Where Are the Non-visiting Patients?

In a clinical trial to increase the screening of inner-city women for breast cancer, some preliminary data suggested that about 40% of the age-eligible women who had been patients at a primary-care clinic might be expected never to visit that clinic again. The distribution of time between patients' clinic visits was modeled by fitting a distribution function to data on the "time to first visit," and indicated that the true defection rate is probably around 31 to 36%. This result, which is consistent with four related observations, suggests that most of the patients who fail to visit their clinics in a year-long interval may have become clients of other clinics. This conjecture implies massive continuing exchanges of patients among clinics, and may mean that explicit outreach by mail or phone to "non-visiting clients" will add few patient contacts to those based on spontaneous clinic visits. Key words: cancer control; community medicine; intervention studies; mass screening; models, statistical; models, the oretical; patient participation, outreach. (Med Decis Making 1994;14:345-356)