INVESTIGACIÓN ORIGINAL ORIGINAL RESEARCH

Reply letter utilization by secondary level specialists in a municipality in Brazil: a qualitative study

Utilización de las cartas de respuesta por los especialistas de nivel secundario de salud en un municipio de Brasil: estudio cualitativo

Matthew HarrisI, *; Ana FerreiraII; Isabela MorãesII; Fernanda de AndradeII; Daise de SouzaII

IOxford University, Department of Public Health and Primary Health Care, Oxford, England
IIFamily Health Program, Camaragibe, Pernambuco, Brazil

ABSTRACT

OBJECTIVES: To identify obstacles impeding the use of the reply letter by secondary level specialists in the municipality of Camaragibe, Pernambuco, Brazil, and to highlight possible solutions for improving communication at the interface between secondary level specialist care and primary care.
METHODS: Conducted in 2003, this qualitative study used semistructured interviews with 10 randomly selected secondary level specialists. The following specialties were represented: neurology, obstetrics, gynecology, psychiatry, otolaryngology, and colposcopy. Interviews were semistructured, using an interview guide, based on the following seven categories: (1) profile of the specialist, (2) description of outpatient clinic services, (3) professional satisfaction, (4) assessment of the municipal health services, (5) assessment of the Family Health Program (Programa Saúde da Família), (6) description of the referral process, and (7) reply letter utilization.
RESULTS: Most of the specialists understood the importance of the reply letter and that this form of communication with the generalist would ensure continuity of care, avoid duplication of efforts or disorganized patient management, and would provide the generalist with the professional support required. However, the study found that most of the specialists do not routinely use the reply letter. In general, the specialists have their own criteria to reply or not to a referral, for example, whether the generalist "deserved" a reply, whether the patient will remain under their care, or whether the patient has a significant pathology. Reasons for low reply letter utilization rates among specialists were categorized into four broad themes: (1) workplace organization, (2) communication inhibition, (3) professional isolation, and (4) medical education.
CONCLUSIONS: Despite a general understanding of its importance, specialists in Camaragibe had difficulty adopting the reply letter as a form of communication at the interface between primary and secondary care levels. The reply letter can be effective in improving communication at the interface, but the broad cultural, historical, and organizational features of secondary level specialists need to be considered in order to improve reply letter utilization.

Key words: Referral and consultation; physicians, family; correspondence; interprofessional relations; national health programs; Brazil.

RESUMEN

OBJETIVOS: Identificar los obstáculos que impiden a los especialistas del nivel secundario de salud utilizar las cartas de respuesta en el municipio de Camaragibe, Pernambuco, Brasil, y presentar algunas posibles soluciones para mejorar la comunicación en la interfase entre los especialistas del nivel secundario y los de atención primaria.
MÉTODOS: Para este estudio cualitativo se realizaron entrevistas semiestructuradas en 2003 a diez especialistas de nivel secundario de salud, seleccionados aleatoriamente, de las siguientes especialidades: neurología, obstetricia, ginecología, psiquiatría, otorrinolaringología y colposcopia. Las entrevistas se realizaron mediante una guía basada en las siguientes categorías: 1) perfil del especialista; 2) descripción de los servicios clínicos ambulatorios; 3) satisfacción profesional; 4) evaluación de los servicios de salud municipales; 5) evaluación del Programa de Salud Familiar (Programa de Saúde da Familia); 6) descripción del proceso de referencia de pacientes; y 7) uso de las cartas de respuesta.
RESULTADOS: La mayor parte de los especialistas comprendían la importancia de las cartas de respuesta y que esta forma de comunicación con los médicos generales podría garantizar la continuidad de la atención, evitaría la duplicación de esfuerzos o la atención desorganizada del paciente, y brindaría al médico general el respaldo profesional necesario. No obstante, según el estudio, la mayoría de los especialistas no utilizan las cartas de respuesta habitualmente. En general, los especialistas tienen sus propios criterios para responder o no a las notas de referencia, por ejemplo, si el médico general "merece" una respuesta, si el paciente permanecerá bajo su cuidado o si el paciente tiene una enfermedad que lo justifica. Las razones para la baja tasa de utilización de las cartas de respuesta por parte de los especialistas se clasificaron en los siguientes grandes temas: 1) organización del trabajo; 2) cohibición a la comunicación; 3) aislamiento profesional; y 4) educación médica.
CONCLUSIONES: A pesar de la comprensión generalizada de su importancia, los especialistas de Camaragibe tienen dificultades para adoptar las cartas de respuesta como forma de comunicación en la interfase entre los niveles de atención primaria y secundaria. Las cartas de respuesta pueden ser un medio eficaz para mejorar la comunicación en esa interfase, pero para mejorar el grado de utilización de las cartas de respuesta se deben tomar en cuenta las características generales de índole cultural, histórica y organizativa de los especialistas del nivel secundario de salud.

Palabras clave: Remisión y consulta, médicos de familia, correspondencia, relaciones interprofesionales, programas nacionales de salud, Brasil.

The use of reply letters to improve communication at the interface between secondary level specialist care and primary care can impact positively on health system efficiency and ensure continuity of patient care. Brazil introduced the Family Health Program (FHP) (Programa Saúde da Família) in the 1990s to provide universal community-based primary care at the municipal level. However, there has been difficulty ensuring the functional integration of these new services with the existing secondary level specialist services. In particular, communication at the interface has been fragile, creating difficulties in patient management and lack of continuity of care. The municipality of Camaragibe, Pernambuco, Brazil introduced the primary care services of the FHP in 1992. A standardized referral and reply pro forma was introduced to encourage specialists to write a reply letter to generalists. However, reply rates have remained unchanged and communication remains poor.

Referral systems

Ambulatory services account for a considerable expenditure both by governments and by the patients themselves (1) and are central components of health systems, which need to be designed as efficiently as possible. Recently, the need to understand how to strengthen health systems, including ambulatory services, has grown in importance (2, 3). Efficiency of ambulatory services is seen to be improved through a referral system that bridges the interface between primary, generalist care and secondary, specialist care (4, 5). A referral system divides responsibilities appropriately and avoids duplication of efforts, thus avoiding the overtreatment and over-investigation characteristic of fragmented and uncoordinated ambulatory services (6, 7). Efficient referral systems ensure continuity of care across the primary/secondary interface and provide the patient with a sense of seamless progress through the health system (8, 9). However, despite consensus that referral systems are needed, the nature of referral systems, including how information flows between the two health care levels, has received little attention.

Although referral systems make empirical sense, there are three underlying assumptions that underpin efforts to improve health system efficiency (10) that should be recognized: (1) the combined efforts of primary care and secondary specialist care in a coordinated manner will bring about better results than if functioning individually, (2) referral linkages actually bring about health benefits, and (3) the case mix and abilities of primary and secondary care are distinct and complementary. Given the preceding assumptions, health systems experts and the World Health Organization agree that referral systems require a cooperative, noncompetitive relationship to exist between the primary and secondary care levels, with a bidirectional flow of information obtained through a unified records system (1116). Difficulties arise in the implementation. How can functional integration of the two care levels be measured? How can referral systems be adapted to local needs? Who is responsible for ensuring communication at the interface?

The role of the first referral level came under scrutiny in the 1980s (4). A variety of country experiences have shown that the drive towards hospital medicine has undermined the coordinating role of primary health care. There were, however, positive experiences of specialists supporting generalists through joint consultations, close supervision, and the use of reply letters. Reply letters, which are used to communicate details of patient care between medical practitioners, brought educational opportunities for the generalists, improved referral patterns, and destabilized the existing power relationships that had historically favored specialists.

Recently, there have been attempts to introduce innovative communication methods at the primary and secondary care interface, specifically in Australia, Honduras, Nepal, and the United States of America (1724). However, communication between primary and secondary care levels, particularly in Latin America, is in the form of written referrals and written replies. The literature places a heavy emphasis on referral patterns and methods to improve generalist letter-writing, whereas the written reply from the specialist to the generalist has received comparatively little attention. However, the limited research has shown that generalists perceive the written reply letter as a useful form of continuing education, because it is directly related to their clinical work (25, 26). The reply letter ensures continuity of medical care between health care levels (8, 9), and can, when sent to the patient, improve patient satisfaction significantly (2729). There is consensus among medical practitioners as to what information should be included in the reply letter (30). However, experience has shown that there are several factors that inhibit the use of the reply letter, such as time constraints (31). Consequently, reply letters are underused (15, 18, 3133) and contain insufficient information (32, 34).

In some countries, specialists routinely reply to referrals from a generalist. Experience in Scandinavia has shown very high reply rates, of up to 90% (35). General practitioners in the United Kingdom, accustomed to receiving written replies from specialists, consider a delay of no more than 3.4 days to be acceptable (12). However, in most countries, particularly in the developing world, if a referral system exists, the expectation of a generalist receiving a reply is not high. For referral systems to function effectively, there is a need to identify and understand the mechanisms underlying the obstacles that impede communication between specialist and generalist.

Brazilian health sector reform

The rewriting of the Brazilian Constitution in 1988 made the provision of health care the duty of the State and the right of all people and was, after years of oppressive military rule, the key event that heralded the creation of the Unified Health System (UHS) (Sistema Único de Saúde). Progressive municipalization of health care services, the creation of a hierarchical system of health care more responsive to local needs, community participation, and intersectoral approaches to health problems were mechanisms by which the new Unified Health System would improve the health of the Brazilian people.

The fragmented, hospital-based health care system that flourished under the military regime was to be replaced by the family-oriented, community-based, comprehensive, primary health care strategy of the FHP. The FHP was to function in health centers and in the home, working at the disease causality level. FHP goals included humanizing health practices, stimulating social control, and working in an integrated manner with other sectors. The FHP, through its equitable distribution throughout the country and its community participation activities, would democratize health services. Furthermore, as the first point of contact for patients requiring medical care, it would help to organize health services in a hierarchical manner across the existing secondary and tertiary services (3638).

FHP teams, located in defined territorial microregions of between 2 500 to 4 000 people, are composed of a medical generalist, a nurse, an auxiliary nurse, and several community health agents. Teams live in or near the designated territories and are responsible for resolving around 90% of basic health problems of the community, through ambulatory care, vaccination programs, outreach work, and health promotion activities. In 1992, the FHP began experimentally in several municipalities in the Northeast of Brazil, but has since expanded exponentially throughout the country. Today, for approximately 45% of the population, all basic health services are now provided through FHP teams.

Unable to bear the burden of all local health needs, complex clinical cases are referred to local secondary level specialists. An effective referral and reply system was recognized as important in order for the FHP to function in close integration with the existing higher service levels (39). The Brazilian Medical Council provides for this through the Brazilian Code of Medical Ethics (5th edition [2000], chapter 5, article 71): "It is forbidden to fail to reply to a referring physician regarding the referral or transfer of care, when this reply has been solicited." However, guidelines and rules regarding the implementation of the referral system have been unclear. Municipalities are responsible for ensuring access to health services for all their inhabitants; however, state-level health authorities determine service distribution and important financing regulations (40).

Despite attempts to transform health service provision into an integrated, hierarchical system, this lack of coordination has resulted in a service that has been described as ". . . a conglomeration of more or less autonomous establishments, instead of a network of services . . ." (41). The FHP exposed an unmet health care demand that it was unable to resolve without the integrated support of the higher health service levels (42). Left isolated by a persistent curative paradigm characteristic of the secondary and tertiary levels, the FHP as a strategy finds itself at risk of being enshrined as a ". . . poor health system, for the poor . . ." (43, 44). A Ministry of Health evaluation of FHP implementation in 10 large urban centers identified serious deficiencies in the integration of the FHP with existing secondary level servicesa concern given the extent with which the FHP has expanded throughout the country. All municipalities studied (including Camaragibe) demonstrated inadequate referral systems, with particular regard to the use of reply letters (39). Recommendations included constructing appointment centers to facilitate consultation bookings, developing consultation protocols, and stimulating secondary level specialists to use the reply letter in response to a referral. However, resistances were identified on the part of the medical community, and the report acknowledged the need for continuous negotiation in order to facilitate the adoption of the new paradigm.

In an analysis of pediatric care in the state of Pernambuco (where Camaragibe is located), Samico (45) identified deficiencies in the communication between FHP generalists and secondary level specialists, which caused a breakdown in the referral system. It was recommended that the referral system should be organized with the participation of specialists, generalists, and the patient community.

Although no formal study has investigated the reasons for the difficulty in integrating the FHP with secondary services, obstacles at the secondary care level have been cited (39, 44). Impediments at the secondary care level, which may include the perceptions and practices of secondary level specialists, have to date received little attention, not only in Brazil but also in the international arena.

The municipality of Camaragibe

The municipality of Camaragibe, located 16 km northwest of Recife (the capital city of the state of Pernambuco), is a useful choice to investigate the attitudes of secondary level specialists. It has enjoyed consistent political administration and leadership over the last three successive terms of office and was one of the first municipalities to implement the FHP in 1992. A municipality considered to have high institutional and administrative capacity, it was also one of the first municipalities to receive full municipalization status, with complete responsibility for health care administration, decentralized to the municipal health authority (39, 46). This municipality has received accolades from the World Bank and the United Nations Children's Fund (UNICEF) for reducing infant mortality from 65.0 per 1 000 live births in 1994 to 17.0 per 1 000 live births in 1997, through the implementation of the basic health care services of the FHP, high profile community participation, and social mobilization (47).

Camaragibe covers an area of 53 km2, has a population of 128 627, and a Human Development Index of 0.585 (48). It has a poor infrastructure, with sewage disposal services provided for 21% of the population, garbage collection provided for 72% of the population, and running water in 67% of households. It is organized into five health territories, with 37 FHP teams serving almost 100% of the population.

FHP generalists are able to resolve, on average, 90% of the basic health problems of the community. Complex clinical cases, in need of specialist advice or management, are referred to local specialists. Specialist consultation is obtained only through referrals from FHP generalists. Referrals are made using a standard referral and reply pro forma (first introduced in 1999) that has a clearly defined section for the specialist to reply to the referring generalist after the consultation has been made (Appendix 1). Patients book appointments at the municipal appointment center, where one copy of the referral form is kept, and the other is given to the specialist at the time of consultation. The municipal referral system is composed of two first-referral specialist centers (staffed by dermatologists, cardiologists, neurologists, otolaryngologists, obstetricians, and gynecologists), a mental health unit with three psychiatrists and four community psychologists, a municipal laboratory, a women's health clinic, a walk-in HIV voluntary testing and counseling center, a low-risk maternity hospital, and a center for physiotherapy.

Despite its evolved array of health care services, studies of maternal and child health care within Camaragibe have shown important deficiencies in the referral system, in particular the noncompliance of secondary level specialists in the use of a reply letter (49, 50). Specialists at first-referral hospitals and at the specialist outpatient centers were performing tasks best suited for the primary health care level, and there was little or no interaction between the two levels (40, 49, 50). Only 4.5% of FHP generalists interviewed agreed that the referral system in Camaragibe was adequate. The rates of reply from specialists have, despite the introduction of the pro forma, generally remained unchanged. Despite almost total coverage of the municipality with FHP teams and almost 10 years since implementation of the first teams, primary health care services continue to function in isolation from the specialistsreferring patients, yet receiving no information in return regarding their management.

The FHP's role as gatekeeper into the Unified Health System can only be fully achieved if a bidirectional flow of information exists with FHP counterparts at the secondary level. Continuity of care can only be maintained if FHP generalists are fully informed as to the management of their referred patients. Given the drive to establish all primary health care in Brazil through the FHP, poor functional integration with existing secondary services will result in two parallel health systems that undermine the specific objectives laid out in the Constitution. Although the government efforts to improve this communication, such as appointment centers and referral forms, have already been implemented in Camaragibe, there has been little improvement in communication. Understanding the difficulties or obstacles confronted by specialists in working under this new paradigm may go some way towards overcoming these difficulties.

Study objectives

This qualitative study aims to identify obstacles that interfere with the use of the reply letter by secondary level specialists in the municipality of Camaragibe, to identify possible solutions for establishing a bidirectional flow of information between the two health care levels, and to improve referral system efficiency.

METHODOLOGY

This research study was done between September and December 2003. To obtain a representative sample of specialists, two to three specialists were randomly selected from each of the following secondary care referral centers in Camaragibe: the mental health unit, specialist outpatient centers, and the women's health center. Pediatricians, general physicians, physiotherapists, occupational therapists, nutritionists, and speech therapists were excluded from the sample because FHP generalists rarely refer patients to them.

Specialists were contacted by phone to arrange a date and time for an interview, which would be carried out in their office, in order to put them more at ease. FHP nurses, who were not involved in the referral process, conducted the interviews. Interviews were semistructured, using an interview guide (Appendix 2), based on the following seven categories: (1) profile of the specialist, (2) description of outpatient clinic services, (3) professional satisfaction, (4) assessment of the municipal health services, (5) assessment of the Family Health Program, (6) description of the referral process, and (7) reply letter utilization. Guides were modified as the interviews progressed to take into account themes that had been revealed in previous interviews.

Informed, written consent was obtained and confidentiality was guaranteed. Interviews were tape-recorded and immediately transcribed by the interviewers, including their own personal reflections of the interview dynamics. Debriefing discussions were also provided to expand on the interviewers' reflections of the process.

The texts were read exhaustively, and manual content analysis was performed using the categories mentioned above to identify common themes that may reveal barriers to reply letter utilization.

RESULTS

Twelve secondary level specialists were randomly selected, but two declined or were repeatedly unavailable for interview. Ten specialists, or 30% of the total number of specialists in Camaragibe, consented to interview. The following specialties were represented: neurology, obstetrics, gynecology, psychiatry, otolaryngology, and colposcopy. To protect confidentiality, the data, where obtainable, have been summarized as a group.

The average age of the specialists interviewed was 45.7 years, and 70% of the interviewees were male. The average number of years of employment in Camaragibe for the specialists interviewed was 8.2 years, typically in repeat temporary contracts. On average, the specialists interviewed were employed at four additional locations outside of Camaragibe, including public hospital emergency departments, private clinics, and other public sector municipal clinics. The average salary (from all employment locations) for the interviewees was US$26 010 per year, which included US$ 4 692 from employment in Camaragibe.

Two of the interviewees did not want their interviews tape-recorded, and monitored what the interviewers could and could not document in their notes. Although this was limiting for the interviewers, the data from these interviews were still sufficient for the analysis.

Although two specialists had difficulty understanding the importance of the reply letter and the circularity of health care within the municipality, the majority of specialists knew that providing the letter would be of benefit to the referring doctor. One interviewee stated, "It is really necessary, it is fundamental [to write a reply.] The management that we perform needs to be continued by the folks at the basic health post, but they can't do that if they don't know what to look out for." In another interview, a specialist said, "It is important that there is a good connection between the professionals for continuity of care . . . otherwise if there is no information it gets all confused; the management of the patient is improved with this exchange."

Furthermore, the specialists understood that the referring doctor would benefit personally from the communication. One specialist mentioned, "In my opinion, at least I think that they [the FHP generalists] will learn a little more . . . something that for me is trivial, simple" Another specialists stated, "It puts the referring doctor more at ease . . . without it, you'll frustrate him and he won't feel able to manage the patient."

The impression that the specialists had of the FHP generalists and the program in general was also positive. Specialists, although varying in their understanding of this federal strategy, unanimously agreed that in terms of improving the workload at the secondary care level, the FHP had brought considerable benefits, through triaging complex cases and through continued care at the local level. Most of the specialists understood that communication at the interface would bring educational opportunities for the generalist, ensure continuity of care, avoid duplication of efforts, and would provide the generalist with the professional support required.

One interviewee said, "The FHP is excellent. The patients come already triaged, organized, with a diagnosis. This is a big help. It if wasn't for the FHP, I reckon this clinic would be bursting at the seams." Another specialist said, "It really has an influence! The big help is through the triage that the FHP generalists do. I feel more relaxed knowing that the patient is looked after in the FHP . . . they [the patients] always show up on the right day [for the specialist consultation.]"

Despite this positive view of the FHP generalists and the program in general, the specialists were not routinely providing reply letters. With the exception of the colposcopist (where the result of a colposcopy is automatically sent to the referring generalist), the other specialists were replying in a sporadic and inconsistent manner.

Specialists appeared to use a variety of their own criteria to reply or not to a referral. One interviewee stated, "I don't send a reply because when they send me a patient it is because I am to sort out the problem. My reply is just to send back the patient well. I am the 'concluder,' if you like, in the final stage of the process. Although, now that you mention it, it would be useful to let the doctors know what I did."

Other specialists commented that they would send replies only if they found a serious pathology in the patient or if they judged the patient to have been well referred. One interviewee stated: "I'll send a reply if I think the patient was well-referred, that the doctor thought carefully about the case . . . there are professionals, and then there are professionals, and they can be different. If the doctor deserves it then I give more, but those that I think don't know what they are doing, well, then I don't reply. I'm quite pragmatic."

Some specialists reflected on their position in the referral system. There was some confusion regarding when a reply letter would be best used. One interviewee said, "If I am to continue the management of the patient, then I don't reply, but if I need to send the patient back to the referring doctor, then I'll write him a letter."

Other specialists found that the reply letter itself was not necessary. One specialist stated, "I'll write to another specialist if he requests some specific information, but generally I don't write back; I don't find it necessary to write back if the case is pretty uncomplicated." Another specialist said, "The patient often comes accompanied by a community health agent, so I often just explain to the agent what the situation is and she passes it back to the doctor."

One specialist was particularly unclear as to the value of the reply, stating, "I never send replies . . . I have no idea what use they would be." However, another specialist was prepared to use the reply as a form of continuing education, albeit sporadically, stating, "If I find that the doctor really didn't have a clue, well, then I'll send a little note back on the referral form . . . I reckon I reply in about 30% of cases."

The sporadic and inconsistent manner in which specialists were in reality replying appears to be due to a variety of obstacles, related to extrinsic factors that they understood to be out of their control, but also due to intrinsic ones. There were several distinct, yet interrelated, themes that emerged from the interviews, which might contribute towards an understanding of the obstacles that specialists confront in writing reply letters. These can be grouped into the following four headings: (1) workplace organization, (2) communication inhibition, (3) professional isolation, and (4) medical education.

These categories were not necessarily present in each interview, but are broad themes that appeared throughout, which provide an understanding of the views and practices of the specialists with regard to the use of the reply letter.

Workplace organization

Two to three specialists were randomly selected from each of the following secondary care referral centers in Camaragibe: the mental heath unit, the specialist outpatient centers, and the women's health center. Specific organizational issues varied from one location to another, but some overall themes were seen among the specialists. Although the specialists interviewed were dissatisfied with their salaries from their employment in Camaragibe (and from other employment sources), many specialists chose to work in Camaragibe because of its reputation as a well-organized municipality that pays salaries regularly. The specialists' annual salaries (including income from employment outside of Camaragibe) are about twice as much as the salaries of FHP generalists.

Specialists work under temporary contracts and do not work with the same degree of formal job security as FHP generalists, who are contracted on a full-time basis and enjoy certain benefits. Despite this, the specialists' contracts are frequently renewed, and the majority of specialists have worked in the municipality for at least four years, and some for well over 15 years. All specialists, including those in this study, work in the municipality on a part-time basis, providing outpatient services two to three times per week. Each specialist sees between 15 and 20 patients per session, depending on the cases and the specialty. There was a general sense that the specialists felt overworked and underappreciated.

Although specialists noted many positive changes in the organization of the workplace since the FHP was implemented, such as the introduction of the appointment center and improved equipment in some clinics, a reason for not providing reply letters with more frequency was, in several instances, explained by lack of time and resources. One interviewee stated, "I just don't get the time to write. There are too many patients." Another complained, "Ahh, if I was to stop and write a letter after each patient, well, I'd be here all day!"

There were complaints of lack of material resources, such as carbon paper to fill in the reply pro forma. However, one specialist was completely unaware that the referral and reply pro forma is to be used as a reply letter. One stated, "I think there should be a special form that can be used to write a reply, but no one has ever told me, 'Look, the doctor referred you a patient and you never sent him back, the patient has kind of disappeared.'" This may be due to inattention, or a missed opportunity by the municipal health authority to set out the guidelines for specialists to use the reply form.

Communication inhibition

Several specialists said that they felt inhibited to communicate with FHP generalists when they did not have a professional or personal relationship with them. The reply letter was not seen as a tool to overcome this lack of familiarity. In fact, several specialists suggested telephone replies, as opposed to written ones. The more personal nature of a telephone reply was perceived to be advantageous. The specialists maintained that in some cases, it is the individual professional's responsibility to bridge the communication gap, and in other cases it is partly the responsibility of the municipal health authority, as the overall coordinator of health services. However, as one interviewee observed, no amount of organizational innovation will create good working relationships. Supportive and communicative relationships are built through natural processes and are not planned. Where prior familiarity existed, communication was greatly facilitated.

One interviewee noted: "When he sends me a patient, he calls me. Then I call him and I say, 'Look, I think this, this, and this.' Because we had a closeness from before, it is easier for us to talk. Doctors that don't know each other, they get kind of inhibited . . . and there is no health care model that can make people integrate and work together . . . either you get along or you don't."

Realizing that this communication rests on familiarity with colleagues, one specialist complained that his proposal to improve relationships had not been acted on by the municipality, stating, "I already put this proposal forward, but I don't know if it is up and running for this year or not. We [specialists and FHP generalists] need to get to know each other."

One specialist openly rejected the possibility of replying. Pointing to the rubbish bin and laughing, he said, "That piece of paper with the referral note? Ahhh, look where I file that!"

Professional isolation

Specialists appear to function in isolation within the municipality. The temporary nature of contractual obligations, the excessive number of alternative outside employment commitments, and the individualistic approach that has characterized specialist training all lend to a disease-focused management that does not consider the circularity and integrity of the municipal health care system. Specialists come, treat, and leave. Their services are employed for a specific purpose and they do not seem to feel like part of the health system. One interviewee noted, "I lend my services. I just come and do my consulting. My function is . . . just technical . . . the rest, the others can sort out."

Further probing revealed a desire by some specialists to be part of the system. Specialists complained of not having had any opportunity to meet generalists to discuss clinical cases; some specialists felt that overcoming this isolation, to an extent, is the municipal health authority's responsibility. A distinct pride at being part of a functioning (and nationally renowned) municipality was evident in some interviews. In others, however, there was a certain resignation to the fact that specialists' work has always been and is supposed to be isolated and fragmented.

Specialists had not been invited to participate in deliberations on the municipal health system organization. One specialist commented, "No, at least as far as I am concerned, I was never invited [to an orientation]. Truth is, I don't understand much about this district health care business. In fact I don't like the whole public health thing at all . . . too many meetings and not enough action." This lack of involvement may be due to little opportunity to familiarize themselves with the organization of the municipality. At the recruitment stage, no explanation or initiation had been provided for specialists to familiarize themselves with the organization of the municipality. One interviewee stated: "Organization of the health care in the municipality? Health policy in the municipality? No, no, no. Nothing was ever explained to me. I went and looked it up on the Internet, and found out it [the FHP] was some sort of a federal program, but I have to say that really when it comes to it, I am a layperson."

Unlike FHP generalists, who meet as a collective entity at least once per month to discuss organizational issues, and have a representative committee with direct access to the municipal secretary of health, the specialists have neither formal representation nor an arena for discussion. They are individual professionals contracted to serve a specific function. However, to an extent, specialists also lacked the drive to participate in municipal health issues. Simple lack of interest in the collective nature of health care in the municipality appears to be another barrier to involvement in municipal health care.

Medical education

Much of what has been observedprofessional isolation, difficulties communicating with colleagues, and inefficiencies in time managementis rooted in the way doctors are trained in Brazil. Heavy emphasis on the specialties throughout medical school and the drive to earn money in the private sector through sub-specialization have created a longstanding tradition of multi-employment and fragmented patient care. For many specialists, work in the public sector is seen as a necessary, but inferior, part of their career, and the ultimate concern is to build a private practice.

Some specialists spontaneously commented that they simply had not been trained to consider the interconnectedness of the primary and the secondary health care levels. Medical training often does not emphasize the public health aspect of medicine, and specialists may not understand their role in the municipal health system.

Some specialists recognize that it is a personal responsibility to seek improved integration. One interviewee stated: "We were trained in an environment that is purely curative, with no public health or collective basis whatsoever, so . . . even if the person has the best intentions, he just wasn't trained this way . . . so it is a personal effort, because at school or in the academic circles this vision isn't there yet."

Another specialist commented that the universities have a greater responsibility to integrate public health matters, such as referral systems, into the medical curriculum. The specialist stated: "I think the universities have to integrate the 'social' bit into the course. They take the social medicine and put it in as an isolated course . . . the student comes out really technical, really dogmatic, wanting to know more about the technical [aspects of medicine] at all costs. But the problem most of the time comes down to social issues . . . the student needs to come out of medical school with a more open mind."

However, another specialist noted that individual work ethic is an important consideration in their role within the municipal public sector. He commented: "From what I hear, there are those who see a couple of patients, and they're already very tired. But when they get into their private consulting room, they'll see upwards of twenty. In the teaching hospital though, they can only manage a couple. That's the kind of thing that people do . . . they pretend to work, and the students pretend that they are learning. They just want to get their diploma, so they can get their fancy notepaper and then they might teach at congresses, and everybody will look at them."

Specialists appeared to understand the importance of communication at the interface. They also offered quite positive opinions of the FHP generalists and their contribution towards streamlining the specialists' own services. Dissatisfaction with the manner or quality with which patients were referred was, with only one or two exceptions, not a prominent feature and did not seem to explain such low rates of reply. So how can we explain such low rates of reply? These interviews begin to reveal that specialists require a high degree of familiarity with referring doctors, a more proactive involvement with municipal health system organization, and that they feel, through their medical training, ill-prepared to consider the circularity of health care.

Limitations of the study

Although 30% of the specialists were selected for an interview, the sample size was reduced considerably by the difficulty with which interviewers had to arrange convenient interview times with the subjects. Two of the subjects were repeatedly unable to be interviewed over a two-month period, and this restricted the already small sample size even further. Finally, the study had a sample size of just 10 subjects. Although some insights were obtained into the attitudes of specialists at the secondary care level, a deeper and broader understanding might have been obtained with a larger sample. In addition, utilizing different qualitative research techniques, including focus groups and in-depth interviews with key informants, such as the municipal secretary of health, would have been extremely useful. Unfortunately, limited by time and resources, the authors were unable to employ these additional research techniques.

Although an effort was made to minimize a possible courtesy bias, by using nurse practitioners not involved in the referral process to be interviewers, it is still possible that a courtesy bias was prominent. Informed at the start of the interview that the interviewers were part of the FHP, the specialists may have felt a pressure to portray nothing other than a positive outlook on the program. The somewhat surprisingly positive outlook of the FHP may have been due to this and responses perhaps should be interpreted in this light.

In addition, due to their own work commitments, it was necessary to use four different FHP nurse practitioners for the interviews. Differences in interview style among the nurses may have affected the output from the specialists. On two occasions, interviews were rendered almost completely unusable by a poor interaction between subject and interviewer. Although this may not be because of interview style and another interviewer may have obtained a similar result, it limited the depth of the study considerably, to just 10 interviews.

To standardize the style among the different interviewers, an interview guide was developed and tested with the nurses. This enabled them to familiarize themselves with the timing and style of the interview and perform the interview in as consistent a manner as possible. Nonetheless, the dynamics between subject and interviewer cannot be accounted for in a guide and may have varied from interviewer to interviewer. Where possible, this was analyzed by feedback discussions between the author and the interviewer.

Finally, the lead author was employed as an FHP generalist from 2000 to 2003 in Camaragibe. Although every attempt has been made to analyze the interviews as objectively as possible, it is important to be aware that as a generalist working with the lack of communication at the interface in Camaragibe, the author may have been influenced by negative experiences.

DISCUSSION

This qualitative study has explored some operational barriers that are interfering with the reply process from the secondary level to the primary health care level in Camaragibe. The use of semistructured interviews provided a window of understanding into the opinions and practices of the secondary level specialists. To date, this study was the first attempt to investigate specialists' views of the referral process in Brazil.

The most apparent feature was that the specialists do not use the reply letter as a matter of routine. In 2002, FHP generalists in Camaragibe referred approximately 10 000 patients to secondary level specialists, but only one out of the 10 specialists who were interviewed in this study was routinely using the reply pro forma as a reply letter to the referring generalists. In general, the specialists have their own criteria to reply or not to a referral, such as whether the generalist "deserved" a reply, whether the patient will remain under their care, or whether the patient has a significant pathology. Replying to a referral should be a routine process irrespective of clinical case or stage of management. In our study, at best, written replies are provided sporadically.

Initially, this study was motivated by the belief that specialists were not yet familiar with the principles of the Family Health Program. Furthermore, it was believed that specialists in Camaragibe were resistant to the use of the reply letter because they might devalue the work of FHP generalists. However, the interviews revealed that all the specialists had some understanding of the FHP strategy. Although there were diverse opinions as to its effectiveness and the way in which it has been implemented, in general, specialists demonstrated enthusiasm for its existence. All of the specialists were able to describe how the patients were referred to the second level of care and, with only one exception, gave positive remarks concerning the usefulness that the FHP has had, not only on the municipal health system, but also on the specialists' own work practices. Neither a lack of understanding as to the origin of the referral nor its value to their work practices explained why a reply was not given routinely. Although the specialists seemed to understand and value the FHP, they still lacked a general understanding of the overall Unified Health System.

Barriers in the organization of the workplace were reasons given by some specialists for not providing a reply letter. More precisely, lack of time or materials was mentioned. Although excessive workloads and lack of resources are features of most developing country health systems, analysis within a broader context is necessary before issues of workplace organization are considered as reasons for not providing a reply letter. However, since specialists complained of lack of time to provide a reply letter, multi-employment could have been partly responsible.

In her analysis of the Brazilian medical profession, Machado (51) noted that nearly 60% of doctors have more than three different professional activities, whether in the public or private sector. Difficulties coping with this workload were cited by nearly 85% of doctors that admitted to having this number of job commitments. Commitment to service in the municipal health system may be weakened due to external obligations that specialists may prioritize, based on better pay or prestige. This is reflected in the historical differences in the quality of care provided in the public and the private sectors in Brazil (44).

The drive towards multi-employment is rooted in the market-oriented overspecialization that has grown more or less freely over the last few decades in Brazil (52). As the number of specialists increases, competition for market niches also increases, and specialists engage in a variety of positions (public sector, private clinic, teaching) to complement their standard of living. At no point has there been an attempt to plan or manage this growth in specialties (53). On the contrary, there has been a steady growth of specialist residency programs, doubling in number from 1985 to 1996 (54). The FHP strategy and the Unified Health System as a whole, however, are founded on low-cost, low technology care best provided by medical generalists. However, only 1.8% of doctors in Brazil are medical generalists (51). This statistic is far removed from the recommendation by the World Health Organization to have three generalists for every specialist.

The Brazilian National Commission for Medical Residencies, which coordinates and regulates medical training in universities throughout the country, has been criticized for stimulating this imbalance. Not only has it been accused of being composed of members who represent the medical elite, but it has also been criticized for having unclear policies on human resources that are out of context with the current national scenario (54).

Competing for an ever-decreasing share of the market, specialists are obliged to seek employment in a variety of locations to increase their earning potential. There have been calls to improve public sector salary structures, incentivize doctors to prioritize their public sector commitments, and reduce the drive to multi-employment (44). However, studies in other Lusophone countries have shown that the culture of combining sources of income is so ingrained that it would remain even if public sector salaries were higher (55).

The drive to specialize begins early in medical school education, and this has been criticized by a number of authors (5660). Brazilian medical education is rooted in the traditional Flexnerian approach, which focuses on biological determinants of disease. Medical students are taught exclusively in the tertiary hospital system, and they are not encouraged to explore the possibility of general practice. The result is fragmented care, with doctors working in spatial and temporal isolation (61), completely dissociated from the epidemiological reality of the population. One interviewee considered himself to have a "concluder" role, working at the end of the line and in isolation. This suggests that specialists do not consider themselves to be accountable to colleagues and patients for their clinical decisions, and they feel as though they operate outside of a system.

Specialists in this study regarded their medical training as a limiting factor in being able to integrate into the new paradigm. They felt ill prepared to adopt the new health system paradigm. In some countries, alternative training programs that place greater emphasis on community medicine and general practice have been attempted, with some degree of success (62). In Brazil, an entire reformulation of the pedagogic, traditional teaching hospital model has been advocated (60) in order to bring medical school education in line with the current health sector reforms. It has been recommended that the ministries of health and education be better integrated into medical school training (58, 63). However, in Brazil, there has been resistance to changes in the medical curriculum, particularly from the corporate medical associations (63). This resistance has hindered the effective transformation of the health care system, despite impressive new legislation and administrative reforms.

The lack of personal contact with the FHP generalists was cited as an inhibitory factor for some specialists. One interviewee commented that there was a feeling of being "inhibited" to communicate with a professional with whom there is no personal or professional relationship. To a certain extent, this observation is related to the function of the municipal health authority as mediator and coordinator of the two health care levels. Considerable investment has been made in the integration and continuing medical education of FHP professionals. In part, this is because the FHP is a high-profile nationwide political reform that needs to attract medical personnel.

In contrast, secondary level specialists frequently operate in isolation. They lack the collective identity of FHP generalists, who are driven by a sense of belonging to a nationwide political and social movement.

Similar efforts on the part of the municipal health authority are needed to integrate the specialists into the new paradigm. However, the legislature and guidelines are unclear as to how either the state or the municipality should ensure this integration. According to the legislature, the municipality is immediately responsible to meet the health needs and demands of its people. Meeting these needs is a necessary condition for the full implementation of the Unified Health System (46). However, municipalities vary in their abilities to perform this function. Camaragibe is one of the more experienced and organized municipalities, with a strong political will to implement the health care reforms. However, not even this municipality has been able to fully integrate the FHP with its secondary level specialists.

As previous authors have noted (46, 64, 65), the FHP was conceptualized in a top-down, centralized manner. Araujo (66) notes that municipalities have not achieved true political or financial autonomy. Contrary to the aim of decentralization, as the FHP has expanded, there has been a progressive devaluing of local autonomy. Local and national obstacles have interfered with the decentralization process. As a result, the responsibility of coordinating the interface between the primary and secondary levels has become blurred.

When the municipal health authority obliges a specialist to use the reply letter, as was the case with the colposcopist in this study, replies are provided as a matter of routine. However, none of the remaining specialists used the reply letter as a matter of routine, even though the use of a reply letter is emphasized in the Brazilian Code of Medical Ethics, the legal directives of the Unified Health System, and even the Brazilian Constitution. It would appear, therefore, that the obligatory sense of the reply letter is not dependent on whether it has ethical or legal importance, but is very much a question of the degree to which the local municipality makes the reply process a necessary part of the job. It would take little in the way of material and human resources for the municipal administration to enforce the reply letter in the same way that has been done for the colposcopy service.

However, the specialists' "inhibition" to communicate with FHP generalists cannot be overcome only by improved municipal enforcement alone. It is also important to understand aspects of local and national political culture that are common in Brazil, such as informal patron relationships and favor exchanges. In this study, these were clearly important factors to consider in optimizing communication at the interface.

In Atkinson's (67) analysis of the role of local political culture in implementing health reforms in the Northeast of Brazil, she notes that decentralization, in the long term, is little more than a legal status. Account of the local organizational arrangements needs to be made in order to ensure proper implementation. Exchanges of favors for political support, the strong patron-client relationship, and the jeitinho1 (68) are characteristics of Brazilian society that require personal, interactive, and relational processes. As one interviewee commented, a prior "knowledge" of a colleague makes replying that much easier and almost expected. The "favor" of replying is not considered communication across an interface, but the expected mutual back-scratching of close colleagues. Although sporadic, unorthodox, unregulated, and inefficient at the health system perspective, for the two parties concerned, replying in this way is extremely effective. It does, however, break the connection that should occur in a health system. The principles of public health and collectivity are irrelevant to the specialist that perceives the communication among "known" colleagues to be satisfactory. Although it would bring obvious benefits to the system as a whole, standardizing replies through the correct use of the reply letter would be difficult to ensure. An understanding of the specialists' need to "know" a colleague and adapt communication innovations accordingly is necessary. Tools that happen to facilitate replies, such as the pro forma in other countries, may need to be entirely redesigned to be more appropriate to the local political and cultural reality.

Camaragibe introduced the reply letter pro forma as a bold attempt to improve communication at the interface between the primary and secondary health care levels. This was a considerably advanced initiative for a resource-poor municipal health authority in a developing country. While it made logical sense to the health system administrators, it did not succeed in improving communication between specialists and generalists. Similar interventions involving health information management in Uganda brought valuable lessons for the implementers (69). They found it necessary to improve the definition of the intervention, and also to understand that the intervention would inevitably bring about organizational changes. Through a stakeholder analysis, they found that to implement changes in the health information management system, it is necessary to first understand the cultural issues related to status, power, and organizational conflicts. Camaragibe did not succeed in improving communication at the interface, in part, for these reasons. The intervention was introduced in a top-down prescriptive manner, paying little attention to the cultural, historical, and organizational features of specialist practice.

Communication at the interface, seamless care for patients, and continuity of care among medical professionals cannot be achieved by introducing a communication innovation in isolation, without a full understanding of the broader issues. The European Working Party on Quality in Family Practice (EQuiP) recommended that for real improvement at the interface to occur, changes are not only needed in the system of care, but also in the ways doctors view their roles and their performance. Providers need to be able to see the system from the patient perspective, the system perspective, the provider perspective, and the medical quality perspective (70). EQuiP prepared a list of 10 key targets that need to be in place in order for the interface between primary and secondary care to function smoothly: (1) to facilitate discussions, address barriers, and distribute tasks; (2) to stimulate communication between professionals and create a dialogue with patients; (3) to create a shift from "my" patient to "our" patient; (4) to describe quality problems at the interface; (5) to create a system that understands patient flows and patient journeys, thereby improving the patients' experiences; (6) to make educational and quality information easily accessible to the public; (7) to train specialists to understand the interaction between the primary and secondary levels; (8) to facilitate team-building at the interface; (9) to develop indicators that can measure the quality of cooperation and communication, such as continuity of care; and (10) to establish an understanding of the need for cost-effectiveness.

The EQuiP guidelines may serve as a useful starting point for leaders at the municipal, state, and federal levels to develop appropriate strategies for improving continuity of care for patients.

CONCLUSION AND RECOMMENDATIONS

Despite a general understanding of its importance, specialists in Camaragibe had difficulty adopting the reply letter as a form of communication at the interface between primary and secondary care levels. Professional isolation, lack of personal contact with the FHP generalists, and excessive work commitments outside of the municipality may be responsible. These barriers may be rooted in medical school curricula that prioritize individualist, market-oriented health care. Doctors need to have a prior acquaintance in order for communication to flow freely, but opportunities for the municipal health authority to better integrate the specialists into the municipal health system were missed. In part, the role of the municipal health authority, as coordinator of communication at the interface, is unclear. The FHP, conceived at the central level, has no clearly defined strategy that can support municipalities' efforts to ensure communication at the interface.

Camaragibe took the initiative by introducing the referral and reply pro forma in order to facilitate communication. However, obstacles to communication are rooted in local and national cultural and political factors that the municipality alone cannot resolve. Nonetheless, the municipality introduced the reply letter in a prescriptive manner, paying little attention to the organizational, political, and historical features of specialist practice. At the local level, strong leadership is required to stimulate discussion between the two care levels, to develop consensus on task division, and to create an atmosphere of teamwork. However, at the national level, there is a similar need to reevaluate the medical school curricula. It is necessary for all doctors to gain a better appreciation of the ongoing health care reforms and the interconnectedness of primary and secondary care levels. This way, perhaps the traditional market-oriented approach to medicine, which has dominated Brazilian health care for decades, may eventually give way to an understanding of the circularity of care in an integrated health system. If this can be done, then ultimately patient care and continuity of patient care will improve.

Acknowledgements. We are grateful to Dr. Monica Franch, formally of the Federal University of Pernambuco, and Prof. Ruairi Brugha, formally of the London School of Hygiene and Tropical Medicine, for their useful comments in the design of the study and the preparation of the manuscript. We would also like to thank the municipality of Camaragibe for its support during the study.

REFERENCES

1. Berman P. Organization of ambulatory care provision: a critical determinant of health system performance in developing countries. Bull World Health Organ. 2000;78(6):791802.

2. Walt G. WHO's World health report 2003: shaping the future depends on strengthening health systems [editorial]. BMJ. 2004;328 (7430):6.

3. World Health Organization. World health report. Geneva: WHO; 2003.

4. World Health Organization. Analysis of the support of first-referral hospitals to primary health care services, including the activities of hospitals in disease prevention and health promotion. Geneva: WHO; 1989.

5. Starfield B. Primary and specialty care interfaces: the importance of disease continuity. Br J Gen Pract. 2003;53(494):7239.

6. Sweeney B. The referral system [editorial]. BMJ. 1994;309(6963):11801.

7. Haggerty JL, Reid RJ, Freeman GK, Starfield BH, Adair CE, McKendry R. Continuity of care: a multidisciplinary review. BMJ. 2003; 327(7425):121921.

8. Preston C, Cheater F, Baker R, Hearnshaw H. Left in limbo: patients' views on care across the primary/secondary interface. Qual Health Care. 1999;8(1):1621.

9. Nielson J, Palshof T, Maniz J, Jensen A, Olesen F. Randomised controlled clinical trial of a shared care programme for newly referred cancer patients: bridging the gap between general practice and hospital. Qual Saf Health Care. 2003;12(4):26372.

10. Van Lerberghe W, de Bethune X, de Brouwere V. Hospitals in Sub-Saharan Africa: why we need more of what does not work as it should. Trop Med Int Health. 1997;2(8): 799808.

11. Collins C. Management and organization of developing health systems. Oxford: Oxford Medical Publication; 1994.

12. Marshall MN. How well do general practitioners and hospital consultants work together? A qualitative study of cooperation and conflict with the medical profession. Br J Gen Pract. 1998;48(432):137982.

13. Hansen JP, Brown SE, Sullivan RJ Jr, Muhlbaier LH. Factors related to an effective referral and consultation process. J Fam Pract. 1982;15(4):6516.

14. Murray SF, Davies S, Phiri RK, Ahmed Y. Tools for monitoring the effectiveness of district maternity referral systems. Health Policy Plan. 2001;16(4):35361.

15. Siddiqi S, Kielmann A, Khan M, Ali N, Ghaffer A, Sheikh U, et al. The effectiveness of patient referral in Pakistan. Health Policy Plan. 2001;16(2):1938.

16. World Health Organization. The hospital in rural and urban districts: report of a WHO study group on the functions of hospitals at the first referral level. Geneva: WHO;1992. (Technical Report Series 819).

17. Graham LE, Zimmerman M, Vassallo DJ, Patterson V, Swinfen P, Swinfen R, et al. Telemedicine: the way ahead for medicine in the developing world. Trop Doct. 2003;33(1): 368.

18. Ohara K, Melendez V, Vehera N, Ohi G. Study of a patient referral system in the Republic of Honduras. Health Policy Plan. 1998;13(4):43345.

19. Blankenship JC, Menapace FJ, Fox LS, Frey CM. Telephone reporting of the results of cardiac procedures: feasibility and primary care physician preferences. Am J Med. 1999; 106(5):5216.

20. Harris MF, Giles A, O'Toole BI. Communication across the divide: a trial of structured communication between general practitioners and emergency departments. Aust Fam Physician. 2002;31(2):197200.

21. Hollins J, Veitch C, Hays R. Interpractioner communication: telephone consultation between rural general practitioners and specialists. Aust J Rural Health. 2000;8(4):22731.

22. Couper ID, Henbest, RJ. The quality and relationship of referral and reply letters. The effect of introducing a pro forma letter. S Afr Med J. 1996;86(12):15402.

23. Ray S, Archbold RA, Preston S, Ranjadayalan K, Suliman A, Timmis AD. Computer-generated correspondence for patients attending an open-access chest pain clinic. J R Coll Physicians Lond. 1998;32(5):4201.

24. Wallace P, Haines A, Harrison R, Barber J, Thompson S, Jacklin P, et al. Joint teleconsultations (virtual outreach) versus standard outpatient appointments for patients referred by their general practitioner for a special- ist opinion: a randomised controlled trial. Lancet. 2002;359(9322):19618.

25. Marshall MN. Qualitative study of educational interaction between general practitioners and specialists. Br Med J. 1998;316(7129): 4425.

26. Gagliardi A. Use of referral reply letters for continuing medical education: a review. J Contin Educ Health Prof. 2002;22(4):2229.

27. Saunders N, Georglas C, Blaney SPA, Dixon H, Topham JH. Does receiving a copy of correspondence improve patients' satisfaction with their out-patient consultation? J Laryngol Otol. 2003;117(2):1269.

28. Jelley D, van Zwanenberg T. Copying general practitioner referral letters to patients: a study of patients' views. Br J Gen Pract. 2000;50 (457):6578.

29. Cowper DM, Lenton SW. Letter-writing to parents following paediatric outpatient consultation: a survey of parent and GP views. Child Care Health Dev. 1996; 22(5):30310.

30. Newton J, Eccles M, Hutchinson A. Communication between general practitioners and consultants: what should their letters contain? Br Med J. 1992;304(6830):8214.

31. Nordberg E, Holmber S, Kuigu S. Exploring the interface between first and second level of care: referrals in rural Africa. Trop Med Int Health. 1996;1(1):10711.

32. Wilson S, Ruscoe W, Chapman M, Miller R. General practitioner-hospital communication: a review of discharge summaries. J Qual Clin Pract. 2001;21(4):1048.

33. McDonnell D, Butow PN, Tattersall MH. Improving the letters we write: an exploration of doctor-doctor communication in cancer care. Br J Cancer. 1999;80(34):42737.

34. Grol R, Rooijackers-Lammers N, von Kaathoven L, Wollersheim H, Mokkink H. Communication at the interface: do better referral letters produce better consultant replies? Br J Gen Pract. 2003;53(488):21719.

35. Lindstrom K, Hagman A, Berg L, Bengtsson C. Referral notes used as a tool for evaluating the cooperation between GPs and hospital physicians. Scand J Prim Health Care. 1994; 12(3):21418.

36. Ministério de Saúde. Saúde de família: uma estratégia para a reorientação do modelo assistencial. Brasília: Ministério de Saúde; 1998.

37. Elias PEM. Estrutura e organização da atenção á saúde no Brasil. In: Cohn A, Elias PEM, eds. Saúde no Brasil: políticas e organização dos serviços. São Paulo: Cartez; 1996. Pp. 57117.

38. Mendes E. Uma agenda para a saúde. São Paulo: Hucitec; 1996.

39. Brasil, Ministério de Saúde. Avaliação da implementação do Programa Saúde de Família em dez grandes centros urbanossíntese dos principais resultados. Brasília: Fundação Oswaldo Cruz/Escola Nacional de Saúde Ministério de Saúde/Secretaria de Políticas de Saúde/Departamento de Atenção Básica; 2002.

40. Brasil, Ministério de Saúde. Regionalização da assistência à saúde: aprofundando a descentralização com equidade no acesso: norma operacional da assistência à saúde. Brasília: Ministério de Saúde; 2001.

41. Paim JS. A reforma sanitária e os modelos assistenciais. In: Rouquayrol MZ, Almeida Filho N, eds. Epidemiologia e saúde. 5th ed. Rio de Janeiro: MEDSI; 1999. Pp. 47387.

42. Lages I. O Programa Saúde da Família: uma apreciação de seus usuários e trabalhadores [masters thesis]. Pernambuco: Universidade Federal de Pernambuco; 2002.

43. Haines A. Health care in Brazil. Br Med J. 1993;306(6876):5036.

44. Ministério da Saúde. 11° Conferência Nacional de Saúde. Brasília: Ministério de Saúde; 2000.

45. Samico I. Avaliação da atenção á saúde da criança: um estudo de caso no Estado de Pernambuco [PhD dissertation]. Pernambuco: Fundação Oswaldo Cruz; 2003.

46. Muniz R. A autonomia municipal na implementação de políticas nacionais de saúde: o PSF em Camaragibe [masters thesis]. Pernambuco: Fundação Oswaldo Cruz; 2003.

47. World Bank. World development report 2004. Washington, D.C.: World Bank; 2004.

48. Município de Camaragibe. Plano municipal de saúde 20002005. Camaragibe: Município de Camaragibe; 2000.

49. Medeiros KR. Modelos assistenciais e equidade em saúde: o caso do município de Camaragibe, Pernambuco [masters thesis]. Pernambuco: Fundação Oswaldo Cruz; 1998.

50. Farias SF. Modelo assistencial e responsabilidade sanitária: os limites e possibilidade da atenção á gestante no município de Camaragibe, Pernambuco [masters thesis]. Pernambuco: Fundação Oswaldo Cruz; 1998.

51. Machado M. Os médicos do Brasil: a realidade. Rio de Janeiro: Editora Fundação Oswaldo Cruz; 1997.

52. Campos G. Análise crítica sobre as especialidades médicas e estratégias para integrá-los ao SUS. Cad Saude Publica. 1997;13(1):1414.

53. Brasil, Ministério de Saúde. Conferência Nacional de Recursos Humanos para a Saúde: contribuição ao debate de uma política de recursos humanos para o setor saúde no Brasil- a situação da categoria médica. Brasília: Ministério de Saúde; 1986.

54. Feuerweker L. A formação de médicos especialistas e a residência médica no Brasil. Saude Debate. 2001;25(57):3954.

55. Ferrinho P, Van Lerberghe W, Julien MR, Fresta E, Gomes A, Dias F, et al. How and why public sector doctors engage in private practice in Portuguese-speaking African countries. Health Policy Plan. 1998;13(3):3328.

56. Beltrão A. A situação dos profissionais médicos e seu mercado de trabalho no estado de Pernambuco [masters thesis]. Bahia: Universidade Federal de Bahia; 2000.

57. Santana J. Reflexões sobre a formação médica para o Programa Saúde de Família. Revista Assoc Saude Publica Piaui. 1999;2(1):846.

58. Neto E. SUS, o ensino medico e os hospitais universitários. Saude Debate. 1996;49/50:79.

59. Andrade AC. Formação de recursos humanos para a estratégia de Saúde de Família em Pernambuco: processo em construção [masters thesis]. Pernambuco: Centro de Pesquisas Aggeu Magalhães/Fundação Oswaldo Cruz; 1999.

60. Maranhão E, Menezes FA. Formação médica: novas diretrizes e um novo modelo pedagógico. In: Arruda B, ed. A educação profissional em saúde e a realidade social. Recife: Instituto Materno-Infantil; 2001. Pp. 15368.

61. Campos G. A Saúde pública e a defesa da vida. 2nd ed. São Paulo: Hucitec; 1994.

62. Worley P, Esterman A, Prideaux D. Cohort study of examination performance of undergraduate medical students learning in community settings. Br Med J. 2004;328(7433): 2079.

63. Cordeiro H. A escola médica e o novo modelo de cuidados de saúde do SUS: saúde da família. In Arruda B, ed. A educação profissional em saúde e a realidade social. Recife: Instituto Materno-Infantil; 2001. Pp. 87119.

64. Rigli L. A estratégia do Programa Saúde da Família: reflexões sobre suas implicações. In: Misocsky MCA, Bech J, eds. Estratégia de organização da atenção à saúde. Porto Alegre: Dacasa Editora; 2002. Pp. 89103.

65. Araujo J. Health sector reform in Brazil, 19951998: a health policy analysis of a developing health system [PhD dissertation]. Leeds: Nuffield Institute for Health; 2000.

66. Araujo J. Attempts to decentralize in recent Brazilian health policy: issues and problems, 19881994. Int J Health Serv. 1997;27(1): 10924.

67. Atkinson S, Madeiros RL, Oliveira PH, de Almeida RD. Going down to the local: incorporating social organization and political culture into assessments of decentralized health care. Soc Sci Med. 2000;51(4):61936.

68. DaMatta R, Hess D. The Brazilian puzzle: culture on the borderlands of the western world. New York: Columbia University Press; 1995.

69. Gladwin J, Dixon RA, Wilson TD. Implementing a new health management information system in Uganda. Health Policy Plan. 2003;18(2):21424.

70. Kvamme OJ, Olesen F, Sammuelsson M. Improving the interface between primary and secondary care: a statement from the European Working Party on Quality in Family Practice (EQuiP). Qual Health Care. 2001; 10(1):339.

Revised version accepted for publication 12 October 2006.

* Send correspondence to: Matthew Harris, Department of Public Health and Primary Health Care, Old Road Campus, Headington, Oxford OX3 7LF, England; e-mail: matthew.harris@green.ox.ac.uk
1 Jeitinho literally means the little way. There is no adequate translation into English other than theunderstanding that when confronted with a problem, resolving it in the orthodox or legitimate manner would be too time-consuming or costly to expect anything other than an unofficial bypassing of this route. It usually requires the asking of a small favor that implicitly will be returned should the need arise. It is the cultural glue that holds Brazilian society together in the face of any adverse scenario, no matter how small.

Organización Panamericana de la Salud Washington - Washington - United States
E-mail: contacto_rpsp@paho.org