Using machine learning involving diagnoses and medications as a risk prediction tool for PASC in primary care, 2025, Lee et al

Using machine learning involving diagnoses and medications as a risk prediction tool for post-acute sequelae of COVID-19 (PASC) in primary care

Seika Lee, Marta A. Kisiel, Pia Lindberg, Åsa M. Wheelock, Anna Olofsson, Julia Eriksson, Judith Bruchfeld, Michael Runold, Lars Wahlström, Andrei Malinovschi, Christer Janson, Caroline Wachtler & Axel C. Carlsson

Abstract
Background
The aim of our study was to determine whether the application of machine learning could predict PASC by using diagnoses from primary care and prescribed medication 1 year prior to PASC diagnosis.

Methods
This population-based case–control study included subjects aged 18–65 years from Sweden. Stochastic gradient boosting was used to develop a predictive model using diagnoses received in primary care, hospitalization due to acute COVID- 19, and prescribed medication. The variables with normalized relative influence (NRI) ≥ 1% showed were considered predictive. Odds ratios of marginal effects (ORME) were calculated.

Results
The study included 47,568 PASC cases and controls. More females (n = 5113) than males (n = 2815) were diagnosed with PASC. Key predictive factors identified in both sexes included prior hospitalization due to acute COVID- 19 (NRI 16.1%, ORME 18.8 for females; NRI 41.7%, ORME 31.6 for males), malaise and fatigue (NRI 14.5%, ORME 4.6 for females; NRI 11.5%, ORME 7.9 for males), and post-viral and related fatigue syndromes (NRI 10.1%, ORME 21.1 for females; NRI 6.4%, ORME 28.4 for males).

Conclusions
Machine learning can predict PASC based on previous diagnoses and medications. Use of this AI method could support diagnostics of PASC in primary care and provide insight into PASC etiology.

Link (BMC Medicine)
https://doi.org/10.1186/s12916-025-04050-w

 

They excluded anyone that did not have any contact with the healthcare system. They acknowledge how this leads to selection bias and limits the generalisability of the findings.

On ME/CFS:

I think this might be influenced by the very broad definitions of PACS.

PACS doesn’t require any one symptom, so this is a bit misleading.

These are the usual studies with huge selection bias, so they are not suitable to estimate the prevalence. It’s a bit concerning that everyone keeps getting this wrong.

This seems to be pure speculation. I don’t know why they felt like they had to include it.

42 is more speculation by Komaroff and Lipkin: https://www.s4me.info/threads/me-cf...to-the-literature-komaroff-lipkin-2023.33588/