Data Collection Notes
Training
From a pool of over 1,200 applicants for the supervisor and interviewer positions, NMCP and LISGIS recruited 56 for the interviewer/supervisor training. They also allowed over 26 observers to attend the training without remuneration, all of whom hoped to do better than those who were officially recruited. The pool of male and female trainees consisted largely of those who had experience in previous surveys such as the 2007 LDHS, the 2005 LMIS, and other social surveys.
These participants attended a two-week training course from December 1-12 at Thinker’s Village Beach on the outskirts of Monrovia. Training of the interviewer/supervisor candidates consisted of reviewing how to fill the Household and Woman’s Questionnaires, mock interviewing, and sessions covering tips on interviewing, how to locate selected households, and how to code interview results. Two quizzes were administered. Trainers included the LMIS Project Director, the Assistant Project Director, and three LISGIS staff, with support from two Macro staff. Despite the large candidate pool, many did not qualify on the basis of tests and practice interviewing and many were not proficient in the major local languages. Of the 82 attendees in the interviewer/supervisor training, twelve were selected as supervisors, 24 were selected as interviewers, and eight were held in reserve.
NMCP also identified over 35 staff with either laboratory or medical experience who were trained in taking blood for the anemia and malaria testing at the same time and place as the interviewer/supervisor candidates. Of these, 24 were selected as health technicians for the biomarker data collection and 7 were further trained as microscopists in the laboratory (see below). The health technicians were trained by a Macro biomarker specialist and a malaria laboratory consultant on how to identify children eligible for testing, how to administer informed consent, how to conduct the anemia and malaria rapid tests, and how to make a proper thick blood smear. They were also trained on how to store the blood slides, how to record test results on the questionnaire, and how to provide results to the parents/caretakers of the children tested. Trainees participated in numerous practice sessions in the classroom.
All trainees participated in two field practice exercises in households living close to the training site. They also received a lecture on the epidemiology of malaria in Liberia and correct treatment protocols by a senior member of the NMCP. Finally, all health technicians, team supervisors, and the nurses/nurse aides on each team received more specific instructions on how to calculate the correct dose of antimalarial medication to leave with the parents/caretakers of children who test positive on the malaria rapid diagnostic test. This included how to use the portable scales to determine the child’s weight. It also included how to record children’s anemia and malaria results on the anemia and malaria brochure that was to be left in every household in which children were tested and on how to fill in the referral slip for any child who was found to be severely anemic.
Fieldwork
Twelve teams were organized for the data collection, each comprised of one supervisor, two interviewers, two health technicians, and one driver. Three senior staff from LISGIS, one from NMCP, and one from the MOH&SW Monitoring and Evaluation Unit were designated as field coordinators and were each assigned a number of teams to monitor. NMCP was able to organize the questionnaire printing on time, and arrange for the fieldwork logistics such as field staff contracts, identification cards with pictures, special survey T-shirts, and other local supplies for the field teams.
Data collection for the LMIS started as scheduled on December 15, 2008. In order to allow for maximum supervision in the first two weeks as well as to allow teams to be home for Christmas, all 12 teams started work in Monrovia, covering two clusters each before moving out of Monrovia just after the holidays. Fieldwork was completed by all teams by the end of February. However, field checking uncovered a situation in which one team had not actually conducted interviews in some four clusters that it claimed it had completed. To rectify the deception, three other teams were sent to complete the four clusters in March 2009.
Laboratory Testing
Prior to the start of the field staff training, a Macro malaria consultant worked with the head of the malaria laboratory at the JFK Hospital compound to inspect the lab, check on supplies, unpack and inventory the supplies sent by Macro, and obtain electrical stabilizers for the microscopes and materials needed for staining the slides. Although the lab was refurbished by the Chinese in 2007, it had not been extensively used.
After the health technician training was completed, the consultant trained the seven identified microscopists at the laboratory. All trainees had participated in the health technician training, so they were fully aware of the objectives and logistics of the survey. The training covered the importance of good laboratory practice such as quality control of reagents, smears, and malaria diagnosis and the consequences of failing to care for and maintain laboratory equipment used in microscopy. Also discussed was the biology of the plasmodium parasite, including describing the red blood cells where the parasites live, the life cycle of each plasmodium species, and their characteristic features. The importance of making good blood smears was emphasized, as were the standard procedures for staining slides. Finally, trainees spent about a week practicing slide reading using blood smears taken during the practice interviewing. One of the trainees was assigned to registering, staining and mounting the slides. The other six microscopists then started to read slides from the actual survey. The purpose of the blood slides was to provide a gold standard for malaria infection and not to ascertain the type of parasite.
The consultant returned to Monrovia in late January to check on the progress of the lab work. During this visit, he conducted a second reading of some 400 slides, including at least 60 from each of the six microscopists. Using his reading as the gold standard, he selected microscopists with the fewest discordant results to be the second readers. If the results of the first and second readings did not match, a third person acted as the tie breaker. Laboratory testing continued for about five months. Macro also provided the computer software for recording the laboratory test results.
After the laboratory testing at the Malaria Center was completed, a systematic sample of 300 slides were sent to the Comprehensive Health Center Laboratory in Saclepea for an independent quality control check.