Barriers to Cancer Evaluation Follow-Up: Reasons and Predictors of Missed Visits in a Rural Population in Botswana

Abstract

Background: Missed clinical appointments contribute to delayed cancer diagnosis, treatment, and ultimately poorer outcomes. In Botswana, a middle income sub-Saharan African country where oncology treatment is available for free to all citizens, median time from first facility visit to treatment initiation is 160 days (IQR: 59-653). Aim: To identify patient and visit characteristics associated with missing appointments, and describe reasons for missed visits among adult patients undergoing evaluation for possible cancer at public facilities in a rural district in Botswana. Methods: Patients ≥ 18 years with symptoms or signs suggestive of cancer were enrolled in the Potlako trial, an ongoing prospective health systems pilot seeking to improve timely diagnosis and treatment of cancer. Study staff tracked patient visits and if a patient missed a visit, study staff administered a brief structured phone interview to 1) determine reasons, 2) reschedule visit and 3) arrange transport support if needed. Patients who enrolled and completed follow-up between May 1, 2016 and March 23, 2018 were included in analysis. We used logistic regression (STATA v12) to explore predictors of missed visits (patient as defaulter vs not, visit being missed vs not). All factors significant in univariate analysis ( P < 0.05) were included in final models. Results: A total of 488 cancer suspects completed their cancer evaluation following a median of 7 visits (IQR: 4-11) and 125 days (IQR: 55-228). Breast and cervical cancer were the most commonly suspected malignancies. Median age was 49 (IQR: 33-64), 358 (73.4%) were female, 143 (28.4%) had an ultimate diagnosis of cancer. There were 172 missed visits, involving 94 (19.3%) patients (defaulters), including 20 (21.3%) found to have cancer, of whom 38 (39.6%) missed more than one visit. Defaulters tended to be less than 40 years (aOR 0.49, 95% CI: 0.31-0.79) and had their first visit for cancer evaluation at a general outpatient unit (aOR 2.1, 95% CI: 1.15-3.87). In the adjusted model, we found no statistically significant association between defaulter status and gender, final diagnosis, and patient's next-of-kin. Missed visits were more likely to occur on Fridays (aOR 1.50, CI: 1.05-2.15) and less likely to occur ±1 week around a public holiday (aOR 0.55, CI: 0.33-0.91). Leading reasons for missed visits were: work obligations (18.0%), family duties including child care and attending funerals (16.3%), prohibitive transport fees (10.5%), not remembering appointment date (8.7%). Conclusion: Multiple visits are required to complete evaluation for cancer suspects and missed visits are common. Our findings suggest possible interventions to improve efficiency, which include: simplifying diagnostic cascade, transport support, visit reminders, intensified counseling for younger patients and those seen at general outpatient clinics, and avoidance of Friday scheduling.