Patients’ preferences in selecting family physician in primary health centers: a qualitative-quantitative approach

Background The role of family physicians (FPs) in the metropolitan area is critical in identifying risk factors for disease prevention/control and health promotion in various age groups. Understanding patients’ preferences and interests in choosing a FP can be an effective and fundamental step in the success of this program. In this study factors affecting the FP selection by Iranian patients referred to health centers in the most populous areas in the south of Tehran were assessed and ranked. Methods A sequential mixed-method (qualitative-quantitative) triangulation approach was designed with three subject groups of patients, physicians, and health officials. The Framework method was used to analyze interviews transcribed verbatim. After implementing an iterative thematic process, a 26-item quantitative questionnaire with high validity and reliability was drafted to evaluate the different factors. A convenient sampling method was used to select 400 subjects on a population-based scale to quantitatively rank the most critical selection factors as a mean score of items. Results The selection factors were divided into six centralized codes, including FPs’ ethics, individual, professional and performance factors; patients’ underlying disease and individual health, and disease-related factors, office’s location and management factors, democracy factors, economic factors, and social factors. After filling out the questionnaires, the most important factors in selecting FP were a specialist degree in family medicine (FM) (4.49 ± 0.70), performing accurate examinations with receiving a detailed medical history (4.43 ± 0.68), and spending enough time to visit patients (4.28 ± 0.75), respectively. However, the parameters such as being a fellow-citizen, being the same gender, and physician’s appearance were of the least importance. Conclusion There is a possibility to screen the most important factors affecting the FP choice through the combination of qualitative and quantitative studies. The first and last patients’ priority was physicians’ specialty in FM and being a fellow-citizen with them, respectively. The clinical and administrative healthcare systems should schedule the entire implementation process to oversee the doctor’s professional commitment and setting the visit times of FP.


Background
Health is the foundation of the social, political, and cultural development of all human societies. As it is of particular importance in the formation of infrastructure in different parts of society, the ultimate goal of any country's healthcare system is to improve the health of individuals so that they can participate in social and economic activities with adequate health [1]. In this regard, attention has been paid to social inequalities in access to health services and followed by the need for reform aiming at increasing the productivity of the healthcare system has begun in many areas of the world [2]. One of the most successful and comprehensive health plans in the developing and developed countries is presenting the package of health care services recommended by family physicians (FPs). The FP is responsible for providing health services within a defined range, without any prejudice to age, sex, and socioeconomic status to the individuals, families, and communities [3][4][5]. In Iran, the FP program was launched in rural areas and cities with fewer than 20,000 populations. It is currently pursuing its output in urban areas in several Iranian provinces [6]. In Iran, Essential Package of Health Services (EPHS) in primary care is related to services provided by a team in the health system in the rural and urban areas, which is usually located close to patients' residences. The EPHS in primary care in Iran includes: (i) prevention: immunization, prevention, and control of communicable and non-communicable diseases, prevention of unwanted pregnancies, oral hygiene, and mental health, (ii) health promotion: health education, and learning healthy lifestyles and life skills, (iii) early treatment and emergency management: visiting the office, diagnosing and treating diseases, performing simple surgeries such as stitches, vasectomy or circumcision, injections, dressing, home visits, and cardio-respiratory resuscitation, (iv) referral: eligible patients for secondary or third level specialty care, and (v) health management: the record of population's health information, advocacy, and monitoring the work of health team members [3,7]. Therefore, the presence of family physicians is a necessity to meet the public health needs, address the unnecessary increase in health costs, and prevent its adverse impact on public health [8,9]. On the other hand, one of the essential principles for the provision and development of health care services is the need to pay attention to the collaboration and participation of patients and physicians in the program. The active involvement of all patients is also essential for its successful implementation and requires effective communication with patients, families, and the community to achieve an integrated and efficient model [10].
Different factors affect the patients' thoughts to choose a primary care doctor. In recent years, the issue of doctor selection, especially in developed countries, has received serious attention from health policymakers and health insurance organizations [7]. Patient-led physician selection leads to greater competitiveness of physicians, improved service quality, better access to medical services, and increased care efficiency [3,11]. However, there is a minimal research background on determinant factors in choosing the FP by patients in Iranian populations. Despite the generalization of the FP program in Iranian urban society, a comprehensive study to determine the most critical personal-social factors in the selection of FPs by patients has not been conducted yet. Therefore, the current study aimed to evaluate factors affecting FP selection in the covered population of health centers in the south of Tehran.

Study design and participants
A qualitative-quantitative mixed study was carried out between 2018 and 2019. Two approaches of relevant literature review and interview process were applied to assess the influencing factors on the FP selection. The selected strategy of interviewees was purposive that included three groups of physicians, health system managers (HSMs), and patients referring to the healthcare centers, who were selected with a triangulation method. This technique is typically used to evaluate an issue by explaining its different aspects. The triangulation method ensures data validation via cross verification from more than two sources [12]. According to the obtained results in this studying phase, the questionnaire was developed based on the internal expert panels' (IEP) and some participants' comments and distributed among 400 participants to identify patients' priorities in choosing an FP. The sample size (n) was calculated using the following formula (Eq. 1): where z, p, and d are the 95% confidence level based on the standard normal distribution (z = 1.96), the estimated proportion of the population presenting the characteristic (assuming at least 50% in the selected priority, p = 0.5), and tolerated margin of error (d = 5%), respectively [13]. Accordingly, the sample size for the current study is~385. However, based on the attrition risk of 10%, the sample size to compensate was reached to 400 subjects. Then, the designed questionnaire was distributed to a population of patients referring to hospitals and health centers in the south of Tehran through available sampling in a field study. The voluntary individuals with at least 18 years old were included in this study. Illiterate patients were asked to assist the questioner or patient companion by reading the questions.

Survey development and administration
Literature search and screening A review of the literature was carried out to extract the most important factors influencing FP selection in books, articles, and texts ( Table 1). The main keywords used to search in different scientific databases within 1987-2018 were: family physician, patient satisfaction, primary care, health care behavior, and continuity of care. About 80 articles, including studies conducted in Iran and other countries, were reviewed, and finally, 57 more relevant articles were selected as the reference.

Research team and reflexivity
Four qualitative researchers involved with the research process. All were medical doctors. One was male and three were female. They had experience in designing qualitative studies, writing qualitative books, and teaching in this field. To reduce bias, the team had not close familiarity with participants. Interviewees were selected based on their experiences with triangular techniques. The research team and also the aims of the study were explained to the subjects and entered into the study if they wish to cooperate and contribute with the team to promoting community health.

Interview procedure
A step-by-step guide on how to properly conduct a face to face interview was applied, and after transcripts confirmed by participants, the content was analyzed according to the complementary viewpoints of an IEP. Interviews in each part were continued until EQIs were satisfied that the data showed saturation. Three population groups including physicians (n = 47), HSMs (n = 5), and patients (n = 30) were interviewed as follows: The group of physicians consisted of three categories: (i) specialist doctors (SDs), (ii) family medicine residents (FMRs), and (iii) general practitioners (GPs) who were active as FPs in the rural and urban area. The interview openended questions of each group were designed by the research team and piloted by the target groups and the necessary corrections were made and finalized. It is worth noting that only two health managers and three physicians due to busy, not cooperate with us for an interview.

Interview with physicians' group
In the first category, FMRs of Tehran University of Medical Sciences (TUMS) were interviewed in a 20person group meeting. At the beginning of the meeting, the research topic and its purposes and applications were explained, and any ambiguities were resolved. The answers were kept confidential, and the FMRs were then asked to provide their comments on the sheet if they were satisfied with the cooperation. Ongoing verbal interviews were developed with openended questions. In the second category, a group of SDs of the hospital from clinical faculty members at TUMS was interviewed in person with open and indepth questions. All interviews were then recorded and transcribed. In the third category, GPs, who acted as FPs, were interviewed. For comprehensive interviews, urban FPs occupied in the national pilot of the FP plan in Mazandaran province (Amol city, Iran) with at least 7 years of work experience, and rural FPs in healthcare centers of the south of Tehran with 3-10 year work experience were considered. The experienced qualitative interviewers (EQIs) conducted interviews for 20-45 min, audio-recorded, and

Interview with health system managers' group
Open and in-depth interviews with HSMs were conducted after multiple coordination and verbal consent. Each interview with HSMs with a management history of 17-28 years lasted from 40 min to an hour in workplaces.

Interview with patients' group
The subjects were interviewed from referrals to Ziaeian Hospital and healthcare centers in the south of Tehran after the oral and written informed consent. According to the expert instructors' opinions, 2-3 general questions of the main factors obtained from the literature review as a result of the content analysis were asked. The in-person, individual, semistructured qualitative interviews for 20-30 min were conducted with patients.

Qualitative data analysis
The 5 main stages thematic framework analysis method was used to analyze the qualitative data obtained from the conducted interviews and included: (i) familiarization, (ii) identifying a thematic framework, (iii) indexing, (iv) charting, and (v) mapping and interpretation [12]. The familiarization phase was based on the principle of immersion, by repeated reading of transcripts, and written notes taken during the interview, line by line, by two coders. This step aimed to list key ideas and recurrent themes and to correctly edit a summary of the content of each interview (Table 2). In this table, items and interviewees' characteristics were displayed in the columns and rows, respectively. After sequential comparisons for each item category, a framework entitled "sub-theme" was selected (Table 3). In the indexing stage, the thematic framework was used in a textual form by explaining the transcripts with the principal and centralized codes from the index. The theme and subtheme were centralized and principal codes, respectively. The codes were discussed by two coders to make coding decisions. For instance, each interviewee in HSMs group was explained as "M" so that "M1" to "M5" indicates the participant number of 1 to 5. Similarly, relevant codes were considered for the rest of the participating groups. Table 4 was then composed of selected summaries of viewpoints in the 'charting' stage. A table was separately tabulated for each group of interviewees by specifying their opinions with an appropriate code in each column, while the principal and centralized codes were defined in the table rows. In the final step of 'mapping and interpretation', tables drawn for each interview group were put together and compared to find correlations between themes with a view to providing clarifications for the results [14,15]. An update was performed in the analysis process of the thematic framework regarding certain coded items that started to cluster, and others separated.

Questionnaire design
All the influencing factors by incorporating the themes mentioned by each of the interviewed groups were identified. These fundamental elements emerged with the principal and centralized codes in terms of 43 questions. These question items were then evaluated by the IEP and some participants to present specialized corrections and dedicated processing strategies. After the revision step, frequent questions with a centralized or principal code were put together. Some of the questions were eliminated due to raising unreasonable expectations. According to the experts' opinions, 26 of the original 43 questions in a short and precise form were arranged. A draft of a questionnaire was then constructed with two main sections of demographic data (e.g., gender, marital status, education level, and health insurance status), and questions related to the various factors affecting the choice of FP. The 26-item questions consisted of four centralized codes and five principal codes. In general, items included questions related to healthrelated physical factors (6 cases), patient-related factors (2 cases), social factors (5 cases), and physician-related factors (13 cases). Answers to questions for each item were presented on a 5-point Likert scale. Items were rated based on a 5-point Likert scale for levels of satisfaction and importance. Answers ranged from labels of "dissatisfied" and "not very important" (scored as 1.0) to "very satisfied" and "very important" (scored as 5.0); with the total score sum of 26-130.

Questionnaire validity and reliability
The qualitative content validity of the developed questionnaire was assessed based on the viewpoints of 20 patients of the studied population and ten individuals of IEP, who had research experience or worked in the field. Twenty participants initially completed the questionnaire to determine whether the designed questions were ambiguous or not? Fortunately, there was not a significant problem in the fluency and understandability of the items so that the response rate ranged from 5 to 85%. Subsequently, 10 IEP members confirmed the re-revised questionnaire by presenting more specific recommendations and better and more obvious questions. In the quantitative content validity, confidence is maintained to select the most important and correct content in a data collection tool. In this assessment, the experts are asked to agree whether an item is essential for operating a construct in a set of items or not? Accordingly, the content validity ratio (CVR) was determined by the following equation (Eq. 2): where the N e and N are the number of panelists indicating "essential" and the total number of panelists, respectively [16]. The Lawshe table was used to estimate the numeric value of CVR. The minimum acceptable CVR (MA-CVR) according to the count of scoring panelists. The MA-CVR for each item should be 0.62 as the number of experts was 10 in this study [17]. In this study, the MA-CVR for each item present in the prepared questionnaire was more than 0.62. The Waltz and Bausell's method was used to assess the content validity index (CVI) for each item by dividing the number of experts who ranked the items as compatible or full compatible for each criterion (relevancy, clarity, and simplicity) to the total   [18]. The mean value of the three criteria was considered as the total CVI for each item.
Overall, a CVI value of more than 0.79 for each item was appropriate to be retained [19]. The mean of CVI for the developed questionnaire was calculated to be 0.88. The questionnaire's reliability was evaluated by calculating the internal consistency reliability coefficient "Cronbach's alpha". Bland explained that this coefficient should be ≥0.8, while alpha equal to 0.7 is also acceptable [20]. The Cronbach's alpha for the constructed questionnaire was 0.845, showing its high reliability.
Code a Physician sub-groups Percent (%)

Quantitative data analysis
The results of completed questionnaires with the data coding in a univariate approach was analyzed using the SPSS software package version 22.0 (SPSS Inc., Chicago, IL, USA).

Qualitative data
A thematic framework was used to analyze the qualitative data in all the groups with 6 themes (centralized codes: F, P, C, D, E, and S) and 10 sub-themes (principal codes). Table 3 shows the definitions and indexes of six centralized codes related to the qualitative data. The centralized code of "F" involves factors relevant to the physician and divided into three subthemes. In the principal code of F1, most of the items mentioned in the study groups included physician's good behavior interaction in interacting with patients. The most frequent physician's individual characteristics in the code of F2 were doctor's neat, clean-cut appearance, gender, and etc. The influential professional parameters from the perspective of the different interviewed groups (F3) were doctor's experience and expertise, good clinical skill, and etc. The severity of illness and need was diagnosed as the main factor in the code of P1, whereas the patient's satisfaction in the previous examination with the physician or the patient's trust in the physician was the key influential factor in P2. Some of the location factors in the subtheme of C2 were geographical proximity, convenient accessibility, and cleanness of the well-equipped and stylish doctor's office. Short waiting time, high availability to the physician in more weekdays, and the nearness of other health services to the doctor's office or clinic were in the sub-theme of C2. The main political factor affecting the FP choice was government decisions on the distribution of physicians in different geographical areas. The code of E shows patients' economic access to doctors such as free or low-cost visit price and private sectors. The most important social factors affecting the patients' choice were being fellow-citizen, having a specialist degree, satisfaction and experiences of other patients. Table 4 shows the charting stage of framework from the viewpoints of HSMs and physicians.

Quantitative data
After obtaining the data codes from the qualitative study section, a new questionnaire was designed with high validity and reliability to evaluate the quantitative data. Four hundred patients in the age range of 18-76 years (age mean of 38.6 years) participated to fill out the questionnaire. The response rate of the questionnaire was 94%.
The demographic characteristics of patients are mentioned in Table 5.
Most participants had a diploma and were married. Overall, the highest score among the 26 questionnaire items belonged to being a family physician specialist (4.49 ± 0.70). After that, performing careful examination with receiving a detailed medical history, and assigning enough time to visit patients were other items influencing the FP selection, respectively. However, being fellow-citizen, the patient's tendency to same-gender doctors and doctor's appearance were ranked with the lowest scores, respectively. The participated women and men similarly mentioned that the physician's specialty, the complete and exact examination, as well as assigning sufficient time to visit were the most important factors influential patients' choice. Women compared to men paid more attention to items like waiting time for a doctor's visit, and the illness follow-up, while men were more important in the physician's office appearance and equipment and her/his characteristics such as trust, confidentiality, and reputability ( Table 6). As majority patients (81.5%) had an educational level lower than bachelor's degree, items with the highest (having the specialist degree) and lowest (doctor's appearance) scores were very similar to the total population of the study. In patients with a bachelor's degree or higher (18.5%), the highest score was given the examination accuracy and receiving the correct medical history, and having the expertise of an FP, respectively. This study showed

Discussion
The most important factors affecting the choice of an FP by patients out of 26 items were the specialist degree in FM, followed by careful examination and history, and assigning sufficient time to visit. Bornstein conducted an exploratory cross-sectional survey of parameters influencing American patients' choice of a primary care doctor. Results revealed that the participants highlighted relevant professional factors (e.g., the validity of FP's degree and office appearance) and management practices (e.g., appointment times (nights and weekends)) more than the FP's inherited characteristics (e.g., race, age, gender, etc.). Also, the most important factors found by 636 patients in choosing an FP are those that have the most significant impact on the quality of health care [21]. Mosadeghrad [22]. Nouronnesa also found that the most important priorities were the physician's skills and expertise, getting a complete medical history under the careful checkup, and how to answer a doctor to the patient's questions [23]. In our study, receiving a detailed medical history with a careful examination was determined as the second priority after having an FP specialty. In contrast to the views of patients referring to target healthcare centers, health system managers in their interviews did not point out having a degree in FM as a priority for the effective implementation of the FP plan. This fact shows the importance of launching a specialty in FM in the country. Results indicated that the least importance belonged to the citizenship, being the same gender, and appearance of FP. As our study area geographically is more populated by immigrants than in other cities in the country, the same dialect, ethnicity, and race from the patients' point of view do not matter much to refer to a doctor. Similar results were reported by Bornstein in the US [21]. However, Bachmann in a qualitative study, realized that Russian-speaking migrants were less satisfied with primary care consultations compared to native Germans [24]. The cause of dissatisfaction with treatment may be attributed to the more inferior patient-physician relationship and frequent physician changes. Therefore, physicians need to be more aware of the cultural expectations of immigrants in order to better understand their needs, improve the relationship between physician and patient, and ensure equal opportunities in health care. It was earlier proved that the leading cause of patients' choice in the UK was the proximity of physician office to patients' home, the suggestion of patient's friends, and previous visits of patients' family members by the selected physician [25]. In the present study, access to the office place had relative importance, whereas the advice of a physician by the patient's friends was of little importance. Oleszczyk in 2017 mentioned that the characteristics of patients and their physicians did not significantly affect the satisfaction and experiences of patients in Polish primary care [26]. However, our qualitative results showed that the FP'(F-code) and patients (P-code) characteristics in all interviews were influencing factors in patient satisfaction and FP selection. This finding revealed that patients' trust in their physician plays a central role in Iran's health system. The trust was also determined as a fundamental factor in the loyalty of French patients [27].
One of the most important factors in this study was sufficient time allocated by an FP to visit patients along with a detailed evaluation of medical history and clinical examination. This goal can be achieved if the workplace conditions and the number of referrals are well controlled. In other words, one of the essential pillars for achieving this purpose is allocating a reasonable number of referrals to each FP. On the other hand, being the same gender or fellow-citizen or having access to a physician in our research is less important than other causes. Concerning the different cultural contexts in Iran's society, patients are willing to accept difficult access times and gender differences but referred to the physician with their preferred priority.
In the present study, there was no difference in the preferences of patients with different education levels. But Aelbrecht reported that participants with lower education were more likely to favor aspects related to emotional issues, while patients with college education paid more concerned with vocational skills issues [10]. In this study, there was a difference in the views of Iranian men and women to choose the FP. For women, the physician's clear expression and the effective follow-up were critical, while for men, reasonable prescription of the test, and medication accompanying with medical services (such as equipped laboratory and pharmacy) was significant. This finding was affirmative to the result of Wolosin, who concluded that American women and men were more satisfied with physician-related items and the service delivery process, respectively [5]. Street in 2014 found that patients were less satisfied with primary care providers who were more likely to look at computers and be more conscientious in counseling. However, this fact was not mentioned by any of our interviewees [28].

Study limitations and strengths
Discrimination of our study from other surveys performed in the field of FP was the design of a 26-item questionnaire using a triangular method according to the population-based data obtained from qualitative evaluations. This design method resulted in effective factors being searched from different angles, and no points were ignored in the questionnaire. On the other hand, given that the FP program and referral system were not implemented in the Tehran metropolitan area, we had to provide verbal explanations about this program and the research aims while filling out questionnaires. Recording an abstract on top of the questionnaires along with giving oral descriptions, could significantly solve this problem, although it was a time-consuming and challenging task. Due to the nascent FP program, this study was limited to the factors influencing FP selection, whereas the most important factors affecting the change or departure of the FP were not examined. Also, it was impossible to exclude desirability bias as participated doctors might have chosen answers thought to be more desirable instead of closer responses to their beliefs. The evaluation result of a broad range of factors showed that what was a top priority for patients was ignored by health managers.

Conclusion
The present qualitative-quantitative study evaluated the patients' preferences to select FP in Iranian primary health centers. A set of influential parameters on the FP selection was assessed after the comprehensive literature review and the face-to-face interview with physicians, patients, and HSMs based on the open-ended questions. A five-stage thematic framework analysis method was first applied to qualitatively analyze the data and then a 26-item quantitative questionnaire was used to rank the most important the influencing factor in choosing the FP according to a 5-point Likert scale. Results showed that the main factor was having a specialist degree in FM, followed by precise examinations, in-depth inquiry into the patient's medical issues, and giving the patient enough visit time to review all outstanding health problems. Also, patients when choosing a FP paid less attention to factors such as physician's appearance, as well as being fellow-citizen and samegender with doctors. Accordingly, expanding this nascent medicine branch with recruiting graduates with high professional commitment in health centers may play a key role in presenting high-quality medical services with patients' satisfaction levels.