Census, Mapping and Demographic Survey in an Urban Area of Uganda Jennifer Davis University of California, San Francisco, MS4

Background Malaria remains a significant global health problem with 1 million deaths each year in sub-Saharan Africa. Control is threatened by growing resistance to anti-malarial drugs and low utilization of prevention measures The burden of malaria is best assessed by longitudinal cohort studies offering a multidisciplinary approach to explore complex interactions between host, parasite and the environment

Background Urbanization is occurring rapidly in sub-Saharan Africa. Differences – Socioeconomic demographics – Improved access to diagnosis and treatment – Lower transmission intensity – Focal breeding sites – Variable prevalence rates Existing strategies used to control and treat malaria may need to be tailored in urban environment.

Census and Survey Purpose: To create a sampling frame for a longitudinal study of malaria incidence and combination anti-malarial treatment in urban Ugandan children (AQ+SP vs. AQ+AS vs. Artemether-lumafantrine)  Provide a detailed characterization of malaria risk  Confer a baseline for future longitudinal studies in this community.  Identify covariates influencing the relationship between predictor variables and clinical outcomes of malaria incidence and response to therapy.

Area of Concentration in Global Health Finding an excellent mentor  Grant Dorsey Designing Clinical Research course – Preparing a study proposal  Census and Survey Global Health Area of Concentration core course Fifth year extension for research project – Including travel abroad, data analysis and manuscript writing in San Francisco Legacy – Survey design and two published manuscripts Optional – MD with Thesis

Methods Description of Study Site Malaria is meso-endemic Mulago III parish:  Typical urban slum  Near Mulago Hospital  Primarily residential, with high population density  Petty commercial activities and small-scale subsistence farming

Census and Survey From July to October 2004, a census was conducted to generate a sampling frame for recruitment Covering the entire parish on foot, all households were identified and enumerated After verbal consent, a demographic survey was administered to collect information on inhabitants, home construction and bednet use.

Mapping the Study Area Systematically searching the area on foot, similar readings were taken for: - boundaries of parish - roads, foot paths - mosquito breeding sites - health facilities - other points of interest Households were mapped using handheld global positioning system receivers linked to pocket personal computers.

Population Density: 18,824 persons/km 2

Census Results Census Results 5171 Households in Mulago III Parish 174 (3%) Households vacant 4997 Households occupied 66 (1%) Households not interviewed 40 (0.8%) Not interested 24 (0.5%) No household member > 18 years 2 (0.04%) Adult unable to answer questionnaire 4931 Household occupied and interviewed (16,172 persons) Households with at least one child < 10 years of age (4,058 children < 10 years)

Census Results K Median family size: 3 persons (1-16)  40% inhabited by single or two adult residents Median # of rooms: 1 room (1-17) Female: 48%

Census Results Among adults, females were significantly more likely than males to use a bednet (54% vs. 41%, P<0.0001) or ITN (16% vs. 11%, P<0.001) 0% 10% 20% 30% 40% 50% 60% 70% < > 18 Age (years) Bednet Use (%) ITN Bednet

Recruitment of Study Participants Probability sampling at the level of the household was used to to recruit a random sample of 600 children aged 1 to 10 years From October 2004 to April 2005, experienced home visitors conducted door-to-door interviews to identify households with at least one child aged 1 to 10 years All children from a single household were eligible for enrollment

Screening of Study Participants Eligibility criteria: – Age 1 to 10 years – Agreement to come to study clinic for any illness – Agreement to avoid outside medications – Agreement to remain in Kampala – Absence of known chronic disease – No history of side effects to the study medications – Informed consent provided by parent or guardian Children underwent a history and physical examination Blood was collected to determine baseline parasitemia and for routine laboratory testing Children with weight<10kg, severe malnutrition, life-threatening laboratory results or homozygous hemoglobin SS were excluded

Household Survey A detailed survey was administered to the primary caregiver of the enrolled child at their home – Detailed demographic information about study participant, primary caregiver, household – Shaped after standardized surveys (DHS) Direct data entry into a into a digitally coded form on handheld PDA/computers

Recruitment 582 Households approached for recruitment 209 (36%) Households not eligible for recruitment 116 (20%) No child between ages 1 to 10 years 65 (11%) Not interested 28 (5%) Vacant or destroyed home 373 Households with at least one child screened for enrollment 322 Households with at least one child enrolled in the study 51 (9%) Households screened with no children enrolled 743 Children screened 601 Children enrolled 142 (19%) Children excluded 2176 Households with at least one child < 10 years of age (4,058 children < 10 years) Random Sampling

Comparison of Eligible and Enrolled Households and Children

Description of Enrolled Children

Multivariate Analysis Factors that decreased risk of mild anemia (Hb <11g/dL) – Increasing age (OR = 0.62 for each 1 yr increase, 95% CI , P < 0.001) – Bednet use (OR = 0.58, 95% CI , P = 0.02) – Gender, G6PD activity, and sickle cell trait were not associated

Multivariate Analysis Factors that decreased risk of parasitemia – Low G6PD activity (OR = 0.33, 95% CI , P = 0.009) – Bednet use (OR = 0.64, 95% CI , P = 0.045) – Gender and sickle cell trait were not associated with parasitemia Increasing age (OR = 1.08 for each yr, 95% CI ) was associated with a increased risk of parasitemia (P = 0.06). Children who were parasitemic at enrollment were much more likely to also be mildly anemic (OR = 3.89, 95% CI , P < 0.001)

Summary Reported net use in this urban area: – Higher than levels commonly reported for rural areas – Higher in children less than five years of age Cross-sectional analysis reveals bednet use lowers the risk of anemia and parasitemia BUT…ITN use is far below the goal set by African leaders at the Roll Back Malaria Summit in Abuja in 2000:  Goal: 60% of those at risk of malaria, especially children under five years of age and pregnant women, to have access to ITNs by 2005

Summary Conducting a census in an urban African setting provides useful descriptive data and a method for recruiting representative cohorts in this increasingly important population – Advantage of GPS – Benefit of Local Council involvement – Obstacles: smaller urban adult households, transient population, lower recruitment efficiency Need a broad study area and sizeable recruitment strategy for population based random sampling for pediatric studies in urban areas

Webale! Acknowledgements: MU-UCSF Grant Dorsey Sarah Stadeke Phil Rosenthal Sarah Kemble Damon Francis