Search

BACKGROUND Digital data collection and the associated mobile health technologies have allowed for the recent exploration of machine learning (ML) as a tool for combatting the HIV epidemic. Machine learning has been found to be useful both in HIV risk prediction and as a decision support tool for guiding pre-exposure prophylaxis (PrEP) treatment. OBJECTIVE This paper reports data from two sequential studies evaluating the viability of using machine learning to predict the susceptibility of adults to HIV infection using responses from a digital survey deployed in a high burden, low-resource setting. METHODS 1036 and 593 participants were recruited across two trials. The first trial was a cross-sectional study in one location and the second trial was a cohort study across three trial sites. The data from the studies were merged, partitioned using standard techniques, and then used to train and evaluate multiple different machine learning models and select and evaluate a final model. Variable importance estimates were calculated using the PIMP and SHAP methodologies. RESULTS Characteristics associated with HIV were consistent across both studies. Overall, HIV positive patients had a higher median age (34 [IQR: 29-39] vs 26 [IQR 22-33], p CONCLUSIONS This study has highlighted the synergies present between mobile health and ML in HIV. It has been demonstrated that a viable ML model can be built using digital survey data from an low-middle income setting with potential utility in directing health resources. CLINICALTRIAL This study is registered on the South African National Clinical Trial Registry (www.sanctr.gov.za; DOH-27-042021-679).

ObjectivesCardiovascular disease (CVD) is a major problem globally. Truck drivers have an increased risk of CVD due to a sedentary lifestyle, irregular working hours and behavioural choices. We aimed to get insight into the contribution of night shift work to CVD risk in long-distance truck drivers in South Africa.DesignA cross-sectional study.SettingEnrolment took place at three South African truck stop locations in two provinces; Bloemfontein (Free State), Pomona Road (Gauteng) and Soweto (Gauteng).Participants607 males aged ≥18 years with full-time employment as a long-distance truck driver were included. The criteria for inclusion were willingness and being able to provide informed consent and to complete the study procedures.Primary and secondary outcome measuresInformation was collected on sociodemographics, occupational and health characteristics. Physical measurements, an ECG and carotid intima–media thickness (CIMT) measurements were taken. A night shift was defined as working at least 3 hours between 22:00 and 6:00 hours once a week. CVD risk was defined with the Framingham Risk Score (FRS), the Atherosclerotic Cardiovascular Disease (ASCVD) risk algorithm, left ventricular hypertrophy (LVH) and CIMT.ResultsIn total, 607 truck drivers were included of which 305 (50.2%) worked in day shifts only and 302 (49.8%) worked day and night shifts. There was a high prevalence of CVD risk factors in both groups as 33% were hypertensive, 28% obese and 37% had abnormal lipid levels. Working day and night shifts compared with working only day shifts did not result in differences in FRS, ASCVD risk or LVH. No difference was found in CIMT measurements, except for the maximum bulb thickness which was higher in day shift workers.ConclusionsCVD risk factors are considerably present in male truck drivers in South Africa. CVD risk does not differ between dayshift and day–night shift workers in this cross-sectional analysis.

Background: Prevalence of HIV, smoking, and pulmonary infections in South Africa are high.Objectives: We investigated the role of smoking and HIV status on lung function.Methods: This is a secondary analysis of a cross-sectional study conducted in South Africa. Data included demographics, pulmonary risk factors and a spirometry test to obtain the forced expiratory volume in one second (FEV1) and the ratio of FEV1/forced vital capacity (FVC). In the initial multivariable regression analysis, the effect of smoking on pulmonary function in HIV-positive adults was assessed. The analysis was repeated, assessing the influence of HIV status on lung function in both HIV-negative and HIV-positive smokers. The models were adjusted for age, sex, body mass index (BMI), time since HIV diagnosis, antiretroviral treatment (ART) use, occupational hazards, history of tuberculosis or pneumonia, indoor smoking and the presence of an indoor fireplace during childhood.Results: This study included 524 people living with HIV (PLWH, 66.7% female, mean age 40.9 years [s.d.; 9.4]) and 79 HIV-negative smokers (77.2% male, mean age 34.4 years [s.d.: 12.1]). Of the PLWH, 118 (22.5%) were past or current smokers and 406 (77.5%) were non-smokers. Smoking was not associated with changes in the FEV1 or FEV1/FVC ratio in multivariable regression analysis. In the second analysis, HIV status was also not associated with reduced pulmonary function following adjustment for confounders.Conclusion: Neither smoking nor being HIV-positive was associated with decreased pulmonary function in this relatively young population. These findings should be confirmed in a longitudinal study, including an older population.

BACKGROUND Mobile technology has helped to advance health programs, and studies have shown that an automated risk prediction model can successfully be used to identify patients who exhibit a high probable risk of contracting human immunodeficiency virus (HIV). A machine-guided tool is an algorithm that takes a set of subjective and objective answers from a simple questionnaire and computes an HIV risk assessment score. OBJECTIVE The primary objective of this study is to establish that machine learning can be used to develop machine-guided tools and give us a deeper statistical understanding of the correlation between certain behavioral patterns and HIV. METHODS In total, 200 HIV-negative adult individuals across three South African study sites each (two semirural and one urban) will be recruited. Study processes will include (1) completing a series of questions (demographic, sexual behavior and history, personal, lifestyle, and symptoms) on an application system, unaided (assistance will only be provided upon user request); (2) two HIV tests (one per study visit) being performed by a nurse/counselor according to South African national guidelines (to evaluate the prediction accuracy of the tool); and (3) communicating test results and completing a user experience survey questionnaire. The output metrics for this study will be computed by using the participants’ risk assessment scores as “predictions” and the test results as the “ground truth.” Analyses will be completed after visit 1 and then again after visit 2. All risk assessment scores will be used to calculate the reliability of the machine-guided tool. RESULTS Ethical approval was received from the University of Witwatersrand Human Research Ethics Committee (HREC; ethics reference no. 200312) on August 20, 2020. This study is ongoing. Data collection has commenced and is expected to be completed in the second half of 2021. We will report on the machine-guided tool’s performance and usability, together with user satisfaction and recommendations for improvement. CONCLUSIONS Machine-guided risk assessment tools can provide a cost-effective alternative to large-scale HIV screening and help in providing targeted counseling and testing to prevent the spread of HIV. CLINICALTRIAL South African National Clinical Trial Registry DOH-27-042021-679; https://sanctr.samrc.ac.za/TrialDisplay.aspx?TrialID=5545 INTERNATIONAL REGISTERED REPORT DERR1-10.2196/30304

The impacts of pandemics are recognized to go beyond infection, physical suffering, and socio-economic disruptions. Other consequences include psychological responses. Using a mental wellbeing lens, we analyzed COVID-19-related stressors in healthcare workers (HCWs) and community members who provided and regularly accessed health services in South Africa, respectively. From February to September 2021, during the second COVID-19 wave we conducted a qualitative study in one urban and one rural district. In-depth interviews and focus group discussions were used to collect data among 43 HCWs and 51 community members purposely and conveniently selected. Most participants experienced mental health challenges regarding multiple aspects of the COVID-19 pandemic and its resulting lockdown, with a few reporting positive adjustments to change. COVID-19 impacts on mental health were consistent among both HCWs and community members in urban and rural alike. Participants’ COVID-19-induced psychological responses included anxiety and fear of the unknown, perceived risk of infection, fear of hospitalization, and fear of dying. Physical effects of the pandemic on participants included COVID-19 infection and associated symptoms, possibilities of severe illness and discomfort of using personal protective equipment. These distresses were exacerbated by social repercussions related to concerns for family wellbeing and infection stigma. Lockdown regulations also intensified anxieties about financial insecurities and social isolation. At times when common coping mechanisms such as family support were inaccessible, cultural consequences related to lack of spiritual gatherings and limited funeral rites posed additional stress on participants. In preparation for future public health emergencies, recognition needs to be given to mental health support and treatment.

BACKGROUND South Africa has the largest HIV treatment program worldwide. Retention in care and medication adherence remain problematic necessitating innovative solutions for improving HIV care. The increasing availability and use of mobile technology can support positive clinical outcomes for persons living with HIV. iThemba Life is a mobile health app designed with input from South African health professionals and patients, promoting engagement with HIV care through access to medical results. OBJECTIVE This study aimed to test the feasibility and acceptability of receiving HIV viral load (VL) results through the app and compare the time to HIV VL result return for study participants before and after app use. METHODS Using convenience sampling, adults having routine VL phlebotomy were recruited from 2 Johannesburg health facilities. After signed consent, the app was downloaded on their Android smartphones, phlebotomy was performed, and the sample barcode was scanned through their phone to link the sample and app. Participants received a notification of the result availability and logged into the app to view results, their explanation and recommended action. RESULTS Overall, 750 people were screened to enroll 500 participants. Of 750, 113 (15.1%) failed eligibility screening. 21.5% (137/637) had smartphone technical limitations preventing enrollment. Results were released to 92.2% (461/500) of participants’ phones. App technical issues and laboratory operational issues limited the number of released results. Approximately 78.1% (360/461) results were viewed in the app. Median time from notification of availability to result viewed being 15.5 hours (0.6; range 0-150 days). Turnaround time from phlebotomy to the result being received was 6 (range 1-167) days for users versus 56 days (range 10-430 days; P1000 copies/mL). Turnaround time for unsuppressed results was 7 days for participants versus 37.5 days before app use (PP CONCLUSIONS Using iThemba Life for VL results was well-received despite limited smartphone access for some participants. App users received results 10 times sooner than before the app and 5 times sooner if their VL >1000 copies/mL. This increased notification speed led to participants wanting to continue using iThemba Life.

BACKGROUND Long-term retention of research participants in studies is challenging. In research in sub-Saharan Africa, phone calls are the most frequently used method to distantly engage with participants. OBJECTIVE We aimed to get insight into the effectiveness of phone calls to retain contact with participants and evaluated determinants of reachability. METHODS A cross-sectional study was performed using the databases of two randomized controlled trials investigating different kinds of antiretroviral therapy in HIV-positive patients. One trial finished in 2018 (study 1), and the other finished in 2015 (study 2). A random sample size of 200 participants per study was obtained. There were up to 3 phone numbers available per participant collected during the studies. Participants received a maximum of 3 phone calls on every available number on different days and at different times. Voicemails were left, and emails sent wherever possible. We documented how many calls were answered, who answered, as well as after how many attempts participants were reached. To further increase our understanding of reachability, we conducted a short questionnaire assessing factors contributing to reachability. The study was approved by the Research Ethics Committee of the University of Witwatersrand, Johannesburg, South Africa (reference number M1811107). RESULTS In our sample size of n=200 per study, study 1, with a median time of 11 months since the last visit at the research site, had a response rate of 70.5% (141/200) participants while study 2, with a median duration of 55 months since the last visit, had a response rate of 50.0% (100/200; PP=.003). The likelihood of reaching a participant decreased with time (odds ratio [OR] 0.73, 95% CI 0.63 to 0.84) for every year since the last face-to-face visit. Having more phone numbers per participant increased reachability (OR 2.32, 95% CI 1.24 to 4.36 for 2 phone numbers and OR 3.03, 95% CI 1.48 to 6.22 for 3 phone numbers compared with 1 number). A total of 141 of 241 reached participants responded to the questionnaire. Of the 93 participants who had changed phone numbers, 5% (50/93) had changed numbers because their phone was stolen. The most preferred method of being contacted was direct calling (128/141) with participants naming this method followed by WhatsApp (69/141). CONCLUSIONS Time since last visit and the number of phone numbers listed were the only determinants of reachability. Longer follow-up time is accompanied with a decrease in reachability by phone while more listed phone numbers increases the likelihood that someone can be reached. CLINICALTRIAL ClinicalTrials.gov NCT02671383; https://clinicaltrials.gov/ct2/show/NCT02671383 and ClinicalTrials.gov NCT02670772; https://clinicaltrials.gov/ct2/show/NCT02670772

BACKGROUND Over the past decade, mobile health has steadily increased in low-income and middle-income countries. However, few platforms have been able to sustainably scale up like the MomConnect program in South Africa. NurseConnect was created as a capacity building component of MomConnect, aimed at supporting nurses and midwives in maternal and child health. The National Department of Health has committed to expanding NurseConnect to all nurses across the country, and an evaluation of the current user experience was conducted to inform a successful scale up. OBJECTIVE This study aims to evaluate the perception and use of NurseConnect by nurses and midwives to produce feedback that can be used to optimize the user experience as the platform continues to scale up. METHODS We conducted focus group discussions and in-depth interviews with 110 nurses and midwives from 18 randomly selected health care facilities across South Africa. Questions focused on mobile phone use, access to medical information and their experience with NurseConnect registration, as well as the content and different platforms. RESULTS All participants had mobile phones and communication through calls and messaging was the main use in both personal and work settings. Of 110 participants, 108 (98.2%) had data-enabled phones, and the internet, Google, and apps (South African National Department of Health Guidelines, iTriage, Drugs.com) were commonly used, especially to find information in the work setting. Of 110 participants, 62 (56.4%) were registered NurseConnect users and liked the message content, especially listeriosis and motivational messages, which created behavioral change in some instances. The mobisite and helpdesk, however, were underutilized because of a lack of information surrounding these platforms. Some participants did not trust medical information from websites and had more confidence in apps, while others associated a “helpdesk” with a call-in service, not a messaging one. Many of the unregistered participants had not heard of NurseConnect, and some cited data and time constraints as barriers to both registration and uptake. CONCLUSIONS Mobile and smartphone penetration was very high, and participants often used their phone to find medical information. The NurseConnect messages were well-liked by all registered participants; however, the mobisite and helpdesk were underutilized owing to a lack of information and training around these platforms. Enhanced marketing and training initiatives that optimize existing social networks, as well as the provision of data and Wi-Fi, should be explored to ensure that registration improves, and that users are active across all platforms.