Geospatial Transmission Hotspots of Recent HIV Infection — Malawi, October 2019–March 2020

Persons infected with HIV are more likely to transmit the virus during the early stages (acute and recent) of infection, when viral load is elevated and opportunities to implement risk reduction are limited because persons are typically unaware of their status (1,2). Identifying recent HIV infections (acquired within the preceding 12 months)* is critical to understanding the factors and geographic areas associated with transmission to strengthen program intervention, including treatment and prevention (2). During June 2019, a novel recent infection surveillance initiative was integrated into routine HIV testing services in Malawi, a landlocked country in southeastern Africa with one of the world’s highest prevalences of HIV infection.^† The objectives of this initiative were to collect data on new HIV diagnoses, characterize the epidemic, and guide public health response (2). New HIV diagnoses were classified as recent infections based on a testing algorithm that included results from the rapid test for recent infection (RTRI)^§ and HIV viral load testing (3,4). Among 9,168 persons aged ≥15 years with a new HIV diagnosis who received testing across 103 facilities during October 2019–March 2020, a total of 304 (3.3%) were classified as having a recent infection. Higher proportions of recent infections were detected among females, persons aged <30 years, and clients at maternal and child health and youth clinics. Using a software application that analyzes clustering in spatially referenced data, transmission hotspots were identified with rates of recent infection that were significantly higher than expected. These near real-time HIV surveillance data highlighted locations across Malawi, allowing HIV program stakeholders to assess program gaps and improve access to HIV testing, prevention, and treatment services. Hotspot investigation information could be used to tailor HIV testing, prevention, and treatment to ultimately interrupt transmission.

During June 2019, a phased approach was used to integrate recent infection surveillance into HIV testing services in 11 of Malawi’s 28 districts. For persons aged ≥13 years who received a new HIV diagnosis and consented to recent infection surveillance, providers performed a finger prick to conduct an RTRI using the Asante HIV-1 Rapid Recency Assay (Sedia Biosciences). If RTRI results indicated recent infection, additional specimens were collected for viral load testing. Using the testing algorithm for recent infections, new diagnoses were classified as recent if RTRI results indicated a recent infection and viral load was ≥1,000 copies/mL.

This analysis included 103 facilities in five districts that carried out surveillance activities during October 2019-March 2020. These districts were selected based on availability of HIV-testing data disaggregated by age and sex at the facility level. Among 9,295 persons with a new diagnosis of HIV during this period meeting eligibility criteria, 127 (1.4%) declined participation. Two persons aged <15 years were excluded to prevent inclusion of persons who might have been infected through mother-to-child transmission; another four persons who self-reported a previous HIV diagnosis or antiretroviral therapy (ART), were also excluded. In addition, 252 persons whose RTRI results indicated recent infection with a viral load <1,000 copies/mL were excluded because viral suppression likely indicated previous ART use and HIV diagnosis.

Transmission hotspots were defined as one or more facilities in which the observed rate of recent infections was significantly higher (p<0.05) than the expected rate. Transmission hotspot analysis was conducted using SaTScan software (version 9.6; Harvard Medical School) via spatial scan statistic in a Poisson probability model to identify clustering of facilities, using facility geographic coordinates, with significantly higher diagnosis rates of recent HIV infection compared with what was expected based on the overall rate of recent infection (5,6). Rates of recent infection were calculated as the number of recent HIV infections per 100,000 persons at risk for HIV (total number of recent infections plus the total number of negative HIV tests). Relative risks were calculated as the risk for recent infection among facilities inside a given hotspot compared with the risk outside of the hotspot. Facilities reported the total number of HIV negative test results quarterly. Because not all facilities were collecting recent infection surveillance data by October 1, 2019, the number of total HIV-negative tests was adjusted proportionally, assuming testing uniformity across quarters. Hotspots were ranked by probability of occurrence based on log-likelihood and reported using the letter “P.” The analysis was adjusted for sex and age (<30 years or ≥30 years), and did not allow cluster overlap.^¶ A maximum cluster radius (20 kilometers [12.4 miles]) was selected to identify smaller hotspots and to allow response efforts to focus on facilities that contributed most to high rates of recent infection. Given its population density, a secondary analysis was conducted in Blantyre district in the southern region of the country, with a maximum radius of 5 kilometers (3.1 miles), to identify potential micro-hotspots within this district. Hotspots from this secondary analysis were also ranked according to the log-likelihood and reported using the letter “S.” Statistical analyses were conducted using R (version 3.5.0; R Foundation) to analyze the percentage of recent HIV infections among total tests performed, by district, age group, sex, HIV testing service entry point, and facility urban-rural classification. This activity was reviewed and approved by the Malawi National Health Science Research Institutional Review Board and was reviewed by CDC and conducted consistent with applicable federal law and CDC policy.**

Among 9,168 new HIV diagnoses, 3.3% (304) were recent infections (Table 1). The number of recent infections was highest in Blantyre district (116). The percentages of new diagnoses that were recent infections was highest in Machinga district (6.9%) and lowest in Blantyre (2.4%) and Mangochi (2.4%) districts. The percentage of new diagnoses that were recent infections was highest among persons aged <30 years (4.6%), females (4.0%), clients at youth clinics (12.8%) and maternal/child health clinics (6.3%) and those who received a diagnosis in rural facilities (3.8%).

Spatial analyses identified six transmission hotspots: three in the primary analysis (P) and three in the secondary analysis (S) (Table 2) (Figure). In the primary analysis, the median age (range = 26–30 years) of persons with recent infection was similar across hotspots. Hotspot P1 was in Blantyre district, a mostly urban area^†† that includes four facilities within a radius of 10.2 kilometers (6.3 miles). The recent infection rate in Blantyre district was 575 per 100,000 persons at risk for HIV (relative risk [RR] = 3.1; p<0.001); the highest percentage of recent infections occurred among females (81.5%). Hotspot P2 included four facilities located across the border of Machinga and Zomba districts within a radius of 16.1 kilometers (9.9 miles), but were primarily in Machinga district, a mostly rural area. The recent infection rate in hotspot P2 was 376 per 100,000 persons at risk (RR = 2.0; p = 0.018). Hotspot P3 was a single facility in Blantyre with a recent infection rate of 818 per 100,000 persons at risk (RR = 4.2; p = 0.025). In the secondary analysis limited to Blantyre district, hotspot S1 included two facilities that were also part of hotspot P1 of the primary analysis, suggesting these facilities contributed most to the high rate of recent infection in their respective localities. Hotspots S2 and S3 included facilities with significantly elevated rates that were not identified in the primary analysis. Median age (range = 25–30 years) and percentage of females (range = 60%–78%) among hotspots in the secondary analysis were similar to hotspots in the primary analysis.

Corresponding author: Andrew F. Auld, [email protected].

¹Division of Global HIV & TB, Center for Global Health, CDC; ²Public Health Institute, Oakland, California; ³International Training and Education Center for Health, Lilongwe, Malawi; ⁴Global Strategic Information, Institute for Global Health Sciences, University of California, San Francisco, California; ⁵International Training and Education Center for Health, Department of Global Health, University of Washington, Seattle, Washington; ⁶Division of Global HIV & TB, Center for Global Health, CDC, Lilongwe, Malawi; ⁷Department of HIV & AIDS, Ministry of Health, Lilongwe, Malawi; ⁸Medical Research Council Centre for Global Infectious Disease Analysis, School of Public Health, Imperial College London, London, United Kingdom.

References

1. Vasylyeva TI, Friedman SR, Smyrnov P, Bondarenko K. A new approach to prevent HIV transmission: Project Protect intervention for recently infected individuals. AIDS Care 2015;27:223–8. https://doi.org/10.1080/09540121.2014.947913external icon PMID:25244688external icon
2. Kim AA, Behel S, Northbrook S, Parekh BS. Tracking with recency assays to control the epidemic: real-time HIV surveillance and public health response. AIDS 2019;33:1527–9. https://doi.org/10.1097/QAD.0000000000002239external icon PMID:31021850external icon
3. Parekh, B. Performance evaluation of Asante™ Rapid Recency Assay for HIV diagnosis and detection of recent infection: potential for surveillance and prevention. IAS 2017 [Abstract A-854–0164–05286]. International AIDS Society Conference on HIV Science; July 23–26, 2017; Paris, France; 2017. https://trace-recency.org/wp-content/uploads/2021/06/Parekh_IAS-2017-Asante-Poster11.pdfpdf iconexternal icon
4. Kassanjee R, Pilcher CD, Busch MP, et al.; Consortium for the Evaluation and Performance of HIV Incidence Assays (CEPHIA). Viral load criteria and threshold optimization to improve HIV incidence assay characteristics. AIDS 2016;30:2361–71. https://doi.org/10.1097/QAD.0000000000001209external icon PMID:27454561external icon
5. Kulldorff M. A spatial scan statistic. Commun Stat Theory Methods 1997;26:1481–96. https://doi.org/10.1080/03610929708831995external icon
6. Latash J, Greene SK, Stavinsky F, et al. Salmonellosis outbreak detected by automated spatiotemporal analysis—New York City, May–June 2019. MMWR Morb Mortal Wkly Rep 2020;69:815–9. https://doi.org/10.15585/mmwr.mm6926a2external icon PMID:32614808external icon
7. Malawi Ministry of Health. Malawi population‐based HIV impact assessment (MPHIA) 2015–16. Lilongwe, Malawi: Malawi Ministry of Health; 2018. https://phia.icap.columbia.edu/wp-content/uploads/2019/08/MPHIA-Final-Report_web.pdfpdf iconexternal icon
8. Nutor JJ, Duah HO, Agbadi P, Duodu PA, Gondwe KW. Spatial analysis of factors associated with HIV infection in Malawi: indicators for effective prevention. BMC Public Health 2020;20:1167. https://doi.org/10.1186/s12889-020-09278-0external icon PMID:32711500external icon
9. Godin A, Eaton JW, Giguère K, et al. Inferring population HIV incidence trends from surveillance data of recent HIV infection among HIV testing clients. AIDS 2021;35:2383–8. https://doi.org/10.1097/QAD.0000000000003021external icon PMID:34261098external icon
10. Wand H, Ramjee G. Targeting the hotspots: investigating spatial and demographic variations in HIV infection in small communities in South Africa. J Int AIDS Soc 2010;13:41. https://doi.org/10.1186/1758-2652-13-41external icon PMID:20979651external icon

Abbreviations: HTC = HIV testing and counseling; VCT = voluntary counseling and testing.

Abbreviation: RR = relative risk.
* Spatial analyses identified six transmission hotspots: three in the primary analysis and three in the secondary analysis.
^† Ranked by probability of occurrence based on log-likelihood.
^§ Cases per 100,000 population. Denominator for rate calculation was total persons at risk for HIV, calculated as the sum of recent infections observed and total negative HIV test results.
^¶ Among recent infections included in a cluster.
** Expected number of recent infections in facilities included in a cluster based on the rate of infection across all facilities.
^†† RR for recent infection among persons tested within a cluster.

FIGURE. Geospatial transmission hotspots of recent HIV infection among health facilities implementing recent HIV infection surveillance in (A) five districts in Malawi and (B) Blantyre district, Malawi* — October 2019−March 2020

* The primary analysis (A) in five districts (Blantyre, Lilongwe, Machinga, Mangochi, and Zomba) in Malawi with a 20-km (12.4-mi) maximum cluster radius identified three HIV transmission hotspots (P1 = Blantyre, P2 = Machinga and Zomba, P3 = Blantyre [one facility]); a secondary analysis (B) focused on Blantyre district alone with a 5-km (3.1-mi) maximum cluster radius identified three additional HIV transmission hotspots (S1, S2, S3 = all Blantyre district).