Introduction Prediction models for functional recovery after stroke can be improved by adjusting for the heterogeneity in functional recovery patterns. This study explored the heterogeneity in functional recovery after stroke using longitudinal latent class analysis and characterised the patients in the different latent classes.
Methods The analyses were performed on a data set from a cohort of 448 stroke survivors participating in a study of outcomes at 1 year. Heterogeneity in functional recovery after stroke was investigated using Longitudinal Latent class analysis of total Barthel scores measured at 1, 6 and 12 months after stroke. Identification of the optimal number of classes was based on BIC, AIC, and Lo-Mendell-Rubin Adjusted Likelihood ratio test. The second analysis characterised the latent classes.
Results A four latent class structure was preferred. All the four latent classes showed a non linear pattern of recovery over time. Persons in the very poor functional recovery group had the largest median length of initial hospital stay 99 (13–257 days), mean age 75±9.27 years and greatest probability of being urinary and bladder incontinence. The group with best functional recovery had the least initial hospital stay 14 (2–147 days), least proportion of people with previous stroke, least proportion of people with urinary and bladder incontinence, the mean age at admission was 68.76±11.72 years.
Conclusion The study showed that there is heterogeneity in functional recovery patterns after stroke. Latent class analysis is a useful method for identifying subgroups of functional recovery after stroke.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.