Brazil's Family Health Strategy in SUS

A territorial team makes a legal right practical

Brazil's Family Health Strategy (Estratégia Saúde da Família, ESF) organizes primary care around multiprofessional teams responsible for defined territories. The Health Ministry currently describes each team as a principal entry point for continuous care and names a physician, nurse, nursing assistant or technician, and community health agents as its minimum composition. Care can occur in a basic health unit, a home, or another community setting; the current population parameter ranges from 2,000 to 3,000 people per team according to municipality size.¹

That operating design sits beneath a larger promise. Articles 196 and 198 of Brazil's 1988 Constitution make health a right of all and a duty of the state, then direct public services into a regionalized, hierarchical system organized through decentralization, comprehensive care, and community participation.² The organizational question is therefore not only whether care is formally available. It is whether a territorial team can notice a need, act on it, and help a person cross the boundaries between home, clinic, laboratory, specialist, and hospital.

A 2015 New England Journal of Medicine perspective makes that chain concrete through Ms Silva, a healthy 23-year-old in a first pregnancy, and Ms Oliveira, her neighbor and community health agent. They are invented figures, not reported patients. In the vignette, Ms Oliveira arranges a prenatal visit, follows up after a missed appointment, and returns with a nurse after mother and baby come home. Tests require a municipal health center, delivery requires a maternity unit, and Ms Silva still bears travel costs even though covered care is free at the point of service.³

The invention describes intended coordination, not typical performance. It also reveals the strategy's central limit: a trusted agent can make absence visible, but cannot create a test slot, medicine supply, ambulance, maternity bed, or specialist appointment. Registration is the beginning of a care journey, not evidence that the journey was completed.

Drought work in Ceará became part of a national model

The household link preceded the national strategy. A historical study of Ceará dates a statewide emergency program to the June 1987–June 1988 drought. It reports that 6,000 women from the rural interior were selected partly to supply paid work to poor households and partly because they were known and respected locally. After three days of training, they sought pregnant people for prenatal care and maternity referral, encouraged breastfeeding and vaccination, and advised on household and child hygiene.⁴

When the emergency ended, Ceará continued the work through a state health-agent program created in 1989. The same history traces the model's movement into the national Community Health Agents Program and then into multidisciplinary family health teams.⁴ The federal Family Health Program began in 1994 and later became ESF.³ What began as paid drought response was thus absorbed into a national attempt to make a universal entitlement present inside households.

That lineage should not be turned into a lone-founder story or a simple causal claim about health gains. The Ceará history combines official records with interviews involving program designers, health officials, and an agent federation president. It reconstructs institutional memory and work design; it does not isolate the program's effect from sanitation, income, vaccination, hospital care, or other changes occurring at the same time.⁴

Agents turn absence into information and absorb the boundary strain

A clinic that reacts only to arrivals sees people able to reach it. A community health agent can notice a missed appointment, ask what prevented attendance, and bring that knowledge to the team. The role joins two kinds of knowledge: the household's circumstances and the clinical judgment of nurses and physicians. Its usefulness depends on whether the team can act on what the agent learns.¹³

A meta-synthesis of 33 Brazilian qualitative studies published from 2004 to 2014 found both sides of this work in agents' own reported perceptions. Agents described recognition by families, bonds with residents, peer support, and working near home as positive. They also described unclear limits on duties, precarious conditions, weak training, bureaucratic work, difficult community and team relationships, and unequal standing in planning and decisions.⁵

Living and working in the same territory creates an additional governance problem. Local familiarity may open a door that would remain closed to an unknown official, yet the same familiarity can make refusal, privacy, and the boundary between care and surveillance harder to negotiate. Visit counts and completed forms can show activity; they cannot by themselves show that a resident trusted the encounter, that a team acted on disclosed need, or that a referral reached its destination. Those questions require worker- and resident-controlled evidence that the reviewed worker studies do not provide.⁵

Outcome evidence is encouraging, observational, and uneven

A 2017 systematic review found 31 studies that met its criteria for assessing selected primary-care-sensitive outcomes. Fourteen associated higher ESF coverage with lower post-neonatal and child mortality. Thirteen hospitalization studies showed no clear pattern, while four studies found no effect on the vaccination or low-birth-weight outcomes they examined.⁶ This supports a narrow conclusion: wider coverage was consistently associated with some child mortality improvements, not with every outcome that primary care might affect.

The review could not pool results because methods and outcomes differed. Twenty-five of the 31 studies were ecological; coverage definitions varied; none of the ecological studies measured actual service use; and ESF expansion was not random. The reviewers therefore called the evidence limited rather than conclusive.⁶

A narrower São Paulo study offers a useful contrast. Among births recorded at one university hospital from 2003 through 2012, residence in an ESF-eligible area was associated with lower child mortality, especially deaths on days 3–27, but not with the measured birth-weight, gestational-length, or stillbirth outcomes.⁷ The rollout was not randomized, eligibility did not establish that a mother used ESF services, socioeconomic status was unavailable, and the hospital served a particular, poorer part of western São Paulo. The association is stronger evidence than a success slogan and weaker than a national causal estimate.⁷

National reach does not make local capacity equal. An OECD health-system review published in 2021 found large regional disparities in access and quality, shortages and maldistribution of primary-care physicians, uneven digital capacity, weak municipal capacity in poorer places, and incomplete coordination with specialist care. It recommended stronger referral pathways, richer patient-level quality measures, more support for low-capacity municipalities, and more consistent workforce arrangements.⁸ The report is a comparative system assessment, not a trial, and its administrative figures predate later policy changes.

Together, these sources shift the useful measure from nominal presence to completed care. A referral issued but never received, a test ordered but not performed, or a hospital episode that never returns information to the local team leaves the person and frontline worker inside a broken chain. Measurement becomes accountable only when someone can see where the chain stopped and has authority and resources to restart it.

Participation is a legal institution, not guaranteed influence

Law 8,142 of 1990 requires health conferences and permanent, deliberative health councils in each sphere of government. It assigns councils roles in formulating strategy and controlling policy execution, including financial matters; includes government, providers, professionals, and users; and requires user representation equal to all other segments combined. Council decisions remain subject to approval by the legally constituted executive authority.⁹

Formal standing does not settle practical influence. A 2021 scholarly viewpoint credits the councils with nationwide reach, inclusive composition, and a legal deliberative role, while identifying recurring challenges in the literature: weak community mobilization, technical and bureaucratic barriers, managerial control of agendas, decisions brought for formal approval after being made elsewhere, and insufficient system funding.¹⁰ Because the article is a viewpoint rather than a national empirical evaluation, it maps a documented debate but cannot establish how often each failure occurs.

A repeated local failure can therefore travel through two uncertain chains. The care chain runs from household to team to diagnostic or specialist service and back. The participation chain runs from lived experience to council agenda to government decision to funded response. Either may stop after making the need visible.

Authority must follow presence

ESF's defining capability is presence: a named team can know a territory over time, and an agent can recognize absence as information. Its defining risk is responsibility without matching authority. Frontline workers may hold the relationship with a person while another organization controls the test, transport, medicine, hospital bed, specialist, data connection, or budget.

These relations are editorial comparisons, not evidence of causal influence or institutional descent. Porto Alegre participatory budgeting offers a contrast in how public voice may connect to appropriation; knowledge, expertise, and professional autonomy helps distinguish household knowledge from clinical authority; governance, stewardship, and accountability asks who must repair a care path that crosses governmental levels; and benefit for all life keeps unequal access, worker burden, and unmeasured long-run effects inside the judgment of success.

The concept fingerprint follows local presence into public learning

Delegation, decentralization, and responsibility locates what national, municipal, team, and community actors can decide. Coordination, communication, and common understanding explains how household presence becomes a shared care pathway. Learning, quality, and reliability asks whether those signals change coverage and care without overloading the agents who carry them.

Paths into deeper study

• Follow individual referrals from household identification through completed specialty care and returned information, stratified by municipality, region, race, income, rurality, disability, and Indigenous status.
• Build resident- and agent-controlled records of consent, privacy, refusal, safety, workload, pay, voice, and what happened after an unmet need was reported.
• Update causal outcome reviews beyond the May 2016 search horizon and separate team registration, actual use, service quality, and access to downstream care.
• Compare council agendas, deliberations, executive approvals, budgets, and delivered services to learn when participation changes care rather than only recording complaint.

The hardest test remains a passage, not a count: did a need noticed at home become a team decision, a completed service, and useful knowledge back at the health post? Universal care becomes practical when presence, knowledge, resources, and authority remain connected for the whole journey.