Search

Abstract Background The first attempt to implement a new tool or practice does not always lead to the desired outcome. Re-implementation, which we define as the systematic process of reintroducing an intervention in the same environment, often with some degree of modification, offers another chance at implementation with the opportunity to address failures, modify, and ultimately achieve the desired outcomes. This article proposes a definition and taxonomy for re-implementation informed by case examples in the literature. Main body We conducted a scoping review of the literature for cases that describe re-implementation in concept or practice. We used an iterative process to identify our search terms, pilot testing synonyms or phrases related to re-implementation. We searched PubMed and CINAHL, including articles that described implementing an intervention in the same environment where it had already been implemented. We excluded articles that were policy-focused or described incremental changes as part of a rapid learning cycle, efforts to spread, or a stalled implementation. We assessed for commonalities among cases and conducted a thematic analysis on the circumstance in which re-implementation occurred. A total of 15 articles representing 11 distinct cases met our inclusion criteria. We identified three types of circumstances where re-implementation occurs: (1) failed implementation, where the intervention is appropriate, but the implementation process is ineffective, failing to result in the intended changes; (2) flawed intervention, where modifications to the intervention itself are required either because the tool or process is ineffective or requires tailoring to the needs and/or context of the setting where it is used; and (3) unsustained intervention, where the initially successful implementation of an intervention fails to be sustained. These three circumstances often co-exist; however, there are unique considerations and strategies for each type that can be applied to re-implementation. Conclusions Re-implementation occurs in implementation practice but has not been consistently labeled or described in the literature. Defining and describing re-implementation offers a framework for implementation practitioners embarking on a re-implementation effort and a starting point for further research to bridge the gap between practice and science into this unexplored part of implementation.

Abstract Background The BetterBirth trial tested the effect of a peer coaching program around the WHO Safe Childbirth Checklist for birth attendants in primary-level facilities in Uttar Pradesh, India on a composite measure of perinatal and maternal mortality and maternal morbidity. This study aimed to examine the adherence to essential birth practices between two different cadres of birth attendants—nurses and auxiliary nurse midwives (ANMs)—during and after a peer coaching intervention for the WHO Safe Childbirth Checklist. Methods This is a secondary analysis of birth attendant characteristics, coaching visits, and behavior uptake during the BetterBirth trial through birth attendant surveys, coach observations, and independent observations. Descriptive statistics were calculated overall, and by staffing cadre (staff nurses and ANMs) for demographic characteristics. Logistic regression using the Pearson overdispersion correction (to account for clustering by site) was used to assess differences between staff nurses and ANMs in the intervention group during regular coaching (2-month time point) and 4 months after the coaching program ended (12-month time point). Results Of the 570 birth attendants who responded to the survey in intervention and control arms, 474 were staff nurses (83.2%) and 96 were ANMs (16.8%). In the intervention arm, more staff nurses (240/260, 92.3%) received coaching at all pause points compared to ANMs (40/53, 75.5%). At baseline, adherence to practices was similar between ANMs and staff nurses (~ 30%). Overall percent adherence to essential birth practices among ANMs and nurses was highest at 2 months after intervention initiation, when frequent coaching visits occurred (68.1% and 64.1%, respectively, p = 0.76). Practice adherence tapered to 49.2% among ANMs and 56.1% among staff nurses at 12 months, which was 4 months after coaching had ended (p = 0.68). Conclusions Overall, ANMs and nurses responded similarly to the coaching intervention with the greatest increase in percent adherence to essential birth practices after 2 months of coaching and subsequent decrease in adherence 4 months after coaching ended. While coaching is an effective strategy to support some aspects of birth attendant competency, the structure, content, and frequency of coaching may need to be customized according to the birth attendant training and competency. Trial registration ClinicalTrials.gov: NCT2148952; Universal Trial Number: U1111–1131-5647.