Specialty Career Preferences among Final Year Medical Students at the School of Medicine, Makerere University College of Health Sciences, Uganda

Background: Uganda has an imbalanced distribution of the health workforce, which may be inuenced by the specialty career preferences of medical students. In spite of this, there is inadequate literature concerning the factors inuencing specialty career preferences. We aimed to determine the specialty career preferences and the factors inuencing the preferences among fth year medical students in the School of Medicine, Makerere University College of Health Sciences (MakCHS). Methods: A sequential explanatory mixed methods study design with a descriptive cross-sectional study followed by a qualitative study was used. A total of 135 nal year medical students in MakCHS were recruited using consecutive sampling. Self-administered questionnaires and three focus group discussions were conducted. Quantitative data was analysed in STATA version 13 (StataCorp, College Station, Tx, USA) using descriptive statistics, chi-square tests and logistic regression. Qualitative data was analysed in NVIVO version 12 (QRS International, Cambridge, MA) using content analysis. Results: Of 135 students 91 (67.4%) were male and their median age was 24 years (IQR: 24, 26). As a rst choice, the most preferred specialty career was obstetrics and gynecology (34/135, 25.2%), followed by surgery (27/135, 20.0%), pediatrics (18/135, 13.3%) and internal medicine (17/135, 12.6%). Non-established specialties such as anesthesia and Ear Nose and Throat (ENT) were not selected as a rst choice by any student. Female students had 63% less odds of selecting surgical related specialties compared to males (aOR=0.37, 95%CI: 0.17-0.84). The focus group discussions highlighted controlled lifestyle, assurance of a good life through better nancial remuneration and inspirational specialists as facilitators for specialty preference. Bad experience during the clinical rotations, lack of career guidance plus perceived poor and miserable specialists were highlighted as barriers to specialty preference.


Background
The World Health Organization (WHO) recommends a doctor to patient ratio of 2.3 doctors per 1000 people for adequate health care services (1). In spite of this, Africa has a doctor patient ratio at 0.25:1000 which is low when compared to the WHO standard (2).This has resulted into uneven distribution of the health workforce within the various medical specialties (2). In Uganda, this uneven distribution of health workers can be demonstrated by the number of specialist doctors registered by the Uganda Medical and Dental Practitioners Council. Among the 2012 registered specialists by 2019, 386 (19.2%) were public health specialists, 376(18.7%) were Pediatricians, 287 (14.3%) were Obstetrics and gynecology specialists, 230 (11.4%) were Internal medicine specialists and 185 (9.2%) were General surgeons (3). There are specialties that had less than 20 (1%) specialists including ENT, Emergency medicine, and Nutrition.
It has been reported that the uneven distribution of the health workforce may occur as a result of specialty career preferences of medical students and their eventual choices of professional training (2,4). Studies conducted in Asia showed that students usually preferred already established clinical specialties like Pediatrics, General Surgery, Internal Medicine and Obstetrics and Gynecology (5,6). This pattern of preference can also be observed among medical graduates in Uganda as re ected in the admissions for post graduate programs in the past 10 years (7). This imbalance in the health workforce may potentially lead to inadequate services offered in health care delivery (4) . Therefore, there is a need to attract medical students to the specialties with inadequate number of personnel so as to improve overall health care (2,8,9).
The medical students' preferences for a particular clinical specialty is likely to be in uenced by both individual and contextual factors (10). An understanding of the career preferences of medical undergraduates can perhaps help to provide important information that can be utilized when planning educational programs. In addition, it can help to set priorities so as to foster the provision of adequate health care through training a wide range of specialists (10). However, there is insu cient literature concerning the factors that in uence specialty career preferences of medical students not only in Uganda, but also in many resource-constrained countries. Therefore, this study was designed to determine specialty career preferences and the factors that in uence these preferences among nal year medical students at Makerere University College of Health Sciences (MaKCHS). Our study could provide important information to be used in setting priorities and planning educational programs so as to attract students to the less preferred specialties.

Methods
This study employed a sequential explanatory mixed study design (quantitative followed by qualitative) so as to triangulate the sources of information. A quantitative cross-sectional study was conducted to determine the specialty career preferences while a qualitative study was conducted to determine the factors in uencing these preferences. The study was conducted among nal year medical students who had given written informed consent. Medical students in their nal year were selected because they had adequate experience through medical school with satisfactory clinical experience. These students were also on the verge of entering the professional practice and had perhaps started re ecting upon the issue of specialty career preferences.

Quantitative component of the study
Regarding sample size determination, the entire population of 178 nal year medical students was considered for the study, therefore, there was no need of sampling. Permission was obtained to access the list of registered fth year medical students during the study period from the Principal of the College of Health Sciences, on submission of the ethical approval letter. The principal investigator (PI) worked with research assistants to mobilize all the fth years to participate in the research. The mobilization was done through messages, phone calls and verbal announcements to the various class groups with the help of class representatives who provided the students' contacts. The students were gathered in a lecture room where the study was explained to them and they were given informed consent forms to read and sign. Those who were not interested were given an opportunity to opt out of the study. Those who gave informed consent were given a self-administered structured questionnaire to collect data from them regarding their specialty career preferences. The students were given 30 minutes to complete the questionnaires and hand them back to the research assistants.
The structured questionnaire included questions on the demographics of the participants, their specialty career preferences ( rst, second and third choices), willingness to practice medicine and career guidance. The data was entered using Epidata and analysis done using STATA Version 13 (StataCorp, College Station, Tx, USA). Specialty career preferences were presented using frequencies, percentages and bar graphs. The rst choice for specialty career preference was reported and compared across the sociodemographic variables using chi-square /Fisher's exact tests in order to determine if there is an effect of these variables on the preferences. Thereafter, odds of selection of surgical versus nonsurgical discipline for rst choice were obtained using multivariable logistic regression.
Qualitative study component of the study Following the quantitative arm of the study, the qualitative arm was carried out to explore the factors in uencing specialty career preferences of medical students. The data collection and analysis was guided by the grounded theory. There were three Focus Group Discussions (FGDs) that were conducted in total. Two of these were homogenous groups (male and female) that comprised of direct entrants. Direct entrants are students who join medical school at University direct from high school. Each of these two FGDs had 10 participants. The third FGD comprised of mature entrants who numbered 11 in total (6 males and 5 females). Mature entrants are students aged 25 years and above who join medical school at University after completion of a diploma or a medical related bachelor's degree. Participants for these FGDs were selected using purposive sampling and mobilized to a quiet place on the college premises where the study objectives and procedure were explained to them face-to-face. Written informed consent was obtained from the participants before commencement of the discussions. Permission to record the discussions using both audio and written means was also obtained from the participants. The discussions were conducted by JK and DM who have expertise in qualitative methods and they lasted approximately one hour each. The discussions were conducted using pretested FGD guides that had semi-structured questions. The questions in the FGD guide aimed at obtaining data on the demographics of the participants, their experiences in the different specialty rotations, their most preferred specialty and the reasons that in uenced this preference. There were no other students enrolled for the FGDs since data saturation had been attained by the third discussion. The audio-recorded responses were transcribed and analysed using Nvivo version 12 (QRS International, Cambridge, MA). Content analysis was used in analysis where data was coded by JK and DM after familiarization, followed by development of an index or conceptual framework of themes and sub-themes. JK then sorted the data by themes and synthesized to provide descriptive and explanatory summaries.

Results
The study was conducted between April and May 2019. There were 178 students registered nal year medical students. However, 33 of these were not available at the university during the study period. Therefore, 145 medical students were eventually recruited into the study. The recruitment of study participants is shown in Figure 1.. Out of these, 135 students returned their questionnaires giving a response rate of 93%.

Sociodemographic Characteristics of Fifth year Medical students at MakCHS
Out of the 135 medical students that participated in the study, 91(67.4%) were males and 44(32.6%) were females, giving a male to female ratio of 2:1. Their ages ranged between 21 years and 47 years, with a median of 24 (IQR: 24, 26). The sociodemographic characteristics of the students are shown in Table 1 and Table 2.

Specialty Career Preferences among nal year Medical students at MakCHS
Obstetrics and Gynecology was the most commonly preferred career as a rst choice, with 34 (25.2%) students selecting it; followed by careers in Surgery, Internal Medicine and Pediatrics respectively as represented in Figure 2. Careers in Family Medicine, Anesthesia and ENT were not selected by any student. The "others" rst choice category consisted of Business, Clinical epidemiology and Biostatistics, Forensic Medicine and voluntary medical work.
Obstetrics and Gynecology was still the most preferred second choice with 27 (20%) students selecting it. The "others" second choice category consisted of Nano medicine/ nuclear medicine and political science/public administration.
Pediatrics was the most preferred third choice with 22 (16.3%) students selecting it. The "others" third choice category consisted of Biogenetics science, health informatics and sports medicine. The details of the various choices as selected by the students are illustrated in Figure 2.
Comparison of rst choice specialty preference across the sociodemographic characteristics The distribution of the rst choice specialty was compared across the various categories of the sociodemographic characteristics using the chi-square test. The test showed that there was a signi cant relationship between the rst choice preference and sex (p = 0.04), marital status (p < 0.01), having children (p < 0.01) and sponsorship (p = 0.04) ( Table 3). On adjustment using multivariate logistic regression, sex was the only factor with a statistically signi cant odds ratio (OR) (OR=0.37; 95%CI (0.17-0.84); p=0.02). Therefore, the female medical students are 63% less likely to select surgical disciplines as a specialty career preference compared to the male medical students. The results for the odds are shown in Table 4.
The female medical students were less likely to select surgical disciplines like Obstetrics and Gynecology and General Surgery when compared to their male counterparts ( Figure 3).
Factors in uencing specialty career preferences of the medical students Three FGDs were conducted in May 2019 among 15 males and 16 females whose ages ranged from 21 to 35. The students were well distributed in terms of places of origin and different backgrounds with one foreign students included. Most of the students had a specialty career preference in mind at the time of the study.
Analysis of these FGDs revealed two major themes which are summarized in Table 5 with the key representative quotes from the students.

Theme 1: Facilitators of specialty career preference
The students selected various careers and based their preferences on the following in uencing factors.

Presence of role models and mentors:
A key facilitator to selecting a specialty career preference among medical students was presence of role models and mentors as illustrated by a participant"… She is not only a surgeon but also a mother. The way she takes you, she respects you even though you are an undergraduate. She is really good and really motivates you to think about more of surgery. Her surgeries are so clean" (Mature entrant FGD). Presence of mentors is also very crucial since these mentors offer guidance for the student as they build their careers.

Organization in the department
Most of the students in the discussions mentioned that an organized department was a sure facilitator for selection of a career preference. A mature student also highlighted the importance of the organization of the discipline "When you look at those two departments, they have a good streamlined program. They try to bring out their lectures very well, associate with students very well, and evaluate well when compared to other departments." (Mature entrant FGD)

Controlled lifestyle:
Ability to have a controlled lifestyle also motivated students to select certain specialties. This controlled lifestyle means that one is able to balance between work, family, leisure and any other activities. A male student re ected this as "I feel like psychiatrists have better working hours, surgeons are called in at 2am because a patient has gone into cardiac arrest or a mother has gone into labour." (Male FDG).
Other facilitators highlighted in the discussions included a positive or good experience during clinical rotation, attitudes of the specialists in the discipline, assurance of a good life due to better nancial remuneration, availability of job opportunities, family in uence, media, and peer in uence.
Theme 2: Barriers to selection of specialty career preference

Perceived poor and miserable specialists
The students' perception of the specialists to be poor and miserable is a major barrier to the selection of certain specialties "…if I look at certain specialty and to me all the doctors look poor and miserable, the residents are broke and suffering…I am not inspired by all that knowledge they have." (Female FGD). Therefore most students tend to avoid specialties where it is assumed that there is a weight of challenges and nancial di culties attached.
The prolonged amount of time spent in training within a certain specialty compared to others is also a barrier. This was especially among female students who had a concern about settling and starting a family before they are too old as shown in this quote "…well we have a biological clock to worry about…I will always go for shorter courses mainly in research with more money." (Mature entrant FGD).

Uncontrollable lifestyle attached to a career
The uncontrollable lifestyle of a specialty career is also a barrier in selection of certain careers since students want to be able to balance between life and work. There are some careers that are very demanding (especially surgical-related careers) and may not allow time for any other activities or for family "Given the kind of mother I would like to be I may not be able to do emergency medicine; maybe because I have not got guidance from a successful emergency medicine specialist. But from what we see on TV, clearly you cannot take care of your kids and be an emergency medicine specialist." (Female FGD)

Inadequate exposure during clinical rotations
Inadequate exposure to a specialty during the clinical years may also be a barrier since a student may not able to appreciate the various aspects of the discipline "I think the challenge there is the small duration we spend or the inadequate duration we are given. "We have to learn and understand because we are going to manage human beings." (Mature entrant FGD) Other barriers to selection of specialty career preference include: Inadequate training methods used in delivering the curriculum, perceived high risk for litigation cases, strict assessments given in a rotation, lack of career guidance and students' negative experience during clinical rotations.

Discussion
The purpose of this study was to determine the specialty career preferences among nal year medical students at MaKCHS as well as to explore factors that in uence the career preferences.
From the ndings of this study, Obstetrics and Gynecology was the most preferred specialty followed by Surgery, Pediatrics and Internal Medicine. This pattern was indeed re ected in the distribution of specialists among the medical personnel registered by the UMDPC by 2019 (3). These specialty careers have been shown to dominate students' choices within most medical schools globally (11). The ndings are similar to those reported in studies done within Gambia and Pakistani which are also low resource countries (12,13). However, they differ slightly from the ndings reported in Japan where internal medicine was the most preferred specialty career with 52.7% (260/493) of the students selecting it (14).
Obstetrics and gynecology was the most preferred specialty career overall across the three choices selected by the students which can be explained by various factors highlighted during the FDGs. The Obstetrics and Gynecology department was perceived by the students to be very organized with inspirational and motivating specialists who were willing to work with students.
Surgery and Internal medicine were selected by students due to the perceived nancial rewards attached to these careers that lead to assurance of living a good life. These careers could have also been selected by students who are driven by passion (15) or are motivated by the prestige attached to these specialties (16). The preference for surgery could also be due to presence of role models as shown in a study done in Jordan among 2 nd , 4 th and 6 th year medical students(17) These role models are powerful in uences and maybe surgeons that are relatives, neighbors, personal physician, family friend, or medical school faculty members. In spite of the fact that these two careers were selected among the top four, some students were strongly against selecting them as a preference due to the demanding work load they experienced when rotating in these departments.
There was a media in uence in selection of internal medicine as a preferred career. Watching certain medical series on television like House inspired some of the students to become internal medicine specialists. House is an American medical drama whose main character leads a team of diagnosticians at a ctional hospital in New Jersey. Internal medicine may also be preferred due to the complex intellectual content required in the process of diagnosis (16).
There was a signi cant difference in the way males and females selected their specialty preferences. The most preferred specialty among male students was Obstetrics and Gynecology followed by Surgery, pediatrics and Internal medicine. The high proportion of males choosing Obstetrics and Gynecology is similar to what was seen in a study in Nigeria (18) where Obstetrics and Gynecology was their top surgical choice by the students. However it is different from te pattern in a study done in Kenya where most male students chose Surgery (40.0%) followed by Internal Medicine (15.0%) (16). Among female respondents, both Obstetrics and Gynecology and internal medicine tied for the preferred rst choice of specialty, followed by pediatrics which is similar to a study done in Jordan (17). This however is very different from that carried out in Kenya where females chose Pediatrics as their rst choice (40.0%) followed by Surgery (28.0%) (16) and a study in Bangladesh where the internal medicine and pediatrics were top choices among females (10).
This pattern of males being drawn to surgical disciplines and the females being drawn to medical disciplines can be explained by the qualitative ndings where most females expressed that they wanted a controllable lifestyle to be able to balance work and other areas of life like family responsibility while the males did not mind having an uncontrolled lifestyle as long as the career came with prestige.
Some specialties were not selected by the students as rst choice such as Anesthesia, ENT and Family Medicine. This nding is in resonance with ndings the study conducted in Jordan where students did not select such careers due to the perception that there are diminished career opportunities (17). Within the current study, the students seemed to allude to the fact that they had limited exposure to some of these disciplines, for instance one week of rotation in anesthesia. The assessment processes in some of these disciplines could also have played out as key factors. For example the students also mentioned that the assessments done were very strict and di cult which left them with low marks that are a barrier to selection of these careers as a rst choice. However, these ndings are contrasted by ndings in a study conducted in British Columbia where Family Medicine was the most preferred specialty career due ready availability of jobs for these specialists (15). In Uganda, medical students during undergraduate training have limited exposure to the Family Medicine and there are also limited job opportunities within the Uganda Health service Commission (19). Therefore, students are probably not motivated enough to select Family Medicine a rst choice specialty career.
There were students who were not interested in pursuing medical-related specialties and were interested in other areas like business or music.. The students believed that they could be able to make more money in other career elds so as to sustain family and other nancial needs.. These ndings are in agreement with those reported in a study carried out in Kenya where a proportion of the students selected nonmedical careers including business, law and music (16). About 2.5% of the students were not yet decided on the preferred specialty career. This proportion is low when compared to ndings in a study done by Maseghe et al where they found that 15.1% of the students had not yet identi ed their future specialty career path (16). The lack of a decision could be due to lack of adequate career guidance as shown by the fact that only 51(62.2%) of the students had ever received any form of career guidance. There is a need for career guidance and mentorship programs as expressed by the students in the FGDs.

Strengths of the study
This study employed a mixed methods design where the qualitative component was used to complement and explain the quantitative results. This provided a more detailed explanation to the various preferences for specialty careers by the medical students.
This study addressed specialty career preferences among nal-year medical students in Uganda and it yielded important ndings, which can be used to plan for future, more comprehensive cohort studies following up the students from the medical school to the actual time they choose the specialties. The ndings will impact the way in which institutions train their students so as to prepare them for selection of careers to avoid imbalance in the health workforce.

Limitations
The results of this study may be viewed in the context of the following limitations. The specialty career preference was measured at one point in time yet literature suggests that the choice does not remain stable over the course of medical education and medical practice after undergraduate study. The preferences may be re ned over the internship period or the time they start practicing as licensed doctors.
Some students may not have developed a preference for any specialty by the time of the study and may have given speculative information. However, this was minimized by selecting medical students in their nal year who had been exposed to the various specialties during their clinical rotations.
The study was conducted in only one medical school within MakCHS therefore the results may not be generalizable to the entire country since the various universities are in different settings. Inspite of this, Makerere University has the biggest intake of medical students compared to other public universities.

Conclusion
The most preferred specialty career among the fth year medical students was Obstetrics and gynecology. The other top overall careers selected include pediatrics, surgery and internal medicine. There were career options that were not selected as a rst choice including anesthesia and ENT due to limited exposure.
The factors in uencing the specialty career preferences include gender where males preferred the surgical disciples compared to females who preferred the medical disciplines that could allow them to balance work and family life. Other factors facilitating career preferences included controlled lifestyle preference, assurance of good life, inspirational specialists. The barriers for selection of some specialty careers included bad experience in clinical rotations, lack of guidance, perceived poor and miserable specialists.
The ndings of this study have several implications on proper distributions of specialists across disciplines thus the following recommendations are suggested. There should be improved planning of the curriculum within all the clinical disciplines to improve students' experiences and foster their exposure to the various specialty career options. Strong programs for career guidance and mentorship should be put in place to ensure that students make a well informed decisions in selection of their preferred careers. Further research is required in form of a follow up study among students form undergraduate through post graduate to post-education phase in order to determine the factors associated with the eventual specialty career choices. with protocol number #REC REF 2019-045. Written informed consent was obtained from all study participants and their willingness to participate was emphasized and they were free to withdraw from the study at any time. All data collected was kept securely in a password protected computer, and the physical data under lock and key. Con dentiality was maintained by restricting access to study data to only the investigators, and not using speci c name identi ers in the data sets.

Consent for publication
Not applicable Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.   was recorded as well. Sponsorship is usually based on who was paying tuition for the student throughout the period of medical school. Government sponsorship is given to a number of students who enter medical school on merit. Those who are not sponsored by the government have to pay the tution on their own or get private organizations to cover them which is also known as Private sponsorship.  Table 2 shows the parents' level of education which is divided into four levels. The rst level is none which is basically those who have never received a formal education. The second level is Primary, third is secondary ad fourth is tertiary. The tertiary level consists of all post-secondary education including universities, colleges, technical training institutes, and vocational schools. The participants were asked if they had ever received any form of career guidance counselling and whether they were interested in a workshop. This was asked so as to assess the students' willingness and interest in counselling. The results were presented as frequency (second column) and percentage (third column).  Table 3 shows the distribution of First choice specialty preferences among the various sociodemographic characteristics of the fth year medical students. The students were asked to select three preferences and rate them as rst second and third choice. The most commonly selected specialty as rst choice was obstetrics and gynecology followed by surgery, internal medicine and pediatrics. About 39 out of 135 selected other specialties including public health, psychiatry, emergency medicine, orthopedics, dermatology, basic sciences, radiology, ophthalmology, Business, Clinical epidemiology and Biostatistics, Forensic Medicine and voluntary medical work. Among these was also a proportion of students who were not yet decided. These were then divided in to ve groups consisting of Obstetrics and gynecology, Surgery, Internal Medicine, Pediatrics and others. The chisquare test was used to determine if the sociodemographic characteristics had a relationship with preference of career specialties. This test was used to compare the selection of careers across the various categories of the sociodemographic characteristics. The following variables were found to have a signi cant relationship with rst specialty career preference: Sex (p = 0.04), marital status (p < 0.01), having children (p < 0.01) and sponsorship (p = 0.04). The factors that had a relationship with selection of rst choice specialty were considered for further analysis using multiple logistic regression. This was done so as to determine the odds of selection of a surgical specialty (obstetrics & gynecology, surgery) vs a non-surgical specialty. The table shows that sex had a signi cant odds ratio (OR) with 63% less likelihood for female medical students to select a surgical speciality when compared with male medical students (OR=0.37; 95%CI (0.17-0.84); p=0.02).