Search

Background. Machine learning (ML) refers to computational algorithms designed to learn from patterns in data to provide insights or predictions related to that data. Objective. Multiple studies report the development of predictive models for triage or identification of critically ill children. In this study, we validate machine learning models developed in South Africa for the identification of critically ill children presenting to a tertiary hospital. Results. The validation sample comprised 267 patients. The event rate for the study outcome was 0.12. All models demonstrated good discrimination but weak calibration. Artificial neural network 1 (ANN1) had the highest area under the receiver operating characteristic curve (AUROC) with a value of 0.84. ANN2 had the highest area under the precision-recall curve (AUPRC) with a value of 0.65. Decision curve analysis demonstrated that all models were superior to standard strategies of treating all patients or treating no patients at a proposed threshold probability of 10%. Confidence intervals for model performance overlapped considerably. Post hoc model explanations demonstrated that models were logically coherent with clinical knowledge. Conclusions. Internal validation of the predictive models correlated with model performance in the development study. The models were able to discriminate between critically ill children and non-critically ill children; however, the superiority of one model over the others could not be demonstrated in this study. Therefore, models such as these still require further refinement and external validation before implementation in clinical practice. Indeed, successful implementation of machine learning in practice within the South African setting will require the development of regulatory and infrastructural frameworks in conjunction with the adoption of alternative approaches to electronic data capture, such as the use of mobile devices. [ABSTRACT FROM AUTHOR]

7‐11 November 2010, Tenth International Congress on Drug Therapy in HIV Infection, Glasgow, UK, Background With the large international mobilization of resources against HIV/AIDS and the great improvement of ARV drug access within the last 7 years in resource‐limited countries, more HIV‐infected people are [...]

Abstract Background Ugandan national guidelines recommend initiation of combination antiretroviral therapy (cART) at CD4+ T cell (CD4) count below 350 cell/μl, but the implementation of this is limited due to availability of medication. However, cART initiation at higher CD4 count increases survival, albeit at higher lifetime treatment cost. This analysis evaluates the cost-effectiveness of initiating cART at a CD4 count between 250–350 cell/μl (early) versus Methods Life expectancy of cART-treated patients, conditional on baseline CD4 count, was modeled based on published literature. First-line cART costs $192 annually, with an additional $113 for patient monitoring. Delaying initiation of cART until the CD4 count falls below 250 cells/μl would incur the cost of the bi-annual CD4 count tests and routine maintenance care at $85 annually. We compared lifetime treatment costs and disability adjusted life-expectancy between early vs. delayed cART for ten baseline CD4 count ranges from 250-350 cell/μl. All costs and benefits were discounted at 3% annually. Results Treatment delay varied from 6–18 months. Early cART initiation increased life expectancy from 1.5-3.5 years and averted 1.33–3.10 disability adjusted life years (DALY’s) per patient. Lifetime treatment costs were $4,300–$5,248 for early initiation and $3,940–$4,435 for delayed initiation. The cost/DALY averted of the early versus delayed start ranged from $260–$270. Conclusions In HIV-positive patients presenting with CD4 count between 250-350 cells/μl, immediate initiation of cART is a highly cost-effective strategy using the recommended one-time per capita GDP threshold of $490 reported for Uganda. This would constitute an efficient use of scarce health care funds.

Abstract Background Nucleic acid amplification tests (NAATs) have offered hope for rapid diagnosis of tuberculosis (TB). However, their efficiency with smear-negative samples has not been widely studied in low income settings. Here, we evaluated in-house PCR assay for diagnosis of smear-negative TB using Lowenstein-Jensen (LJ) culture as the baseline test. Two hundred and five pulmonary TB (PTB) suspects with smear-negative sputum samples, admitted on a short stay emergency ward at Mulago Hospital in Kampala, Uganda, were enrolled. Two smear-negative sputum samples were obtained from each PTB suspect and processed simultaneously for identification of MTBC using in-house PCR and LJ culture. Results Seventy two PTB suspects (35%, 72/205) were LJ culture positive while 128 (62.4%, 128/205) were PCR-positive. The sensitivity and specificity of in-house PCR for diagnosis of smear-negative PTB were 75% (95% CI 62.6-85.0) and 35.9% (95% CI 27.2-45.3), respectively. The positive and negative predictive values were 39% (95% CI 30.4-48.2) and 72.4% (95% CI 59.1-83.3), respectively, while the positive and negative likelihood ratios were 1.17 (95% CI 0.96-1.42) and 0.70 (95% CI 0.43-1.14), respectively. One hundred and seventeen LJ culture-negative suspects (75 PCR-positive and 42 PCR-negative) were enrolled for follow-up at 2 months. Of the PCR-positive suspects, 45 (60%, 45/75) were still alive, of whom 29 (64.4%, 29/45) returned for the follow-up visit; 15 (20%, 15/75) suspects died while another 15 (20%, 15/75) were lost to follow-up. Of the 42 PCR-negative suspects, 22 (52.4%, 22/42) were still alive, of whom 16 (72.7%, 16/22) returned for follow-up; 11 (26.2%, 11/42) died while nine (21.4%, 9/42) were lost to follow-up. Overall, more PCR-positive suspects were diagnosed with PTB during follow-up visits but the difference was not statistically significant (27.6%, 8/29 vs. 25%, 4/16, p = 0.9239). Furthermore, mortality was higher for the PCR-negative suspects but the difference was also not statistically significant (26.2% vs. 20% p = 0.7094). Conclusion In-house PCR correlates poorly with LJ culture for diagnosis of smear-negative PTB. Therefore, in-house PCR may not be adopted as an alternative to LJ culture.

Background: Sexually transmitted infections (STIs) significantly increase HIV transmission. Sexually transmitted infections may be asymptomatic and therefore remain undiagnosed in HIV-positive persons. Routine screening and treatment of STIs in HIV-positive high-risk populations in sub-Saharan Africa have not been described previously., Methods: We reviewed data from an HIV-positive high-risk population at the Infectious Diseases Institute, Makerere University, a large urban HIV clinic, between July 2011 and April 2012. Our high-risk population cohort included female sex workers, long-distance drivers, barmaids, taxi drivers, commercial motorcycle "boda-boda" riders, soldiers, police officers, prison officers, security guards, prisoners, and fishermen., Results: Of 355 participants enrolled in the high-risk population's program, 21.4% were diagnosed with an STI either clinically or microbiologically. The STIs diagnosed in this population were syphilis, hepatitis B, genital herpes, human papilloma virus infection (condylomata acuminata), nongonococcal urethritis (NGU), and gonorrhea. Rates of syphilis, hepatitis B, genital herpes, condylomata acuminata, NGU, and gonorrhea were 8.5%, 7.0%, 5.4%, 1.4%, 1.4%, and 0.3%, respectively., Conclusion: Clinical and microbiologically diagnosed STIs were diagnosed in nearly one-fourth of the HIV-positive high-risk population. HIV care programs should note our high rates of STIs among HIV-positive high-risk populations and consider routine screening and treatment algorithms for these populations in their own settings., (© The Author(s) 2013.)

Background: Short-medium term studies from sub-Saharan Africa show that, despite high early mortality, substantial loss to program, and high rates toxicity, patients on antiretroviral treatment have achieved outcomes comparable to those in developed settings. However, these studies were unable to account for long term outcomes of patients as they stayed longer on treatment., Objectives: We aim to describe ten years outcomes of one of the first cohort of HIV positive patients started on antiretroviral treatment (ART) in Sub-Saharan Africa., Methods: We report 10-years outcomes including mortality, retention, CD4-count response, virological outcomes, ART regimens change from a prospective cohort of 559 patients initiating ART and followed up for 10 years Uganda., Results: Of 559 patients, 69.1% were female, median age (IQR) was 38 (33-44) years, median CD4-count (IQR) 98 (21-163) cell/μL; 74% were started on stavudine, lamivudine and nevirapine, 26% on zidovudine, lamivudine and efavirenz. After 10 years 361 (65%) patients were still in the study; 127 (22.7%) had died; 30 (5%) were lost to follow-up; 27 (5%) transferred; 18 (3%) withdrew consent. The probability of death was high in the first year (0.15, 95%, CI 0.12-0.18). The median CD4 count increased from 98 to 589 cell/μL (IQR: 450-739 cell/μL) with a median increase of 357 cells/μL (IQR: 128-600 cells/μL); 7.4% never attained initial viral suppression and of those who did 31.7% experienced viral failure. Three hundred and two patients had at least one drug substitution while on first line after a median of 40 months; 66 (11.9%) of the patients were switched to a second line PI-based regimen due to confirmed treatment failure., Conclusions: Despite the high rate of early mortality due to advanced disease at presentation the outcomes from this cohort are encouraging, particularly the remarkable and incremental immune-recovery and a satisfactory rate of virologic suppression.

Introduction: Due to a limited health workforce, many health care providers in Africa must take on health leadership roles with minimal formal training in leadership. Hence, the need to equip health care providers with practical skills required to lead high-impact health care programs. In Uganda, the Afya Bora Global Health Leadership Fellowship is implemented through the Makerere University College of Health Sciences (MakCHS) and her partner institutions. Lessons learned from the program, presented in this paper, may guide development of in-service training opportunities to enhance leadership skills of health workers in resource-limited settings., Methods: The Afya Bora Consortium, a consortium of four African and four U.S. academic institutions, offers 1-year global health leadership-training opportunities for nurses and doctors. Applications are received and vetted internationally by members of the consortium institutions in Botswana, Kenya, Tanzania, Uganda, and the USA. Fellows have 3 months of didactic modules and 9 months of mentored field attachment with 80% time dedicated to fellowship activities. Fellows' projects and experiences, documented during weekly mentor-fellow meetings and monthly mentoring team meetings, were compiled and analyzed manually using pre-determined themes to assess the effect of the program on fellows' daily leadership opportunities., Results: Between January 2011 and January 2015, 15 Ugandan fellows (nine doctors and six nurses) participated in the program. Each fellow received 8 weeks of didactic modules held at one of the African partner institutions and three online modules to enhance fellows' foundation in leadership, communication, monitoring and evaluation, health informatics, research methodology, grant writing, implementation science, and responsible conduct of research. In addition, fellows embarked on innovative projects that covered a wide spectrum of global health challenges including critical analysis of policy formulation and review processes, bottlenecks in implementation of national HIV early infant diagnosis and prevention of mother-to-child HIV-transmission programs, and use of routine laboratory data about antibiotic resistance to guide updates of essential drug lists., Conclusion: In-service leadership training was feasible, with ensured protected time for fellows to generate evidence-based solutions to challenges within their work environment. With structured mentorship, collaborative activities at academic institutions and local health care programs equipped health care providers with leadership skills.

Background: Syphilis infection during pregnancy leads to avoidable morbidity and mortality and remains a significant problem in sub-Saharan Africa. Despite global initiatives to increase the proportion of pregnant women screened, implementation has been slow. We sought to investigate the feasibility of adding syphilis screening within an integrated antenatal HIV clinic., Methods: Pregnant women attending the HIV antenatal clinic were sequentially enrolled and consenting participants answered a questionnaire on sexual behavior and previous pregnancies, provided sociodemographic data, and were tested using rapid plasmin reagin (RPR). If positive, participants were treated with benzathine penicillin. All were given a partner notification slip and were followed up after delivery to determine birth outcomes., Results: 584 of 606 (95.7%) women approached and consented to test for syphilis. 570 women were enrolled (median age 29 (IQR 25-32) with a median (IQR) CD4 of 372 (257-569) cells/μL). Of the 5.1% (29/570) with a positive RPR, all were asymptomatic, were successfully contacted, and treated with benzathine penicillin without adverse reactions. Overall, 61 (12.1%) of the participants had an adverse birth outcome. In the bivariate analysis, only age was significantly different between those with and without a positive RPR (RR = 1.15, 95% CI 1.065-1.248; p < 0.001). Partners of only 10 (34.5%) participants returned for treatment., Conclusions: Structural interventions such as opt-out testing for syphilis within integrated HIV-antenatal care clinics are feasible and capitalize on the excellent care programs that have already been established for HIV care. Novel approaches are required for partner notification.

Background: We describe older (>50 years) HIV-infected adults after antiretroviral therapy (ART) initiation, evaluating immunological recovery by age category, considering individual trajectories based on the pretreatment CD4. We also describe mortality on ART and its risk factors by age category including the contribution of poor immunological recovery at a large urban clinic in Kampala, Uganda., Methods: We performed a cohort analysis of adult (>18 years) HIV-infected patients who initiated ART between January 1, 2004 and January 3, 2012. Immunological response was evaluated using mixed-effects linear regression. We described mortality using Kaplan-Meier survival methods analyzing for risk factors of mortality using multivariate Weibull survival regression stratified by age category., Results: Among 9806 individuals who initiated ART, mean age was 37 years (SD: 8.8), average follow-up 5.7 years (SD: 1.7), and median baseline CD4 was 115 cells per cubic millimeter (interquartile range: 42-184). Adults younger than 50 years had on average a higher CD4 increase of 45 cells per cubic millimeter (95% confidence interval: 17 to 72; P = 0.001) compared with counterparts aged 60 years and older. Mortality was highest among older adults compared with younger counterparts. Only CD4 count <100 cells per cubic millimeter after 1 year on ART and a CD4 count less than baseline were associated with a statistically significant higher rate of death among older adults., Conclusions: Older adults had a slower immunological response, which was associated with mortality, but this mortality was not typically associated with opportunistic infections. Future steps would require more evaluation of possible causes of death among these older individuals if survival on ART is to be further improved.

Background: The purpose of this study was to assess patients' experience and satisfaction with pharmacy-only refill program (PRP) and to compare those who were removed from the PRP with those still in the program., Methods: A sample of 446 patients was selected from 1503 patients on antiretroviral therapy that had been enrolled in the PRP for at least 24 months. The study used interviewer-administered questionnaires to assess patients' experience and satisfaction with PRP., Results: Of the 446 patients, 133 (29.8%) were removed from the PRP. By multivariate analysis, it was found that wanting to see a clinician before their scheduled clinic visit, Christian religion, and not understanding why they were enrolled in PRP were associated with having been removed from the PRP. Patients felt that the greatest benefit from the program was the time that they saved to do other activities. Patients preferred to collect their medication every 3 months instead of every month., Conclusion: All patients interviewed scored the program high, and all recommended that the PRP should continue. Stable patients prefer to see clinicians less frequently and visit clinic less often.

Background: Brucellosis is one of the most common zoonotic infections globally. Lack of knowledge about brucellosis may affect the health-seeking behavior of patients, thus leading to sustained transmission in these communities. Our study assessed knowledge and perceptions of brucellosis among pastoral communities adjacent to Lake Mburo National Park (LMNP), Kiruhura District, Uganda., Methods: A community cross-sectional questionnaire survey involving 371 randomly selected household heads from three sub-counties neighboring LMNP were interviewed between June and August 2012. Data collected included communities' knowledge on causes, symptoms, transmission, treatment, prevention and risk factors of brucellosis. Multivariable logistic regression analysis was performed to explore strength of association between overall knowledge of brucellosis and various individual factors using odds ratios and 95% confidence intervals., Results: Only 70 (19%) knew the symptoms of brucellosis in animals, and three quarters (279, 75.5%) mentioned joint and muscle pain as a common symptom in humans. Almost all participants (370, 99.3%) had ever heard about brucellosis, majority (311, 84.7%) believed it affects all sexes and two thirds (67.7%) of the respondents believed close proximity to wildlife contributes to the presence of the disease. Almost all (352, 95.4%) knew that brucellosis in humans could be treatable using modern drugs. The main routes of infection in humans such as consumption of unpasteurized dairy products were known by 97% (360/371); eating of half-cooked meat by 91.4% and eating contaminated pasture in animals by 97.4%. There was moderate overall knowledge of brucellosis 197 (53.1%). Factors associated with higher overall knowledge were being agro-pastoralists (aOR: 2.08, CI: 1.17-3.71) compared to pure pastoralists while those who reported that the disease was a health problem (aOR: 0.18, CI: 0.06-0.56) compared to those who said it was not were less likely to be knowledgeable., Conclusions: There was moderate overall knowledge of human and animal brucellosis among the participants. Majority of the participants believed that close proximity to wildlife contributes to the presence of the disease in the area. There is a need for collaboration between the public health, veterinary and wildlife sectors to provide health education on brucellosis for better management of the disease in the communities.

Introduction: The existing diagnostic algorithms for sputum smear-negative tuberculosis (TB) are complicated, time-consuming, and often difficult to implement. The decision to initiate TB treatment in resource-limited countries is often largely based on clinical predictors. We sought to determine the clinical predictors and accuracy of empiric TB treatment initiation in HIV-infected sputum smear-negative TB suspects using sputum culture as a reference standard., Setting: Out-patient HIV-TB integrated urban clinic in Kampala, Uganda., Methods: HIV-infected TB suspects were screened using sputum smear microscopy, and mycobacterial sputum liquid and solid cultures were performed. Smear results were made available to the clinician who made a clinical decision on empiric TB treatment initiation for sputum smear-negative patients. Clinic records were reviewed for patients whose sputum smears were negative to collect data on socio-demographics, TB symptomatology, chest X-ray findings, CD4 cell counts and TB treatment initiation., Results: Of 253 smear-negative TB suspects, 56% (142/253) were females, median age 38 IQR (31-44) years, with a median CD4 cell count of 291 IQR (150-482) cells/mm(3). Of the 85 (33.6%) smear-negative patients empirically initiated on TB treatment, 35.3% (n = 30) were sputum culture positive compared to only 18 (10.7%) of the 168 untreated patients (p<0.001). Abnormal chest X-ray [aOR 10.18, 95% CI (3.14-33.00), p<0.001] and advanced HIV clinical stage [aOR 3.92, 95% CI (1.20-12.85), p = 0.024] were significantly associated with empiric TB treatment initiation. The sensitivity and specificity of empiric TB treatment initiation in the diagnosis of TB in HIV-infected patients after negative smear microscopy was 62.5% and 73.7% respectively., Conclusion: In resource-limited settings, clinically advanced HIV and abnormal chest X-ray significantly predict a clinical decision to empirically initiate TB treatment in smear-negative HIV-infected patients. Empiric TB treatment initiation correlates poorly with TB cultures. Affordable, accurate and rapid point-of-care diagnostics are needed in resource-limited settings to more accurately determine which HIV-infected TB suspects have smear-negative TB.

Background: Light emitting diode (LED) fluorescence microscopy (FM) is an affordable, technology targeted for use in resource-limited settings and recommended for widespread roll-out by the World Health Organization (WHO). We sought to compare the operational performance of three LED FM methods compared to light microscopy in a cohort of HIV-positive tuberculosis (TB) suspects at an urban clinic in a high TB burden country., Methods: Two spot specimens collected from TB suspects were included in the study. Smears were stained using auramine O method and read after blinding by three LED-based FM methods by trained laboratory technicians in the Infectious Diseases Institutelaboratory. Leftover portions of the refrigerated sputum specimens were transported to the FIND Tuberculosis Research Laboratory for Ziehl Neelsen (ZN) smear preparation and reading by experienced technologist as well as liquid and solid culture., Results: 174 of 627 (27.8%) specimens collected yielded one or more positive mycobacterial cultures. 94.3% (164/174) were M. tuberculosis complex. LED FM was between 7.3-11.0% more sensitive compared to ZN microscopy. Of the 592 specimens examined by all microscopy methods, there was no significant difference in sensitivity between the three LED FM methods. The specificity of the LED FM methods was between 6.1% and 7.7% lower than ZN microscopy (P<0.001), although exclusion of the single poor reader resulted in over 98% specificity for all FM methods., Conclusions: Laboratory technicians in routine settings can be trained to use FM which is more sensitive than ZN microscopy. Despite rigorous proficiency testing, there were operator-dependent accuracy issues which highlight the critical need for intensive quality assurance procedures during LED FM implementation. The low sensitivity of FM for HIV-positive individuals particularly those with low CD4 T cell counts, will limit the number of additional patients found by LED FM in countries with high rates of HIV co-infection.

Introduction. While the introduction of highly active antiretroviral therapy decreased HIV-related morbidity and mortality rates in the sub-Saharan Africa, a subsequent increase in metabolic abnormalities has been observed. We sought to determine the prevalence of HIV-associated metabolic abnormalities among patients on first-line antiretroviral therapy (ART) in an ART clinic in Kampala, Uganda. Methods. Four hundred forty-two consecutive patients on first-line ART for at least 12 months were screened for eligibility in a cross-sectional study, and 423 were enrolled. Pre-ART patient characteristics were abstracted from medical charts, examinations included anthropometric measurement and physical assessment for lipodystrophy. Results. The prevalence of hyperglycemia and dyslipidemia was 16.3% (69/423) and 81.5% (345/423), respectively. Prevalence of dyslipidemia between stavudine- and zidovudine-based regimens (91% versus 72%; P < 0.001). Being on stavudine (aOR 4.79, 95%, 2.45-9.38) and peak body weight (aOR 1.44, 95% CI 1.05-1.97) were independent risk factors for dylipidemia. Stavudine (aOR 0.50, 95% CI  0.27-0.93) use was associated with lower risk for hyperglycemia while, and older age (aOR 1.31, 95% CI 1.11-1.56) and having a family history of DM (aOR 2.18, 95% CI 1.10-4.34) were independent risk factors for hyperglycemia. Conclusions. HIV-associated metabolic complications were prevalent among patients on thymidine analogue-containing ART regimens. Screening for lipid and glucose abnormalities should be considered in ART patients because of cardiovascular risks.

Objective: To model the cost-effectiveness in Uganda of combination antiretroviral therapy (ART) to prevent mother-to-child transmission of human immunodeficiency virus (HIV)., Methods: The cost-effectiveness of ART was evaluated on the assumption that ART reduces the risk of an HIV-positive pregnant woman transmitting HIV to her baby from 40% (when the woman is left untreated) to 25.8%, 17.4% and 3.8%, respectively, when the woman is given: (i) single-dose nevirapine (at an estimated total drug cost of 0.06 United States dollars [US$]); (ii) dual therapy with zidovudine and lamivudine for 7 weeks (at a total drug cost of US$ 15.63); or (iii) ART for 18 months (at a total annual cost of US$ 469.77). Lifetime ART (US$ 6883), recommended for pregnant women with < 350 CD4+ T lymphocytes per mm(3), was assumed to give the same reduction in transmission risk in each subsequent pregnancy., Findings: Compared with single-dose nevirapine, dual therapy and no therapy, 18 months of ART averted 5.21, 3.22 and 8.58 disability-adjusted life years (DALYs), respectively, at a cost of US$ 46, US$ 99 and US$ 34 per DALY averted. The corresponding figures for lifetime ART are, respectively, 19.20, 11.87 and 31.60 DALYs averted, at a cost of US$ 205, US$ 354 and US$ 172 per DALY averted., Conclusion: In Uganda, ART appears highly cost-effective for the prevention of mother-to-child HIV transmission, even if continued over the patients' lifetimes. Given the additional public health benefits of ART, efforts to ensure that all HIV-positive pregnant women have access to lifelong ART should be intensified.

Our objective was to evaluate outcomes in patients with sustained viral suppression compared to those with episodes of viremia. Methods. In a prospective cohort of patients started on ART in Uganda and followed for 48 months, patients were categorized according to viral load (VL): (1) sustained-suppression: (VL ≤1,000 copies/mL) (2) VL 1,001-10,000, or (3) VL >10,000. Results. Fifty-Three (11.2%) and 84 (17.8%) patients had a first episode of intermediate and high viremia, respectively. Patients with sustained suppression had better CD4+ T cell count increases over time compared to viremic patients (P < .001). The majority of patients with viremia achieved viral suppression when the measurement was repeated. Only 39.6% of patients with intermediate and 19.1% with high viremia eventually needed to be switched to second line (P = .008). Conclusions. The use of at least one repeat measurement rather than a single VL measurement could avert from 60% to 80% of unnecessary switches.