Dr. Katherine Halliday is Research Fellow at the London School of Hygiene & Tropical Medicine and trial coordinator for the Tumikia Project. Here she explains the baseline surveys, which took place between March and May 2015 and covered over 24,000 households in Kwale County, in the South Coast of Kenya.
How long did the baseline surveys take?
We recruited 101 field officers on 10th February, conducted training on community sensitisation meetings and conducted 586 sensitsation meetings between 12-27th February. The field officers were then trained in the baseline survey methodologies and we conducted piloting of the surveys between 2nd and 17th March. On the 25th March we started the surveys and took a ten day break for Easter, but otherwise continued until 22nd May. Overall the surveys took 2 months.
How many people were surveyed? Communities? Districts?
We covered 4 subcounties, 84 sublocations and 120 clusters during the surveys. A team of 9 field officers covered a cluster (equivalent to 1,000 households) in three days, sampling 225 households at random from the cluster. In total the 101 field officers visited just over 24,000 households. We conducted surveys in approximately 23,000 households and collected and analysed stool from just over 21,000 individuals of all ages.
How much territory did you cover per day? How far away from Tumikia base in Ukunda is the most remote community or schools?
The TUMIKIA Project covers the whole of Kwale County. Every day we had about 14 vehicles for the field officers and 6 for the lab technicians. The 11 teams of 9 field officers would set out at 6.30 every morning and would drive for up to five hours in some cases to get to their field area for the day, sometimes using ferries or sailing dhows to get there. Some of our clusters were a 260km roundtrip from the office and this trip would be done in a day. Then once the field officer team arrived they would be faced with extremely dispersed households where they were sent to collect samples, meaning a long day of walking between households. When possible they could take motorbikes (piki pikis) between households. Meanwhile the 11 related technician teams comprising three technicians would be based in local labs near each of the clusters being sampled and would wait for the stool samples to be delivered for analysis.
What information and data did you collect?
In a random sample of individuals we collected information on household composition and members, household sociodemographic information, construction, latrine availability and usage, through a questionnaire with the head of the household. We also collected information on individuals’ latrine use and hygiene behaviours with a randomly sampled individual in the house. We also collected a stool sample to investigate STH infection.
What technology or devices did you use?
We administered the questions using a survey programmed on an android application called Survey CTO. Each of the 101 field officers went to the field with a Samsung GT7580 smartphone with the survey programmed. They were taken through a household census and general household questionnaire first and then the smartphone was programmed with a random number generator, which selected one individual at random to ask more in depth individual-level behaviour questions. This person was also requested to provide a stool sample. The smartphone was enabled with a GPS device with which to establish the geographical position of the sampled household. At the end of each day at the office the data was downloaded onto the central server.
What kind of questions do the surveys ask?
We asked questions such as how many people live in the house, how many attend school, main source of income for the household, and school enrolment of children in the household. In addition we asked whether there was a latrine and handwashing facilities in the household, and where individuals go to the toilet when out and about.
How do you make sure answers are truthful and accurate?
The questions are asked to the household head in their compound and relate to household construction, latrines and handwashing facilities. The field officer is required to ask to observe these facilities for themselves to verify these answers. The questions on hygiene behaviours are asked in a few varying ways so as to really verify routine behaviours.
What language are the questions asked in?
The questions are primarily asked in Kiswahili, the National language of Kenya. However, where the field officer spoke the mother tongue language of the area the surveys could be conducted in Digo, Duruma, Kamba, Giriama etc.
How do you analyse the samples in the lab?
The stool pots (containing approximately 50g of stool) were taken to the lab and analysed using the Kato Katz technique, whereby the stool sample is sieved and a small quantity is passed through a template onto a microscope slide. This is covered by a square of cellophane soaked in malachite green and pressed firmly to spread the stool out evenly across the microscope slide surface. The slide is then read under the microscope and examined for ascaris, trichuris and hookworm eggs. Number of eggs per slide are then multiplied up to give the eggs per gram (a measure of intensity of infection).
What do you do with the results now?
We are now double entering all the parasitology data and cleaning and merging all the data so we can begin analysis of prevalence of STH infection at baseline, examining key heterogeneities and risk factors. We shall also provide deworming treatment to all those individuals found infected in the surveys.
What were the biggest challenges? How did you overcome them?
The biggest challenges were the sheer distances the teams had to cover in the vehicles to get to the communities every day and the distances they had to walk between sampled households everyday. They did such a fantastic job, delivering the stool samples to the technicians waiting at the designated laboratories all over the county. It was a tremendous team effort with 101 field officers, 6 supervisors, 33 technicians and 18 drivers all pulling together to cover the 24,000 households!
Were there any surprises along the way?
We encountered transport issues – with flooded roads, ferry jams and punctures due to the extremely rough terrain! On a positive note we were pleasantly surprised with how incredibly welcoming and supportive all the communities were considering we were asking for stool samples!
What is the next step?
The implementing partners - Ministry of Health, Ministry of Education, Science and Technology and Deworm the World at Evidence Action - are currently preparing to treat over half a million people in the 80 clusters due to receive community-based treatment. Last week 55 community health assistants were trained and next week 700 community health volunteers will be trained in how to administer the deworming medicine and fill in the treatment forms. The community-based deworming will be carried out between 20th and 30th July (once Ramadan has ended). This will be a great undertaking for the Kwale County MOH and we are very ready to support teams in this task.