Improvement of Adequate Use of Warfarin for the Elderly Using Decision Tree-based Approaches

Liu, Kang Ernest; Lo, Chi-Fang; Hu, Ya-Han

doi:10.3414/me13-01-0027

Cited by 26 publications

(12 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our literature search was limited to journal articles published before May 31, 2014. From among the publications on WSD prediction, Six reported machine learning techniques were selected, namely ANN [ 26 ], regression tree (RT) [ 29 ], multivariate adaptive regression splines (MARS) [ 3 ], BRT, SVR and RFR [ 25 ]. In addition to the six machine learning based techniques mentioned above, two classical machine learning techniques, namely, lasso regression (LAR) and Bayesian additive regression trees (BART) and the most widely used MLR were included in our study for comparison.…”

Section: Methodsmentioning

confidence: 99%

Comparison of Nine Statistical Model Based Warfarin Pharmacogenetic Dosing Algorithms Using the Racially Diverse International Warfarin Pharmacogenetic Consortium Cohort Database

Liu

Zhang

et al. 2015

PLoS ONE

View full text Add to dashboard Cite

ObjectiveMultiple linear regression (MLR) and machine learning techniques in pharmacogenetic algorithm-based warfarin dosing have been reported. However, performances of these algorithms in racially diverse group have never been objectively evaluated and compared. In this literature-based study, we compared the performances of eight machine learning techniques with those of MLR in a large, racially-diverse cohort.MethodsMLR, artificial neural network (ANN), regression tree (RT), multivariate adaptive regression splines (MARS), boosted regression tree (BRT), support vector regression (SVR), random forest regression (RFR), lasso regression (LAR) and Bayesian additive regression trees (BART) were applied in warfarin dose algorithms in a cohort from the International Warfarin Pharmacogenetics Consortium database. Covariates obtained by stepwise regression from 80% of randomly selected patients were used to develop algorithms. To compare the performances of these algorithms, the mean percentage of patients whose predicted dose fell within 20% of the actual dose (mean percentage within 20%) and the mean absolute error (MAE) were calculated in the remaining 20% of patients. The performances of these techniques in different races, as well as the dose ranges of therapeutic warfarin were compared. Robust results were obtained after 100 rounds of resampling.ResultsBART, MARS and SVR were statistically indistinguishable and significantly out performed all the other approaches in the whole cohort (MAE: 8.84–8.96 mg/week, mean percentage within 20%: 45.88%–46.35%). In the White population, MARS and BART showed higher mean percentage within 20% and lower mean MAE than those of MLR (all p values < 0.05). In the Asian population, SVR, BART, MARS and LAR performed the same as MLR. MLR and LAR optimally performed among the Black population. When patients were grouped in terms of warfarin dose range, all machine learning techniques except ANN and LAR showed significantly higher mean percentage within 20%, and lower MAE (all p values < 0.05) than MLR in the low- and high- dose ranges.ConclusionOverall, machine learning-based techniques, BART, MARS and SVR performed superior than MLR in warfarin pharmacogenetic dosing. Differences of algorithms’ performances exist among the races. Moreover, machine learning-based algorithms tended to perform better in the low- and high- dose ranges than MLR.

show abstract

Section: Methodsmentioning

confidence: 99%

Comparison of Nine Statistical Model Based Warfarin Pharmacogenetic Dosing Algorithms Using the Racially Diverse International Warfarin Pharmacogenetic Consortium Cohort Database

Liu

Zhang

et al. 2015

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…An RF is an ensemble learning method that is developed by constructing multiple DTs (Jiang et al., ; Liu et al., ). In the training process, an RF applies a bagging technique to bootstrap instances and selects a random subset of features.…”

Section: Related Workmentioning

confidence: 99%

Predicting the prolonged length of stay of general surgery patients: a supervised learning approach

Chuang

2016

Int Trans Operational Res

View full text Add to dashboard Cite

Determining the likelihood of a prolonged length of stay (LOS) for surgery patients can improve medical resource management. This study was aimed at developing predictive models for determining whether patient LOS is within the standard LOS after surgery. This study analyzed the complete historical medical records and lab data of 896 clinical cases involving surgeries performed by general surgery physicians. The cases were divided into urgent operation (UO) and non-UO groups to develop a prolonged LOS prediction model using several supervised learning techniques. Several critical factors for the two groups were identified using the gain ratio technique. The results indicated that the random forest method yielded the most accurate and stable prediction model. Additionally, comorbidity, body temperature, blood sugar, and creatinine were the most influential variables for prolonged LOS in the UO group, whereas blood transfusion, blood pressure, comorbidity, and the number of ICU admissions were the most influential variables in the non-UO group. This study shows that supervised learning techniques are suitable for analyzing patient medical records in accurately predicting a prolonged LOS; thus, the clinical decision support system developed based on the prediction models may serve as reference tools for communicating with patients before surgery. The system may also assist physicians when making decisions regarding whether patients require more clinical care, thereby improving patient safety.

show abstract

“…And again the methodological roots came from informatics (e.g. [45,46]) as well as from biostatistics (e.g. [47,48]) and sometimes from both sides (e.g.…”

Section: Major Topics Discussed In Mimmentioning

confidence: 99%

Improving Bridging from Informatics Theory to Practice

Haux

Koch

2015

Appl Clin Inform

View full text Add to dashboard Cite

SummaryBackground: In 1962, Methods of Information in Medicine (MIM) began to publish papers on the methodology and scientific fundamentals of managing data, information, and knowledge in biomedicine and health care. Meeting an increasing demand for research about practical implementation of health information systems, the journal Applied Clinical Informatics (ACI) was launched in 2009. Both journals are official journals of the International Medical Informatics Association (IMIA). Objectives: Based on prior analyses, we aimed to describe major topics published in MIM during 2014 and to explore whether theory of MIM influenced practice of ACI. Our objectives were further to describe lessons learned and to discuss possible editorial policies to improve bridging from theory to practice. Methods: We conducted a retrospective, observational study reviewing MIM articles published during 2014 (N=61) and analyzing reference lists of ACI articles from 2014 (N=70). Lessons learned and opinions about MIM editorial policies were developed in consensus by the two authors. These have been influenced by discussions with the journal's associate editors and editorial board members. Results:The publication topics of MIM in 2014 were broad, covering biomedical and health informatics, medical biometry and epidemiology. Important topics discussed were biosignal interpretation, boosting methodologies, citation analysis, health-enabling and ambient assistive technologies, health record banking, safety, and standards. Nine ACI practice articles from 2014 cited eighteen MIM theory papers from any year. These nine ACI articles covered mainly the areas of clinical documentation and medication-related decision support. The methodological basis they cited from was almost exclusively related to evaluation. We could show some direct links where theory impacted practice. These links are however few in relation to the total amount of papers published. Conclusions: Editorial policies such as publishing systematic methodological reviews and clarification of possible practical impact of theory-focused articles may improve bridging.

show abstract

Improvement of Adequate Use of Warfarin for the Elderly Using Decision Tree-based Approaches

Cited by 26 publications

References 42 publications

Comparison of Nine Statistical Model Based Warfarin Pharmacogenetic Dosing Algorithms Using the Racially Diverse International Warfarin Pharmacogenetic Consortium Cohort Database

Comparison of Nine Statistical Model Based Warfarin Pharmacogenetic Dosing Algorithms Using the Racially Diverse International Warfarin Pharmacogenetic Consortium Cohort Database

Predicting the prolonged length of stay of general surgery patients: a supervised learning approach

Improving Bridging from Informatics Theory to Practice

Contact Info

Product

Resources

About