Validating the CirCom Score in the English CPRD population

Date of Approval: 
2016-08-16 00:00:00
Lay Summary: 
Liver cirrhosis is a condition where the liver has accumulated too much scar tissue to function normally. Studies to understand differences in outcomes between patient groups are important to identify what factors are important in survival. However, some of the differences in outcomes are due to other diseases that the patients may have that are not directly related to the liver. Measurement scores that try to summarise the effect of these other diseases can help to adjust for differences in case mix between patients. CirCom is one such score that was developed in one Danish group of liver patients and also tested in another group of Danish patients. To understand if this score would be applicable outside Denmark in other groups of patients it is important to test how it performs in cohorts such as the Clinical Practice Research Datalink. We have previously defined a group of patients with liver cirrhosis within the Clinical Practice Research and published detailed descriptions of their disease and outcomes, and will therefore use this cohort to test the performance of the Danish CirCom score.
Technical Summary: 
Objectives To validate the CirCom score within an English cirrhosis population as a measure of a patient’s co-morbidity burden associated with mortality, and assess whether it is a better measure than the Charlson index. Methods and Data analysis We have previously defined and validated a cohort of liver cirrhosis patients in the Clinical Practice Research Datalink. We will use this cohort to assess the performance of a Cox Proportional Hazards model predicting all-cause one-year mortality in the linked Office of National Statistics death register. The survival model will consist of age, gender and the CirCom score. Comparisons will be made with the Charlson index using the Bayesian Information Criterion as a measure of goodness of fit, the Brier’s score as a measure of calibration, and the Harrell’s C Statistic and the Net Reclassification index as a measure of discrimination. The performance of the score will be compared to that in the original Danish paper. The results will also be stratified by age, gender, socioeconomic status, cirrhosis aetiology, cirrhosis stage and follow up year from the 1st year to the 5th year following the first diagnosis date of cirrhosis.
Health Outcomes to be Measured: 
CirCom score Charlson Index
Application Number: 

Colin Crooks - Chief Investigator - University of Nottingham
Colin Crooks - Corresponding Applicant - University of Nottingham
Joe West - Collaborator - University of Nottingham
Peter Jepsen - Collaborator - Aarhus University Hospital

HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation