Table 2.

Dose and measurement of intervention fidelity

Intervention	Proposed dose	Delivered dose	Fidelity measures and sources	Outputs
Data strengthening	• Initial training • Monthly data collection and review sessions • DQAs every 6 months	• Initial training • Monthly data sessions, approx. 20 h/year • 2 DQAs in Uganda, 3 in Kenya	• Training rosters • Monitored completeness of monthly data on project dashboard • DQA reports	• Published paper documenting increased completion rates after initial data strengthening publisheda • Paper in press on data strengthening over the life of the project in Kenya
Modified WHO Safe Childbirth Checklist	• Initial training • Checklists supplied for all births	• Initial training • Checklists supplied for all births	• Training rosters • Program records • Periodic audits to assess completion rates by pause point	• Completion audits showed completion rates of 75–90% by study end • mSCC-specific paper forthcoming
PRONTO simulation training and mentorship	• 5 trainings in Uganda • 12 weeks of in situ training/mentoring in Kenya	• 7 trainings in Uganda plus 4 beside mentoring sessions • 12 weeks of in situ training/mentoring in Kenya	• Training rosters • Pre-post knowledge tests at initial and final training sessions • Key simulations video-recoded and coded for uptake of evidence-based practices	• All facilities completed required simulations • PRONTO-specific paper on changes in knowledge and evidence-based practice in recorded simulations forthcoming
Quality improvement Collaborative	• Initial training • Coaching visits/ team meetings at facilities every 2 weeks • 3–6 total Collaborative learning sessions per country	• Initial training • 5 QI Collaborative sessions per country	• Program records • Coaches logged visits and topics during meetings • Collaboratives tracked indicators across facilities at each meeting	• All teams had indicators to present and change efforts to report at each collaborative. • Indicators improved in each country collaborative • QI-specific paper on QI team process and results forthcoming

^aKeating et al. [29]