Methods of estimating losses to follow-up from a diabetic clinic

Abstract

Continuity of follow-up is extremely important for the diabetic, to ensure that treatable complications such as retinopathy are detected at an early stage. Therefore, it is important to ensure that patients are not lost to follow-up and to know when they have been. Defining waht is meant by lost to follow-up, and identifying patients who have been lost, from traditionally organised outpatient clinics is not always easy. We have retrospectively studied 3537 patients who at some time attended a diabetes outpatient clinic over a seven year period, to assess how many were lost to follow-up. During this period, a computer system was established which included the booking of appointments and the maintenance of a register of patients. We have compared losses from the traditional clinic (19 per month) with the new computer-assisted clinic (seven per month). Estimates of the proportion of those who were lost to follow-up vary depending on the criteria adopted. Data from a patient survey suggest that in excess of 17% were lost from the traditional clinic. An audit of case notes suggests that the proportion of those leaving the computer-assisted clinic who were lost was lower after the implementation of the computer system (29% vs 7%) (X2 = 4.1; 1df; p<0.05). Of those last seen at a clinic more than three years ago, 30% had not had an eye examination within three years. The process of identifying and following-up patients who had left the traditional clinic took nine person-months compared to a few days for patients leaving the computer-assisted clinic. Even in the absence of special efforts for the maintenance of the continuity of care, an appropriately designed clinical information system and improved clinic organisation will reduce losses to follow-up. However, our aim should be to eliminate losses to follow-up, and such systems should be used to promote a population approach to the maintenance of continuity of care through, for example, community-based registers and purpose designed shared-care schemes.