Answering important questions in HIV with mismeasured clinical endpoints

About This Grant

ABSTRACT Routinely collected clinical data are often used for high impact research in HIV and elsewhere. But, as life expectancy for people with HIV improves, many endpoints that are not captured well in routine clinical data are increasingly important. Cardiometabolic outcomes and cancers, leading drivers of morbidity and mortality among an aging global cohort of people with HIV, may be particularly prone to mismeasurement in routine clinical data. The goal of this application is to answer previously intractable questions in HIV clinical research that involve the timing of acute, chronic, and terminal endpoints, which are often measured with error in clinical HIV data. This project will construct new statistical estimators to account for mismeasurement in these important clinical outcomes and evaluate the performance of these estimators using a series of simulation experiments. The new methods will be developed and applied in US-based clinical cohorts, specifically those in North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD), and clinical cohorts operating in resource limited settings, including those in the International Epidemiology Database to Evaluate AIDS (IeDEA). The aims are applications of the new estimators to address important current problems in HIV. Specific aim #1 estimates the effect of switching to an integrase-inhibitor containing regimen on risk of myocardial infarction among virally suppressed participants in the NA-ACCORD, where the occurrence and timing of events using routine health records collected from HIV clinics may be recorded with error. Specific aim #2 estimates the effect of initiating an integrase inhibitor containing antiretroviral therapy regimen on cancer incidence in the NA-ACCORD, where identification of cancers in the HIV clinical record may be incomplete. Specific aim #3 describes the risk of mortality after entry into HIV care in the IeDEA East Africa cohort, where vital status data is not routinely available for people out of care. The assembled team includes expertise in epidemiology, statistics, and HIV medicine. Products will include publicly available software to implement the new approaches and publications leveraging the new methods to answer important questions in HIV. The impact of this work is to improve the body of evidence available to inform decision-making in HIV care and treatment, thereby improving quality of care for people with HIV.