This was a hospital-based, cross-sectional study, which included diabetic patients on treatment with oral hypoglycaemic agents or insulin, in the outpatient clinics or inpatient wards of our hospital, a tertiary eye care centre in South India. The study was conducted from June 2013 to November 2013. Children (age less than 18 years), patients who did not speak or understand English, Tamil or Hindi, mentally challenged patients who were unable to give informed consent or respond meaningfully to the questions administered, patients with hazy media in both eyes precluding adequate visualization of the fundus for grading of diabetic retinopathy, and patients with retinal vein occlusion or ocular ischemic syndrome in one or both eyes were excluded from the study.

Diabetic patients, who met the eligibility criteria of the study, were enrolled after obtaining informed consent. The study was started after obtaining the approval of the Institutional Review Board. Data were collected using a clinical research form. The socioeconomic and educational status of each patient was graded using Modified Kuppuswamy classification [14]. The presence and level of diabetic retinopathy were assessed by dilated fundus examination using slit lamp binocular indirect ophthalmoscopy. Diabetic retinopathy was classified according to the Modified Airlie House Classification (Early Treatment Diabetic Retinopathy Study) [15].

Data regarding KAP patterns, and barriers to compliance with follow up and treatment regimes for diabetes and diabetic retinopathy were collected using a 45-point, verbally administered questionnaire in the clinical research form [Appendix-1].

The questionnaire was formulated by the investigators after conducting a thorough literature search. To minimize bias due to ‘leading’ questions, most of the questions in the knowledge and practice sections of the questionnaire were constructed as open-ended questions. The questions in the attitude section were framed as statements, and the patient was asked whether he or she agreed or disagreed with the statement, or was undecided. The questionnaire was reviewed for adequacy, appropriateness and relevance of content by five subject matter experts. It was then translated to Tamil and Hindi. To ensure uniformity of administration, the questionnaire was administered in all cases by one of two investigators; both investigators were trained to administer the questionnaire in a standard manner. A pilot study was then conducted to familiarize the investigators with the questionnaire, and to identify practical problems with its administration. The questionnaire was further refined, based on the lessons learned from the pilot study. A final 45-point questionnaire was thus formulated, with 13 questions in the knowledge section, 8 questions in the attitude section, and 24 questions in the practice section. This questionnaire (in English, Tamil or Hindi) was verbally administered to the patient to assess his or her knowledge, attitude and practice patterns regarding diabetes and diabetic retinopathy.

The answers to the questions were scored. The total score achieved by the patient in each section was calculated. On the basis of the number of correct responses to ‘must know’ questions in the knowledge section of the questionnaire and ‘must do’ questions in the practice section, each patient in the study was categorized as having ‘good’ or ‘poor’ knowledge, and ‘good’ or ‘poor’ practice pattern. In the attitude section of the questionnaire, the responses best indicative of a positive attitude were scored, and the patients were categorized as having ‘positive’ or ‘negative’ attitude.

Data were analysed using SPSS, Version 22.0. Chi-square/Fisher-exact test was done to check the association for categorical variables. Binary logistic regression was done with statistically significant variables at 25% level of significance for univariate analysis, and 5% for multivariate analysis. Duration of diabetes, gender, educational status, socio-economic status and presence of diabetic retinopathy were identified as potential confounders. Information regarding these potential confounders was meticulously documented using the clinical research form. Potential confounders were addressed by using multiple logistic regression analysis.

Two hundred and eighty eight patients who fulfilled the eligibility criteria were recruited into the study. The demographic characteristics and retinopathy status of the study population are given in [Table/Fig-1]. Of the 288 patients recruited, 121 (42.0%) had good knowledge of diabetes, while only 84 (29.2%) had positive attitude towards diabetes; 158 patients (54.9%) were found to have good practice patterns [Table/Fig-2].

Among the 288 patients in our study, 207 (71.9%) were ‘aware’ that eyes could be affected by diabetes, but only 49 patients (17.01%) were ‘aware’ of diabetic retinopathy as an ocular complication of diabetes. The questions to assess knowledge of diabetic retinopathy were administered only to the 49 patients who were aware of retinopathy. Among these 49 patients, five had not had an eye check up prior to the study visit, and therefore, had never come into contact with an eye doctor or an eye health care system. As the questions to assess attitude towards diabetic retinopathy were designed and follow up guidelines prescribed by their eye doctor, that could not be administered to these five patients. Therefore, attitude towards diabetic retinopathy was assessed only in 44 patients in the study. All the patients in the study (n=288) were administered questions regarding practice patterns, based on their retinopathy status [Table/Fig-3]. Patients with good knowledge of retinopathy constituted only 4.51% of the total number of patients in the study (n=288), while positive attitude towards retinopathy was found in 9.38% of the total number of patients in the study.

Among the 121 patients with good knowledge of diabetes, only 12 (9.9%) had good knowledge of retinopathy; 79 out of these 121 patients (65.3%) were not even aware of diabetic retinopathy.

To identify the source of information about diabetic retinopathy, we asked those patients who were aware of retinopathy (n=49) about how they first came to know that diabetes could cause retinopathy. Doctors (both ophthalmologists and physicians) constituted the most important source of information (35 patients, 71.4%). Media, books, family and friends were the other sources of information for the patients in the study.

The odds of patients with good knowledge of diabetes having positive attitude towards diabetes were 4.2 (2.21-7.82) times those of patients with poor knowledge of diabetes, after adjusting for educational status, socio-economic status, duration of diabetes and gender, with p <0.01 [Table/Fig-4]. Similarly, the odds of patients in the higher socio-economic status group having positive attitude towards diabetes were 3.3 (1.73-6.35) times those of patients in the lower socio-economic status group, after adjusting for educational status, duration of diabetes, gender and knowledge of diabetes, with p <0.01 [Table/Fig-4].

The odds of patients with good knowledge of diabetes having good practice patterns regarding retinopathy were 3.9 (1.97-7.94) times those of patients with poor knowledge of diabetes, after adjusting for educational and retinopathy status, with p<0.01[Table/Fig-5].

The odds of patients with awareness of retinopathy having good practice patterns regarding retinopathy were 3.6 (1.67-7.69) times those of patients who were unaware of retinopathy, after adjusting for educational and retinopathy status, with p=0.01[Table/Fig-6].

Out of the 288 patients in the study, 41 (14.2%) were not compliant with regular follow up with their physicians for the management of diabetes. The barriers to compliance are listed in [Table/Fig-7].

The most common reasons that the patients gave for poor compliance were ‘did not find time’ and ‘do not think it is important’.

One hundred and seventy six patients (61.1%) in the study did not go for a periodic eye examination.

The barriers to compliance are given in [Table/Fig-8]. The most common barrier identified was the fact that the patients did not know that they should go for a periodic eye check up (111 patients, 38.54%). The second most common reason cited was that the patients did not feel the necessity for an eye check up as they had good vision.

This was a hospital-based, cross-sectional study, which documented the KAP patterns of diabetic patients regarding diabetes and diabetic retinopathy. The study included 215 patients from Tamil Nadu and 73 patients from different parts of India. One hundred and twenty one patients (42%) in the study had good knowledge about diabetes. This is similar to the results of other studies conducted in South India by Hussain R et al., and Rani PK et al., who reported good knowledge in 40.7% and 49.9% respectively, of the subjects of their studies [16,17]. However, in another study done in South India by Babu N et al., only 28% of the population was ’aware’ of diabetes [4]. We found that only 84 patients (29.2%) in our study had a positive attitude towards diabetes. In contrast to this, Hussain R et al., found a positive attitude towards diabetes in 53.8% of the diabetic patients in their study [16]. Good practice patterns with respect to diabetes were found in 158 patients (54.9%). In comparison, 57.6% and 48.45% respectively, of the subjects in the studies by Hussain R et al., and Rani PK et al. were reported to have good practice patterns [16,17].

The questions to assess knowledge of diabetic retinopathy in our study were designed to assess both awareness and knowledge of diabetic retinopathy. Just having heard about the disease is awareness, while having understood the disease is knowledge [18]. Among the KAP studies done on diabetes and diabetic retinopathy in India, Mahesh G et al., have also documented both knowledge and awareness of diabetic retinopathy [18]. Koshy J et al., and Dandona R et al., have reported awareness of diabetic retinopathy [19,20], while Hussain R et al., and Rani PK et al. have documented knowledge of diabetes and retinopathy [16,17]. In the study published by Babu N et al., the terms ‘awareness’ and ‘knowledge’ have been used interchangeably [4].

We felt that it was important to differentiate between awareness and knowledge of diabetic retinopathy. While awareness of the disease is important, having good knowledge of the disease is probably more important in influencing attitude and practice patterns regarding the disease. We therefore, documented both awareness and knowledge of diabetic retinopathy among our patients, and looked for the association of both awareness and knowledge of diabetic retinopathy with attitude and practice patterns regarding retinopathy.

Among the 288 patients in our study, 207 (71.9%) were aware that eyes could be affected by diabetes, but only 49 patients (17.01%) were aware of diabetic retinopathy as an ocular complication of diabetes. Babu N et al., and Dandona R et al., have also reported poor awareness of diabetic retinopathy (7% and 27% respectively) among the subjects in their studies done in South India [4,20]. However, in the study done in South India by Mahesh G et al., 36.31% felt that they were well educated about retinopathy, while 30.9% of the patients in the study done in North India by Koshy J et al., knew that diabetes could lead to retinal disease [18,19]. In our study, only thirteen out of the 49 patients (26.5%), who were aware of retinopathy, had good knowledge of retinopathy. This constituted only 4.51% of the total number of patients in the study. Das T et al., also reported poor knowledge of retinopathy among the patients in their study conducted in Eastern India [21]. In contrast to this, 37.1% had ‘knowledge’ of retinopathy in the study by Rani PK et al., [17].

Even among the 121 patients who had good knowledge about diabetes, only 12 (9.9%) had good knowledge of diabetic retinopathy. We also found that 65.29% of the 121 patients who had good knowledge of diabetes were not even aware of retinopathy. In spite of the fact that diabetic retinopathy is the most serious, potentially blinding complication of diabetes in the eye, the majority of the patients were completely unaware of the existence of such an entity. This indicates the poor state of patient education measures regarding diabetic retinopathy, as it was the same subgroup of patients who had good knowledge of diabetes.

Strategies to educate diabetic patients about this potentially blinding complication of diabetes should be evolved. This would have to done at all points of patient contact with the health care system. General practitioners, physicians, endocrinologists, ophthalmologists and optometrists should be made aware of the sad lack of knowledge about diabetic retinopathy among diabetic patients, and should all be involved in the planning and implementation of both hospital-based and community-based patient education strategies.

Health education measures should be implemented at primary, secondary and tertiary levels of health care. Health education through mass media, pamphlets, posters and diabetic retinopathy screening camps on special days like World Diabetes Day and World Sight Day would help in creating awareness of diabetic retinopathy, especially among people in the lower educational and socio-economic status groups.

We found a statistically significant association between good knowledge of diabetes and positive attitude towards diabetes. The odds of patients with good knowledge of diabetes having positive attitude towards diabetes were 4.2 (2.21-7.82) times those of patients with poor knowledge of diabetes, after adjusting for educational status, socio-economic status, duration of diabetes and gender, with p <0.01. This shows that, as the knowledge that a patient has about his or her disease increases, the attitude towards the disease also becomes positive. Therefore, imparting knowledge about the disease to the patient is of paramount importance, and it is the duty of the treating physician to ensure that this is done. This is especially important in the Indian health care scenario. The average Indian patient may not get a lot of information about diabetes from books or mass media, and the doctor is often the only individual in the health care system that the patient comes into contact with on a regular basis. Data from our study also corroborate this surmise. Doctors (both ophthalmologists and physicians) constituted the most important source of information (35 patients, 71.4%) for the patients who were aware of retinopathy in our study.

We found that awareness of diabetic retinopathy (p=0.01) and good knowledge of diabetes (p<0.01) were significantly associated with good practice patterns regarding diabetic retinopathy. The odds of patients with awareness of retinopathy having good practice patterns regarding retinopathy were 3.6 (1.67-7.69) times those of patients who were unaware of retinopathy, after adjusting for educational and retinopathy status, with p=0.01. Mahesh G et al., also found a statistically significant association between awareness of retinopathy and good practice regarding retinopathy [18]. The odds of patients with good knowledge of diabetes having good practice patterns regarding retinopathy were 3.9 (1.97-7.94) times those of patients with poor knowledge of diabetes, after adjusting for educational and retinopathy status, with p<0.01. Knowledge about the disease and its complications is a powerful tool, which helps patients in developing good practice patterns that will ultimately help them in keeping the disease under good control.

Out of the 288 patients in our study, 41 (14.24%) were not compliant with regular follow up visits for the management of diabetes, with only three patients (1.04%) saying that they did not know that regular follow up for diabetes was necessary. However, 176 patients (61.1%) in the study did not go for a periodic eye examination; the most common barrier identified was the fact that the patients did not know that they should go for a periodic eye check-up (111 patients, 38.54%). Most diabetic patients seem to know that regular follow up is necessary for their systemic disease; however, the majority do not know that they need to have a periodic eye check up to look for ocular complications of diabetes. The facts that diabetic retinopathy is a silently blinding disease, and ‘good vision’ is not an indicator of the status of the retina in a diabetic patient need to be emphasized to the patient. It is the duty of the ophthalmologists to educate diabetic patients in their clinics about these basic facts.

Most of the questions in the questionnaire were constructed as open-ended questions to minimize bias due to ‘leading’ questions. However, we could not avoid a few closed-ended questions. These may have been ‘leading,’ which may have resulted in falsely high scores in certain sections of the questionnaire. Among the 288 patients in our study, only 49 were aware of diabetic retinopathy. The questions to assess knowledge of diabetic retinopathy were administered only to these 49 patients. After excluding another five patients who had not had an eye check up prior to the study visit, the questions to assess attitude towards retinopathy were administered only to 44 patients in the study. Many of the associations between KAP of retinopathy may not have been statistically significant due to the small sample size for analysis of KAP of retinopathy.

Visual impairment and blindness due to diabetic retinopathy are almost entirely preventable with early detection and timely treatment. Awareness and knowledge about diabetic retinopathy were very poor among the patients in our study. Lack of knowledge concerning the need for screening for diabetic retinopathy was found to be a major barrier to compliance with regular screening. Good knowledge about diabetes was significantly associated with positive attitude towards diabetes and good practice patterns regarding retinopathy. Awareness of diabetic retinopathy was significantly associated with good practice patterns regarding retinopathy. Therefore, there is an urgent need to evolve strategies to educate diabetic patients about this potentially blinding complication of diabetes.