A Holistic Evaluation of Clinical Competence and Professional Attitudes in Diabetes Health Care: Insights From a Multinational Cohort

Abstract

This article reports on a pilot study of the Joint Diabetes and Quality assessment tool, which evaluates clinical competence, attitudes, and beliefs about diabetes care among physicians, in the Kingdom of Saudi Arabia. Participant responses revealed a significant diabetes care proficiency gap, a disconnect between physicians’ self-perceived competence and their actual clinical abilities in diabetes management, and high physician stress levels, all of which merit deeper analysis and further evaluation. Expanding this research to include larger and more diverse national and international cohorts could provide comprehensive insights into competency levels and attitudes about diabetes care in the global medical workforce.

The International Diabetes Federation reports that 365 million adults worldwide have diabetes, of whom >90% have type 2 diabetes, and this number is expected to reach 700 million by 2045 (1). Diabetes contributes to morbidity, premature death, and excess mortality, consuming about 12% of global health spending (1–3). This global issue highlights specific regional contexts, such as in the Kingdom of Saudia Arabia (KSA), where diabetes has a significant impact.

Physicians treating diabetes must stay updated on rapidly evolving evidence, medications, and technologies in this advancing field. Doing so is particularly challenging for primary care and internal medicine physicians (PCPs) who manage diabetes with other complex conditions. The fast pace of developments, especially in areas affecting diabetes care, such as cardiovascular and renal outcomes, has led some diabetes guidelines to adopt a “Living Standards of Care” model for updating recommendations in a timely manner (4). This model’s necessity is clear in regions such as the KSA, where diabetes management complexities are magnified.

Although physicians are traditionally viewed as lifelong learners, the accuracy of their self-assessment abilities is not well established. Studies from the United States offer some insights (5,6), but a broader, multinational perspective, including in regions such as the KSA, is essential for a more comprehensive understanding.

Diabetes clinical guidelines are vital for training health care professionals and shaping health care planning and delivery. However, consensus is lacking on the most effective approach to guideline implementation for chronic disease states such as diabetes (7,8). A thorough needs assessment and standardized method for evaluating clinical competence and attitudes toward patient care are needed (7,8). Studies often omit clinical competence evaluation (9–11), and those that include it rarely validate their assessment tools (12). This article seeks to bridge this gap, with a focus on regions such as the KSA, where the high diabetes burden necessitates effective guideline implementation.

The Case of the KSA: High Diabetes Disease Burden, Suboptimal Performance, and Workforce Shortages

The Joint Diabetes and Quality (JDQ) assessment tool evaluates clinical competence, attitudes, and beliefs about diabetes care among physicians. The KSA provides a crucial context for assessing the JDQ assessment framework within the public sector, given the region’s substantial diabetes burden (13–15). In the KSA, factors such as high obesity rates, rapid economic growth, urbanization, and increased life expectancy contribute to a diabetes prevalence of 15.8–31.6% (1,15), with up to 40% of cases undiagnosed (13–15). Management challenges are compounded by many patients failing to meet key health targets (13,16).

In the KSA’s public sector, which operates 60% of the nation’s clinics and hospitals, delivering optimal diabetes care is challenging because of a shortage of trained PCPs (17). This situation has led to efforts to recruit internationally trained physicians from diverse educational and training backgrounds, whose competence and attitudes toward diabetes care in the KSA are not yet fully understood. The Saudi Commission for Health Specialties regulates medical licensure, reflecting global practices (18). Continuing medical education (CME) programs must identify and address gaps in physician competency; this effort is crucial in the KSA, where diabetes care complexities are heightened.

Given the global shortage of endocrinologists and diabetologists (19), PCPs predominantly handle diabetes care in the KSA. Public sector initiatives aim to improve diabetes metrics and reduce specialized referrals (19). Collaborating with a United States–based academic center, the KSA public sector has launched a diabetes guideline implementation training program to enhance workforce capabilities in managing the diabetes epidemic. In this study, we assessed a sample of physicians at the start of this initiative, focusing on clinical competence, self-efficacy, professional satisfaction, and attitudes toward diabetes care and education.

Research Design and Methods

Survey

In this study, 450 physicians providing diabetes care at primary care centers or diabetes centers (primary care services that may offer more support for people with diabetes) participated in a pre-program survey after an informational session. These centers were staffed by PCPs with expertise in diabetes. Participation was voluntary. A total of 249 physicians completed the survey, yielding a response rate of 55%. No financial incentives were offered for participation.

We obtained permission to revise and adapt a previously validated survey tool, initially designed for another chronic disease (20), for use in this study. This adaptation focused on assessing physicians’ attitudes and perceptions regarding diabetes care. It included measures of self-reported self-efficacy or competence, feelings of burden/anxiety in patient care, professional isolation, satisfaction with professional life, access to CME opportunities, access to expert opinion, knowledge transfer to clinical team members, and perceived patient benefits from the physicians’ CME participation and resulting reduced referrals to specialized clinics. The survey modifications were reviewed by diabetes experts before distribution. The adapted survey is provided in the Supplementary Material. This tool was chosen because of a lack of other validated surveys in chronic conditions that would assess competence alongside other relevant domains for this study.

We segmented diabetes care into 23 clinical domains based on this survey (Figure 1), aligning with Knowles’ educational theory framework (21). This categorization aimed to tailor and target the educational program to the specific needs of this physician group and to assess the relevance of the program content. This approach also provided an opportunity for participants to engage in the planning of subsequent educational activities based on this initial needs assessment.

Figure 1.

Percentage of correct answers by clinical domain and percentage of respondents reporting moderate to high burden by domain. DM, diabetes; Eval, evaluation; Mod-High, moderate to high; Patho, pathophysiology; QI, quality improvement; T1DM, type 1 diabetes; T2DM, type 2 diabetes.

The survey required participants to rate their self-efficacy on a Likert scale from 1 to 7 (with 1 indicating no knowledge or skill and 7 indicating expertise and teaching ability) and their level of burden or stress in diabetes patient care on a Likert scale from 1 to 5 (with 1 indicating not at all burdened and 5 indicating highly burdened).

Clinical Competence Assessment Exam

For the evaluation of physicians’ clinical competence, our study used a validated, open-book, case-based, multiple-choice question exam. This format was chosen to emulate real-world scenarios in which clinicians often have access to online resources or clinical references in their practice. The case-based method, focusing on clinical decision-making, was used as a cost-effective and practically feasible approach to achieve a near-authentic assessment experience (22,23).

The exam content was developed and peer-reviewed by diabetes experts, guided by learning objectives that align with the standards of care set forth by the American Diabetes Association and the local health care system (24,25). An advisory group composed of endocrinologists/diabetologists and PCPs from both the United States and the KSA was involved in this process. Diabetes clinical domains were assigned varying degrees of importance, and the exam questions were weighted accordingly. Greater weights were given to domains crucial for everyday primary care practice, such as noninsulin and insulin pharmacological therapy, whereas specialized areas more suitable for specialist management were given lower weights (22,23).

The exam questions, crafted by the principal investigators (N.A.E. and M.A.) and reviewed for technical accuracy, style, readability, and potential bias toward subgroups, underwent statistical validation by Integrity Castle Rock Research of Ontario, Canada. The validation process yielded a difficulty mean of 53%, a Kuder Richardson Formula 20 reliability of 87%, and a fair negative discrimination value of 0.224 (26). No collusion was detected during the exam.

The exam was administered via the StudyTRAX research portal, with the passing cutoff score determined by the modified Angoff method, a systematic approach used to set a passing score for exams. It involved a group of experts estimating the probability of a minimally competent candidate correctly answering each exam question. These estimates were averaged to determine a recommended passing score, aiming to reflect a fair and accurate threshold of minimal competence. Participants completed this time-bound, open-book exam remotely.

This study was approved by the Harvard Medical School Institutional Review Board (IRB), the KSA Ministry of Health (MOH) IRB/Ethics Committee, and the Joslin Diabetes Center IRB.

Statistical Analysis

We calculated descriptive statistics for all variables as mean ± SD or median (interquartile range [IQR]) and assessed the normality of continuous data. Based on the distributions and types of variables involved, we tested associations using a Pearson correlation analysis, Wilcoxon rank sum test, independent samples t test, or Kruskal-Wallis test, as appropriate.

Data Availability

The datasets generated and/or analyzed during this study are not publicly available due to KSA MOH regulations but may be made available after prior permission is obtained from the MOH upon reasonable request to the corresponding author.

Results

Physician Characteristics

Table 1 delineates the self-reported demographics and professional backgrounds of the participating physicians. A substantial proportion (96%) were 30–60 years of age, and males constituted 64% of the cohort. Notably, 97% of the physicians had >5 years of professional experience, and more than half had more than a decade in the field.

Table 1.

Participant Characteristics (n = 249)

Characteristic	n (%)
Age, years  20–29  30–39  40–49  50–59  ≥60	7 (2.8) 128 (51.4) 77 (30.9) 35 (14.1) 2 (0.8)
Sex  Male  Female	159 (63.9) 90 (36.1)
Years since medical school graduation  <5  5–10  >10	8 (3.2) 66 (26.5) 175 (70.3)
Years of practice  <5  5–10  >10	21 (8.4) 89 (35.7) 139 (55.8)
Postgraduate training  Fellowship (diabetes/endocrinology)  Residency (family practice/internal medicine)  Ongoing training (residency)  No formal training	24 (9.6) 145 (58.2) 2 (0.8) 78 (31.3)

With regard to educational background, 58% had completed residency training in family medicine or internal medicine. Ten percent reported completing a fellowship in diabetes or endocrinology, whereas only a minor fraction (0.8%) had not yet completed their residency training. Interestingly, a significant portion (31%) lacked formal training in internal medicine, family medicine, diabetes/endocrinology, or other relevant fields. Our study included a diverse array of physicians from 22 countries, reflecting varied medical school systems and training backgrounds (Table 1).

We observed a positive correlation between postgraduate training completion and higher self-efficacy (P <0.01), as well as enhanced clinical competence scores (P <0.01). Moreover, a stronger association was evident between increased clinical experience and improved self-efficacy (r = 0.263, P <0.001) (Figures 2 and 3).

Figure 2.

Relationships between training and self-efficacy score (A) and between training and knowledge exam score (B). max, maximum.

Figure 3.

Relationship between years of practice and self-efficacy score.

Clinical Competence Scores

The percentage of physicians correctly answering at least 70% of the questions on the clinical competence exam was relatively low at 12%, and the percentage of those correctly answering at least 65% of the questions was also low at 23% (Figure 1). Diabetes units scored higher on average than primary care centers.

Our results showed pronounced gaps in clinical competence among physicians with regard to certain areas of diabetes care. The most significant challenges were found in managing prediabetes and diabetes prevention, understanding the pathophysiology of type 2 diabetes, and conducting comprehensive evaluations of patients with diabetes. Other areas with noticeable gaps in competence included managing sexual dysfunction related to diabetes, addressing diabetic retinopathy, providing effective weight management, applying lifestyle medicine in diabetes care, and attending to mental health issues associated with diabetes. The key domains for which clinical competence gaps were identified are detailed in Figure 1. Physicians who had received training performed slightly better than those who had not received any form of training (mean scores 56 and 51%, respectively; P <0.01) (Figure 2).

Self-Efficacy

Approximately half of the participants reported high self-efficacy in various domains of diabetes care. They felt competent, very competent, or expert at least 50% of the time in areas such as prediabetes management, pathophysiology of type 1 and type 2 diabetes, establishment of diabetes goals, diagnosis and therapy goals, care for elderly patients with diabetes, diabetes education, insulin and noninsulin therapies, dyslipidemia management, weight management, and hypertension (Figure 1). Diabetes diagnosis had the highest self-efficacy rate at 87%, followed by diabetes prevention at 69% and hypertension management at 62% (Figure 4).

Figure 4.

Percentage of respondents reporting higher levels of competence. DM, diabetes; Patho, pathophysiology; QI, quality improvement; T1DM, type 1 diabetes; T2DM, type 2 diabetes.

However, an average of 18% of physicians reported lower self-efficacy, indicating no, vague, or slight knowledge and skill in diabetes care. More than 30% of physicians expressed low self-efficacy in managing transitions to and from inpatient care, mental health/counseling in diabetes, and transitional care. Additionally, more than 20% reported low self-efficacy in managing cardiovascular disease, male sexual dysfunction, retinopathy, and neuropathy associated with diabetes (Figure 4).

Sense of Burden and Stress in Caring for People With Diabetes

A substantial proportion of physicians (>30%) reported experiencing moderate to high levels of burden and stress across all 23 identified clinical domains of diabetes care. Furthermore, in 12 of these clinical domains, >40% of physicians indicated experiencing moderate to high stress levels (Figure 1). Notably, the domain associated with the highest reported burden and stress was managing transitions to and from inpatient diabetes care.

Figure 1 also presents a detailed analysis of the relationship between the perceived sense of burden and clinical competence scores across various clinical domains. Our findings revealed no significant association between the perceived sense of burden and factors such as sex, training status, clinical competence score, or years of experience (Table 1).

Professional Satisfaction

More than half of the participants (56%) reported moderate to high levels of professional satisfaction. The mean score of professional satisfaction was 3.5 ± 0.9 on a 5-point Likert scale.

Professional Isolation

Professional isolation rates were reported as moderate to high in 27% of participants and were low in 50%. The mean score of professional isolation was low at 2.6 ± 1.2 on a 5-point Likert scale.

Satisfaction With Access to CME and Expert Opinion

Physicians expressed high levels of satisfaction with CME access, with 64% moderately to highly satisfied and 16% minimally or not satisfied. Satisfaction with access to expert opinion was also positive, with a mean score of 3.68 ± 1.27 and 60% moderately to highly satisfied. However, 23% were not satisfied. No significant correlation was found between satisfaction level and clinical competence score (P = 0.37 and P = 0.41, respectively).

These findings highlight a generally positive perception among physicians regarding the accessibility of CME and expert opinion, suggesting these resources meet their educational and professional needs. However, the lack of association between satisfaction level and clinical competence score indicates that factors influencing clinical competence are multifaceted and may extend beyond CME and expert consultations.

Perceived Benefit From CME With Regards to Patient Safety and Reduction in Referrals

Physicians reported high levels of benefit from their CME participation with regard to its effect on patient safety and reductions in referrals to diabetes units and endocrinology specialty centers. The majority (83%) agreed to a moderate or large degree that CME access improves patient safety. Additionally, 78% reported that CME access reduced the need to refer patients for specialty diabetes care.

Transfer of Knowledge From CME

The majority of physicians (54%) reported a moderate to a large degree of knowledge transfer when asked about transferring knowledge from their CME participation to their clinic staff. Less than 20% of physicians reported limited to no knowledge transfer after CME participation.

Use of Radar Plots for Multidimensional Analysis of Medical Centers’ Performance

We used radar plots to gain a comprehensive and multidimensional perspective of performance across various medical centers in the tested domains. These plots were instrumental in visualizing the results from different centers, enabling us to identify patterns in their performance. As illustrated in Figure 5, certain primary care centers exhibited similar, albeit not identical, result patterns. This observation suggests that common challenges encountered by these centers could potentially be addressed through a unified, global strategy. Conversely, a few centers displayed unique patterns in their radar plots, indicating distinct issues. Such unique profiles necessitate a more tailored approach, focusing on the specific needs and challenges identified in these individual centers. The use of radar plots thus proved effective in not only delineating the performance variances among different medical centers, but also guiding the formulation of strategies—be they global or targeted—based on the unique characteristics and needs of each center.

Figure 5.

Radar plots of outcomes for two primary care centers.

Discussion

The JDQ assessment conducted in our study unveiled crucial insights for customizing diabetes care training and guideline implementation, highlighting the unique needs of physicians within the KSA health care system. The KSA served as an excellent case study because of its diverse physician workforce and the significant burden of diabetes in the region. These factors make the KSA a microcosm of global challenges in diabetes care, offering insights with wide-reaching implications.

Our findings on the misalignment between self-perceived and actual clinical competence are particularly striking. Although 50% of physicians self-identified as competent or expert in diabetes care, their exam scores did not align with this perception, underscoring a global issue of inadequate self-assessment among health care professionals. This observation aligns with similar trends observed in other regions, including the United States (27,28), highlighting a universal need for enhanced support and training in diabetes care, particularly in prevention and management. These data likely reflect a need for more support for PCPs in the rapidly changing field of diabetes, especially with respect to its prevention and complications.

The study also sheds light on the high levels of burden and stress experienced by physicians in all domains of diabetes care, a concern that transcends geographical boundaries. This high burden level, despite having no direct correlation with demographic characteristics or clinical competence, suggests a need for broader investigations into systemic, patient-related, and other contributing factors. The phenomenon of physician burnout, while highlighted in the KSA context, resonates globally, emphasizing the need for effective strategies to address this challenge worldwide. Physician burnout presents serious risks for physician wellness and performance (29–32). We identified specific areas of burden to help tailor future interventions and training (Figure 1). Thirty percent of our physician population reported high levels of burden in caring for people with diabetes in all domains of diabetes. Although we did not find an association between the burden and stress experienced by the diabetes care providers and their demographic characteristics or their clinical competence, we feel that this is only a starting point. High levels of the burden alone are of great concern and warrant a follow-up study to investigate more broadly (e.g., system factors, patient factors, and others).

Other unexpected findings captured by this assessment were the proportion of the workforce (∼30%) reporting no formal clinical training in the form of a residency (or beyond), their performance on testing, and their reported attitudes. The majority of the workforce (∼92%) is quite experienced, indicating that some of the training for this physician population started on the job rather than in a traditionally structured training program. This finding, along with others, prompt attention to education and training needs that may be unique to this group and should be addressed before diabetes guideline implementation.

High CME satisfaction despite varying competence scores suggests a gap in using these resources. Health care systems need to align CME content with clinical needs and ensure practical skill application. Evaluating CME quality, effectiveness, and alignment with diabetes care goals is essential. Scrutinizing utilization rates, content relevance, and dissemination practices is crucial, especially for integrating diabetes guidelines across regions. Addressing these areas will enhance diabetes care practices and outcomes.

Physicians reported high satisfaction with accessibility of expert opinion, although this finding needs analysis to prevent overreliance on specialist referrals. More than half expressed moderate to high professional satisfaction despite stress and competence gaps, and low professional isolation suggests a positive work environment. However, variations in satisfaction and isolation across locations require further investigation.

Our study positions CME and other forms of professional education as pivotal elements that extend beyond mere credentialing necessities (33). It highlights the often limited feedback available in postgraduate training, which can lead physicians to prioritize time and responsibilities over institutional needs and current guidelines (33). This situation underscores the importance of providing structured and relevant educational opportunities that enable physicians to effectively assimilate and apply new knowledge, ultimately benefiting patient care and reducing the rates of complications associated with diabetes (34).

Furthermore, our findings emphasize the significance of aligning educational strategies with performance improvement measures within health care systems. Such strategic alignment is crucial for effectively implementing clinical guidelines in diabetes care, with the potential to enhance the overall functionality and efficacy of the health care system. In response to identified clinical competence gaps and quality improvement priorities, our study advocates for the creation of targeted CME and training programs. Such programs should be specifically designed to address these gaps, ensuring that the training is not only relevant, but also conducive to building capacity and improving clinical outcomes in diabetes care.

To aid in guideline implementation and decision-making, we used radar plots. These tools enable efficient allocation of resources and interventions to centers with similar performance patterns, enabling more effective comparative assessments. The data from this study informed the design of the JDQ training program that followed this assessment and are likely to influence other diabetes CME programs.

However, the results of our study must be interpreted with caution. The self-reported nature of our survey data introduces a potential for response bias, despite using a validated tool for data collection. The clinical competence scores, expertly crafted and validated, were not proctored, which may affect their reliability. Although our response rate was 55%, the diverse regional representation within the KSA provides a substantial and generalizable insight into the physician population involved in diabetes care. Moreover, the diverse origins of our physician cohort, spanning 22 countries, enhance the generalizability of our findings to primary care settings in regions such as the Middle East, North Africa, Southeast Asia, and beyond. Our methodology and findings are also globally applicable.

Conclusion

We demonstrated that JDQ assessment is a useful and simple, yet validated and evidence-based, approach to clinical competence and attitude assessment in a diabetes physician workforce. This assessment framework can be used in diabetes care workforces worldwide and can be adapted for other primarily outpatient chronic conditions.

Our assessment demonstrated a case study and pilot with valuable information to plan and design this health care system’s diabetes guideline implementation. It clearly demonstrated what gaps in clinical competence participants had on a validated tool, which can inform educational design. It also clearly demonstrated areas for which physicians are reporting high levels of burden. Moreover, our findings underscore the necessity run for health care systems to adopt a more holistic approach to guideline implementation using professional education or CME, moving beyond traditional credentialing to address the dynamic challenges in diabetes management.

Based on the insights gained from this assessment, a guideline implementation program was initiated, with the findings from the assessment playing a crucial role in shaping its design and planning. By highlighting specific areas for improvement and professional development, this study sets a precedent for future research and policy-making aimed at enhancing diabetes care quality and physician well-being worldwide.

This article contains supplementary material online at https://doi.org/10.2337/figshare.28319096.

Funding Statement

The KSA MOH provided funds to the Joslin Diabetes Center to be used for research as per Joslin Diabetes Center/Harvard research proceedings. Co-principal investigator M.A. contributed to the trial design, data interpretation, and writing of the manuscript in a researcher capacity. No other agency or company provided funding for this study.