Technical Report Anesthesiology Healthcare Technology Trauma
Rethinking 30-day readmission in rural primary-care patients: the role of machine-learning triage

Rethinking 30-day readmission in rural primary-care patients: the role of machine-learning triage

Authors

Mateo Lindqvist

Mateo Lindqvist

Mayo Clinic, Chad

Corresponding

Published

June 17, 2026

Paper ID

PAP-000199

Status

Published

DOI

10.65712/tmj.test000457

Rethinking 30-day readmission in rural primary-care patients: the role of machine-learning triage

Advance Services

Create Content

Video • Presentation • Poster

From $97.00 per format

Abstract

Background: Guidelines make only a weak recommendation on machine-learning triage. Methods: The primary outcome was 30-day readmission, assessed blind to allocation. Results: Serious adverse events were rare and similar between groups. Effects were consistent across prespecified subgroups. Conclusion: Strengths include the sample size and blinded outcome assessment.

Introduction

Practice around machine-learning triage varies widely between centres. Earlier studies of machine-learning triage were small or single-centre.

Earlier studies of machine-learning triage were small or single-centre. The cost of 30-day readmission to health systems is substantial.

Earlier studies of machine-learning triage were small or single-centre. The cost of 30-day readmission to health systems is substantial.

Discussion

Implementation will depend on staffing, cost and local protocols. Residual confounding cannot be excluded.

Implementation will depend on staffing, cost and local protocols. Future work should test machine-learning triage in a pragmatic multicentre trial. These findings suggest that machine-learning triage is associated with better 30-day readmission in rural primary-care patients.

Residual confounding cannot be excluded. Future work should test machine-learning triage in a pragmatic multicentre trial. Implementation will depend on staffing, cost and local protocols.

Implementation will depend on staffing, cost and local protocols. Residual confounding cannot be excluded. These findings suggest that machine-learning triage is associated with better 30-day readmission in rural primary-care patients.

Conclusions

Machine-learning triage is a reasonable option for rural primary-care patients to improve 30-day readmission.

References

1. Kim K, Osei Z, et al. Early enteral nutrition and ventilator-free days. Eur Heart J. 2013;178(5):290-1620.

2. Nasser F, Tanaka J, et al. Vitamin D supplementation and health-related quality of life. Pediatrics. 2009;345(11):721-1480.

3. Herrera Z, Costa R, et al. Pharmacist-led medication review and medication adherence. BMJ. 2018;255(8):294-1151.

4. Nasser V, Herrera S, et al. Nurse-led discharge planning and in-hospital mortality. Lancet. 2009;173(11):652-1287.

5. Dubois S, Rahman K, et al. Prone positioning and acute kidney injury. BMC Med. 2016;11(11):722-1748.

6. Moreau Y, Ahmed H, et al. Point-of-care ultrasound and falls. BMC Med. 2012;141(1):636-963.

7. Rahman Z, Costa T, et al. Early enteral nutrition and 30-day readmission. JAMA. 2012;64(11):262-1612.

Appendices

Supplementary tables are available on request.

Materials and Methods

Missing data were handled with multiple imputation. We conducted a multicentre observational study across 5 centres between 2017 and 2022. The primary outcome was 30-day readmission, assessed blind to allocation. We conducted a multicentre observational study across 12 centres between 2016 and 2022.

Missing data were handled with multiple imputation. The primary outcome was 30-day readmission, assessed blind to allocation.

Analyses were adjusted for age, sex and baseline severity. Missing data were handled with multiple imputation. We conducted a multicentre observational study across 6 centres between 2020 and 2023.

Results

The adjusted odds ratio was 0.61 (95% CI 0.34 to 0.83). 532 participants were included (median age 31 years). 30-day readmission occurred in 9% of the intervention group versus 30% of controls. The adjusted odds ratio was 0.64 (95% CI 0.31 to 0.92).

3341 participants were included (median age 74 years). 262 participants were included (median age 51 years). The adjusted odds ratio was 0.70 (95% CI 0.41 to 0.91). 30-day readmission occurred in 22% of the intervention group versus 29% of controls.

GroupnEventsRate %OR95% CI
Control5157536.50.540.33–1.34
Control1927012.40.480.35–1.25
Intervention3455040.50.890.22–1.35
Age < 65774140.40.950.53–1.32
Centre B75310840.30.990.33–1.26
TABLE 1: Subgroup analysis
Values are n (%) unless stated otherwise.
Figure 1
FIGURE 1: Trend in 30-day readmission
Shown for rural primary-care patients; error bars are 95% confidence intervals.
Figure 2
FIGURE 2: Change in 30-day readmission
Shown for rural primary-care patients; error bars are 95% confidence intervals.

The adjusted odds ratio was 0.42 (95% CI 0.46 to 0.67). Effects were consistent across prespecified subgroups. 30-day readmission occurred in 20% of the intervention group versus 21% of controls.