We conducted a cluster-randomized controlled trial in 24 primary healthcare facilities in rural Tanzania We implemented a four component intervention (training in basic emergency obstetric and newborn care, supportive supervision every six weeks, infrastructure support, and peer outreach) in the 12 randomly selected intervention facilities.

We found that use of health facilities for childbirth did increase, especially among women who's last delivery was at home: 

The intervention did NOT effect quality of obstetric care. We assessed 18 indicators of quality and found improvement on only one.