Lot Quality Assurance Sampling in Liberia
MEASURE Evaluation began providing technical assistance in Liberia for a pilot study using the lot quality assurance sampling (LQAS) methodology in early 2011. The purpose of the study was to provide information for programmatic decision making on key health indicators at both the national and county levels. The four counties selected for the pilot test were Bomi, Bong, Lofa and Nimba counties; combined, they comprise one third of Liberia’s total population.
A team from MEASURE Evaluation made an initial visit to Liberia in January, during which preliminary work, such as indicator and county selection, took place. Stakeholders chose 26 indicators from six program areas – water, sanitation and hygiene; immunization and child health; nutrition; malaria; maternal health; and family planning. They also selected three subgroups to include in the study – women of reproductive age; children aged 0-59 months; and children aged 0-59 months who had a diarrheal episode, a febrile episode and/or cough and difficulty breathing in the previous two weeks.
Training in Liberia began in March, with fieldwork slated to start in April – a timeline dictated largely by the impending rainy season. “We had to do as much of the survey as possible before Liberia’s rainy season because, if not, we would run into problems,” explained Scott McKeown, who served as activity lead. “Potholes would be even bigger and trenches would form in the roads.”
Before setting out for the counties, McKeown and his team conducted several weeks of training for the interviewers and data-entry teams – who were selected by sub-contracting agency Subah-Belleh Associates.
“We had four teams of interviewers, with four people each – three served as interviewers and one served as a supervisor,” McKeown said. And, importantly, all interviewers were Liberian and experienced in the counties they worked in, as McKeown went on to explain: “It was necessary that the surveyors knew how to talk to the people they were interviewing, and that they had a bond. We wanted them to stick to the survey instrument, of course, but they also needed to be able to slip into local slang when necessary.”
After sufficient training and tweaking of the surveys, the interview teams launched into the field on April 10, scattering as planned to the various counties. The Liberian team used a traditional LQAS design, with five lots (supervision areas) in each county, and 19 interviews (from randomly selected households) in each lot, for a total of 380 questionnaires.
In addition to the interview teams, a Quality Control Team, made up of McKeown, Jack Hazerjian of MEASURE Evaluation and Sumo Zeze of Liberia, traveled from county to county to check on the teams. “We put at least 2,500 miles on our car in 2-3 weeks,” remembered McKeown. “That’s a lot of travel on mediocre roads!”
Fieldwork lasted two-and-a-half weeks and went smoothly. Interviewers were well received by people in the different communities, with very few households refusing to participate in surveys.
One of the most striking aspects of the LQAS pilot study was the quick turnaround from data collection to data entry. “On April 10th we went into the field, and the first round of data entry for the 380 questionnaires was done by April 29th,” McKeown sad. “That’s pretty impressive.”
As a result of the pilot study’s success, Liberia’s Ministry of Health has requested additional activities using the LQAS methodology in other counties. “From the ministry side, the LQAS method is recognized as being proactive,” McKeown said. “Most data they get is facility-based. This method, though, involves going out and talking to people in the communities. It’s a great instrument to add to their information arsenal.”