The national level questionnaire was completed by a panel of 11 experts from the government, universities and professional association. Statistical data have been provided by the Ministry of Health.
The pilot implementation included 78 FDs; 37 in Bolu and 41 in Eskişehir. In both provinces most physicians were from urban family health centres, but in Eskişehir this proportion was greater (81%) than in Bolu (68%). In both provinces two thirds of the FDs were male and one third was female. Respondents were relatively young; on average 36 years in Bolu and 41 in Eskişehir. Since family medicine had recently been introduced, physicians had little experience as FDs (on average 1.5 and 2.5 years in Bolu and Eskişehir respectively).
In total 1548 patients filled in a questionnaire; 738 in Bolu and 810 in Eskişehir. The average age was around 40 years and the majority were women. Almost half of the patients only had primary education. Almost all respondents lived in a family setting; living alone was extremely rare. Three quarters of patients were from urban family health centres.
Table provides a summary overview of a selection of proxy indicators by PC function for Bolu and Eskişehir. We will discuss the main results in the next sections. Differences between the two provinces will be indicated where appropriate.
Overview of selected proxy indicators by primary care function for Bolu and Eskişehir
Policy development (national level questionnaire)
PC was acknowledged as being important a long time ago in Turkey but its implementation has only recently become effective. The concept of integrated PC was introduced in 1961. Plans launched in the early 1990s, including those for decentralization, partial gatekeeping by FDs and better training programmes were not successful. It was not until 1996 that family medicine was adopted as a more comprehensive model for PC. Since 2003, this model has been implemented in 13 provinces (by early 2008) out of 81 overall in Turkey [14
Despite decentralization, the role of the Ministry of Health in PC was still strong. To a large extent, the management and provision of PC services in Turkish provinces was uniform. The Ministry retains a strong influence on staff appointments at provincial health directorates and directorates take technical decisions in line with central guidelines. Furthermore, they assume major responsibilities for the management of estate and human resources in their province.
The government's vision of PC has been published in various laws and documents. These cover the specification of PC disciplines and their tasks and responsibilities, education and accreditation requirements, norms for availability of doctors and facilities, medical record requirements and requirements related to performance monitoring. Organizations of professionals and patients were rarely involved in the policy making process but rather in the implementation of policies [15
]. It is expected that the roles (in the policy making process) of these organizations will become more formalized in future. The position and rights of patients was formally acknowledged but this position has not yet been fully translated into practice. For instance, patient complaint procedures were only present in 78% of the family health centres included in the family doctor survey.
Topics that were debated by health policy experts during the period of the survey's implementation were, for example, shortages of physicians and nurses and the (unequal) provincial distribution of physicians in the country; the improvement of coordination of care between levels of care through gatekeeping and multidisciplinary teams; and physical improvement of health care premises and equipment.
Financial incentives for providers (national level and family doctor questionnaire)
Almost all FDs (90%) were state employed, receiving fixed salaries, with additional capitation elements of payment. This included bonuses for working in disadvantaged areas and additional payments for the performance of predefined preventive services. The recent introduction of a more performance-related payment scheme (mixed scheme) seemed to be a major step towards implementing a more comprehensive, efficient and responsive PC system. Incentives needed to be fine-tuned in order to avoid overproduction and to stimulate quality of care.
Professional development (national level and family doctor questionnaire)
The implementation of family medicine in the 13 provinces (as of early 2008) was well underway. A total of 27 500 physicians were working in PC nationally. In these provinces with newly introduced family doctors, a majority of PC doctors were now FDs. Nationwide (81 provinces), however, the proportion of FDs was only 10%. Three quarters of all medical universities in Turkey had departments for family medicine where FDs were trained. However, the capacity in the residential programmes (about 500 places per year in 40 medical universities) was not fully used. Only 80% of places were filled. In light of the current severe shortages of physicians and nurses in PC, everything should be done to ensure full use of capacity. Registers of PC professionals were in place, but it was not clear whether they were up-to-date and how they were used for workforce planning.
There was a national organization of FDs (TAHUD) with a broad range of activities; however, its role in the policy-making process was not formalized. In addition, GP and FD organizations were developing in eight provinces in early 2008.
Many FDs (54% and 70% in Bolu and Eskişehir respectively) reported having difficulty in keeping up with the latest professional developments. FDs in Eskişehir reported spending much more time on professional reading (12 hours/month) than their colleagues in Bolu (7 hours/month).
Quality management (national level questionnaire)
Quality improvement mechanisms such as obligatory re-certification schemes or periodic knowledge and skills tests were not yet in use and formalized. There were few requirements concerning the quality and confidentiality of medical records. Formal and informal mechanisms of performance assessment, such as practice inspections and medical audits, were infrequently applied. A positive change here may be the introduction of the performance-related payment scheme. However, its focus seemed to be more on the quantitative side of performance than on the quality of care. So far, clinical guidelines in PC were developed and implemented under the exclusive responsibility of the Ministry and drawn up by assigned medical specialists with minor inputs from FDs. The use of the guidelines was not formally evaluated. The responding FDs reported that they did not frequently use clinical guidelines.
Financial and geographical access for patients (patient questionnaire)
Although PC was officially free of charge, this was not true for prescribed medicines or injections. Half of the patients reported co-payments for these services. Some patients (12%) also reported co-payments for home visits and for visits to a specialist after referral from the FD. Co-payments seemed to be an obstacle to the utilization of health care services. A significant minority of patients answered they had abstained from a visit to their FD (9%) or a medical specialist (17%) for financial reasons.
The national level questionnaire showed that PC physicians were unevenly distributed between provinces in Turkey. This suggested provincial differences in the availability of PC services. In Bolu and Eskişehir, however, patients had no difficulty to reach family health centres, pharmacies and hospitals.
Organizational access to services (family doctor and patient questionnaires)
Compared to the European situation, practices were very large with an average of 2484 people per physician in those provinces where family medicine implementation had not yet started, but there were also variations across the country. In provinces where the family medicine reforms had already been implemented, the average population per family physician was around 3500; in Bolu and Eskişehir for example the average was 3700. As a result, the number of consultations per day was high (47 on average). Home visits were rarely made (on average 1.7 per day).
Most patients were satisfied with the current opening hours of the family health centres (84% on average; which varied significantly between centres), the availability of medical staff during these hours (83%) or getting the opportunity to speak to a doctor on the telephone (49%). Almost all patients reported that it was usually possible to visit a FD the same day and waiting times were acceptable, even if making an appointment was unusual. Visiting a FD outside the normal office hours, in the evening or on weekend, was only rarely possible.
Family health centres hardly ever used the Internet for their communication with and information to the patients. Consultation time per patient was relatively short (11 minutes on average).
Cohesion within primary care (family doctor questionnaire)
Lack of coordination of care seemed to be a major problem. For instance, multidisciplinary teamwork, for the benefit of patients with chronic diseases (such as diabetes) hardly existed in Bolu and Eskişehir. The majority of FDs worked in teams of three or more FDs. In addition to FDs, family health centres consisted of practice nurses and, in most cases, midwives as well. Other PC disciplines, like physiotherapists, dentists and pharmacists were not usually part of the family health centre. Cooperation was not strongly formalized between team members. Regular meetings were not usual between FDs and nurses and even less so between FDs and midwives.
Coordination with other care levels (family doctor and patient questionnaires)
There were no mechanisms to promote coordination between the primary and the secondary care levels. The policy on the gatekeeping role for FDs was not clear to patients and, in daily practice, gatekeeping was not well maintained. Despite this lack of clarity, most people first visited their FD with new health problems. It was not usual to refer patients back to PC after hospitalization. Working relations between FDs and medical specialists and hospitals left much to be desired. Consultation or asking advice from medical specialists were infrequently reported (on average only by 8% and 4% of the FDs in Bolu and Eskişehir respectively).
Referral letters were poorly used (only frequently used by 56% and 12% of the FDs in Bolu and Eskişehir respectively) and medical specialists did not inform FDs properly about their treatment. Discharge reporting was not formalized.
Informational continuity (family doctor and patient questionnaires)
Conditions for clinical and other information were good in the family health centres. 97% of respondent FDs had a computer at their disposal, which was used for keeping patients' medical records.
However, these possibilities were not optimally used because records were not kept routinely. Furthermore, it was difficult to use computer records to produce lists of patients on the basis of common diagnosis or elevated health risks. Most patients (45% on average) were not sure whether they could see their own medical files if they wanted. Many patients (63% on average) felt the exchange of information between physicians could be better. Patients' expectations of the communication between their own FD and other physicians were also modest.
Longitudinal continuity (patient questionnaire)
Patients had visited the family health centres about seven times during the previous year according to their estimates. In Eskişehir, the visiting rate with FDs was much higher in the urban family health centres than in the rural ones. Patients thought that it was not possible to choose their doctor. They had usually been assigned to their current FD.
Patients saw restrictions in changing from one doctor to another. Since family medicine had been introduced rather recently, patients had been with their doctors for a rather short period.
Interpersonal continuity (family doctor and patient questionnaires)
Despite the fact that FDs worked in groups, patients would generally (93%) see their own FD during each visit. Consultations were relatively short and FDs did not always have the patient's medical file at hand.
Patients were satisfied about the way they were treated by their FD (95%), although they were not generally convinced that the FD was aware of their personal situation (40%) and the details of their medical history (44%). Patients found that FDs took sufficient time (86%), listened (94%) and communicated (90%) well and kept to promises and appointments (84%). Patients were reserved about their FD's preparedness to make a home visit (45% of patients did not know). Many patients (40%) were also not sure if their FD would be the right person for discussing non-medical problems that impacted on health.
Services delivery (family doctor and patient questionnaires)
FDs had a strong position as the doctor of first contact for health problems of children (except for hearing problems), and women (except for menstruation problems). For problems with strong social and psychological components, the first contact role was less developed.
For sexual problems, psychiatric or relationship problems, FDs were not the first choice to contact. FDs reported to be moderately involved in the provision of preventive care and medical technical procedures. Expansion of these tasks could include insertion of intrauterine devices and minor surgical procedures.
Activities of FDs aimed at specific patient groups or other public health related tasks mainly covered the areas of mother and child health and family planning. FDs did not conduct much screening for sexually transmitted infections, HIV/AIDS, tuberculosis or cervical cancer.
Family health centres were reasonably well equipped, especially with computers. With regard to medical equipment, the situation in Eskişehir was slightly better than in Bolu. Typically absent were peakflow meters, tuning forks and ultrasound equipment. FDs in Eskişehir were better equipped for gynaecological services than their colleagues in Bolu. A general problem as perceived by FDs was insufficient access to external X-ray facilities; access to laboratory facilities however was felt to be very good.
Links with the community turned out to be weak. FDs in Bolu mentioned community connections more frequently than FDs in Eskişehir, e.g. meeting with local authorities, social workers or religious groups.