Background: Over the last five years, international development organizations began to modify and adapt the conventional Performance Improvement Model for use in low-resource settings. This model outlines the five key factors believed to influence performance outcomes: job expectations, performance feedback, environment and tools, motivation and incentives, and knowledge and skills. Each of these factors should be supplied by the organization in which the provider works, and thus, support is considered as an overarching element for analysis. Little research, domestically or internationally, has been conducted on the actual effects of each of the factors on performance outcomes and most PI practitioners assume that all the factors are needed in order
for performance to improve. This study presents a unique exploration of how the factors, individually as well as in combination, affect the performance of primary reproductive health providers (nurse-midwives) in two regions of Armenia. Methods: Two hundred and eighty-five nurses and midwives were observed conducting real or
simulated antenatal and postpartum/neonatal care services and interviewed about the presence or absence of the performance factors within their work environment. Results were analyzed to compare average performance with the existence or absence of the factors; then, multiple regression analysis was conducted with the merged datasets to obtain the best models of "predictors" of performance within each clinical service.

Results: Baseline results revealed that performance was sub-standard in several areas and several performance factors were deficient or nonexistent. The multivariate analysis showed that (a)
training in the use of the clinic tools; and (b) receiving recognition from the employer or the client/ community, are factors strongly associated with performance, followed by (c) receiving
performance feedback in postpartum care. Other – extraneous – variables such as the facility type (antenatal care) and whether observation was on simulated vs. real patients (postpartum care) also
had a role in observed performance.

Conclusion: This study concludes that the antenatal and postpartum care performance of health providers in Armenia is strongly associated with having the practical knowledge and skills to use everyday tools of the trade and with receiving recognition for their work, as well as having performance feedback. The paper recognized several limitations and expects further studies will illuminate this important topic further.