Browsing by Author "Engelhard, Matthew M"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Incorporating informatively collected laboratory data from EHR in clinical prediction models.
    (BMC medical informatics and decision making, 2024-07) Sun, Minghui; Engelhard, Matthew M; Bedoya, Armando D; Goldstein, Benjamin A

    Background

    Electronic Health Records (EHR) are widely used to develop clinical prediction models (CPMs). However, one of the challenges is that there is often a degree of informative missing data. For example, laboratory measures are typically taken when a clinician is concerned that there is a need. When data are the so-called Not Missing at Random (NMAR), analytic strategies based on other missingness mechanisms are inappropriate. In this work, we seek to compare the impact of different strategies for handling missing data on CPMs performance.

    Methods

    We considered a predictive model for rapid inpatient deterioration as an exemplar implementation. This model incorporated twelve laboratory measures with varying levels of missingness. Five labs had missingness rate levels around 50%, and the other seven had missingness levels around 90%. We included them based on the belief that their missingness status can be highly informational for the prediction. In our study, we explicitly compared the various missing data strategies: mean imputation, normal-value imputation, conditional imputation, categorical encoding, and missingness embeddings. Some of these were also combined with the last observation carried forward (LOCF). We implemented logistic LASSO regression, multilayer perceptron (MLP), and long short-term memory (LSTM) models as the downstream classifiers. We compared the AUROC of testing data and used bootstrapping to construct 95% confidence intervals.

    Results

    We had 105,198 inpatient encounters, with 4.7% having experienced the deterioration outcome of interest. LSTM models generally outperformed other cross-sectional models, where embedding approaches and categorical encoding yielded the best results. For the cross-sectional models, normal-value imputation with LOCF generated the best results.

    Conclusion

    Strategies that accounted for the possibility of NMAR missing data yielded better model performance than those did not. The embedding method had an advantage as it did not require prior clinical knowledge. Using LOCF could enhance the performance of cross-sectional models but have countereffects in LSTM models.
  • Thumbnail Image
    ItemOpen Access
    Validation of the Sleep Regularity Index in Older Adults and Associations with Cardiometabolic Risk.
    (Scientific reports, 2018-09-21) Lunsford-Avery, Jessica R; Engelhard, Matthew M; Navar, Ann Marie; Kollins, Scott H
    Sleep disturbances, including insufficient sleep duration and circadian misalignment, confer risk for cardiometabolic disease. Less is known about the association between the regularity of sleep/wake schedules and cardiometabolic risk. This study evaluated the external validity of a new metric, the Sleep Regularity Index (SRI), among older adults (n = 1978; mean age 68.7 ± 9.2), as well as relationships between the SRI and cardiometabolic risk using data from the Multi-Ethnic Study of Atherosclerosis (MESA). Results indicated that sleep irregularity was associated with delayed sleep timing, increased daytime sleep and sleepiness, and reduced light exposure, but was independent of sleep duration. Greater sleep irregularity was also correlated with 10-year risk of cardiovascular disease and greater obesity, hypertension, fasting glucose, hemoglobin A1C, and diabetes status. Finally, greater sleep irregularity was associated with increased perceived stress and depression, psychiatric factors integrally tied to cardiometabolic disease. These results suggest that the SRI is a useful measure of sleep regularity in older adults. Additionally, sleep irregularity may represent a target for early identification and prevention of cardiometabolic disease. Future studies may clarify the causal direction of these effects, mechanisms underlying links between sleep irregularity and cardiometabolic risk, and the utility of sleep interventions in reducing cardiometabolic risk.