Center for Biomedical Informatics Works

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 4 of 4
  • Item
    Informatics Interventions for Maternal Morbidity: A Scoping Review
    (National Library of Medicine, 2023-06-23) Inderstrodt, Jill; Stumpff, Julia C.; Smollen, Rebecca; Sridhar, Shreya; El-Azab, Sarah A.; Ojo, Opeyemi; Haggstrom, David A.
    Individuals of childbearing age in the U.S. currently enter pregnancy less healthy than previous generations, putting them at risk for maternal morbidities such as preeclampsia, gestational diabetes mellitus (GDM), and postpartum mental health conditions. These conditions leave mothers at risk for long-term health complications that, when left unscreened and unmonitored, can be deadly. One approach to ensuring long-term health for mothers is designing informatics interventions that: (a) prevent maternal morbidities, (b) treat perinatal conditions, and (c) allow for continuity of treatment. This scoping review examines the extent, range, and nature of informatics interventions that have been tested on maternal morbidities that can have long-term health effects on mothers. It uses MEDLINE, EMBASE, and Cochrane Library to chart demographic, population, and intervention data regarding informatics and maternal morbidity. Studies (n=79) were extracted for analysis that satisfied the following conditions: (a) tested a medical or clinical informatics intervention; (b) tested on adults with a uterus or doctors who treat people with a uterus; and (c) tested on the following conditions: preeclampsia, GDM, preterm birth, severe maternal morbidity as defined by the CDC, and perinatal mental health conditions. Of the 79 studies extracted, 38% (n=30) tested technologies for GDM, 38% (n=30) tested technologies for postpartum depression, and 15.2% (n=12) tested technologies for preeclampsia. In terms of technologies, 35.4% (n = 28) tested a smartphone or tablet app, 29.1% (n=23) tested a telehealth intervention, and 15.2% (n=12) tested remote monitoring technologies (blood pressure, blood glucose). Most (86.1%; n=68) of the technologies were tested for patient physical or mental health outcomes. This scoping review reveals that most tested informatics interventions are those aimed at three conditions (GDM, preeclampsia, mental health) and that there may be opportunities to treat other common causes of maternal mortality (i.e. postpartum hemorrhage) using proven technologies such as mobile applications.
  • Item
    Identifying Biases in Clinical Decision Models Designed to Predict Need of Wraparound Services
    (AMIA Informatics summit 2021 Conference Proceedings, 2021-03) Kasthurirathne, Suranga N.; Vest, Joshua R.; Grannis, Shaun J.
    Investigation of systemic biases in AI models for the clinical domain have been limited. We re-created a series of models predicting need of wraparound services, and inspected them for biases across age, gender and race using the AI Fairness 360 framework. AI models reported performance metrics which were comparable to original efforts. Investigation of biases using the AI Fairness framework found low likelihood that patient age, gender and sex are introducing bias into our algorithms.
  • Item
    Generative Adversarial Networks for Creating Synthetic Free-Text Medical Data: A Proposal for Collaborative Research and Re-use of Machine Learning Models
    (AMIA Informatics summit 2021 Conference Proceedings., 2021-03) Kasthurirathne, Suranga N.; Dexter, Gregory; Grannis, Shaun J.
    Restrictions in sharing Patient Health Identifiers (PHI) limit cross-organizational re-use of free-text medical data. We leverage Generative Adversarial Networks (GAN) to produce synthetic unstructured free-text medical data with low re-identification risk, and assess the suitability of these datasets to replicate machine learning models. We trained GAN models using unstructured free-text laboratory messages pertaining to salmonella, and identified the most accurate models for creating synthetic datasets that reflect the informational characteristics of the original dataset. Natural Language Generation metrics comparing the real and synthetic datasets demonstrated high similarity. Decision models generated using these datasets reported high performance metrics. There was no statistically significant difference in performance measures reported by models trained using real and synthetic datasets. Our results inform the use of GAN models to generate synthetic unstructured free-text data with limited re-identification risk, and use of this data to enable collaborative research and re-use of machine learning models.