2012 Liberia Health Outcome Monitoring Report


PDF document icon tr-12-88.pdf — PDF document, 11,675 kB (11,956,076 bytes)

Author(s):

Year: 2012

2012 Liberia Health Outcome Monitoring Report Abstract:

The Ministry of Health and Social Welfare (MOHSW) in Liberia is developing an annual health behavior and health outcome monitoring system using Lot Quality Assurance Sampling (LQAS).

LQAS is a relatively rapid and inexpensive approach to data collection for monitoring and evaluation purposes. It can be used to empower program managers to assess program performance, enabling them to determine whether program objectives and targets have been achieved within a specific unit of interest (a geographical area, a facility, an organization, or any other catchment area). The LQAS data collection method and simplified data analysis provides a viable alternative to traditional surveys. It allows for smaller sample sizes than standard probability surveys, and the lower associated costs allow for more frequent sampling. Thus, the speedy collection and ready availability of data from LQAS can help program managers use data as evidence to inform decisions, help in the planning and budgeting process and in formulating targeted interventions.

With LQAS, the entire program area (or catchment area) is divided into meaningful sub-divisions or “lots.” The lot is usually defined as a program supervision area, and the measure is binary (e.g., yes/no, or acceptable/not acceptable) for each indicator included in the study. For example, to determine the status of immunization coverage, “acceptability” is determined by whether the “lot” (supervision area) meets the target for immunization coverage—yes/no. Information from each lot can then be aggregated to provide a coverage estimate for the entire catchment area.

LQAS provides three key pieces of information. It identifies 1) what the problem is because of indicators selected and their coverage estimates, 2) how big the problem is because of the comparison of responses to target levels set for selected indicators, and 3) where the problem is because of the results available by geographic sub-division. However, LQAS does not offer information on why there is problem. Other sources of information are needed to explain the underlying reasons for quantitative results from an LQAS survey, and to identify strategies for improvement.

It is important to clarify that the LQAS approach is not intended to measure incremental change over time. It is designed to assess whether a target has been “met” or “not met” in a designated program supervision area. The small sample size that is required for providing binary (met/not met, acceptable/unacceptable) estimates is a key feature of an LQAS. However, the small sample size at the lot level means that confidence intervals around point estimates at the aggregate level will be wide, such that changes in these point estimates do not register as statistically significant—unless they are very large (e.g., 40–50% increase or decrease).