Impact of an optimized care model for advanced HIV disease: a non-randomized cluster study in Malawi

Despite significant advances in HIV diagnosis and access to ART, many patients still present with advanced HIV disease (AHD). We assessed the effect of an optimized AHD care package on the screening and diagnosis of opportunistic diseases among clients enrolled in AHD care.

This non-randomized cluster design was conducted using a hub-and-spoke model. Twenty-two health facilities, including hub facilities (eight hub sites) and their associated spoke facilities (14 spoke sites), were purposively selected as intervention sites (IS) across three districts in Malawi. The optimized AHD package implemented in these IS included enhanced CD4 testing, tuberculosis (TB) and cryptococcal antigen (CrAg) screening, and appropriate treatment of opportunistic diseases delivered through a hub-and-spoke model. Thirteen non-intervention sites (NIS) (five hub and eight spoke sites) were chosen from four districts that did not implement the intervention and were matched with the IS based on rural/urban settings and health facility types. We abstracted individual-level data from routine clinical records of clients meeting the World Health Organization's definition of AHD between June and December 2021.

Of 963 patients with AHD, 57.4% were seen at IS, and 42.6% at NIS. The IS showed higher proportions of AHD clients identified at (44.3% vs. 36.8%, p = 0.020) and increased screening of children under five years old (7.1% vs. 2.7%, p = 0.004). Additionally, IS diagnosed more cases of WHO stage 3 or 4 disease (47.6% vs. 40.5%, P = 0.029). Patients seen at IS were significantly more likely to receive TB symptom screening (Adjusted Relative Risk [ARR]: 1.13, 95% Confidence Interval [CI]: 1.06-1.21), urine lateral flow lipoarabinomannan test administration (ARR: 1.94, 95% CI: 1.18-3.20), and TB diagnosis (ARR: 2.64, 95% CI: 1.47-4.75). Screening for neurological signs in IS was also improved (ARR: 1.07, 95% CI: 1.02-1.13), as was the diagnosis of cryptococcal meningitis (ARR: 4.28, 95% CI: 1.58-11.70), compared to NIS. There was no difference in retention and mortality in the care of patients after twelve months of follow-up between IS and NIS.

Our study underscores the vital role of improving screening and diagnostic efforts for advanced HIV disease (AHD), notably targeting AHD-related opportunistic infections, including TB and Cryptococcal diseases. Word count: 4,832 words, excluding references.