Background Hypertension is a major risk factor for cardiovascular disease, which is a common cause of death in Zambia. Data on hypertension prevalence in Zambia are scarce and limited to specific geographic areas and/or populations. We measured hypertension prevalence among persons living with HIV (PLHIV) in Zambia using a national electronic health record (EHR) system. Methods We did a cross-sectional study of hypertension prevalence among PLHIV aged >=18 years in Zambia during 2021. Data were extracted from the SmartCare EHR, which covers ~90% of PLHIV on treatment in Zambia. PLHIV with >=2 recorded blood pressure (BP) readings in 2021 were included. Hypertension was defined as >=2 elevated BP readings (i.e., systolic BP of >=140 mmHg or diastolic BP of >=90 mmHg) during 2021 and/or on anti-hypertensive medication recorded in their EHR in the past five years. Multivariable logistic regression was used to assess associations between hypertension and independent variables. Results Among 750,098 PLHIV aged >=18 years with >=2 visits in SmartCare during 2021, 101,363 (13.5%) had >=2 blood pressure readings recorded in their EHR. Among these PLHIV, 14.7% (95% confidence interval [CI]: 14.5-14.9) had hypertension during 2021. Only 8.9% of PLHIV with hypertension had an antihypertensive medication recorded in their EHR. The odds of hypertension were greater in older age groups compared to PLHIV aged 18-29 years (adjusted odds ratio [aOR] for 30-44 years: 2.6 [95% CI: 2.4-2.9]; aOR for 45-49 years: 6.4 [95% CI: 5.8-7.0]; aOR for >=60 years: 14.5 [95% CI: 13.1-16.1]), urban areas (aOR: 1.9 [95% CI: 1.8-2.1]), and persons prescribed ART for >=6-month at a time (aOR: 1.1 [95% CI: 1.0-1.2]). Discussion Hypertension was common among a cohort of PLHIV in Zambia, with few having documentation of being on antihypertensive treatment. Most PLHIV were excluded from the analysis because of missing BP measurements in their EHR. Strengthening integrated management of non-communicable diseases in ART clinics might help to diagnose and treat hypertension in Zambia. Data completeness needs to be improved to routinely capture cardiovascular disease risk factors, including blood pressure readings consistently for PHLIV in their EHRs. Copyright The copyright holder for this preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Hypertension is a major risk factor for cardiovascular disease, which is a common cause of death in Zambia. Data on hypertension prevalence in Zambia are scarce and limited to specific geographic areas and/or populations. We measured hypertension prevalence among persons living with HIV (PLHIV) in Zambia using a national electronic health record (EHR) system. We did a cross-sectional study of hypertension prevalence among PLHIV aged ≥18 years during 2021. Data were extracted from the SmartCare EHR, which covers ~90% of PLHIV on treatment in Zambia. PLHIV with ≥2 clinical visits in 2021 were included. Hypertension was defined as ≥2 elevated blood pressure readings (systolic ≥140 mmHg/diastolic ≥90 mmHg) during 2021 and/or on anti-hypertensive medication recorded in their EHR ≤5 years. Logistic regression was used to assess for associations between hypertension and demographic characteristics. Among 750,098 PLHIV aged ≥18 years with ≥2 visits during 2021, 101,363 (13.5%) had ≥2 recorded blood pressure readings. Among these PLHIV, 14.7% (95% confidence interval [CI]: 14.5-14.9) had hypertension. Only 8.9% of PLHIV with hypertension had an anti-hypertensive medication recorded in their EHR. The odds of hypertension were greater in older age groups compared to PLHIV aged 18-29 years (adjusted odds ratio [aOR] for 30-44 years: 2.6 [95% CI: 2.4-2.9]; aOR for 45-49 years: 6.4 [95% CI: 5.8-7.0]; aOR for ≥60 years: 14.5 [95% CI: 13.1-16.1]), urban areas (aOR: 1.9 [95% CI: 1.8-2.1]), and on ART for ≥6-month at a time (aOR: 1.1 [95% CI: 1.0-1.2]). Hypertension was common among PLHIV in Zambia, with few having documentation of treatment. Most PLHIV were excluded from the analysis because of missing BP measurements. Strengthening integrated management of non-communicable diseases in HIV clinics might help to diagnose and treat hypertension in Zambia. Addressing missing data of routine clinical data (like blood pressure) could improve non-communicable diseases surveillance in Zambia.

HIV pre-exposure prophylaxis (PrEP) is being scaled-up in Zambia, but PrEP continuation data are limited by paper-based registers and aggregate reports. Utilization of Zambia's electronic health record (EHR) system, SmartCare, may address this gap. We analyzed individuals aged ≥ 15 years who initiated PrEP between October 2020 and September 2021 in four provinces in Zambia in SmartCare versus aggregate reports. We measured PrEP continuation using Kaplan-Meier survival analysis and Cox proportional hazards models. SmartCare captured 29% (16,791/58,010) of new PrEP clients; 49% of clients continued at one month, and 89% discontinued PrEP by February 2022. Women were less likely than men to discontinue PrEP (adjusted hazard ratio [aHR]: 0.89, 95% CI 0.86-0.92, z = - 6.99, p < 0.001), and PrEP clients aged ≥ 50 years were less likely to discontinue PrEP compared to clients 15-19 years (aHR: 0.53, 95% CI 0.48-0.58, z = - 13.04, p < 0.001). Zambia's EHR is a valuable resource for measuring individual-level PrEP continuation over time and can be used to inform HIV prevention programs.

BACKGROUND: Measuring clinically relevant opioid-related problems in health care systems is challenging due to the lack of standard definitions and coding practices. Well-defined, opioid-related health problems (ORHPs) would improve prevalence estimates and evaluation of clinical interventions, crisis response, and prevention activities. We sought to estimate prevalence of opioid use disorder (OUD), opioid misuse, and opioid poisoning among inpatients at a large, safety net, health care institution. METHODS: Our study included events documented in the electronic health records (EHRs) among hospitalized patients at Denver Health Medical Center during January 1, 2017 to December 31, 2018. Multiple EHR markers (ie, opioid-related diagnostic codes, clinical assessment, laboratory results, and free-text documentation) were used to develop diagnosis-based and extended definitions for OUD, opioid misuse, and opioid poisoning. We used these definitions to estimate number of hospitalized patients with these conditions. RESULTS: During a 2-year study period, 715 unique patients were identified solely using opioid-related diagnostic codes; OUD codes accounted for the largest proportion (499/715, 69.8%). Extended definitions identified an additional 973 unique patients (~136% increase), which includes 155/973 (15.9%) who were identified by a clinical assessment marker, 1/973 (0.1%) by a laboratory test marker, and 817/973 (84.0%) by a clinical documentation marker. CONCLUSIONS: Solely using diagnostic codes to estimate prevalence of clinically relevant ORHPs missed most patients with ORHPs. More inclusive estimates were generated using additional EHR markers. Improved methods to estimate ORHPs among a health care system's patients would more fully estimate organizational and economic burden to more efficiently allocate resources and ensure capacity to provide clinical services.

BACKGROUND: Timely, precise, and localized surveillance of nonfatal events is needed to improve response and prevention of opioid-related problems in an evolving opioid crisis in the United States. Records of naloxone administration found in prehospital emergency medical services (EMS) data have helped estimate opioid overdose incidence, including nonhospital, field-treated cases. However, as naloxone is often used by EMS personnel in unconsciousness of unknown cause, attributing naloxone administration to opioid misuse and heroin use (OM) may misclassify events. Better methods are needed to identify OM. OBJECTIVE: This study aimed to develop and test a natural language processing method that would improve identification of potential OM from paramedic documentation. METHODS: First, we searched Denver Health paramedic trip reports from August 2017 to April 2018 for keywords naloxone, heroin, and both combined, and we reviewed narratives of identified reports to determine whether they constituted true cases of OM. Then, we used this human classification as reference standard and trained 4 machine learning models (random forest, k-nearest neighbors, support vector machines, and L1-regularized logistic regression). We selected the algorithm that produced the highest area under the receiver operating curve (AUC) for model assessment. Finally, we compared positive predictive value (PPV) of the highest performing machine learning algorithm with PPV of searches of keywords naloxone, heroin, and combination of both in the binary classification of OM in unseen September 2018 data. RESULTS: In total, 54,359 trip reports were filed from August 2017 to April 2018. Approximately 1.09% (594/54,359) indicated naloxone administration. Among trip reports with reviewer agreement regarding OM in the narrative, 57.6% (292/516) were considered to include information revealing OM. Approximately 1.63% (884/54,359) of all trip reports mentioned heroin in the narrative. Among trip reports with reviewer agreement, 95.5% (784/821) were considered to include information revealing OM. Combined results accounted for 2.39% (1298/54,359) of trip reports. Among trip reports with reviewer agreement, 77.79% (907/1166) were considered to include information consistent with OM. The reference standard used to train and test machine learning models included details of 1166 trip reports. L1-regularized logistic regression was the highest performing algorithm (AUC=0.94; 95% CI 0.91-0.97) in identifying OM. Tested on 5983 unseen reports from September 2018, the keyword naloxone inaccurately identified and underestimated probable OM trip report cases (63 cases; PPV=0.68). The keyword heroin yielded more cases with improved performance (129 cases; PPV=0.99). Combined keyword and L1-regularized logistic regression classifier further improved performance (146 cases; PPV=0.99). CONCLUSIONS: A machine learning application enhanced the effectiveness of finding OM among documented paramedic field responses. This approach to refining OM surveillance may lead to improved first-responder and public health responses toward prevention of overdoses and other opioid-related problems in US communities.

BACKGROUND: Assessment of people affected by opioid-related problems and those receiving care is challenging due to lack of common definitions and scattered information. We sought to fill this gap by demonstrating a method to describe a continuum of opioid addiction care in a large, public safety-net institution. METHODS: Using 2017 clinical and administrative data from Denver Health (DH), we created operational definitions for opioid use disorder (OUD), opioid misuse (OM), and opioid poisoning (OP). Six stages along a continuum of patient engagement in opioid addiction care were developed, and operational definitions assigned patients to stages for a specific time point of analysis. National data was used to estimate the Denver population affected by OUD, OM and OP. RESULTS: In 2017, an estimated 6688 people aged >/=12 years were affected by OUD, OM, or OP in Denver; 48.4% (3238/6688) were medically diagnosed in DH. Of those, 32.5% (1051/3238) were in the medication assisted treatment stage, and, of those, 59.8% (629/1051) in the adhered to treatment stage. Among that latter group, 78.4% (493/629) adhered at least 90 days and 52.3% (329/629) for more than one year. Among patients who received medication assisted treatment, less than one third (31.3%, 329/1051) were adherent for more than one year. CONCLUSIONS: A health-system level view of the continuum of opioid addiction care identified improvement opportunities to better monitor accuracy of diagnosis, treatment capacity, and effectiveness of patient engagement. Applied longitudinally at local, state and national levels, the model could better synergize responses to the opioid crisis.

BACKGROUND: In many Latin American countries, official influenza reports are neither timely nor complete, and surveillance of influenza-like illness (ILI) remains thin in consistency and precision. Public participation with mobile technology may offer new ways of identifying nonmedically attended cases and reduce reporting delays, but no published studies to date have assessed the viability of ILI surveillance with mobile tools in Latin America. We implemented and assessed an ILI-tailored mobile health (mHealth) participatory reporting system. OBJECTIVE: The objectives of this study were to evaluate the quality and characteristics of electronically collected data, the user acceptability of the symptom reporting platform, and the costs of running the system and of identifying ILI cases, and to use the collected data to characterize cases of reported ILI. METHODS: We recruited the heads of 189 households comprising 584 persons during randomly selected home visits in Guatemala. From August 2016 to March 2017, participants used text messages or an app to report symptoms of ILI at home, the ages of the ILI cases, if medical attention was sought, and if medicines were bought in pharmacies. We sent weekly reminders to participants and compensated those who sent reports with phone credit. We assessed the simplicity, flexibility, acceptability, stability, timeliness, and data quality of the system. RESULTS: Nearly half of the participants (47.1%, 89/189) sent one or more reports. We received 468 reports, 83.5% (391/468) via text message and 16.4% (77/468) via app. Nine-tenths of the reports (93.6%, 438/468) were received within 48 hours of the transmission of reminders. Over a quarter of the reports (26.5%, 124/468) indicated that at least someone at home had ILI symptoms. We identified 202 ILI cases and collected age information from almost three-fifths (58.4%, 118/202): 20 were aged between 0 and 5 years, 95 were aged between 6 and 64 years, and three were aged 65 years or older. Medications were purchased from pharmacies, without medical consultation, in 33.1% (41/124) of reported cases. Medical attention was sought in 27.4% (34/124) of reported cases. The cost of identifying an ILI case was US $6.00. We found a positive correlation (Pearson correlation coefficient=.8) between reported ILI and official surveillance data for noninfluenza viruses from weeks 41 (2016) to 13 (2017). CONCLUSIONS: Our system has the potential to serve as a practical complement to respiratory virus surveillance in Guatemala. Its strongest attributes are simplicity, flexibility, and timeliness. The biggest challenge was low enrollment caused by people's fear of victimization and lack of phone credit. Authorities in Central America could test similar methods to improve the timeliness, and extend the breadth, of disease surveillance. It may allow them to rapidly detect localized or unusual circulation of acute respiratory illness and trigger appropriate public health actions.