Assessing Supervisors’ Fidelity Ratings of a Preventive Intervention for Perinatal Open Access Deposited
Downloadable ContentDownload PDF
Intervention Fidelity (IF) is often used in treatment trials to examine the extent to which an intervention is delivered in the way it was intended to be, in order to assess effectiveness of treatment outcomes. The current study examined IF in the context of evaluating an intervention to prevent perinatal depression among pregnant and postpartum women in rural villages in Kenya. The intervention was delivered by community health workers/CHWs who were supervised by two groups of raters: (1) case managers (CMs) who had experience with this intervention in a previous intervention trial; and (2) sisters from a Catholic organization who in turn were trained by the CMs (“sisters”). The goal of this study was to examine if intervention fidelity differed between CMs and sisters: that is, if the two groups provided different and/or similar ratings on how CHWs delivered the intervention. The study sample included CHWs who were the facilitators of a 13-session group intervention (n = 7), and supervisors (2 CMs and 2 sisters). In preparation for the intervention trial, the supervisors observed seven practice groups and completed fidelity rating forms for these 7 groups. In each session, supervisors (CMs or sisters) observed the session and provided ratings on seven session topics across three factors: content (extent to which the CHW covered all topics in the session), comprehension (how well the participants understood content), and participation (group engagement level) from 1-10 (1 = least observed; 10 = most observed). The total rating for each of the three factors used in the analyses was calculated by adding the ratings from each of the 7 topics. The analytical sample included 29 CM ratings and 9 sister ratings over 6 sessions in which both groups provided ratings. Independent samples t-tests were conducted to compare the two supervisor groups on three factors. Results indicate no significant differences in ratings between CMs and sisters on content (CM: M=49.76, SD=7.06; Sisters: M=53.56, SD=6.09); t(36)= -1.45; comprehension (CM: M=49.90, SD=6.98; Sisters: M=53.56, SD=4.16); t(37)=-1.49; and participation (CM: M=53.41, SD=7.55; Sisters: M=56, SD=4.53); t(36)=-.97, all ps =ns. Overall, CMs and sisters rated the facilitators as equally effective in delivering the intervention in the way that it was intended. These findings suggest that sisters, with less mental health experience, can be trained to learn the intervention and supervise CHWs with fidelity. Implications for effective disseminations will be discussed.