Factors affecting the survival of stroke patients in the last eight years using parametric and non-parametric models: A study in the northwest of Iran

Background. The aim of this study was to evaluate factors affecting the survival of 5,430 stroke patients in the northwest of Iran using parametric and non-parametric models of survival analysis. Methods. This longitudinal school-based study was retrospectively conducted using the registered data of 5,430 stroke patients in Tabriz. The Akaike criterion and Cox-Snell residual diagram were used to evaluate the parametric models, and the Kaplan-Meyer curve was employed for assessing the non-parametric model. Analyses were performed using STATA17. Results. The findings demonstrated that different factors were associated with an increased risk of death from stroke, including the age group of 65 and older (proportional hazard [PH]= 1.75 and P<0.001), snoring (PH=1.34 and P= 0.005), addiction (PH= 1.87 and P= 0.007), smoking (PH= 5.94 and P< 0.001), and hookah use (PH = 1.81 and P= 0.005). The other related factors were diabetes (PH= 1.46 and P= 0.001), aspiration pneumonia (PH = 2.09 and P < 0.001), atrial fibrillation (PH= 1.59 and P= 0.002), systolic hypertension (PH= 3.15 and P< 0.001), and high triglyceride (PH = 1.33 and P= 0.012). Considering the frailty, the fitness of the Weibull model was better than that of the other models. Conclusion. According to the findings, snoring and having aspiration pneumonia were two important risk factors related to the survival of stroke patients. It is also recommended that the parametric models of survival analysis be used if the relevant assumptions are established. Practical Implications. The risk factors determined in the present study can be utilized in determining relevant interventions. It is recommended that identification programs such as screening and treatment of patients be implemented in this respect.