Understanding the importance of key risk factors in predicting chronic bronchitic symptoms using a machine learning approach

被引:11
作者
Deng, Huiyu [1 ]
Urman, Robert [1 ]
Gilliland, Frank D. [1 ]
Eckel, Sandrah P. [1 ]
机构
[1] Univ Southern Calif, Dept Prevent Med, 2001 N Soto St,MC-9234, Los Angeles, CA 90089 USA
基金
美国国家卫生研究院;
关键词
Bronchitic symptoms; Air pollution; Machine learning; Gradient boosting model; Prediction model; SOUTHERN CALIFORNIA CHILDREN; AIR-POLLUTION; ASTHMA EXACERBATIONS; CLASSIFICATION; TREES; COST;
D O I
10.1186/s12874-019-0708-x
中图分类号
R19 [保健组织与事业(卫生事业管理)];
学科分类号
摘要
BackgroundChronic respiratory symptoms involving bronchitis, cough and phlegm in children are underappreciated but pose a significant public health burden. Efforts for prevention and management could be supported by an understanding of the relative importance of determinants, including environmental exposures. Thus, we aim to develop a prediction model for bronchitic symptoms.MethodsSchoolchildren from the population-based southern California Children's Health Study were visited annually from 2003 to 2012. Bronchitic symptoms over the prior 12months were assessed by questionnaire. A gradient boosting model was fit using groups of risk factors (including traffic/air pollution exposures) for all children and by asthma status. Training data consisted of one observation per participant in a random study year (for 50% of participants). Validation data consisted of: (1) a random (later) year in the same participants (within-participant); (2) a random year in participants excluded from the training data (across-participant).ResultsAt baseline, 13.2% of children had asthma and 18.1% reported bronchitic symptoms. Models performed similarly within- and across-participant. Previous year symptoms/medication use provided much of the predictive ability (across-participant area under the receiver operating characteristic curve (AUC): 0.76 vs 0.78 for all risk factors, in all participants). Traffic/air pollution exposures added modestly to prediction as did body mass index percentile, age and parent stress.ConclusionsRegardless of asthma status, previous symptoms were the most important predictors of current symptoms. Traffic/air pollution variables contribute modest predictive information, but impact large populations. Methods proposed here could be generalized to personalized exacerbation predictions in future longitudinal studies to support targeted prevention efforts.
引用
收藏
页数:12
相关论文
共 32 条
[1]   Classification of repeated measurements data using tree-based ensemble methods [J].
Adler, Werner ;
Potapov, Sergej ;
Lausen, Berthold .
COMPUTATIONAL STATISTICS, 2011, 26 (02) :355-369
[2]  
[Anonymous], 2007, Analysis of racial disparities in the New York City Police Departments stop, question, and frisk practices
[3]  
[Anonymous], 2002, HEALTH-LONDON
[4]   Cost of near-roadway and regional air pollution-attributable childhood asthma in Los Angeles County [J].
Brandt, Sylvia ;
Perez, Laura ;
Kuenzli, Nino ;
Lurmann, Fred ;
Wilson, John ;
Pastor, Manuel ;
McConnell, Rob .
JOURNAL OF ALLERGY AND CLINICAL IMMUNOLOGY, 2014, 134 (05) :1028-1035
[5]   Chronic effects of air pollution on respiratory health in Southern California children: findings from the Southern California Children's Health Study [J].
Chen, Zhanghua ;
Salam, Muhammad T. ;
Eckel, Sandrah P. ;
Breton, Carrie V. ;
Gilliland, Frank D. .
JOURNAL OF THORACIC DISEASE, 2015, 7 (01) :46-+
[6]   Factors associated with asthma exacerbations during a long-term clinical trial of controller medications in children [J].
Covar, Ronina A. ;
Szefler, Stanley J. ;
Zeiger, Robert S. ;
Sorkness, Christine A. ;
Moss, Mark ;
Mauger, David T. ;
Boehmer, Susan J. ;
Strunk, Robert C. ;
Martinez, Fernando D. ;
Taussig, Lynn M. .
JOURNAL OF ALLERGY AND CLINICAL IMMUNOLOGY, 2008, 122 (04) :741-747
[7]   A working guide to boosted regression trees [J].
Elith, J. ;
Leathwick, J. R. ;
Hastie, T. .
JOURNAL OF ANIMAL ECOLOGY, 2008, 77 (04) :802-813
[8]   Predicting asthma exacerbations in children [J].
Forno, Erick ;
Celedon, Juan C. .
CURRENT OPINION IN PULMONARY MEDICINE, 2012, 18 (01) :63-69
[9]   Random forests and stochastic gradient boosting for predicting tree canopy cover: comparing tuning processes and model performance [J].
Freeman, Elizabeth A. ;
Moisen, Gretchen G. ;
Coulston, John W. ;
Wilson, Barry T. .
CANADIAN JOURNAL OF FOREST RESEARCH, 2016, 46 (03) :323-339
[10]   Greedy function approximation: A gradient boosting machine [J].
Friedman, JH .
ANNALS OF STATISTICS, 2001, 29 (05) :1189-1232