Regency EF
Longitudinal Parallel-Process Modeling: An Application to a Study of Pediatric Stem Cell Transplant Patients (304052)
*Paula M. Murray, Children's Hospital Los Angeles
Keywords: longitudinal data analysis, symptom modeling, multilevel modeling
Longitudinal parallel-process (LPP) modeling is a special case of latent growth curve modeling which may be used to study the relationship between two or more longitudinal outcomes that occur in tandem. Despite its applicability to data collected across many areas of research and industry, LPP remains an underutilized method for longitudinal data analysis.
We present an application of LPP to study the association between parent anxiety, depression, and stress with symptom burden and quality of life in children undergoing hematopoietic stem cell transplantation. By examining the estimated intercepts and slopes, we can explore the relationship between pre-transplant health status and the change in health outcomes over time.
Practical considerations for performing LPP will be discussed including methods for implementing LPP in standard statistical software packages.