The Impact of Positive Fluid Balance on Sepsis Subtypes: A Causal Inference Study

Table 3

Refutation results from DoWhy library for sepsis subtypes: comparison of estimated effects with new effects and associated values.

	Estimated effect	New effect	value
Sepsis all
Using a subset of data	0.042	0.042	0.920
Adding a random common cause	0.042	0.042	2.0
Pulmonary sepsis
Using a subset of data	0.046	0.047	0.92
Adding a random common cause	0.046	0.046	2.0
Urinary sepsis
Using a subset of data	−0.014	−0.014	1.0
Adding a random common cause	−0.014	−0.014	2.0

values >0.05 indicate no statistically significant difference between original and refutation ATE estimates, supporting the validity of the original estimates.