Risk factors for health-related quality of life deterioration following pediatric critical illness

Elizabeth Killien | 2019

Advisor: Noel Weiss

Research Area(s): Epidemiologic Methods, Global Health, Maternal & Child Health, Social Determinants of Health


Background: Many children experience declines in health-related quality of life (HRQL) after critical illness, but it is unknown what components of HRQL are most affected. We aimed to evaluate risk factors for deterioration between baseline and post-discharge HRQL status in the individual components of common HRQL measures among pediatric survivors of critical illness. Methods: We conducted a retrospective cohort study of children ≤18 years admitted to the Seattle Children’s Hospital pediatric and cardiac intensive care units from 12/2011-02/2017 on whom pre-admission baseline, admission, and post-discharge follow-up (median 42.0 days) PedsQL or FS II-R HRQL assessments had been obtained. We determined associations between patient and illness characteristics with failure to recover within 4.5 points of baseline HRQL at follow-up (the minimum clinically significant difference between two scores) using multivariable generalized linear Poisson regression. We determined the individual score components of the PedsQL Infant Scales (age <2), PedsQL Generic Core Scales (age 2-18), and FS II-R (for children with severe developmental or functional disabilities) with the highest prevalence of decline from baseline to follow-up, and assessed factors associated with decline for each question. Results: Failure to recover within 4.5 points of baseline HRQL status occurred in 24.7% of 736 study patients. Factors independently associated with failure to recover were older age (adjusted RR 1.02/year, 95% CI 1.00-1.05), primary admission diagnosis of an oncologic condition (aRR 1.89 relative to respiratory, 95% CI 1.07-3.34) or orthopedic surgery (aRR 1.68, 95% CI 1.04-2.70), and longer duration from discharge to follow-up HRQL assessment (aRR 0.91/week, 95% CI 0.86-0.95). Individual score components most affected for the PedsQL Infant Scales were primarily emotional (crying/fussing, difficulty sleeping, difficulty self-soothing, feeling tired, and feeling afraid). Children 2-18 years assessed with the PedsQL Generic Core Scales most commonly experienced declines in physical functioning (participating in active play/exercise, hurting, lifting something heavy, being unable to do things other children their age can do) and feeling angry. Children for whom the FS II-R was completed most commonly experienced declines in both emotional and physical domains (reacting by crying, not sleeping through the night, irritability, poor energy, and acting moody). Patient demographic characteristics, underlying medical complexity, primary admission diagnosis, and severity of illness were commonly associated with declines in individual score components. Conclusions: One-quarter of children surviving critical care experienced a clinically significant deterioration in HRQL when assessed a median of six weeks post-discharge, and the individual components of HRQL that were most affected varied by patient age and diagnosis. Potential targets for intervention to improve HRQL outcomes for critically ill children include minimizing deconditioning, improving sleep quality, addressing fear, and adequately managing pain.