Background: The coronavirus disease 2019 (COVID-19) pandemic necessitated the replacement of in-person physician consultations with telemedicine. During the pandemic, Medicaid covered the cost of telemedicine visits. Objectives: The aim was to measure the adoption of telemedicine during the pandemic. We focus on key patient subgroups including those with chronic conditions, those living in urban versus rural areas, and different age groups. Methods: This study examined the universe of claims made by Florida Medicaid beneficiaries (n=2.4 million) between January 2019 and July 2020. Outpatient visits were identified as in-person or telemedicine. Telemedicine visits were classified into audio-visual or audio-only visits. Results: We find that telemedicine offsets much of the decline in in-person outpatient visits among Florida’s Medicaid enrollees, however, uptake differs by enrollee type. High utilizers of care and beneficiaries with chronic conditions were significantly more likely to use telemedicine, while enrollees living in rural areas and health professional shortage areas were moderately less likely to use telemedicine. Elderly Medicaid recipients (dual-eligibles) used audio-only telemedicine visits at higher rates than other age groups, and the demand for these consultations is more persistent. Conclusions: Telemedicine offset the decline in health care utilization among Florida’s Medicaid-enrolled population during the novel coronavirus pandemic, with particularly high uptake among those with prior histories of high utilization. Audio-only visits are a potentially important method of delivery for the oldest Medicaid beneficiaries.

Models of protein evolution currently come in two flavors: generalist and specialist. Generalist models (e.g. PAM, JTT, WAG) adopt a one-size-fits-all approach, where a single model is estimated from a number of different protein alignments. Specialist models (e.g. mtREV, rtREV, HIVbetween) can be estimated when a large quantity of data are available for a single organism or gene, and are intended for use on that organism or gene only. Unsurprisingly, specialist models outperform generalist models, but in most instances there simply are not enough data available to estimate them. We propose a method for estimating alignment-specific models of protein evolution in which the complexity of the model is adapted to suit the richness of the data. Our method uses non-negative matrix factorization (NNMF) to learn a set of basis matrices from a general dataset containing a large number of alignments of different proteins, thus capturing the dimensions of important variation. It then learns a set of weights that are specific to the organism or gene of interest and for which only a smaller dataset is available. Thus the alignment-specific model is obtained as a weighted sum of the basis matrices. Having been constrained to vary along only as many dimensions as the data justify, the model has far fewer parameters than would be required to estimate a specialist model. We show that our NNMF procedure produces models that outperform existing methods on all but one of 50 test alignments. The basis matrices we obtain confirm the expectation that amino acid properties tend to be conserved, and allow us to quantify, on specific alignments, how the strength of conservation varies across different properties. We also apply our new models to phylogeny inference and show that the resulting phylogenies are different from, and have improved likelihood over, those inferred under standard models.

A central issue in the literature on health insurance is how its impact on recipients varies across individuals. Causal estimates of this impact are obtained by comparing a randomly or quasi-randomly assigned treatment and control group, with one group receiving differential access to health insurance, and then examining heterogeneity in the estimated treatment effect. Identifying how the effect of access varies by subgroup is difficult since membership of that subgroup might itself be affected by better access to medical services. For example, better access to testing improves the rate at which SARS-COV2 infections are detected. If we naively compared the death rate from these infections among insured individuals to that among uninsured individuals, we will be overestimating the effect of access to insurance. This will be because uninsured individuals will have fewer detected cases of SARS-COV2, artificially shrinking the denominator when dividing the number of deaths by the number of cases.In this paper, I propose methods for bounding or point-identifying these subgroup-specific treatment effects, depending on the data available to the researcher. Bounds can be obtained via imposing monotonicity assumptions in the spirit of Manski and

We investigate the consequences of discreteness in the assignment variable in regressiondiscontinuity designs for cases where the outcome variable is itself discrete. We nd that constructing condence intervals that have the correct level of coverage in these cases is sensitive to the assumed distribution of unobserved heterogeneity. Since local linear estimators are improperly centered, a smaller variance for unobserved heterogeneity in discrete outcomes actually requires larger condence intervals, since standard condence intervals become narrower around a biased estimator, leading to a higher-than-nominal false positive rate. We provide a method for mapping structural assumptions regarding the distribution and variance of unobserved heterogeneity to the construction of honest condence intervals that have the correct level of coverage. An application to retirement behavior reveals that the spike in retirement at age 62 in the United States can be reconciled with a wider range of values for the variance of unobserved heterogeneity (due to reservation wages or oers) than the spike at age 65.