Living at higher altitude may have a protective effect on IHD and a harmful effect on COPD. At least in part due to these two opposing effects, living at higher altitude appears to have no net effect on life expectancy.

Purpose of review: Electronic health records (EHRs) contain valuable data for identifying health outcomes, but these data also present numerous challenges when creating computable phenotyping algorithms. Machine learning methods could help with some of these challenges. In this review, we discuss four common scenarios that researchers may find helpful for thinking critically about when and for what tasks machine learning may be used to identify health outcomes from EHR data. Recent findings: We first consider the conditions in which machine learning may be especially useful with respect to two dimensions of a health outcome: 1) the characteristics of its diagnostic criteria, and 2) the format in which its diagnostic data are usually stored within EHR systems. In the first dimension, we propose that for health outcomes with diagnostic criteria involving many clinical factors, vague definitions, or subjective interpretations, machine learning may be useful for modeling the complex diagnostic decision-making process from a vector of clinical inputs to identify individuals with the health outcome. In the second dimension, we propose that for health outcomes where diagnostic information is largely stored in unstructured formats such as free text or images, machine learning may be useful for extracting and structuring this information as part of a natural language processing system or an image recognition task. We then consider these two dimensions jointly to define four common scenarios of health outcomes. For each scenario, we discuss the potential uses for machine learning – first assuming accurate and complete EHR data and then relaxing these assumptions to accommodate the limitations of real-world EHR systems. We illustrate these four scenarios using concrete examples and describe how recent studies have used machine learning to identify these health outcomes from EHR data. Summary: Machine learning has great potential to improve the accuracy and efficiency of health outcome identification from EHR systems, especially under certain conditions. To promote the use of machine learning in EHR-based phenotyping tasks, future work should prioritize efforts to increase the transportability of machine learning algorithms for use in multi-site settings.

Objective To demonstrate the feasibility of integrated screening for cryptococcal antigenemia and tuberculosis (TB) prior to antiretroviral therapy (ART) initiation and to assess disease specific and all-cause mortality in the first 6 months of follow-up. Methods We enrolled a cohort of HIV-infected, ART-naïve adults with CD4 counts ≤ 250 cells/µL in rural Uganda who were followed for 6 months after ART initiation. All subjects underwent screening for TB; those with CD4 ≤ 100 cells/µL also had cryptococcal antigen (CrAg) screening. For those who screened positive, standard treatment for TB or preemptive treatment for cryptococcal infection was initiated, followed by ART two weeks later. Results Of 540 participants enrolled, pre-ART screening detected 10.6% (57/540) with prevalent TB and 6.8% (12/177 with CD4 count ≤ 100 cells/µL) with positive serum CrAg. After ART initiation, 13 (2.4%) patients were diagnosed with TB and one patient developed cryptococcal meningitis. Overall 7.2% of participants died (incidence rate 15.6 per 100 person years at risk). Death rates were significantly higher among subjects with TB and cryptococcal antigenemia compared to subjects without these diagnoses. In multivariate analysis, significant risk factors for mortality were male sex, baseline anemia of hemoglobin ≤ 10 mg/dL, wasting defined as body mass index ≤ 15.5 kg/m2, and opportunistic infections (TB, positive serum CrAg). Conclusion Pre-ART screening for opportunistic infections detects many prevalent cases of TB and cryptococcal infection. However, severely immunosuppressed and symptomatic HIV patients continue to experience high mortality after ART initiation.