Factors associated with HIV infection among children in Larkana District, Pakistan: a matched case-control study

Fatima Mir, Apsara Ali Nathwani, Victoria Simms, Syed Hani Abidi, Amna Rehana Siddiqui, Aneeta Hotwani, Sikander Ali Memon, Saqib Ali Shaikh, Jamila Soomro, Sharaf Ali Shah, Baseer Achakzai, Sofia Furqan, Quaid Saeed, Palwasha Khan, Helen A. Weiss, Syed Faisal Mahmood, Rashida Abbas Ferrand

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)


Background: In April, 2019, an HIV outbreak predominantly affecting children occurred in Larkana District, Sindh, Pakistan. By December, 2019, 881 (4·0%) of 21 962 children screened for HIV had tested positive. We aimed to assess factors associated with HIV infection in this outbreak. Methods: In this individually matched case-control study, we sampled 406 cases (individuals aged <16 years who had registered for paediatric HIV care at the HIV Treatment Centre at Shaikh Zayed Children's Hospital in Larkana City, Pakistan) and 406 controls (individuals without HIV matched by age, sex, and neighbourhood residence, recruited through doorknocking at houses adjacent to case participants). An interviewer-administered questionnaire was used to collect data on possible risk factors for HIV acquisition and a blood sample was collected from all participants for hepatitis B and hepatitis C serology. Mothers of all participants underwent HIV testing. Odds ratios were estimated using conditional logistic regression to assess factors associated with HIV infection. Findings: 406 case-control pairs were recruited between July 3 and Dec 26, 2019. Five pairs were excluded (three pairs had an age mismatch and two pairs were duplicate cases) and 401 were analysed. The prevalence of hepatitis B surface antigen was 18·2% (95% CI 14·5–22·3) among cases and 5·2% (3·3–7·9) among controls, and the prevalence of hepatitis C antibodies was 6·5% (95% CI 4·3–9·4) among cases and 1·0% (0·3–2·5) among controls. 28 (7%) of 397 mothers of cases for whom we had data, and no mothers of 394 controls, were HIV positive. In the 6 months before recruitment, 226 (56%) of 401 cases and 32 (8%) of 401 controls reported having more than ten injections, and 291 (73%) cases and 78 (19%) controls had received an intravenous infusion. At least one blood transfusion was reported in 56 (14%) cases and three (1%) controls in the past 2 years. HIV infection was associated with a history of more injections and infusions (adjusted odds ratio 1·63; 95% CI 1·30–2·04, p<0·0001), blood transfusion (336·75; 23·69–4787·01, p<0·0001), surgery (399·75, 13·99–11 419·39, p=0·0005), the child's mother being HIV positive or having died (3·13, 1·20–8·20, p=0·020), and increased frequency of private clinic (p<0·0001) and government hospital visits (p<0·0001), adjusting for confounders. Interpretation: The predominant mode of HIV transmission in this outbreak was parenteral, probably due to unsafe injection practices and poor blood safety practices. General practitioners across Pakistan need training and systems support in reducing injection use, and in providing safe injections and transfusions only when necessary. Funding: Department of Pediatrics and Child Health, the Aga Khan University, Karachi, Pakistan.

Original languageEnglish
Pages (from-to)e342-e352
JournalThe Lancet HIV
Issue number6
Publication statusPublished - Jun 2021


Dive into the research topics of 'Factors associated with HIV infection among children in Larkana District, Pakistan: a matched case-control study'. Together they form a unique fingerprint.

Cite this