Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Applying statistical methods in a registry dataset of cardiopulmonary resuscitation to predict probability of survival by chest compression time in children

Huang, Hsin-Hui (2014) Applying statistical methods in a registry dataset of cardiopulmonary resuscitation to predict probability of survival by chest compression time in children. Master's Thesis, University of Pittsburgh. (Unpublished)

Submitted Version

Download (709kB) | Preview


The focus of this thesis was to apply advanced statistical methods to the American Heart Association Get With The Guidelines-Resuscitation (AHA GWTG-R) registry, a registry data set derived from a prospective multi-sites observational study, the American Heart Association’s National Registry of Cardiopulmonary Resuscitation (NRCPR). The data comprise comprehensive information related to the cardiopulmonary resuscitation (CPR) process, patients’ outcome, and characteristics of both the patients and the hospitals. The purpose of the registry data is to provide information that can be used to improve the outcomes of sudden cardiac arrest (SCA) patients and updates protocol of CPR.
This thesis has two purposes. The first one is to investigate the relationship between the patients’ disease and survival for SCA patients receiving different durations of chest compression. The second one is to establish a model for predicting the probability of survival according to the duration of CPR. In the clinical setting, a categorized variable may provide more meaningful inferences. To explore this option, a Generalized additive model (GAM) was used to identify cutoff points for the categorization of chest compression duration. This categorized variable was then used for the development of prediction models for survival and the Net reclassification index (NRI) was used to select the appropriate predictors for this model. Logistic regression, generalized estimating equations (GEE), and a generalized linear mixed model (GLMM) were performed to obtain the estimates of parameters. Thereafter, the probability of survival was estimated based on the results of the regression model.
Comprehensive registry data have been established for many healthcare problems, which include many observations and variables. A systematic process to analyze registry data is necessary. This thesis used multiple statistical techniques to create meaningful variables, select appropriate predictors, fit regression models, and predict the probabilities of outcome. The public health significance of this thesis is the identification of subgroups of SCA patients who may benefit from prolonged CPR duration and to assess significance of cluster effects in the registry data.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Huang, Hsin-Huihsh16@pitt.eduHSH16
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorWeissfeld, Lisa Alweis@pitt.eduLWEIS
Committee MemberChang, Chung-Chou H.changj@pitt.eduCHANGJ
Committee MemberBertolet, Marniebertoletm@edc.pitt.eduMHB12
Date: 30 September 2014
Date Type: Publication
Defense Date: 28 July 2014
Approval Date: 30 September 2014
Submission Date: 9 July 2014
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 57
Institution: University of Pittsburgh
Schools and Programs: School of Public Health > Biostatistics
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: GAM, NRI, Logistic regression, GEE, GLMM
Date Deposited: 30 Sep 2014 13:47
Last Modified: 15 Nov 2016 14:22


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item