The online chat last week (full transcript .txt) hosted by Abt Associates covered a wide range of topics in OBA: utilization, fraud control, quality assurance, targeting, demand versus supply side programming, clinic management of medical supplies, service substitution, patient empowerment, clinic accreditation, transportation (and rural versus urban divide), staff pay and correlation of facility level incentives in staff turnover, training/ capacity building…
I want to address the most salient issues in separate posts. In this post, I’ll focus on the discussion about utilization. The chat considered two key points: 1) whether vouchers can increase utilization of services and 2) whether that utilization leads to better health outcomes.
In Uganda OBA project, utilization at contracted clinics increased over the year prior to program launch. It is likely some patients sought care who would have not sought care without the voucher. Also likely some patients opted to visit OBA clinics over other facilities. This “substitution effect” is less desireable than giving “new” patients access to care. In the general population, STI utilization increased between baseline survey (2006) and the follow-up survey (2007). The increase was evident in the control areas and the intervention areas, which raises a few concerns about quality of survey implementation. However, the significantly lower syphilis prevalence in 2007 is consistent with the improved use of STI treatment services. In Uganda, we have evidence that vouchers increase utilization and are associated with better health otucomes.
Generally, we measure utilization at two levels: facilities and the general population. The evidence for observed utilization increases is discussed below with links to reports on both evaluations of both patient and general population evaluations.
Clinic evidence
In the Uganda OBA program, seven contracted facilities had lab books available from the year prior to OBA. Among the seven clinics surveyed, non-OBA client visits for STI-related laboratory tests increased by an average of 32% in the first year of OBA compared to the year prior. Some clinics saw considerable increases while others recorded a decrease in non-OBA client numbers but on average there was a 32% increase. Total client visits (OBA and non-OBA) increased by 226% for the same period, with all clinics experiencing an increase in the first year of OBA contracts. (See previous post on THDBlog and download the original report [pdf])
We also compared utilization at contracted facilities to the use of public sector STI treatment services. In Mbarara, Kiruhura, Ibanda and Isingiro districts there were approximately 65,000 patient visits in 2005 at 143 facilities (ranging from rural health center IIs to centralized located health center IVs and regional referral hospitals). In the first year of OBA, the 18 contracted OBA facilities saw more than 9000 patients – representing the equivalent of roughly 14% of the total patient load in the district’s public sector the year before.
General population evidence
We carried out a general population evaluation with a pair of surveys in 82 villages and urban cells in 2006 and 2007. Highlights from the executive summary of the preliminary final report are listed below (download report [pdf]).
• Between 2006 and 2007, awareness of the STI voucher increased more than 25 percent.
• Recognition of STI symptoms among the general population improved 10 percent between 2006 and 2007. Sixty-nine percent of respondents in 2006 and 79 percent of respondents in 2007 were able to recognize two or more STI symptoms.
• Although the most common reason for not seeking healthcare when reporting a health complaint was “lack of money” in both years, 10 percent fewer women cited “lack of money” in the 2007 survey. The second most common reason for not seeking healthcare both years was “distance to provider”.
• Evidence is suggestive but by no means conclusive that syphilis prevalence fell between baseline and follow-up surveys. Syphilis prevalence, based on TPHA test results alone, fell six percent between baseline and follow-up surveys. 18 percent of respondents had reactive results in 2006 and 12 percent in 2007. Syphilis prevalence, based on VDRL lab results alone, fell two percent between baseline and follow-up surveys. Six percent of respondents had reactive results in 2006 and 4 percent in 2007. Syphilis prevalence remained unchanged when TPHA and VDRL results are combined – only 3 percent of respondents were reactive on both tests in the 2006 and 2007 surveys.