Title

Late diagnosis of HIV among adults in Zimbabwe, Malawi and Zambia: Risk factors and burden from three population-based HIV impact assessments (PHIAs)

Presenter

Elizabeth Radin

Authors

E. Radin¹, S. Birhanu², A. Low¹, S. Saito¹, K. Sachathep¹, S. Balachandra³, J. Manjengwa¹, Y. Duong¹, S. Jonnalagadda², G. Bello⁴, A. Hakim², N. Ahmed¹, J. Cuervo-Rojas¹, A. Auld⁵, H. Patel², B. Parekh², A. Voetsch², D. Williams², D. Barradas⁶, D. Payne⁵, O. Mugurungi⁷, L. Mulenga⁸, J. Justman¹

Institutions

¹ICAP at Columbia University, New York, United States, ²U.S. Centers for Disease Control & Prevention, Atlanta, United States, ³U.S Centers for Disease Control & Prevention, Harare, Zimbabwe, ⁴Ministry of Health, Government of Malawi, Lilongwe, Malawi, ⁵U.S. Centers for Disease Control & Prevention, Lilongwe, Malawi, ⁶U.S. Centers for Disease Control & Prevention, Lusaka, Zambia, ⁷Ministry of Health & Child Care, Government of Zimbabwe, Harare, Zimbabwe, ⁸Ministry of Health, Government of Zambia, Lusaka, Zambia

Background: Late diagnosis (LD) of HIV increases the risk of morbidity, mortality, and disease transmission. However, the group(s) at greatest risk for LD are not clearly identified in Zimbabwe, Malawi and Zambia. We describe the distribution, burden and correlates of LD among people living with HIV (PLHIV) using the 2015-16 PHIA household surveys.
Methods: The Zimbabwe, Malawi and Zambia PHIAs respectively recruited 20,577 adults aged 15-64, 17,187 adults aged 15-64 and 19,115 adults aged 15-59 for interviews and HIV rapid testing using each country''s national algorithm. All HIV+ adults underwent CD4 count testing. HIV+ participants who reported being HIV- or unaware of their status were defined as undiagnosed. Undiagnosed PLHIV with CD4cell count< 350 cells/µl were defined as meeting LD criteria. Correlates of LD were assessed using multivariate logistic regression. All analyses were weighted to account for the complex survey design of the PHIAs.
Results: Across countries, 26% of PLHIV were undiagnosed. Among undiagnosed PLHIV, 48% met LD criteria: 54% of men and 43% of women. This corresponds to 371,000 undiagnosed adults, 189,000 men and 182,000 women, meeting LD criteria in the total population. Among undiagnosed PLHIV aged 15-24, 35%, or 58,000 people, met LD criteria, while 49% of undiagnosed PLHIV aged 25-40, or 196,000 people, and 55% of undiagnosed PLHIV aged 41-64, or 116,000 people, met LD criteria.
In multivariate analysis, being male (Adjusted Odds Ratio (aOR) 1.54), being aged 25-40 (aOR 1.62), being aged 41-64 (aOR 1.89) and being widowed (aOR 1.76) were associated with LD. Wealth, education, country, urban residence, never having an HIV test and high-risk sexual behaviors (having multiple partners in 12 months, buying or selling sex, and having non-marital, non-cohabiting partners) were not associated with LD.

Characteristic	Proportion with CD4 cell count <350 cells/µl	Number in Total Adult Population w/ CD4 cell count <350 cells/µl	aOR (95% CI) = p≤.05, =p≤0.01, **p≤0.001
Zimbabwe	50.3% (46.6%-54.1%)	157,000 (137,000-177,000)	ref
Malawi	47.1% (41.2%-53.1%)	93,000 (77,000-109,000	0.88 (0.66-1.19)
Zambia	45.1% (41.1%-49.1%)	121,000 (105,000-138,000)	0.86 (0.68-1.07)
Female	42.6% (39.6%-45.7%)	182,000 (164,000-202,000)	ref
Male	54.0% (49.9%-58.0%)	189,000 (167,000-210,000)	1.54 (1.25-1.90) ***
Age 15-24y	34.7% (29.6%-40.3%)	58,000 (47,000-70,000)	ref
Age 25-40y	49.2% (45.7%-52.7%)	196,000 (175,000-218,000)	1.62 (1.22-2.16) ***
Age 41-64y	55.2% (50.6%-59.7%)	116,000 (102,000-131,000)	1.89 (1.38-2.59) ***
Widowed	61.1% (52.3%-69.2%)	35,000 (27,000-43,000)	1.76 (1.18-2.64) **

[Table: Proportion, Number and Correlates of CD4 cell count <350 cells/µl among Undiagnosed PLHIV in Zimbabwe, Malawi and Zambia (N=1934)]

Conclusions: Timely diagnosis remains a challenge, particularly among men and PLHIV over 40 in Zimbabwe, Malawi and Zambia. In addition to increasing risk for morbidity, mortality and transmission, LD may compound difficulties faced by older adults in immune system recovery after treatment initiation. To improve patient outcomes and control the epidemic, it is critical to reach the large numbers of undiagnosed PLHIV with low CD4 counts in this African sub-region.

Download the e-Poster