Lessons learned from stakeholders in a facilitation intervention targeting neonatal health in Quang Ninh province, Vietnam

BACKGROUND: In northern Vietnam the Neonatal health - Knowledge Into Practice (NeoKIP, Current Controlled Trials ISRCTN44599712) trial has evaluated facilitation as a knowledge translation intervention to improve neonatal survival. The results demonstrated that intervention sites, each having an assigned group including local stakeholders supported by a facilitator, lowered the neonatal mortality rate by 50% during the last intervention year compared with control sites. This process evaluation was conducted to identify and describe mechanisms of the NeoKIP intervention based on experiences of facilitators and intervention group members. METHODS: Four focus group discussions (FGDs) were conducted with all facilitators at different occasions and 12 FGDs with 6 intervention groups at 2 occasions. Fifteen FGDs were audio recorded, transcribed verbatim, translated into English, and analysed using thematic analysis. RESULTS: Four themes and 17 sub-themes emerged from the 3 FGDs with facilitators, and 5 themes and 18 sub-themes were identified from the 12 FGDs with the intervention groups mirroring the process of, and the barriers to, the intervention. Facilitators and intervention group members concurred that having groups representing various organisations was beneficial. Facilitators were considered important in assembling the groups. The facilitators functioned best if coming from the same geographical area as the groups and if they were able to come to terms with the chair of the groups. However, the facilitators' lack of health knowledge was regarded as a deficit for assisting the groups' assignments. FGD participants experienced the NeoKIP intervention to have impact on the knowledge and behaviour of both intervention group members and the general public, however, they found that the intervention was a slow and time-consuming process. Perceived facilitation barriers were lack of money, inadequate support, and the function of the intervention groups. CONCLUSIONS: This qualitative process evaluation contributes to explain the improved neonatal survival and why this occurred after a latent period in the NeoKIP project. The used knowledge translation intervention, where facilitators supported multi-stakeholder coalitions with the mandate to impact upon attitudes and behaviour in the communes, has low costs and potential for being scaled-up within existing healthcare systems.

Response process and test-retest reliability of the Context Assessment for Community Health tool in Vietnam

BACKGROUND: The recently developed Context Assessment for Community Health (COACH) tool aims to measure aspects of the local healthcare context perceived to influence knowledge translation in low- and middle-income countries. The tool measures eight dimensions (organizational resources, community engagement, monitoring services for action, sources of knowledge, commitment to work, work culture, leadership, and informal payment) through 49 items. OBJECTIVE: The study aimed to explore the understanding and stability of the COACH tool among health providers in Vietnam. DESIGNS: To investigate the response process, think-aloud interviews were undertaken with five community health workers, six nurses and midwives, and five physicians. Identified problems were classified according to Conrad and Blair's taxonomy and grouped according to an estimation of the magnitude of the problem's effect on the response data. Further, the stability of the tool was examined using a test-retest survey among 77 respondents. The reliability was analyzed for items (intraclass correlation coefficient (ICC) and percent agreement) and dimensions (ICC and Bland-Altman plots). RESULTS: In general, the think-aloud interviews revealed that the COACH tool was perceived as clear, well organized, and easy to answer. Most items were understood as intended. However, seven prominent problems in the items were identified and the content of three dimensions was perceived to be of a sensitive nature. In the test-retest survey, two-thirds of the items and seven of eight dimensions were found to have an ICC agreement ranging from moderate to substantial (0.5-0.7), demonstrating that the instrument has an acceptable level of stability. CONCLUSIONS: This study provides evidence that the Vietnamese translation of the COACH tool is generally perceived to be clear and easy to understand and has acceptable stability. There is, however, a need to rephrase and add generic examples to clarify some items and to further review items with low ICC.