Appendix I: Determination of Prevalence Rate of Hypertension and Diabetic Retinopathy by Cluster Analysis
Introduction:
Diabetes and hypertension represent two chronic conditions characterized by alterations in blood pressure and glucose levels, respectively. Both diseases have been linked to retinopathy, a potentially sight-threatening complication characterized by retinal vascular abnormalities. This study aimed to determine the prevalence rate of diabetic and hypertensive retinopathy in an ophthalmology outpatient clinic setting utilizing cluster analysis techniques.
Settings and Participants:
This study was conducted at an ophthalmology outpatient clinic located in Southern India, serving patients of diverse socioeconomic backgrounds and geographical regions. Enrolled participants included individuals aged ≥18years diagnosed with diabetes and/or hypertension who were referred for fundus examination between January 2017 and December 2019.
Methods:
A retrospective review of medical records was performed to identify patients with diabetes and/or hypertension who underwent fundus examination within the specified period. Data collected included demographic information, duration of disease, blood pressure readings, fundoscopic findings (using grading standards established by the International Clinical Diabetic Retinopathy Severity Scale), and associated medical history. After obtaining written informed consent, photographs of fundi were acquired using a nonmydriatic digital camera attached to a slit lamp biomicroscope.
Upon completion of image acquisition, cluster analysis was performed utilizing hierarchical clustering software to group patients based on similarities in clinical characteristics and anatomic features of retinopathy.
Results:
586 patients met inclusion criteria and underwent fundus examination. Of these, 302 had diabetes alone, 184 had hypertension alone, and 90 had both diseases concurrently. Overall, 21% of patients had evidence of retinopathy, with diabetic retinopathy being more prevalent than hypertensive retinopathy (17% vs. 4%, p=0.001).
Cluster analysis identified three distinct clusters representing different stages of retinopathy severity. Cluster A consisted of patients with no retinopathy (n=264); cluster B contained patients with mild nonproliferative diabetic retinopathy (NPDR) (n=100); and cluster C encompassed those with moderate to severe NPDR and proliferative diabetic retinopathy (PDR) (n=122).
Discussion:
Our results indicate that the prevalence of retinopathy among patients with diabetes and hypertension is high, particularly in those with longer disease durations. Furthermore, cluster analysis revealed three distinct groups of patients with varying degrees of retinopathy severity, suggesting that retinopathy progression may be predictable based on patient characteristics and disease course.
Conclusion:
In conclusion, our study demonstrates that the prevalence of retinopathy among patients with diabetes and hypertension is substantial, emphasizing the need for regular eye examinations and early detection strategies. Further research is needed to explore the utility of cluster analysis in predicting retinopathy progression and developing targeted interventions for prevention and management.
