HIV Outbreak Investigation Larkana (HOIL) Pakistan
Description
We hypothesized that there would be differences in exposure to parenteral, vertical and/or sexual transmission risk factors in children with HIV compared to age and sex matched neighbourhood controls in Larkana Pakistan. Our data showed association of HIV infection (cases) to history of more injections, blood transfusion, surgery, mothers being positive for HIV/dead and increased number of health facility visits. The predominant mode of transmission in these cases was parenteral, likely due to unsafe injections and poor blood safety. Data on exposure variables was collected through an interview (questionnaire) conducted by our team visiting the child at home. The mother of cases and controls were retested at home through rapid tests. Blood was collected or Hepatitis B and C serology (both cases and controls), HIV rapid testing (from controls only to confirm negative status) and HIV phylogeny (from cases only).
Files
Steps to reproduce
We designed a matched case–control study. Cases were 0-16 year old children registered at the government-run treatment centre. Controls were HIV-uninfected children matched 1:1 to cases by age, sex and neighbourhood. We collected information on likely exposures using a structured questionnaire. For methods, see protocol paper (ref). Statistical Analysis: Statistical analysis was be done using STATA (V.14 or later). Continuous variables were presented as means (SD) for normally distributed and median (IQR) for skewed. Categorical data was presented as proportions. Our dataset includes cases individually pair-matched to controls of the same age, sex and neighborhood (usually living on the same street). The bivariate analysis was done to determine crude ORs and 95% CIs for the association of HIV with key exposures related to parenteral (e.g., receipt of injections, history of injection drug use, history of blood transfusion, history of surgery, hepatitis B or C antibody positivity and so on), vertical (maternal and/or sibling HIV status) or sexual transmission. Multivariable conditional logistic regression was conducted to control for confounding and adjusted ORs for exposures in cases compared with controls with 95% CIs. Conditional logistic regression was required because each case was individually pair-matched to a control of the same age, sex and neighborhood. Adjusting for age and gender was still be required as the matching itself can introduce a bias.