos programme was implemented in July 2013. We studied 1708 municipalities with remote and deprived populations, which were those with 20% or more of the population living in extreme poverty (per capita family monthly income below 70 Brazilian reais, corresponding to 19.4 United States dollars) in 2013, and those located in the border areas.

Sources: DATASUS health portal, Ministry of Health; Evaluation and Information Management website, Ministry of Social Development.

Avoidable hospitalizations

In 2012 the percentage of hospitalizations for ambulatory care-sensitive conditions in the municipalities enrolled in the programme was 44.9% (342 908 of the total clinical admissions of 764 342). This decreased to 41.2% (298 566 of 724 921 admissions) by 2015, a decrease of 8.8% (Table 3). By contrast, the rate remained unchanged over this period in the municipalities not enrolled in the programme.

Table 3

Rate of potentially avoidable hospitalizations in municipalities with remote and deprived populations, before and after implementation of the Mais Médicos programme in Brazil, 2011–2015

Year	Municipalities enrolled(n = 1450)a			Municipalities not enrolled(n = 258)a
	Total no. of clinicaladmissions	No. (%) ofavoidablehospitalizationsb		Total no. ofclinicaladmissions	No. (%) ofavoidablehospitalizationsb
2011	833 111	376 182 (45.2)		74 549	33 227 (44.6)
2012	764 342	342 908 (44.9)		69 078	30 831 (44.6)
2013	792 385	356 695 (45.0)		69 918	31 710 (45.4)
2014	765 845	332 498 (43.4)		69 247	31 882 (46.0)
2015	724 921	298 566 (41.2)		65 249	29 771 (45.6)

a Of the 1450 municipalities initially enrolled in the programme 39 (2.7%) had dropped out by September 2015. The 258 municipalities not enrolled were eligible for the programme but never enrolled at the time of the study.

b Rate of potentially avoidable hospitalizations was the number of hospitalizations due to ambulatory care-sensitive conditions as a percentage of the total number of clinical admissions. A set of 17 ambulatory care-sensitive conditions were defined in Brazilian Ministry of Health directive MS/SAS no. 221 of 17 April 2008.

Notes: The Mais Médicos programme was implemented in July 2013. Municipalities with remote and deprived populations were those with 20% or more of the population living in extreme poverty (per capita family monthly income below 70 Brazilian reais, corresponding to 19.4 United States dollars) in 2013, and those located in the border areas.

Sources: DATASUS health portal, Ministry of Health; Evaluation and Information Management website, Ministry of Social Development.

Infrastructure investments

Investments in basic health units totalled 1.10 billion Brazilian reais in the period 2012–2015, mostly concentrated in 2013, the year the programme was created. These included funds for the construction of 3496 new basic health units and refurbishment of 3417 units. Investments in the municipalities that enrolled in the programme totalled 4275 million Brazilian reais in 2013: 83.4% higher than in those that did not enrol (2331 million Brazilian reais; Fig. 3).

Fig. 3

Investment in the construction and renovation of basic health units in municipalities with remote and deprived populations, before and after implementation of the Mais Médicos programme in Brazil, 2012–2015

Notes: The Mais Médicos programme was implemented in July 2013. Municipalities with remote and deprived populations were those with 20% or more of the population living in extreme poverty (per capita family income below 70 Brazilian reais, equivalent to 19.4 United States dollars) in 2013 or those in the border areas. Of the 1450 municipalities initially enrolled in the programme 39 (2.7%) had dropped out by September 2015. The 258 municipalities not enrolled were eligible for the programme but never enrolled at the time of the study. Data were not available for investments before 2012.

Source: Support unit for strategic management website, Brazilian Ministry of Health.

Medical school places

In 2013,109 (40.7%) of the 268 medical schools in Brazil were in the richer south-east, whereas only 22 (8.2%) were in the north and 25 (9.3%) in the mid-west. From 2013 to 2016 a total of 70 new medical schools were created (public and private), offering 5540 new places for medical students. More than half of the new schools (38 schools; 54.3%) were in the more deprived regions of the north-east, mid-west and north, thus contributing to decreased regional differences.

We observed a similar pattern with respect to the distribution of undergraduate places in medical schools in 2013. The south-east had 10 639 (43.6%) of the total of 24 401 places and the north-east had 6051 (24.8%) places. There was an increasing availability of undergraduate medical places in Brazil over the years 1994–2015 (Fig. 4). Between 2012 and 2015, however, there was a marked increase in availability of medical school places in all regions of the country, especially the priority areas in the north-east and mid-west. The number of undergraduate places in medical schools in Brazil rose from 0.83 per 10 000 inhabitants in 2012 to 1.07 per 10 000 in 2015.

Fig. 4

Number of undergraduate places in medical schools per 100 000 inhabitants, by region of Brazil, 1994–2015

Note: The Mais Médicos programme was implemented in July 2013 in eligible municipalities in all regions.

Sources: Secretariat of Labour Management and Health Education, Ministry of Health. Medical schools of Brazil website.

Discussion

The Mais Médicos programme provides physicians to areas where deprived populations are living in conditions of extreme poverty. Working towards a goal of guaranteeing the right to universal health care, the programme facilitates access to health services for people who have seldom, or never, had access to that right.,

The municipalities that we studied were remote, and in the case of the Amazon region may be accessible only by boat or aeroplane. These regions face problems of recruitment and retention of physicians, primarily because they are far from large urban centres., Previous interventions such as PROVAB – based on successes in other countries – did not achieve the desired results, especially in the remote areas, due to low participation by Brazilian physicians. The Mais Médicos programme was set up to make more effective changes in the provision and attachment of physicians to primary health care. There was some resistance at first, especially from the Federal Medical Council, which headed a petition in the Supreme Court alleging that the programme was unconstitutional., The request was not approved, but disputes continue, due to controversy about hiring foreign doctors. The general public, however, has recognized the need for more physicians as many areas in Brazil have had insufficient numbers of physicians for long periods. Opinion polls conducted by the Brazilian National Transport Confederation, disclosed that the programme was approved by 49.7%, 73.9% and 84.3% of the population of 2022 respondents in July, September and November of 2013 respectively.

The data from our study showed that the programme succeeded in providing greater numbers of primary health-care physicians to remote and deprived populations and an increase in primary health-care coverage.

Introduction of the programme also coincided with important changes in health care, such as the reduction in hospital admissions due to ambulatory care-sensitive conditions, which could reflect positive changes in primary health-care practices. The programme could also be linked to mid- and long-term measures, such as increased investments in basic health units, infrastructure and equipment. An increase in the number of medical school places in Brazil was recorded in the period after implementation of the programme, resulting in a more equitable geographical distribution of places.

Similar programmes to expand professional training have been conducted in other countries with remote areas., The need to expand the number of medical courses in Brazil was based on the low supply of undergraduate medical places, which was 0.8 per 10 000 inhabitants in 2011, half that of other countries with universal health systems, such as the United Kingdom of Great Britain and Northern Ireland (1.6 per 10 000 inhabitants). In countries such as Australia, Canada and the United States of America, training programmes to serve populations in remote and rural areas where it was difficult to recruit and retain physicians were implemented in medical schools in the 1960s. Medical schools adopted measures to promote the selection of students from rural communities and enacted curricular reforms that provided theoretical and practical learning, with a focus on primary health care. In Australia, for example, the Rural Clinical Training and Support programme finances medical training aimed at remote and rural areas.

A limitation of our study is that the short time elapsed between the before and after evaluations precludes analysis of its impact on the health of the population. Further studies are needed to assess the impact of the programme on the population in Brazilian municipalities with extreme poverty.

Keeping up the recruitment and provision of physicians in the future is a major challenge for Brazil. The first phase of the programme had no long-term perspective. It was a temporary policy, effective for three years according to the law passed in October 2013. A new provisional measure to extend the temporary foreign exchange visa scheme for physicians for three more years was approved in September 2016. The health ministry has indicated that the programme will continue for this period. The provision of physicians to remote areas in the long term, however, will require measures such as the creation of a career in the public system that will provide greater job security and access to continuing education. On the other hand, this will not be possible if the chronic underfunding of the national unified health system is not solved. Less than one-third of physicians recruited by the programme up to September 2015 were Brazilians. Maintaining the programme in Brazil with foreign medical collaboration may be essential until sufficient new physicians have graduated to meet the need of all municipalities.

In conclusion, the Mais Médicos programme was successful in the immediate provision of physicians to work in primary health care with the most remote and deprived populations. In the longer term the programme assured investments, refurbishment and construction of basic health units and expansion of the network of medical schools, following a more equitable geographical distribution. Similar strategies could be established in other countries with a shortage of physicians, especially underdeveloped nations, to contribute towards achieving universal health access and coverage for all as proposed by the SDGs.