Objective: To test the hypothesis that pretreatment fluorescein angiography (FA) is not necessary for effective laser treatment of patients with clinically significant diabetic macular edema (CSME).
Design: Prospective, randomized, controlled treatment simulation.
Participants: Six fellowship trained retina specialists.
Intervention: The authors compared the ability of four retina specialists (observers) to plan laser treatment with and without the use of FA. One hundred consecutive cases of CSME were selected, each case consisting of a stereo pair of color photographs and a corresponding fluorescein angiogram. These cases were first read by two retina specialists who reached consensus on a treatment plan for each case (standard map). Each of the 4 observers reviewed 50 of these cases on 2 occasions and plotted 2 sets of treatment maps, 1 set created with and 1 without the aid of FA. Each observer's 100 treatment maps were graded for accuracy by comparing them to the corresponding standard maps. The role of FA in improving the accuracy of treatment maps was evaluated using logistic regression analysis to control for different observers, different cases, and different posterior pole characteristics.
Main outcome measures: Accuracy was defined as the proportion of standard treatment correctly treated by the observer.
Results: For the observers as a group, the use of FA improved treatment planning accuracy from 49% to 54.5% (P = 0.02); however, there was significant interobserver variation in performance (P < 0.001). Treatment planning accuracy without and with FA was as follows: observer 1, 40.8% and 40.2%; observer 2, 49.8% and 72%; observer 3, 56.1% and 59.5%; and observer 4, 49.2% and 46.4%.
Conclusion: The use of FA improves the accuracy of treatment planning for CSME. The authors' study supports the use of FA in laser treatment of patients with CSME.