[Non mydriatic retinal camera: cost-effectiveness study for early detection of diabetic retinopathy].

BACKGROUND AND
OBJECTIVE: Cost-effectiveness analysis (ACE) for application of fundoscopic photograph with non mydriatic retinal camera (Ffo-CNM) in early diagnostic of diabetic retinopathy (RD) compared with ophthalmoscopic view by pupillary dilation. PATIENTS AND
METHOD: diabetic patients, older than 14 years, attended on three Health Primary Care Areas (n=1495). Effectiveness measurement: predictive value of a positive test (VPD) and correctly diagnosed cases. Cost measurement: total cost by patient. ACE: defined as expected cost by VPD case and as expected cost by correctly diagnosed case. The results were submitted to an analysis of sensitivity for the study main variables.
RESULTS: Ffo-CNM presented 90.91% of sensitivity [95% CI, 69.4-98.4%], 78.21% of specificity [95% CI, 67.1-86.4%], 54.05% positive predictive value [95% CI, 37.1-70.2%] and 96.83% of negative predictive value [95% CI, 88-99.4%]. Effectiveness, defined as VPD case, was 15.4% for ophthalmoscopic view and 19.5% for Ffo-CNM, and defined as correctly diagnosed case, was 70% and 79.8%, respectively. Cost-effectiveness ratio: a) for health care system, the cost by VPD case was 52.62 euros for ophthalmoscopic view and 28.44 euros for Ffo-CNM and cost by correctly diagnosed case was 11.58 euros and 6.95 euros, respectively, and b) for the society, cost by VPD case was 100.13 euros for ophtalmoscopic view and 34.54 for Ffo-CNM and the cost by correctly diagnosed case was 22.03 euros and 8.44 euros respectively.
CONCLUSIONS: If an introduction of a early detection of RD program for the entire diabetic population was decided the option to make it using Ffo-CNM would be the most efficient.