Challenges Facing Undergraduate Medical Education in Ambulatory Care Clinics at Tertiary Care Hospitals

Background: Medical education has been rapidly growing and transforming due to the enormous evolution of medicine. There have been many proficient ways to learn in medicine, but academic lectures, attending wards, and ambulatory care clinics (ACC) remain the three main ways of gaining clinical knowledge and experience for medical students. Over the past decade, there has been a dramatic shift in care by focusing on ambulatory care rather than inpatient settings, which provides a golden opportunity to reinforce medical education. Purpose: Most of the published studies that have focused on the teaching barriers in ACC were descriptive rather than analytic studies. Herein, we aim to detect and determine the barriers to teaching in ACC settings using qualitative analysis. Methods: This is a cross-sectional, observational study, involving medical students in their clerkship years (i.e., fourth and fifth) from two different medical colleges in Riyadh, Saudi Arabia. Faculty who are involved in undergraduate medical education in both colleges were also included. Main Results: A total of 387 medical students studying at the two universities were enrolled in the study. Most of the participants preferred attending outpatient clinics with consultants (44.2%) and the majority preferred attending internal medicine (IM) and IM subspecialties clinics (40.4%). Regarding the challenges, students believe the top three barriers are related to: faculty (39%), environment (34.8%), and patients (14.8%). Faculty on the other hand see that the top three barriers are related to environment (55.6%), patients (24.4%), and faculty (20%). Conclusion: Undergraduate medical education in outpatient settings has many challenges. In our study, the most significant challenges were COVID-19-related restrictions, patient refusal, and insufficient time for teaching. Future studies are needed to investigate these barriers and explore potential solutions that can decrease their burden on undergraduate medical education.


Introduction
Understanding the best means to enhance medical education has been a primary goal for many educators and researchers. During the past decade, many changes in the healthcare system have been developed. One of the most significant changes was the dramatic shift of care towards ambulatory rather than the traditional inpatient care settings [1]. Since inpatient cases are more likely to be critically ill or under subspecialty conditions, they are less representative of common case presentation in clinical practice. This emphasizes the importance of clinical education in ambulatory care settings in addition to inpatient settings. Unfortunately, many programs have been described as flawed in outpatient training and required major changes [2].
The ways that stand the most are academic lectures, attending wards, and ambulatory care clinics (ACC). All of these ways positively impact the educational experience of medical This is an observational cross-sectional study that was conducted at two major medical colleges in Riyadh, Saudi Arabia namely: Alfaisal University (AU) with King Faisal Specialist Hospital & Research Center (KFSH&RC) as the clinical training center, and King Saud University (KSU) with King Khaled University Hospital (KKUH) as the clinical training center. This study was conducted between 1 March 2020 and 30 April 2021. The study involved two types of participants: medical students and faculty members from both colleges. The inclusion criteria for the medical students included 4th and 5th year of undergraduate program attending the college of medicine at either AU or KSU who completed a minimum of 12 weeks of clinical clerkship. The inclusion criteria for the faculty members included faculty members of AU or KSU who are actively involved in ACC education.

Sample Size
A convenience sampling technique of 405 and 375 participants for the medical student sample was determined from AU and KSU medical students, respectively. As for the faculty, a convenience sampling technique of 297 and 180 participants was recruited from AU and KSU, respectively. The sample size was calculated by Asher Sample Size Calculator.

Data Collection/Instrument
The email list of all eligible medical students and faculty members from both medical schools was obtained from administration offices at both sites. An email template for the medical student questionnaire and another email template for faculty members were prepared. Since the current literature lacks a validated tool that could be used specifically to assess challenges that students might face in ACC, we used the literature to build this tool based on potential barriers. Questionnaire validation was subject to a panel of experts for content and face validity. An email was sent to five out of the six experts who evaluated the instrument and had agreed on the value and the usefulness of the used instrument. The experts forwarded their feedback via emails and it is beyond our ability to retrieve and perform Kendall's W test at this stage. Unfortunately, because the survey was being amended as the email feedback was received from the panel, emails were not kept in record.
The questionnaire for medical students included three main domains: demographic information, perception of general barriers to ACC education, and perception of specific barriers. Emails were sent out to eligible medical students and complete responses were included in the study. Questionnaires emailed to faculty members included three main domains: socio-demographic information, perception of ACC education, and perception of barriers to ACC teaching. Emails were sent out to eligible faculty members and complete responses were included in the study. Both student and faculty surveys can be found in the Supplementary Material.

Data Analysis
Data were analyzed using SPSS statistical software. Descriptive statistics (mean, standard deviation, frequencies, and percentages) were used to describe the quantitative and categorical variables. The Kolmogorov-Smirnov test of statistical normality and the Histograms were used to assess the statistical normality of the measured metric variables, and Levene's test was applied to assess the statistical equality of variance assumption. The tested metric variables were found approximately normally distributed with no violation of homogeneity of variance statistical assumption.
Pearson's chi-square test was used to assess the association between the categorical variables. Student's t-test was used to compare mean values. A p-value ≤ 0.05 and 95% confidence intervals were used to report the statistical significance and precision of the results (Figure 1). medical student questionnaire and another email template for faculty members were prepared. Since the current literature lacks a validated tool that could be used specifically to assess challenges that students might face in ACC, we used the literature to build this tool based on potential barriers. Questionnaire validation was subject to a panel of experts for content and face validity. An email was sent to five out of the six experts who evaluated the instrument and had agreed on the value and the usefulness of the used instrument. The experts forwarded their feedback via emails and it is beyond our ability to retrieve and perform Kendall's W test at this stage. Unfortunately, because the survey was being amended as the email feedback was received from the panel, emails were not kept in record.
The questionnaire for medical students included three main domains: demographic information, perception of general barriers to ACC education, and perception of specific barriers. Emails were sent out to eligible medical students and complete responses were included in the study. Questionnaires emailed to faculty members included three main domains: socio-demographic information, perception of ACC education, and perception of barriers to ACC teaching. Emails were sent out to eligible faculty members and complete responses were included in the study. Both student and faculty surveys can be found in the Supplementary Material.

Data Analysis
Data were analyzed using SPSS statistical software. Descriptive statistics (mean, standard deviation, frequencies, and percentages) were used to describe the quantitative and categorical variables. The Kolmogorov-Smirnov test of statistical normality and the Histograms were used to assess the statistical normality of the measured metric variables, and Levene's test was applied to assess the statistical equality of variance assumption. The tested metric variables were found approximately normally distributed with no violation of homogeneity of variance statistical assumption.
Pearson's chi-square test was used to assess the association between the categorical variables. Student's t-test was used to compare mean values. A p-value ≤ 0.05 and 95% confidence intervals were used to report the statistical significance and precision of the results (Figure 1).

Ethical Considerations
The study received approval from the office of research affairs and ethics committees of KFSH&RC (Project Number: 2211197) and KSU (Project Number: E-22-6592) and was conducted according to the hospitals' bylaws. Participation in this study is voluntary, and

Ethical Considerations
The study received approval from the office of research affairs and ethics committees of KFSH&RC (Project Number: 2211197) and KSU (Project Number: E-22-6592) and was conducted according to the hospitals' bylaws. Participation in this study is voluntary, and no personal identifiers were collected. Consent to participate is incorporated in the survey cover page, and data were stored under password protection.

Results
In this section, we will discuss the results of both student and faculty surveys on the learning process at ACC. In each of the two sections (Sections 3.1 and 3.2), we will go over the characteristics, general perception, in addition to environment-, faculty-, patient-, and student-related barriers. In the student survey, we will also evaluate how receptive students are to virtual clinics (VC), and we will discuss their preferred clinic (based on specialty).

Demographic Factors
A total of 405 eligible students attending AU and 375 eligible students attending KSU were invited to participate in the survey. The response rate for each medical school was 37.8% and 62.4%, respectively. The total number of medical students who agreed to participate and completed the survey was 387 (153 and 234 from AU and KSU, respectively). Female participants were seen in 45.7% of the whole sample with 62.1% and 35% were female students from AU and KSU respectively. Fourth-year medical students participated in 49.1% (54.2% and 45.7% from AU and KSU, respectively) ( Table 1). Most students preferred attending clinics with consultants (44.2%) followed by senior residents (25.3%) and fellows (15.5%). Thirty-nine percent considered faculty relatedbarriers to be the main source of challenges. While institutional-related barriers were perceived as a challenge by 34.8% of the sample. Participants attending KSU perceived significantly (p = 0.022) less faculty-related barriers to outpatient ACC learning compared to participants from AU. Participants from KSU were more likely to perceive COVID-19related factors as a barrier than AU participants (p = 0.018). Perceived patient-related barriers, medical student-related barriers, and other factors did not show significant differences between the two groups.

Virtual Clinics and Preferred Clinics
When asked to rate how much they support involving medical students in VCs using a 1-5 Likert-like scale, with 5 denoting a great extent of support, the results showed that the students' overall mean support level was measured with 3.1/5 points, SD = 1.44 points, signifying that these students had between a moderate-to-great extent of agreement with learning engagement in VCs. In an independent sample, the t-test showed that the AU medical students expressed greater support to the virtual clinical sessions (mean = 3.28) as compared to the KSU medical students (mean = 2.94, p = 0.026). Attending internal medicine (IM) and IM subspecialties clinics were preferred by 40.4% followed by Ob-Gyn (20.9%), then pediatrics (20.1%) ( Table 2).

Environmental Barriers
Looking closer at environmental factors (Table 3), "restrictions due to COVID19" was perceived as a barrier by more than half (52.5%) of the participants. Lack of structured teaching objectives was the second most perceived barrier to ACC learning according to 24.3% of the participants. Nineteen percent of the sample also perceived "inadequate distribution of students among ambulatory clinics" as a barrier to ACC learning, with only (4.1%) considering "inappropriate or small clinic rooms" as a challenge.

Faculty-Related Barriers
The most perceived barrier by the sample in the faculty-related section was the "inadequate supervision and teaching by faculty" (48.6%). This is followed by "lack of time for teaching by doctors due to intense patient agenda" and "inappropriate or absent feedback", which was perceived as a barrier by (33.9%) and (12.1%) of the participants, respectively. The variability between the two groups in the perception of faculty-related barriers did not yield a statistical significance, as (49.1%) of KSU participants and (47.7%) of AU participants perceived "inadequate supervision and teaching by faculty" as a barrier to ACC learning. The ranking of the rest of the perceived faculty-related factors was similar in participants from both universities.

Patient-Related Factors
Around 49% of the entire sample perceived "patient refusal to be seen by medical students" as a patient-related factor limiting their ACC learning experience. While "no follow-up/continuity of care for cases attended" and "lack of suitable patients for teaching" were considered a barrier by (25.3%) and (25.1%) of the sample, respectively. In a comparison between the two universities, (51.3%) of KSU participants and (47.1%) of AU participants perceived patient refusal as a barrier. Participants from both universities had similar perceptions in terms of patient-related factors.

Student-Related Factors
The most perceived student-related barrier was "the increasing number of students attending the ambulatory clinic" (39%) and (30.5%) of the sample perceived "not enough time to attend clinics" as a barrier to their ACC learning experience. Some students believed "lack of student commitment" (18.6%) and "no additional information gained compared to inpatient setting" (11.9%) to be a barrier to their ACC learning.
The four different groups of barriers significantly differed between students in the two universities, as relates to their unique characteristics (Table 4).

Demographic Factors
Ninety faculty members from both KSU and AU agreed to participate and completed the questionnaire (68.9% of the participants were from AU, while the rest were from KSU). Males were 83.3% of the participants. Most of the responses were from associate professors (35.6%), followed by assistant professors (33.3%). Professors shaped 26.7% of the sample, with the least participants being "lecturers" (4.4%). The most common specialty of the participants was internal medicine and its subspecialties (43.3%), followed by pediatrics (23.3%), then neurosciences (6.7%) ( Table 5).

Faculty's General Perception
When participants were asked how comfortable they were with medical students attending clinics with them, 35.6% of participants answered "uncomfortable" and 25.6% answered, "somewhat comfortable". The rest of the sample (38.9%) ranged from "moderately comfortable" to "very comfortable". The overall comfort rating was rated with 2.42 points out of 5 points, which highlights a moderate comfort perceived by participants. Participants from KSU and AU did not differ in their comfort toward medical students attending clinics and 93.3% from both universities believe that outpatient experience is important for medical students and should be integrated into their curriculum.
The environmental-and institutional-related barriers were believed to be the main barrier to proper and efficient medical student education in outpatient settings (55.6%) ( Figure 2). While patient-related variables were considered a barrier by 24.4% of the participants, followed by faculty-related variables, which were considered a barrier by 20% of the sample. The two groups did not differ in the general perception of barriers toward medical student education in outpatient settings. Around one-third of the faculty (36.7%) support the involvement of medical students in virtual clinics to a very great extent. The overall support rating was 3.44 out of 5, indicating that the sample supported medical students attending virtual clinics to a great extent. An independent sample t-test suggested that the medical teachers at KSU may have significantly lower levels of support for involving students in virtual clinics (mean = 2.96) compared to the teachers in AU (mean = 3.66), p = 0.045 (Table 6). medical student education in outpatient settings. Around one-third of the faculty (36.7%) support the involvement of medical students in virtual clinics to a very great extent. The overall support rating was 3.44 out of 5, indicating that the sample supported medical students attending virtual clinics to a great extent. An independent sample t-test suggested that the medical teachers at KSU may have significantly lower levels of support for involving students in virtual clinics (mean = 2.96) compared to the teachers in AU (mean = 3.66), p = 0.045 (Table 6).

Environmental Factors
The most common environmental factor (Table 7) perceived as a barrier by the participants was restrictions due to COVID-19 (26.7%). In addition, 21.1% of the sample considered inadequate structuring and distribution of students in ACC as a barrier, and another 21.1% perceived "inappropriate or small clinics rooms" as a barrier toward medical teaching. Comparing the environmental factors between the two sites did not show statistical significance.

Faculty Related Factors
Seventy percent of participants considered "insufficient time for teaching due to intense patient agenda" to be a barrier to medical teaching in outpatient settings. While 20% believed that medical teaching not being part of their key performance indicators (KPI) as a barrier toward medical teaching.

Patient Related Factors
Patient refusal was perceived as a barrier toward medical teaching in outpatient settings by 41.1% of participants, while 24.4% of participants perceived "fear of patient dissatisfaction" as a barrier toward medical teaching. When the participants from the two sites were compared, no statistical significance was found.

Student Related Factors
Half (51.1%) of participants believed that the "lack of student commitment and interest in learning" is a barrier toward medical teaching in outpatient settings. While another 48.9% believed that "the increasing number of students" is a barrier toward medical teaching in outpatient settings. When comparing between the two universities, perceived student-related barriers did not significantly differ between the two groups.
The differences in the perceived barriers between the faculty members of both universities are summarized in (Table 8).

Discussion
Several medical schools around the world have expanded and further developed the education process through ACC [9]. In the US, medical students spend at least a third of their total clinical experience in ACC [10]. While all methods of medical education are needed and each contributes differently, the teaching methods used in ACC are different from those commonly exercised in medical wards [11,12]. In most wards, the teaching rounds of inpatients are usually scheduled and are limited to a specific duration that is usually focused on a specific agenda [13]. Medical students are usually asked to examine patients who have already been diagnosed and have a treatment plan established. However, encounters in ACC can be related to either acute or chronic issues, giving students a broader perspective and exposing them to basic, yet essential clinical scenarios. This can aid in enhancing the learning experience for students and expose them to a wide variety of medical conditions [14,15].
The inadequate time of outpatient teaching and the reliance on unsuitable methods of instruction can be extremely hampering [16]. Adams et al. classified the ambulatory teaching costs that have been cited in the literature into seven categories: (1) increased office time for teaching purposes, (2) reduced productivity of the teaching site or teaching physician, (3) elevated operation costs, (4) trainee support direct cost, (6) losses of revenue due to the patients' types seen in sites of teaching, (5) patients' loss, and (7) clinical costs elevation due to patterns of students practice and the use of extra medical resources [17].
While there is existing literature on the different barriers to effective learning in ambulatory care clinics, many simply describe them and divide these issues into broader categories. This study aims to identify these barriers using qualitative analysis to investigate which specific factors were the most significant. The study was conducted in AU and KSU, two of the largest medical colleges in Saudi Arabia providing us with an adequate number of participants. By surveying both medical students and faculty members in these two major academic centers, we ensured factors affecting all parties involved in the explored learning process were mostly addressed.

Environmental Factors
The most significant environmental barrier to effective ambulatory clinic learning mentioned by both students and faculty members were restrictions put in place due to COVID-19. While this may seem to be a temporary barrier due to the current pandemic, it serves as a reminder of the considerable burden the pandemic or other infectious disease outbreaks has had on the medical education system [18][19][20]. It has been challenging for medical colleges to adapt to this unprecedented situation, with many opting for virtual learning. Harries et al. found that students recognized significant disruption to their medical education, with many willing to accept the risk of acquiring the infection and return to previous guidelines [21]. While the safety of both students and patients should be of utmost importance, innovative solutions must be developed to help minimize the disruption in the medical education system.
Medical students and faculty members also highlighted the need of optimizing the distribution of students and the importance of structured teaching objectives in clinics [14,22,23]. While these can be significant barriers, adjustments to both teaching curriculums and student schedules can ensure these barriers are reduced, providing a more structured and productive clinical teaching sessions to medical students. Studies suggest that allowing students the opportunity to talk to patients and perform clinical examinations alone as part of the learning objectives was greatly beneficial. Students reported increased learning opportunities as well as more confidence in clinical skills such as history taking and clinical examination [24,25].

Faculty Related Factors
Under faculty-related factors, many students and faculty members stated that the relatively insufficient time for teaching due to intense patients' agenda was an important barrier to effective outpatient clinic teaching. Inadequate supervision by faculty members was another significant barrier perceived by students, which can be explained by the time constraints during most clinic visits. This is consistent with several studies on the topic which suggest that the time pressures of ambulatory clinics often hindered learning, especially in health institutions with high patient demands [26][27][28][29]. With their own clinical workload increasing and continued pressures to increase clinical productivity, balancing both patient care and student education can be extremely challenging for faculty members. Despite this, the literature report that most faculty members receive very little to no explicit training on the processes of teaching. Faculty development in the form of workshops and conferences could serve as an essential tool to aid in dealing with the complexities of medical teaching today [30]. In addition, a study looking into protecting teaching time in an outpatient clinic recommended that academic medical centers must reaffirm education as one of their central missions and ensure all policies are in line with this vital goal [31]. Emphasis must be placed on the importance of spending time on teaching during clinic visits to provide high-quality medical education to young physicians.
Some faculty members also mentioned the fact that teaching was not integrated as part of their KPI, and annual evaluation was another important barrier to effective teaching. This was an interesting finding as there are little to no studies that have cited the lack of ACC teaching on KPI as a significant barrier to effective teaching. However, this presents another opportunity for medical schools to clearly communicate and highlight the importance of ambulatory clinic teaching to their faculty members, in addition to finding a suitable way to minimize the negative impact of time spent on teaching on KPI. More research must be done to investigate whether adding key performance indicators related to teaching would improve outpatient education.

Patient Related Factors
Most students and faculty members agreed that patient refusal to be seen by medical students was an important barrier to effective ambulatory clinic education. However, this opposes other studies which suggest that only a minority of patients refuse or have negative feelings towards the involvement of medical students in their care [32][33][34][35]. Marwan et al. investigated factors that affect patient refusal and found that only a minority of participants would refuse students to take their medical history with the presence of a supervising doctor; however, the refusal rate was higher when the patients were asked if they would permit medical students to take their history unsupervised [32]. Faculty members and students should learn to master essential communication skills to aid in convincing patients and ensure that their refusal is not a barrier to learning.
Fear of patient dissatisfaction was another significant barrier that faculty members constantly faced. Patient satisfaction has become one of the cornerstones of modern medicine and has always been a top priority for all physicians and healthcare providers. Concerns for patient dissatisfaction have contributed to the unwillingness of many institutions and faculty members to embrace medical education [36]. Despite popular belief, studies looking into the effects of medical student teaching in an outpatient setting found that patient satisfaction was not affected by the presence of the students [36,37]. This further confirms that medical student education in ambulatory clinics should always be encouraged, and that patient refusal or fear of dissatisfaction should not be considered a barrier to effective learning.
Continuity of care is a fundamental component of primary healthcare and is associated with improved clinical outcomes and adherence to treatment regimens [38]. Despite this, maintaining adequate continuity for students to experience in ambulatory clinics remains a challenge, most notably, due to students' division of time between inpatient wards, outpatient experiences, and other specialties [39]. We found that students and faculty members agreed that lack of continuity heavily affected learning in an ambulatory clinic setting which is consistent with the current literature [38][39][40].
Another significant barrier that students faced in clinics was the lack of suitable patients for teaching. This was an expected result since our study was centered around teaching in tertiary care hospitals, where most patients tend to have subspecialty conditions that may be less suitable for undergraduate teaching. Several studies also stressed the importance of selecting suitable patients in accordance with the student's level and learning objectives to maximize effective learning [15,22].

Student Related Factors
We found that an increasing number of students attending clinics was an important barrier that significantly compromised ambulatory clinic education which is consistent with many studies [3,41,42]. A higher number of students in a given clinic will affect the time and exposure for each individual student to spend with patients, an essential component of outpatient learning. This will also compromise the ability of faculty members to provide precise feedback and evaluations, which are an important part of effective learning. Logistical solutions must be put into place to ensure an adequate number of students are distributed to each clinic to facilitate a positive learning environment for both faculty members and students.
Lack of student's commitment and interest in learning was a significant barrier that was mentioned by our faculty members. This was also noted in another study which found that students who did not take charge of their learning tend to show little interest in learning [41]. Azher et al. supported a style of outpatient teaching where students are not just observers but are actively involved in patient care, which was the preferred participation style from a student's perspective and significantly improved their interest [43].

Study Implications
Learning in ambulatory care clinics has become an essential component of medical education that provides many unique learning opportunities. Compared to inpatient wards, ambulatory clinics provide students with an opportunity to encounter a plethora of typical conditions and presentations and provide students with vital training in communication skills, preventive medicine, and psychosocial elements of disease [27].
Despite these benefits, ambulatory clinics host several barriers which can greatly hinder the overall learning experience. While there are several studies that have described these barriers, few have taken an analytic approach to identify which factors have the single biggest impact on learning from both the tutor and the student's perspective. Our study in two of the largest tertiary care hospitals in Saudi Arabia aimed to fill this gap and identified some of the biggest challenges faced when learning in ambulatory care clinics. Insufficient time, patient refusal, increased number of students, and restrictions during the COVID-19 pandemic were the biggest barriers we observed during our analysis. As the field of medicine is constantly evolving, the content and delivery of medical education continue to be globally refined. With these significant barriers identified, our findings can serve as a guide to medical educators and curriculum developers to ensure the necessary changes are made, to facilitate effective clinical teaching and the overall improvement of the medical education system.

Limitations
The two contributing teaching hospitals have a wide range of world-class facilities and infrastructure tailored for complex cases requiring tertiary care. This could serve as a point of bias, as most outpatient teaching is conducted in less specialized centers. Surveying students and faculty members from primary and secondary health care facilities would provide more generalizable results. The convenience sampling could have biased the collected samples, which may not necessarily be representative of the other students or faculty from either university. Moreover, a larger study may reveal other significant barriers which may not be present in these advanced tertiary care centers. Additionally, future studies should aim to include a larger sample size. Another limitation is that our study was carried out during the COVID-19 crisis, and quantitative data were not compiled from the experts' feedback regarding the used instrument; therefore, the Q-sorting or Kendall's W analysis was not used because much of the correspondence was lost during the pandemic crisis, so future research could explore if the reported barriers change after the pandemic is over.

Conclusions
Our cross-sectional study identified ambulatory clinic teaching barriers as reported by medical students and faculty members in two tertiary care hospitals. Insufficient time for teaching, patient refusal, and restrictions due to COVID-19 were the most significant barriers. Faculty members also highlighted that teaching was not integrated as part of the KPI. Identifying these factors paves the way for future studies to investigate these specific issues and help develop innovative solutions to eliminate these barriers.
While it is difficult to be generalized, the outcomes of this study give an idea about possible challenges that medical students might face when attending ACC. This study paves the way for future work, which can be conducted to evaluate these challenges on a bigger scale and in different hospitals all over the globe. Moreover, and based on our results, we think it is important for medical schools to understand all the challenges that are faced by their students through a validated and concise tool. This can highlight the major challenges and allow decision-makers to target them individually; aiming to improve medical education in ACC settings.