Individualized treatment rule characterization via a value function surrogate.

Precision medicine is a promising framework for generating evidence to improve health and health care. Yet, a gap persists between the ever-growing number of statistical precision medicine strategies for evidence generation and implementation in real-world clinical settings, and the strategies for closing this gap will likely be context-dependent. In this paper, we consider the specific context of partial compliance to wound management among patients with peripheral artery disease. Using a Gaussian process surrogate for the value function, we show the feasibility of using Bayesian optimization to learn optimal individualized treatment rules. Further, we expand beyond the common precision medicine task of learning an optimal individualized treatment rule to the characterization of classes of individualized treatment rules and show how those findings can be translated into clinical contexts.

Use of machine learning to identify characteristics associated with severe hypoglycemia in older adults with type 1 diabetes: a post-hoc analysis of a case-control study.

INTRODUCTION: Severe hypoglycemia (SH) in older adults (OAs) with type 1 diabetes is associated with profound morbidity and mortality, yet its etiology can be complex and multifactorial. Enhanced tools to identify OAs who are at high risk for SH are needed. This study used machine learning to identify characteristics that distinguish those with and without recent SH, selecting from a range of demographic and clinical, behavioral and lifestyle, and neurocognitive characteristics, along with continuous glucose monitoring (CGM) measures. RESEARCH DESIGN AND METHODS: Data from a case-control study involving OAs recruited from the T1D Exchange Clinical Network were analyzed. The random forest machine learning algorithm was used to elucidate the characteristics associated with case versus control status and their relative importance. Models with successively rich characteristic sets were examined to systematically incorporate each domain of possible risk characteristics. RESULTS: Data from 191 OAs with type 1 diabetes (47.1% female, 92.1% non-Hispanic white) were analyzed. Across models, hypoglycemia unawareness was the top characteristic associated with SH history. For the model with the richest input data, the most important characteristics, in descending order, were hypoglycemia unawareness, hypoglycemia fear, coefficient of variation from CGM, % time blood glucose below 70 mg/dL, and trail making test B score. CONCLUSIONS: Machine learning may augment risk stratification for OAs by identifying key characteristics associated with SH. Prospective studies are needed to identify the predictive performance of these risk characteristics.

Assessing risk of bias in the meta-analysis of round 1 of the Health Care Innovation Awards.

BACKGROUND: Systematic reviews of observational studies can be affected by biases that lead to under- or over-estimates of true intervention effects. Several tools have been reported in the literature that attempt to characterize potential bias. Our objective in this study was to determine the extent to which study-specific bias may have influenced intervention impacts on total costs of care (TCOC) in round 1 of the Health Care Innovation Awards. METHODS: We reviewed 82 statistical evaluations of innovation impacts on Medicare TCOC. We developed five risk-of-bias measures and assessed their influence on TCOC impacts using meta-regression. RESULTS: The majority of evaluations used propensity score matching to create their comparison groups. One third of the non-randomized interventions were judged to have some risk of biased effects due largely to the way they recruited their treatment groups, and 35% had some degree of covariate imbalance remaining after propensity score adjustments. However, in the multivariable analysis of TCOC effects, none of the bias threats we examined (comparison group construction method, risk of bias, or degree of covariate imbalance) had a major impact on the magnitude of HCIA1 innovation effects. Evaluations using propensity score weighting produced larger but imprecise savings effects compared to propensity score matching. DISCUSSION: Our results suggest that it is unlikely that HCIA1 TCOC effect sizes were systematically affected by the types of bias we considered. Assessing the risk of bias based on specific study design features is likely to be more useful for identifying problematic characteristics than the subjective quality ratings used by existing risk tools.