TY - JOUR
T1 - Prognostic factors of first-ever stroke patients in suburban Malaysia by comparing regression models
AU - Wan-Arfah, Nadiah
AU - Muzaimi, Mustapha
AU - Naing, Nyi Nyi
AU - Subramaniyan, Vetriselvan
AU - Wong, Ling Shing
AU - Selvaraj, Siddharthan
N1 - Publisher Copyright:
© 2023 by Author/s.
PY - 2023/11/1
Y1 - 2023/11/1
N2 - Introduction: The aim of this study was to compare regression models based on the parameter estimates of prognostic factors of mortality in first-ever stroke patients. Methods: A retrospective study among 432 first-ever stroke patients admitted to Hospital Universiti Sains Malaysia, Kelantan, Malaysia, was carried out. Patient’s medical records were extracted using a standardized data collection sheet. The statistical analyses used for modelling the prognostic factors of mortality were Cox proportional hazards regression, multinomial logistic regression, and multiple logistic regression. Results: A total of 101 (23.4%) events of death were identified and 331 patients (76.6%) were alive. Despite using three different statistical analyses, the results were very similar in terms of five major aspects of parameter estimates, namely direction, estimation, precision, significance, and magnitude of risk assessment. It was reported slightly better in Cox proportional hazards regression model, especially in terms of the precision of the results. Conclusions: Given that this study had compared the findings from three different types of advanced statistical methods, this research has clearly yielded that with data of high quality, the selection of appropriate statistical method should not be a worrisome problem for researchers who may not be of expertise in the field of medical statistics.
AB - Introduction: The aim of this study was to compare regression models based on the parameter estimates of prognostic factors of mortality in first-ever stroke patients. Methods: A retrospective study among 432 first-ever stroke patients admitted to Hospital Universiti Sains Malaysia, Kelantan, Malaysia, was carried out. Patient’s medical records were extracted using a standardized data collection sheet. The statistical analyses used for modelling the prognostic factors of mortality were Cox proportional hazards regression, multinomial logistic regression, and multiple logistic regression. Results: A total of 101 (23.4%) events of death were identified and 331 patients (76.6%) were alive. Despite using three different statistical analyses, the results were very similar in terms of five major aspects of parameter estimates, namely direction, estimation, precision, significance, and magnitude of risk assessment. It was reported slightly better in Cox proportional hazards regression model, especially in terms of the precision of the results. Conclusions: Given that this study had compared the findings from three different types of advanced statistical methods, this research has clearly yielded that with data of high quality, the selection of appropriate statistical method should not be a worrisome problem for researchers who may not be of expertise in the field of medical statistics.
KW - Cox proportional hazards regression
KW - first-ever stroke
KW - health risk
KW - multinomial logistic regression
KW - prognostic factors
UR - http://www.scopus.com/inward/record.url?scp=85178901686&partnerID=8YFLogxK
U2 - 10.29333/ejgm/13717
DO - 10.29333/ejgm/13717
M3 - Article
AN - SCOPUS:85178901686
SN - 2516-3507
VL - 20
JO - Electronic Journal of General Medicine
JF - Electronic Journal of General Medicine
IS - 6
M1 - em545
ER -