Nonrandom Return: Lessons from Women Returning for Follow-Up Results of Cervical Cancer Screening in Rural Haiti

Abstract

Purpose: Few studies have evaluated approaches to improve patient follow-up in screening programs within resource-limited areas. This study investigated the patterns of return among women presenting for results from a cervical cancer screening program implemented in a resource-poor region to highlight areas for future research and potential interventions. Methods: In the development of a cervical cancer screening program in rural Haiti, women aged 18-45 years were recruited from local churches to undergo cervical cancer screening and follow-up for results. A dot plot was used to analyze the distribution of women returning during the results phase compared to the random return rates based on the screening phase. Results: Of the 250 women tested, 121 (48%) returned for follow-up. On the 2 result days with the most returnees, 46 and 63% of the women had been tested during only 3 of the 11 testing days which themselves accounted for 41% of the total number of women who returned. The highest volume return days, a Monday and a Friday, gave results to women tested on Monday, Wednesday and Friday, and on Monday, Tuesday and Thursday, respectively. The 2 result days with the lowest return numbers had 80% (8/10) of those who returned coming as the only returnee from their day of testing. Conclusion: Result days corresponding with the highest return rates had increased clustering of women from the screening phase, and the result days with fewer women returning had less clustering. These findings are among the first to implicate the importance of defining these patterns of return and the potential for recruitment techniques that exploit such groupings, potentially described by social connections, in limited-resource settings to improve follow-up for screening programs.