Trajectory Study On Rehabilitation Exercise Compliance In First-episode Stroke Patients Based On The Latent Growth Mixture Model

Research Purpose:（1）The latent growth mixture modeling was used to describe the general change trajectory of rehabilitation exercise compliance in the first stroke patients within 6 months,and the heterogeneous development trajectory of different subgroups was identified based on the latent class growth model.（2）Based on the bio-psycho-social medicine model,the predictive factors of different trajectories of rehabilitation exercise compliance of stroke patients were explored in a multi-dimensional manner,providing a basis for the formulation of precise nursing interventions.Research Methods:A longitudinal study design was used to collect 282 first-episode stroke patients who met the inclusion criteria in a grade III,Class A general hospital in Zhejiang Province from January2021 to December 2021.The patient general information questionnaire（self-made）,modified Barthel Index（MBI）,Simple Cognitive Function Scale（MMSE）,self-rating Depression Scale（SDS）,and Social support Rating Scale（SSRS）were distributed on the day of admission（T0）to complete the patient’s inclusion.The functional exercise Compliance Scale（EAQ）for stroke patients was collected at 2 weeks（T1）,6 weeks（T2）,8 weeks（T3）,12 weeks（T4）,and 24 weeks（T5）,at the bedside of patients in hospital,and by telephone follow-up after discharge.SPSS25.0 and Mplus8.3 software were used to collect and input data.First,the latent variable growth mixed model was used to describe the general development trend of rehabilitation exercise compliance of stroke patients.Then,the latent category growth model was used to fit the potential categories of rehabilitation exercise compliance development trajectory of stroke patients.Log Likelihood value（LL）,AIC（Akaike Information Criterion,AIC）,BIC（Bayesian Information Criterion,BIC）,Sample corrected BIC（Sample-size Adjusted BIC）,Entropy,Bootstrap-based likelihood ratio test（BLRT）,and LMR(Lo-Mendell-Rubin,LMR and other fitting indexes were used to evaluate the fitting results and determine the best fitting model.From the general demographic data,disease-related data and other biological factors,combined with psychological and social factors,the predictive factors affecting the trajectory of rehabilitation exercise compliance of stroke patients were analyzed.First,single factor analysis was used to screen statistically significant independent variables,and then multi-category logistic regression was used to analyze the predictors of different patient compliance trajectory.Research Results:（1）Development trend of compliance with rehabilitation exercise: Repeated measurement an OVA was used to conduct statistical tests on the compliance index scores at five time points,and the results showed that the total score of the patient’s compliance index and the compliance index scores of all dimensions were statistically significant over time（P < 0.001）.The total scores of stroke patients’ compliance with t1-T5 rehabilitation exercise were（63.01 ± 12.94）points,（63.89 ± 12.28）points,（63.06 ± 12.51）points,（62.83 ± 12.42）points,（61.13 ± 11.88）points,respectively.At T2,compliance increased to（63.89±12.28）points,reaching the peak;T4 time point began to show a downward trend,T5 time point was（61.13±11.88）,the lowest point.（2）Model fitting: the latent variable growth mixed model was used to fit the general change trajectory of stroke patients’ compliance at five time points,and it was found that.According to the latent category growth model,the data fitting results of the three latent category models were determined to be the best,with a low BIC value of 7114.325,entropy value of 0.880 and STATISTICALLY significant BLRT（P < 0.05）.Combined with the intercept and slope changes,the three types of development trajectory were named as "high level slow decline group"（C1,18.3%）,"low level sustained stability group"（C2,73.7%）and "medium level rapid decline group"（C3,8.0%）.（3）Predictors of exercise compliance trajectory:Univariate analysis: statistically significant independent variables included gender（P=0.023）,occupation（P=0.048）,number of children（P=0.018）,and number of chronic diseases（P=0.008）.Multiple logistic regression: positive single factor regression indicators,social support and cognitive function were included.The results showed that compared with C2 group,female patients were the main predictors of patient compliance type development into C1.Compared with C3 group,the main predictors of stroke patients’ development into C2 group included:employment,number of chronic diseases and social support.Research Conclusion:（1）Overall status of rehabilitation exercise compliance of stroke patients: rehabilitation exercise compliance of stroke patients within 6 months changed with the course of disease,and the change was non-linear.（2）There were three different development trajectories of rehabilitation exercise compliance in stroke patients within 6 months: high level slow decline group（18.3%）,low level sustained stability group（73.7%）and medium level rapid decline group（8.0%）.Rehabilitation exercise compliance of stroke patients reached the peak at 6 weeks of onset,and decreased significantly from 12 weeks to 24 weeks of onset.Track exploration was conducted for each dimension of compliance,and the score of the dimension of actively seeking advice was the lowest,and most patients showed a significant downward trend.This warns clinical staff to change the traditional passive education in health education.（3）Gender,occupation,number of chronic diseases,number of children and degree of social support of patients with stroke all affect the development trajectory of rehabilitation exercise compliance.Among them,gender,number of complicated chronic diseases,occupation and social support play an important role in distinguishing and predicting the development trajectory of different compliance,which can provide reference for early identification and screening of patients with low compliance in the future.