Back

Causal Machine Learning for Comparative Effectiveness of GLP-1 RA versus SGLT2i in Heart Failure Using Real-World EHR Data

Han, G. Y.; Kalogeropoulos, A. P.; Butzin-Dozier, Z.; Wong, R.; Wang, F.

2026-04-07 cardiovascular medicine
10.64898/2026.04.06.26350259 medRxiv
Show abstract

Clinicians lack precision medicine tools to estimate individualized treatment effects for patients with heart failure (HF). Causal machine learning leveraging electronic health records can estimate both average and individualized treatment effects, enabling estimation of treatment heterogeneity. Using Stony Brook University Hospital data, we compared the effectiveness of glucagon-like peptide-1 receptor agonists (GLP-1 RA) versus sodium-glucose cotransporter 2 inhibitors (SGLT2i) in patients with HF. Under a doubly robust framework, we found a stable population-average effect: GLP-1 RA was associated with a lower risk than SGLT2i for a 1-year composite outcome of all-cause mortality or HF-related hospitalization. Heterogeneity analyses provided limited evidence for individualized treatment selection, although subgroup tests identified loop diuretic use, body mass index, and estimated glomerular filtration rate as potential effect modifiers. While these models hold promise for translating observational data into actionable precision care, careful assessment of causal assumptions and rigorous validation are essential before clinical implementation.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

1
Circulation
66 papers in training set
Top 0.3%
10.2%
2
Nature Medicine
117 papers in training set
Top 0.1%
9.9%
3
Nature Communications
4913 papers in training set
Top 19%
9.9%
4
Nature Genetics
240 papers in training set
Top 0.7%
9.9%
5
npj Digital Medicine
97 papers in training set
Top 0.8%
6.2%
6
European Heart Journal - Digital Health
15 papers in training set
Top 0.1%
6.2%
50% of probability mass above
7
Journal of the American Medical Informatics Association
61 papers in training set
Top 0.6%
4.2%
8
eLife
5422 papers in training set
Top 23%
3.9%
9
Circulation: Genomic and Precision Medicine
42 papers in training set
Top 0.5%
3.2%
10
Scientific Reports
3102 papers in training set
Top 48%
2.3%
11
Journal of the American College of Cardiology
12 papers in training set
Top 0.3%
2.0%
12
Journal of Clinical Investigation
164 papers in training set
Top 3%
1.7%
13
Genome Medicine
154 papers in training set
Top 5%
1.6%
14
The Lancet Digital Health
25 papers in training set
Top 0.6%
1.3%
15
Diabetologia
36 papers in training set
Top 0.7%
1.3%
16
PLOS ONE
4510 papers in training set
Top 61%
1.2%
17
Journal of the American Heart Association
119 papers in training set
Top 3%
1.2%
18
Nature
575 papers in training set
Top 14%
0.9%
19
Nature Machine Intelligence
61 papers in training set
Top 3%
0.8%
20
Nature Biomedical Engineering
42 papers in training set
Top 2%
0.8%
21
Epidemiology
26 papers in training set
Top 0.5%
0.8%
22
Cell Genomics
162 papers in training set
Top 6%
0.8%
23
Circulation: Heart Failure
14 papers in training set
Top 0.5%
0.8%
24
European Heart Journal
16 papers in training set
Top 0.9%
0.7%
25
PLOS Digital Health
91 papers in training set
Top 3%
0.7%
26
JACC: Clinical Electrophysiology
11 papers in training set
Top 0.4%
0.7%
27
The American Journal of Human Genetics
206 papers in training set
Top 4%
0.7%
28
PLOS Medicine
98 papers in training set
Top 5%
0.6%
29
BMC Medical Research Methodology
43 papers in training set
Top 2%
0.6%
30
Science Translational Medicine
111 papers in training set
Top 7%
0.6%