Effectiveness of self-managed continuous monitoring for maintaining high-quality early essential newborn care compared to supervision visit in Lao PDR: a cluster randomised controlled trial

Abstract Background Thousands of neonatal deaths are expected to be averted by introducing the Early Essential Newborn Care (EENC) in the Western Pacific Region. In Lao People’s Democratic Republic (Lao PDR), the government adopted the EENC programme and expanded it to district hospitals. With the expansion, maintaining the quality of EENC has become difficult for the government. Methods A cluster randomised controlled trial with four strata based on province and history of EENC coaching was implemented to evaluate the effectiveness of self-managed continuous monitoring compared with supervisory visit in Lao PDR between 20 July 2017 and 2 April 2019. Health workers who were routinely involved in maternity care were recruited from 15 district hospitals in Huaphanh (HP) and Xiangkhouang (XK) provinces. The primary endpoint was the score on the determinants of EENC performance measured by the Theory of Planned Behaviour (TPB). Secondary endpoints were set as the knowledge and skill scores. A linear mixed-effects model was applied to test the effects of intervention over time on the endpoints. Results Among 198 recruited health workers, 46 (23.2%) did not complete the final evaluation. TPB scores were 180.9 [Standard Deviation: SD 38.6] and 182.5 [SD 37.7] at baseline and 192.3 [SD 30.1] and 192.3 [SD 28.4] at the final evaluation in the intervention and control groups, respectively. There was no significant difference in changes between the groups in the adjusted model (2.4, p = 0.650). Interviews with participants revealed that district hospitals in HP regularly conducted peer reviews and feedback meetings, while few hospitals did in XK. Accordingly, in stratified analyses, the TPB score in the intervention group significantly increased in HP (15.5, p = 0.017) but largely declined in XK (− 17.7, p = 0.047) compared to the control group after adjusting for covariates. Skill scores declined sharper in the intervention group in XK (− 8.78, p = 0.026), particularly in the practice of managing nonbreathing babies. Conclusions The study indicates that self-managed continuous monitoring is effective in improving behaviour among district health workers; however, additional measures are necessary to support its proper implementation. To maintain resuscitation skills, repeated practice is necessary. Trial registration This trial was registered at UMIN Clinical Trials Registry on 15/6/2017. Registration number is UMIN000027794 .