Toward a multivariate prediction model of pharmacological treatment for women with gestational diabetes mellitus: algorithm development and validation
<p><strong>Background: </p></strong> Successful management of gestational diabetes mellitus (GDM) reduces the risk of morbidity in women and newborns. A woman’s BG readings and risk factors are used by clinical staff to make decisions regarding the initiation of pharmacologic...
| Main Authors: | , , , , , |
|---|---|
| Format: | Internet publication |
| Language: | English |
| Published: |
JMIR Publications
2020
|
| Summary: | <p><strong>Background:
</p></strong>
Successful management of gestational diabetes mellitus (GDM) reduces the risk of morbidity in women and newborns. A woman’s BG readings and risk factors are used by clinical staff to make decisions regarding the initiation of pharmacological treatment in women with GDM. Mobile-Health (mHealth) solutions allow the real-time follow-up of women with GDM and allow timely treatment and management. Machine learning offers the opportunity to quickly analyse large quantities of data to automatically flag women at risk of requiring pharmacological treatment.
<p><strong>
Objective:
</p></strong>
We sought to assess whether data collected through a mHealth system can be analysed to automatically evaluate the switch to pharmacological treatment from diet-based management of GDM.
<p><strong>
Methods:
</p></strong>
We collected data from 3,029 patients to design a machine-learning model that can identify when a woman with GDM needs to switch to medications (Insulin or Metformin) by analysing the data related to blood glucose and other risk factors.
<p><strong>
Results:
</p></strong>
Through the analysis of 411,785 blood glucose (BG) readings we have designed a machine learning model that can predict the timing of initiation of pharmacological treatment. After one hundred experimental repetitions we have obtained an average performance of 0.80 AUC and an algorithm that allows the flexibility of setting the operating point rather than relying on a static heuristic method, currently used in clinical practice.
<p><strong>
Conclusions:
</p></strong>
Using real-time data collected via a mHealth system may further improve the timeliness of intervention and potentially improve patient care. Further real-time clinical testing will enable validating our algorithm using real-world data. |
|---|