Obstetric Comorbidity Index (OB-CMI) Scoring¶

Problem¶

The U.S. has the highest maternal mortality ratio when compared to 11 other developed nations in the world- despite being one of the highest income nations in the world [1]. The maternal mortality rate in the U.S. has been on the rise since 1987 and literature suggests that the numbers have continued to rise since the start of the COVID-19 pandemic[2].

Solution¶

One of the issues that keeps maternal morbidity/mortality high in the U.S. is inadequate prenatal care. Prenatal care includes discussions with a healthcare provider about how to have a healthy pregnancy. Illnesses such as diabetes, hypertension, and heart disease are extremely common among Americans and these comorbidities can impact the course of a person's pregnancy. Prenatal visits may be able to address the implications of comorbidities, thus a comorbity based screening tool (OB-CMI) was created to predict maternal morbidity at the time of delivery[3].

In order for physicians to use the OB-CMI tool more efficiently to assess pregnant individuals earlier in their pregnancy, a machine learning model can be developed using pregnant patients' OB-CMI scores and obstetric outcomes to predict the obstetric outcomes of a patient who may be at risk of a maternal morbidity.

Data¶

Unfortunately, I am unable to provide the dataset at the moment because I am waiting to see if I can get access to the dataset I was using during my co-op at Brigham and Women's Hospital, but here is a snippet of what it might look like: Screen-Shot-2023-02-27-at-11-57-21-PM

How will the data be used?¶

Since the OB-CMI score is a continuous variable from 0-24, we could use a regression model to predict an OB-CMI score based on other patient variables. The variables in the data set would include:

  • Obstetric comorbidity index (OB-CMI) score
  • Maternal age
  • Gestational age at delivery
  • Number of previous pregnancies
  • History of pre-existing medical conditions (e.g., diabetes, hypertension)
  • History of pregnancy-specific conditions (e.g., preeclampsia, placenta previa)
  • Obstetric outcomes (e.g., preterm delivery, cesarean delivery)

References¶

  1. Morrissey S, Blumenthal D, Osborn R, Curfman GD, Malina D. International Health Care Systems. New England Journal of Medicine 2015;372(1):75-76 doi: 10.1056/NEJMe1415036

  2. Easter SR, Bateman BT, Sweeney VH, et al. A comorbidity-based screening tool to predict severe maternal morbidity at the time of delivery. American Journal of Obstetrics and Gynecology 2019;221(3):271.e1-71.e10 doi: https://doi.org/10.1016/j.ajog.2019.06.025[published

  3. March of Dimes. (2011). Your first prenatal care checkup. Retrieved Feb 27, 2023, from http://www.marchofdimes.org/pregnancy/your-first-prenatal-care-checkup.aspx

  4. DL H. Maternal mortality rates in the United States, 2020. 2022 doi: https://dx.doi.org/10.15620/cdc:113967