Effect of machine-learning triage on acute kidney injury among pregnant women
Published
March 13, 2025
Paper ID
PAP-000302
Status
Published
DOI
10.65712/tmj.test000560
Advance Services
Abstract
Background: Practice around machine-learning triage varies widely between centres. Methods: Analyses were adjusted for age, sex and baseline severity. Results: Serious adverse events were rare and similar between groups. 1177 participants were included (median age 69 years). Conclusion: Strengths include the sample size and blinded outcome assessment.
Introduction
The cost of acute kidney injury to health systems is substantial. acute kidney injury remains a leading cause of avoidable harm among pregnant women. Guidelines make only a weak recommendation on machine-learning triage.
Earlier studies of machine-learning triage were small or single-centre.
Guidelines make only a weak recommendation on machine-learning triage. Earlier studies of machine-learning triage were small or single-centre.
Guidelines make only a weak recommendation on machine-learning triage. Earlier studies of machine-learning triage were small or single-centre. The cost of acute kidney injury to health systems is substantial.
Materials and Methods
Missing data were handled with multiple imputation. Analyses were adjusted for age, sex and baseline severity.
Missing data were handled with multiple imputation. Eligible participants were pregnant women. Analyses were adjusted for age, sex and baseline severity.
Results
2186 participants were included (median age 59 years). acute kidney injury occurred in 18% of the intervention group versus 26% of controls. acute kidney injury occurred in 7% of the intervention group versus 26% of controls. Serious adverse events were rare and similar between groups.
acute kidney injury occurred in 16% of the intervention group versus 13% of controls. The adjusted odds ratio was 0.68 (95% CI 0.20 to 0.79). The adjusted odds ratio was 0.61 (95% CI 0.46 to 0.93). 3514 participants were included (median age 45 years).
| Group | n | Events | Rate % |
|---|---|---|---|
| Centre A | 748 | 78 | 26.0 |
| Age < 65 | 542 | 7 | 36.0 |
| Age ≥ 65 | 600 | 22 | 24.6 |
| Age ≥ 65 | 432 | 33 | 14.5 |
| Centre A | 804 | 1 | 33.5 |
| Male | 644 | 55 | 10.7 |
| Age < 65 | 842 | 115 | 12.4 |
Values are n (%) unless stated otherwise.

Shown for pregnant women; error bars are 95% confidence intervals.
The adjusted odds ratio was 0.50 (95% CI 0.22 to 0.90). The adjusted odds ratio was 0.74 (95% CI 0.27 to 0.65).
The adjusted odds ratio was 0.63 (95% CI 0.41 to 0.96). The adjusted odds ratio was 0.71 (95% CI 0.39 to 0.82). Effects were consistent across prespecified subgroups. acute kidney injury occurred in 20% of the intervention group versus 31% of controls.
Discussion
These findings suggest that machine-learning triage is associated with better acute kidney injury in pregnant women. Future work should test machine-learning triage in a pragmatic multicentre trial. Implementation will depend on staffing, cost and local protocols.
Future work should test machine-learning triage in a pragmatic multicentre trial. Residual confounding cannot be excluded. Implementation will depend on staffing, cost and local protocols.
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 acute kidney injury in pregnant women. Strengths include the sample size and blinded outcome assessment.
Residual confounding cannot be excluded. Strengths include the sample size and blinded outcome assessment.
Conclusions
Machine-learning triage is a reasonable option for pregnant women to improve acute kidney injury.
References
1. Kowalski R, Herrera G, et al. Machine-learning triage and post-operative delirium. Pediatrics. 2011;260(11):745-993.
2. Rossi Y, Mueller Q, et al. Tranexamic acid and health-related quality of life. Pediatrics. 2019;179(1):585-1390.
3. Moreau B, Tanaka X, et al. Delayed cord clamping and length of stay. Pediatrics. 2008;75(5):607-1635.
4. Dubois G, Mueller Q, et al. Frailty screening and post-operative delirium. BMC Med. 2012;151(10):855-1581.
5. Nakamura G, Volkov Y, et al. A sepsis alert system and medication adherence. Eur Heart J. 2009;261(4):639-1136.
6. Kaur O, Nakamura P, et al. Low-dose aspirin and ventilator-free days. BMC Med. 2016;102(6):272-1064.
7. Farouk S, Rossi W, et al. Early mobilisation and glycaemic control. Ann Intern Med. 2025;262(11):148-1583.
Appendices
Supplementary tables are available on request.