Use this URL to cite or link to this record in EThOS:
Title: Persisting inflammation after critical illness : prevalence, risk factors and association with physical recovery
Author: Griffith, David Morgan
ISNI:       0000 0004 6425 3365
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Introduction: Survivors of critical illness suffer physical, psychological, and social problems. The factors hindering recovery and the best rehabilitation interventions remain illusive. Critical illness is associated with inflammation that persists in some individuals and this might affect recovery. The hypotheses for the thesis are 1. Persistent inflammation is common after critical illness. 2. Persistent inflammation is associated with functional recovery. 3. Persistent inflammation is associated with critical illness-induced viral reactivation. 4. Biomarkers may help predict functional disability. The main part of the project was based on 197 of the 240 patients enrolled in the RECOVER trial who also consented to take part in an inflammation sub study. Systematic Review. Aims: Prevalence of systemic inflammation in ICU survivors; association of inflammation with physical recovery. 7433 references identified. 208 full text articles were reviewed. 57 were eligible. 22 studies included the relevant data. CRP at ICU discharge was elevated ( > 10mg/L) in most cases (70% of mixed medical/surgical patients and 100% of severe sepsis survivors). Lower CRP observed in trauma patients (23mg/L), VAP (46mg/L), > 6 days in ICU (45mg/L), and medical ICU patients (36mg/L). CRP was higher in sepsis (107mg/L) and surgical ICU patients (99mg/L). IL-6, TNF-a, and PCT were elevated in most patients at ICU discharge. Ninety percent of acute COPD exacerbations admitted to ICU had elevated CRP at hospital discharge and 43% general adult ICU patients fulfilled SIRS criteria 3 days after ICU discharge. Anaemic ICU survivors had elevated CRP and IL-6 at 6 months. There were no studies that measured both inflammation and physical function after ICU discharge. Association of inflammation with functional outcome after critical illness. Aims: Prevalence of inflammation in heterogeneous ICU cohort, association of CRP with Rivermead Mobility Index (RMI) and other outcome measures at 3 months. At ICU discharge, 173 patients (94%) had elevated serum CRP with a median concentration of 27 (11-60) mg/L. At hospital discharge 169 patients (90%) had elevated CRP with a median concentration of 21 (8-42) mg/L, At 3 months 72 patients (59%) had elevated CRP with a median concentration of 4 (1-12) mg/L. CRP was associated with RMI (p < 0.01), and percentage of predicted handgrip strength (HGS) (p=0.03) at 3 months. CMV infection, systemic inflammation and the post ICU syndrome. Aims: Prevalence of active and latent CMV at ICU discharge; association between CMV, inflammation, and recovery. 115 patients (62.8%) had latent CMV. 13 (11.4%) had active CMV (11.4% of those with prior CMV and 7.2% of ICU survivors). Active CMV associated with longer hospital length of stay (57 days v 28 days p=0.016), poorer baseline physical function (HGS 12 v 16 p=0.032; RMI 1 v 2 p=0.018). At 3 months, patients with latent CMV infection had higher CRP (5.4 v 2.8mg/L p=0.06), higher HNE (118 v 91.3 pg/mL p < 0.00), lower TGF β (9.2 v 11.4 ng/mL p=0.01), and were slower on 2 min timed up and go test (p=0.03). Active CMV infection at ICU discharge was not associated with inflammation or physical function at 3 months. Prediction of physical disability after ICU discharge. Aims: To identify risk factors for poor physical function; to derive a prognostic index to identify for poor functional outcome. Age, Functional Comorbidity Index, Scottish Index of Multiple Deprivation quintile, CMV IgG status, ventilator days, baseline RMI, physical component of SGA, CRP, and SLPI met statistical criteria for consideration in the multivariable models. 2 linear multivariable models with reasonable fit (R2=0.175; 0.193) were constructed. AUCs were 0.759 for the clinical model and 0.725 for the model incorporating biochemical markers. The models did not perform any better than a baseline assessment of mobility (RMI).
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (M.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available