Knowledge, attitude and practices of physicians on diabetic retinopathy

Study design and setting

A descriptive cross-sectional study was conducted between January and December 2023 at Jawahar Lal Nehru Medical College and its associated hospitals in Ajmer, Rajasthan, India. Institutional ethics committee approval was obtained before commencement, and verbal informed consent was secured from all participants.

Study population

The target population included non-ophthalmologist physicians involved in the management of adult diabetic patients. Eligible participants were general practitioners, residents and consultants from the departments of internal medicine and paediatrics, as well as other physicians engaged in diabetes care. Exclusion criteria were ophthalmologists, surgeons, anaesthetists and obstetricians/gynaecologists, as their roles did not directly involve routine diabetic follow-up.

Sample size

Based on an estimated physician population of 280 in the hospital and its affiliated units, the required sample size was calculated using Cochran’s formula with finite population correction. Assuming a 95% confidence interval and a 5% margin of error, the minimum sample size was 280. Thus, all 280 eligible physicians were invited to participate, and all completed the survey, ensuring complete enumeration.

Study instrument

Data were collected using a validated, self-administered 28-item questionnaire adapted from published KAP studies and international guidelines on DR. The questionnaire consisted of four sections:

1. Demographics: Age, gender, speciality and years of practice.

2. Knowledge domain (12 items): Focused on risk factors, screening guidelines, clinical features and treatment of DR. Multiple-choice questions were used, with one correct option per question (except for three questions that allowed multiple correct responses). Each correct answer was scored as 1; incorrect or ‘don’t know’ responses were scored 0. The maximum possible score was 23. Knowledge levels were graded as:

   • Excellent: ≥75% (≥9 correct answers)

   • Good: 50–75% (6–8 correct answers)

   • Poor: <50% (<6 correct answers).

3. Attitude domain (8 items): Physicians’ perceptions about the importance of screening, feasibility of fundoscopy and responsibility for DR detection were assessed using a two-point Likert scale (agree/disagree). A positive attitude was defined as ≥6 correct responses, while ≤5 was considered negative.

4. Practice domain (8 items): Explored current practices regarding fundoscopy, use of mydriatic drops, referral to ophthalmologists and involvement in public awareness. Responses were dichotomous (yes/no). Scoring was classified as:

   • Good: ≥75% (≥6 correct)

   • Fair: 50–75% (4–5 correct)

   • Poor: <50% (<4 correct).

The questionnaire was initially piloted on four physicians to ensure clarity, and their responses were validated against ophthalmologists’ clinical standards. Minor revisions were made before final administration.

Data collection procedure

Participants were recruited by random sampling within their departments. Questionnaires were distributed during departmental meetings and were completed under the supervision of investigators to prevent consultation of textbooks or colleagues. On average, each physician required 20–25 mins to complete the form. Completed forms were checked for completeness, anonymised and coded before data entry.

Data management and statistical analysis

Data were entered in Microsoft Excel 2019 and analysed using the statistical package for the social sciences version 22.0 (IBM Corp., Chicago, IL, USA). Descriptive statistics were used to summarise demographic variables and KAP responses. Frequencies and percentages were reported for categorical variables and means with standard deviations for continuous variables.

The Chi-square test was used to assess associations between demographic characteristics and KAP levels. To ensure construct validity, principal component analysis with varimax rotation was employed to group statistically related variables. Sampling adequacy was checked using the Kaiser–Meyer–Olkintest, and sphericity was assessed using Bartlett’s test. A P< 0.05 was considered statistically significant.

Demographic characteristics

A total of 280 physicians participated in the study, yielding a 100% response rate. The mean age was 37.06 ± 9.96 years (range: 24–62 years). Among them, 152 (54.3%) were male and 128 (45.7%) female. Most respondents were from internal medicine (171; 61.1%), followed by paediatrics (76; 27.1%) and other specialities (33; 11.8%). A majority had ≤5 years of practice experience (161; 57.5%), while 72 (25.7%) had 6–10 years and 47 (16.8%) had >10 years [Table 1].

Knowledge assessment

Knowledge regarding DRwas mixed across domains. A large proportion (87.1%) correctly identified that type 2 diabetic patients should undergo ophthalmic evaluation immediately after diagnosis, whereas only 57.9% knew that type 1 diabetics without visual symptoms should be examined after 5 years of diagnosis.

Awareness about screening intervals was good, with 86.1% recommending annual ophthalmic review for type 2 diabetics. Similarly, most respondents (81.4%) correctly identified diabetes-related ocular conditions, and 89.3% recognised high-risk groups for DR, such as patients with uncontrolled diabetes, hypertension or pregnancy.

The majority (90.4%) knew that a dilated fundus examination is the preferred diagnostic method, and 71.1% correctly identified microaneurysms as the earliest clinical sign. Knowledge of treatment modalities was also high (86.1%). However, only 28.2% reported knowing how to use an ophthalmoscope, highlighting a critical skills gap [Table 2].

Attitudes toward DR screening

Attitudinal responses revealed an overall positive orientation toward DR prevention. The majority (82.1%) agreed that all diabetic patients should be referred to ophthalmologists, and 89.6% believed that blindness from DR could be prevented if diabetes were adequately controlled. Furthermore, 90.7% agreed that fundus examination should be confirmed by an ophthalmologist.

However, only 64.3% considered fundoscopy feasible for non-ophthalmologists in routine practice, suggesting hesitancy among physicians to adopt this role themselves.

Practices regarding DR care

Despite good knowledge and favourable attitudes, practices remained suboptimal. Only 28.2% of participants reported being able to use an ophthalmoscope, and a majority (71.8%) had never attempted fundoscopy in diabetic patients. Routine referral of all diabetic patients was not consistently implemented in practice, despite most physicians acknowledging its importance.

Availability of mydriatic drops and working ophthalmoscopes in outpatient departments was limited. Less than one-third of respondents reported participating in public awareness programmes for DR [Table 3].

Key findings

• Knowledge: Adequate in areas of screening intervals, high-risk groups and treatment, but weak regarding ophthalmoscope use.

• Attitude: Generally positive toward referral and prevention, but uncertainty about the feasibility of physician-performed fundoscopy.

• Practice: Lagged behind knowledge and attitude, with low fundoscopy rates, inconsistent referral practices and limited engagement in awareness activities.

Knowledge gaps despite awareness of DR burden

Our findings indicate that physicians had reasonable awareness about the significance of DR, its risk factors and treatment modalities. Nearly 90% of respondents recognised uncontrolled diabetes, long duration of disease and associated comorbidities as key risk factors. Similarly, 86% identified standard treatment options such as laser photocoagulation, intravitreal anti-VEGF agents, corticosteroids and vitrectomy. These results align with findings from Saudi Arabia, Oman and Nepal, where general awareness about DR as a complication of diabetes was adequate.^[1-3]

However, several gaps were identified. Only 57.9% correctly identified the appropriate referral interval for type 1 diabetes without visual symptoms, suggesting incomplete adherence to international screening guidelines.^[4] This is concerning since type 1 patients are at risk of progressive DR, particularly after 5 years of disease onset. Similar deficiencies in knowledge of screening intervals have been reported in studies from Nigeria and Pakistan reflecting a global trend of inconsistent familiarity with guidelines among nonophthalmologists.^[5,6]

Another critical gap was the limited knowledge and use of ophthalmoscopy. While 90.4% recognised dilated fundus examination as the gold standard for DR detection, only 28.2% reported knowing how to use an ophthalmoscope. This discrepancy between theoretical knowledge and practical competence has been documented in multiple KAP studies, including those from Oman to Kathmandu, where many physicians endorsed the importance of fundoscopy but lacked the necessary skills to perform it reliably.^[2,7]

Attitudes: Positive but passive

The study revealed an overall positive attitude among physicians regarding DR prevention and management. More than four-fifths of respondents endorsed referral of all diabetic patients to ophthalmologists, and nearly 90% believed that blindness from DR could be prevented with good glycaemic control. These responses reflect an encouraging awareness of the preventable nature of DR.

Nevertheless, attitudes regarding the role of general physicians in fundoscopy were less favourable. Only 64.3% believed that it was feasible for non-ophthalmologists to conduct fundoscopy in outpatient settings. This cautious approach may stem from a perceived lack of training or confidence, as also reported by Al Rasheed and Al Adel^[1] in Saudi Arabia, where only 56% of primary care physicians routinely attempted ophthalmoscopy. Such reluctance perpetuates dependence on specialist ophthalmologists, which in resource-limited settings often leads to delayed detection and referral.

Practices: The weakest domain

Despite fair knowledge and positive attitudes, practices were suboptimal. Only 28.2% of physicians reported being able to use an ophthalmoscope, and 71.8% had never attempted fundoscopy on diabetic patients. Furthermore, routine referral of all diabetic patients was inconsistent, and participation in community awareness programmes was limited to 13^rd.

This knowledge-practice gap has been observed globally. A Nigerian study by Abdulsalam et al.^[5] highlighted that while over 75% of physicians knew the importance of DR screening, only 40% actually performed eye examinations. Similarly, a study in Saudi Arabia found that although most physicians agreed with early referral, only 24% correctly referred type 1 diabetics at diagnosis.^[8] In India, a survey by Vashist et al.^[9] revealed that <40% of diabetic patients had ever undergone eye examination, reflecting systemic deficiencies in referral practices.

The gap between knowledge and practice may be explained by several factors: Lack of equipment (ophthalmoscopes and mydriatic drops), inadequate training, time constraints in busy clinics, and the belief that fundoscopy is the domain of ophthalmologists. Without addressing these barriers, the potential benefits of physician awareness cannot be translated into patient outcomes.

Comparison with regional studies

Our findings are consistent with KAP studies across South Asia. In Nepal, Pradhan et al.^[7] reported good physician knowledge and attitudes toward DR, but fundoscopy practices were poor. In Pakistan, Memon et al.^[6] demonstrated that while 70% of physicians were aware of DR as a cause of blindness, <½ routinely referred diabetic patients for screening. In India, regional studies haveemphasised similar deficiencies in referral practices despite increasing awareness campaigns.^[9,10]

These parallels highlight that the issue is not one of knowledge alone, but of implementation and system integration. Physicians in tertiary care hospitals often face patient overload, which discourages detailed fundoscopy or preventive counselling. Unless DR screening is institutionalised as part of standard diabetic care protocols, practices are unlikely to improve.

Implications for policy and practice

The study underscores an urgent need for capacity building and structural support. Several strategies could help bridge the knowledge-practice gap:

1. CME: Regular CME sessions focusing on DR screening guidelines, ophthalmoscope use and referral protocols can enhance competence. Evidence from Saudi Arabia and Indonesia suggests CME significantly improves physician practices.^[1,11]

2. Provision of equipment: Ensuring the availability of ophthalmoscopes and mydriatic eye drops in nonophthalmology outpatient departments is essential. In our study, lack of resources was a major barrier, similar to findings from northwestern Nigeria.^[5]

3. Integration of DR screening into diabetes care: Institutional protocols mandating annual eye referral at diagnosis for type 2 and 5 years after type 1 diabetes should be enforced. Linking endocrinology or internal medicine clinics with ophthalmology services could facilitate routine referral.

4. Task shifting and teleophthalmology: Training general practitioners or trained technicians in fundus photography and coupling this with telemedicine consultations with ophthalmologists could overcome the shortage of specialists. This model has shown promise in India and other low-resource settings.^[12]

5. Community engagement: Physicians should be encouraged to participate in public awareness campaigns. Awareness of DR remains low among diabetic patients in India and physicians are often the most trusted source of information.^[10]

6. It further needs training to identify the diabetic pathology as well as the time and inclination on physician part to use it. Good, reliable ophthalmoscopy effectively falls out of their domain.

7. Use of ophthalmoscopy by non-ophthalmologists can be one of the many approaches to screen for DR. Other more effective approaches being systematic referral mechanisms which are easy to adopt, non-mydriatic, AI-enabled retinopathy screening devices which are easy to use and increasing the general awareness of diabetic patients about periodic check-ups for DR.

This study reveals a paradox: while physicians at a tertiary hospital in Rajasthan demonstrate adequate knowledge and positive attitudes toward DR, their practices are inadequate, particularly regarding ophthalmoscopy and consistent referrals. This knowledge–practice gap is not unique to India but reflects a global challenge in integrating DR screening into routine diabetes care. Targeted training, better resource allocation and institutionalised referral protocols are critical to ensuring early detection and prevention of vision-threatening DR.

Strengths and limitations

The strengths of this study include its complete enumeration of eligible physicians at a major tertiary hospital, use of a validated questionnaire and comprehensive analysis of KAP domains. The 100% response rateminimises selection bias and ensures representativeness.

However, limitations must be acknowledged. First, the study was conducted in a single tertiary hospital, which may limit generalisability to other regions of Rajasthan or India. Second, self-reported practices may not accurately reflect actual behaviour in clinical settings, introducing potential reporting bias. Third, the cross-sectional design captures KAP at one point in time but cannot assess changes after interventions. Future multicentre studies with observational components are needed to validate and expand these findings.