Setting: 100 bed medical ward in referral hospital, Lilongwe, Malawi.Objective: HIV positive patients admitted to hospital often have advanced HIV disease (AHD) and are at risk for mortality. WHO guidelines suggest a package of care for AHD; these are often not implemented, especially in inpatient settings. We describe an implementation model for AHD care, its outcomes in routine care and provide cost estimates.Design: An "AHD care room" was established staffed by HIV counselor, nurse, and clinical officer allowing Provider Initiated Testing and Counseling, diagnostic testing for AHD and ensuring availability of HIV and TB drugs for rapid treatment initiation. Results:In the observation period from January to December 2020, a total of 1549 medical inpatients were tested for HIV (coverage 77.1%); 69 tested positive (yield 4.5%). The total proportion of HIV positive was 32.3% (638 already on ART and 69 newly diagnosed). CD4+ testing was done in 460 medical inpatients (65.1%); 245 (53.2%) were below 200 cells/ml and thus met definition of AHD. A total of 238 received S-CrAg tests; 39 (16.3%) were positive; 62 (28.3%

No abstract

Background Cryptococcal meningitis (CM) and tuberculosis (TB) remain leading causes of hospitalization and death amongst people living with HIV, particularly those with advanced HIV disease. In hospitalized patients, prompt diagnosis of these diseases may improve patient outcomes. The advanced HIV rapid diagnostic tests such as determine TB urine lipoarabinomannan lateral flow assay (urine LAM), urine X-pert MTB/RIF assay (urine X-pert), and serum/blood cryptococcal antigen test (serum CrAg) are recommended but frequently not available in many resource-limited settings. We describe our experience providing these tests in a routine hospital setting. Method From 1 August 2016 to 31 January 2017, a prospective cohort study to diagnose TB and Cryptococcal meningitis using point of care tests was conducted in the medical wards at Kamuzu Central Hospital, in Lilongwe, Malawi. The tests offered were PIMA CD4 cell count, serum CrAg, urine LAM, and urine X-pert. The testing was integrated into an existing HIV/TB treatment room on the wards and performed close to admission time. Patients were followed until discharge or death in the ward. Results We included 438 HIV-positive patients; 76% had a previously known HIV diagnosis (87% already on ART). We measured CD4 count in 365/438 (83%), serum CrAg in 301/438 (69%), urine LAM in 363/438 (83%), and urine X-pert in 292/438 (67%). The median CD4 count was 144 cells/ml (IQR 46–307). Serum CrAg positivity rate was 23 /301 (8%) and CM was confirmed by CSF Crag in 13/23 (56%). The majority of CM patients 9/13 (69%) started antifungal therapy within two days of diagnosis. Urine LAM and urine X-pert positivity rates were 81/363(22%) and (14/292 (5%) respectively. The positivity rate of urine LAM was higher in patients with low CD4 cell counts (< 100 cells/ml) and low BMI (< 18.5). Most patients with positive urine LAM started TB treatment on the same day. Despite the early diagnosis and treatment of TB and CM, the inpatient mortality was high; 30% and 25% respectively. Conclusion Although advanced HIV rapid diagnostic tests are recommended, one key challenge in implementation is the limited trained personnel administering the tests. Despite the effective use of the point of care tests in the clinical care of hospitalized TB and CM patients, mortality among these patients remained unacceptably high. Henceforth we need to train other cadres apart from nurses, clinicians, and laboratory technicians to conduct the tests. There is an urgent need to identify and modify other risks of death from TB and CM. Trial registration: Malawi National Health Science Research committee: Protocol # 1144. Registered 2 July 2014 and University Of North Carolina IRB #: UNCPM 21412, approved 13th October 2014.

Background The spleen is frequently scanned in workup of infections. Hypoechoic splenic micro-abscesses are known signs of disseminated tuberculosis in HIV co-infected patients. The spleen of HIV patients is thus often scanned using high-frequency transducers. Methods and findings We describe a reticulo-nodular “sponge pattern” in the spleen of an HIV-positive patient with Hodgkin’s lymphoma. Disseminated throughout the spleen, very small (1.5–2.0 mm) hypoechoic lesions having a branching reticulo-nodular distribution were seen. The lesions partly, but not entirely, follow splenic vasculature. Review of stored images of other patients identified 15 more cases showing a similar pattern. All patients were HIV positive, almost all with CD4 counts below 200 cells/mm3. Seven (44%) were additionally diagnosed with HHV-8-associated diseases, but the pattern was seen with various underlying opportunistic infections. Discussion and conclusion After comparison with spleen microscopic anatomy, we hypothesize that the white pulp of spleens in our patients is hyperplastic or otherwise changed in consistency to be better visible by high-frequency ultrasound. Concomitant human herpesvirus-8 infection may be another cause of this visible white pulp. While we can only speculate about the etiology of the splenic “sponge pattern,” it needs to be recognized as it may be misinterpreted as splenic micro-abscesses of disseminated infections, like tuberculosis in severely immune-compromised patients.