Using a mobile nurse mentoring and training program to address a health workforce capacity crisis in Bihar, India: Impact on essential intrapartum and newborn care practices

Abstract

Background: To address a health workforce capacity crisis, in coordination with the Government of Bihar, CARE India implemented an on-the-job, on-site nurse mentoring and training intervention named - Apatkalin Matritva evam Navjat Tatparta (AMANAT, translated Emergency Maternal and Neonatal Care Preparedness) - in public facilities in Bihar. AMANAT was rolled-out in a phased manner to provide hands-on training and mentoring for nurses and doctors offering emergency obstetric and newborn care (EmONC) services. This study examines the impact of the AMANAT intervention on nurse-mentees' competency to provide such services in Bihar, India during 2015-2017.

Methods: We used data from three AMANAT implementation phases, each covering 80 public facilities offering basic EmONC services. Before and after the intervention, CARE India administered knowledge assessments to nurse-mentees; ascertained infection control practices at the facility level; and used direct observation of deliveries to assess nurse-mentees' practices. We examined changes in nurse-mentees' knowledge scores using χ² tests for proportions and t tests for means; and estimated proportions and corresponding 95% confidence intervals for routine performance of infection control measures, essential intrapartum and newborn services. We fitted linear regression models to explore the impact of the intervention on nurse-mentees' knowledge and practices after adjusting for potential confounders.

Results: On average, nurse-mentees answered correctly 38% of questions at baseline and 68% of questions at endline (P < 0.001). All nine infection control measures assessed were significantly more prevalent at endline (range 28.8%-86.8%) than baseline. We documented statistically significant improvements in 18 of 22 intrapartum and 9 of 13 newborn care practices (P < 0.05). After controlling for potential confounders, we found that the AMANAT intervention led to significant improvements in nurse-mentees' knowledge (30.1%), facility-level infection control (30.8%), intrapartum (29.4%) and newborn management (24.2%) practices (all P < 0.05). Endline scores ranged between 56.8% and 72.8% of maximum scores for all outcomes.

Conclusion: The AMANAT intervention had significant results in a health workforce capacity crisis situation, when a large number of auxiliary nurse-midwives were expected to provide services for which they lacked the necessary skills. Gaps in intrapartum and newborn care knowledge and practice still exist in Bihar and should be addressed through future mentoring and training interventions.

Study registration: ClinicalTrials.gov number NCT02726230.

Figure 1

Implementation of the AMANAT intervention between September 2015 and January 2017. AMANAT – Apatkalin Matritvaevam Navjat Tatparta, translated emergency obstetrical and neonatal readiness. Blue dots indicate start of data collection; orange dots indicate end of data collection for each phase; each of the three phases covered 80 facilities offering basic emergency obstetric care. The evaluation was on-going, with baseline data collection occurring before the intervention and endline data collection occurring about one month after the end of the intervention.

Figure 2

Nurse-mentee knowledge before and after the AMANAT intervention. Panel A. Changes in the proportion of nurses responding correctly to knowledge questions between baseline and endline. Panel B. Changes in proportion of questions correctly answered between baseline and endline by content. AMANAT – Apatkalin Matritvaevam Navjat Tatparta, translated emergency obstetrical and neonatal readiness. All shown differences between baseline and endline proportions are statistically significant at P < 0.001.

Figure 3

Infection control practices before and after the AMANAT intervention. AMANAT – Apatkalin Matritvaevam Navjat Tatparta, translated emergency obstetrical and neonatal readiness. Columns are proportions and corresponding 95% confidence intervals of facilities performing each of the infection control measure shown. All differences between baseline and endline are statistically significant at P < 0.006 or better.