Improving risk prediction in heart failure using machine learning
(2020) In European Journal of Heart Failure 22(1). p.139-147- Abstract
Background: Predicting mortality is important in patients with heart failure (HF). However, current strategies for predicting risk are only modestly successful, likely because they are derived from statistical analysis methods that fail to capture prognostic information in large data sets containing multi-dimensional interactions.
Methods and results: We used a machine learning algorithm to capture correlations between patient characteristics and mortality. A model was built by training a boosted decision tree algorithm to relate a subset of the patient data with a very high or very low mortality risk in a cohort of 5822 hospitalized and ambulatory patients with HF. From this model we derived a risk score that accurately... (More)
Background: Predicting mortality is important in patients with heart failure (HF). However, current strategies for predicting risk are only modestly successful, likely because they are derived from statistical analysis methods that fail to capture prognostic information in large data sets containing multi-dimensional interactions.
Methods and results: We used a machine learning algorithm to capture correlations between patient characteristics and mortality. A model was built by training a boosted decision tree algorithm to relate a subset of the patient data with a very high or very low mortality risk in a cohort of 5822 hospitalized and ambulatory patients with HF. From this model we derived a risk score that accurately discriminated between low and high-risk of death by identifying eight variables (diastolic blood pressure, creatinine, blood urea nitrogen, haemoglobin, white blood cell count, platelets, albumin, and red blood cell distribution width). This risk score had an area under the curve (AUC) of 0.88 and was predictive across the full spectrum of risk. External validation in two separate HF populations gave AUCs of 0.84 and 0.81, which were superior to those obtained with two available risk scores in these same populations.
Conclusions: Using machine learning and readily available variables, we generated and validated a mortality risk score in patients with HF that was more accurate than other risk scores to which it was compared. These results support the use of this machine learning approach for the evaluation of patients with HF and in other settings where predicting risk has been challenging.
(Less)
- author
- organization
- publishing date
- 2020-01
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- Heart failure, Machine learning, Outcomes
- in
- European Journal of Heart Failure
- volume
- 22
- issue
- 1
- pages
- 9 pages
- publisher
- Elsevier
- external identifiers
-
- scopus:85075021624
- pmid:31721391
- ISSN
- 1388-9842
- DOI
- 10.1002/ejhf.1628
- language
- English
- LU publication?
- yes
- id
- 3b82e946-9976-42b5-9c3b-12d5f9b4a56b
- date added to LUP
- 2019-12-06 09:53:30
- date last changed
- 2024-06-26 06:47:43
@article{3b82e946-9976-42b5-9c3b-12d5f9b4a56b, abstract = {{<p>Background: Predicting mortality is important in patients with heart failure (HF). However, current strategies for predicting risk are only modestly successful, likely because they are derived from statistical analysis methods that fail to capture prognostic information in large data sets containing multi-dimensional interactions. </p><p>Methods and results: We used a machine learning algorithm to capture correlations between patient characteristics and mortality. A model was built by training a boosted decision tree algorithm to relate a subset of the patient data with a very high or very low mortality risk in a cohort of 5822 hospitalized and ambulatory patients with HF. From this model we derived a risk score that accurately discriminated between low and high-risk of death by identifying eight variables (diastolic blood pressure, creatinine, blood urea nitrogen, haemoglobin, white blood cell count, platelets, albumin, and red blood cell distribution width). This risk score had an area under the curve (AUC) of 0.88 and was predictive across the full spectrum of risk. External validation in two separate HF populations gave AUCs of 0.84 and 0.81, which were superior to those obtained with two available risk scores in these same populations. </p><p>Conclusions: Using machine learning and readily available variables, we generated and validated a mortality risk score in patients with HF that was more accurate than other risk scores to which it was compared. These results support the use of this machine learning approach for the evaluation of patients with HF and in other settings where predicting risk has been challenging.</p>}}, author = {{Adler, Eric D. and Voors, Adriaan A. and Klein, Liviu and Macheret, Fima and Braun, Oscar O. and Urey, Marcus A. and Zhu, Wenhong and Sama, Iziah and Tadel, Matevz and Campagnari, Claudio and Greenberg, Barry and Yagil, Avi}}, issn = {{1388-9842}}, keywords = {{Heart failure; Machine learning; Outcomes}}, language = {{eng}}, number = {{1}}, pages = {{139--147}}, publisher = {{Elsevier}}, series = {{European Journal of Heart Failure}}, title = {{Improving risk prediction in heart failure using machine learning}}, url = {{http://dx.doi.org/10.1002/ejhf.1628}}, doi = {{10.1002/ejhf.1628}}, volume = {{22}}, year = {{2020}}, }