Bayesian machine learning algorithms can improve discrimination of risk stratification in pulmonary arterial hypertension (PAH), according to study results published in the European Respiratory Journal.
Researchers sought to determine the utility of Bayesian networks based machine learning through a Tree Augmented Naive Bayes model (titled PHORA) by predicting 1-year survival in patients with PAH included in the United States Registry to Evaluate Early and Long-Term PAH Disease Management (REVEAL) using the same variables and cut-points as REVEAL 2.0, an existing state-of-the-art risk stratification tool. PHORA models were validated internally (within the REVEAL registry) and externally with 2 other registries.
PHORA had an area under the curve (AUC) of 0.80 for predicting 1-year survival, which was an improvement over REVEAL 2.0 (AUC=0.76) and REVEAL 1.0 (AUC=0.71), and 1-year survival rates were greater for patients with lower risk scores and poorer for those with higher risk scores (P <.001), with excellent separation between low-, intermediate-, and high-risk groups in all 3 registries.
In addition, PHORA had a specificity of 0.76, sensitivity of 0.79, a negative predictive value of 0.30, and a positive predictive value of 0.97 for 1-year survival.
“Our [Bayesian network] derived risk prediction model, PHORA, demonstrated an improvement in discrimination over existing models” the study authors concluded. “Hence machine learning based risk modeling can provide PAH clinicians with a greater level of confidence for making medical decisions in this complex, progressive disease.”
Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.
Kanwar MK, Gomberg-Maitland M, Hoeper M, et al. Risk stratification in pulmonary arterial hypertension using Bayesian analysis [published online May 4, 2020]. Eur Respir J. doi:10.1183/13993003.00008-2020