ISQUA17-2076GETTING FROM 22 TO 125: SCALING UP INTERVENTIONS TO IMPROVE OUTCOMES OF HIV-POSITIVE MOTHER-BABY PAIRS IN NORTHERN UGANDA

Abstract

Objectives: Scale-up, described as the increased coverage of successfully proven health innovations tested in few sites to a large volume of sites[1] is one of the goals of improvement work. Although the aim of scaling up is to achieve increased coverage and benefits of interventions thus achieving impact on a larger scale[2], the process is often challenging. Challenges to successful scale up include the lack of a clearly defined mechanism for scale up and spread; resource constraints and lack of buy-in. With PEPFAR support, in January 2016, the USAID ASSIST Project initiated scale-up of proven interventions with the objectives of rapidly spreading changes which contribute to a reduction of mother-to-child transmission of HIV to less than 5% and improving the quality of prevention of mother- to child transmission (PMTCT) services to 125 health facilities in 16 districts in Northern Uganda. Method(s): Innovations piloted in 22 demonstration sites (April 2014 - September 2015), through a continuous quality improvement (QI) collaborative model, were organized into concepts and change packages. Scale-up was implemented through a wave-sequence approach to 109 sites; 16 high-volume sites were included in an improvement collaborative to facilitate intensive adaptation of the proven interventions. The first 3 waves constituted 25 sites each, while the 4th had 36 facilities. By December 2016, 102 sites were reached through coaching visits. Spread teams include champions from pilot sites, district and regional improvement coaches and ASSIST staff. Facility teams selected changes from change packages and applied them, with minor adaptations; a log of changes to track implementation was kept. The scale-up work focused on improving 4 key areas, including mother-baby pair retention, provision of a standard care package, rapid testing to monitor exposed infants' status, and data quality. Result(s): Baseline performance of retention of mother-baby pairs (MBP) ranged from 49% in wave 1 sites, 53% in wave 2 sites, 28% in wave 3 sites and 55 % in collaborative sites. By September 2016, retention of MBP pairs was 73% in wave 1, 75% in wave 2, 45% in wave 3 and 75% in collaborative sites. The total number of MBP accessing care monthly in 91 sites increased from 1,704/ 3408 (50 %) to 4,585/6485 (71%), compared to 89% in pilot sites. All scale up sites have functional MBP care points and provide a comprehensive package of services. At baseline, proportion of babies discharged as HIV positive was an average of 16% (12/75) in all sites, as of September 2016, 1.5% of HEI (3/188) were discharged as positive. Conclusion(s): Whilst the assumption was that improvement would occur rapidly and scale-up concluded in 9 months, the scale-up is ongoing. Spread sites proved to be atypical of the demonstration sites and improvement did not occur instantly; sites were not ready, data quality, QI team formation and functionalization of MBP care points had to be done first. However, using a wave-sequencing approach, change packages, coaching visits, and support from QI champions all contributed to successful scale-up.