Background & aims

Large, population-based studies that have included the full spectrum of cirrhosis estimating survival, taking into account time-at-risk are lacking. We aimed to report 1- and 5-year average survival rates for people with cirrhosis to be used in a clinical and healthcare policy setting.

Methods

We used the Clinical Practice Research Datalink and linked English Hospital Episode Statistics to identify adult cases of cirrhosis from January 1998 to December 2009. We estimated 1- and 5-year survival according to whether time-at-risk was ambulatory or followed an emergency hospital admission related to liver disease, stratified by age, sex, and aetiology to be used in a clinical setting. We used a multivariate Cox-proportional hazards model with a time-varying variable, adjusted for Baveno IV stage of cirrhosis at diagnosis, age, aetiology, and sex.

Results

We identified 5118 incident cases. Average survival probabilities at 1- and 5-years were 0.84 (95% CI 0.83-0.86) and 0.66 (95% CI 0.63-0.68) for the ambulatory group and 0.55 (95% CI 0.53-0.57) and 0.31 (95% CI 0.29-0.33) following hospitalisation, respectively. A hospital admission at diagnosis or subsequently for liver disease substantially impaired prognosis independent of stage of cirrhosis (HR=2.78, 95% CI 2.53, 3.06).

Conclusions

Emergency hospitalisation for liver disease heralds a downturn in a patient's outlook independent of their stage of cirrhosis. Our results provide population-based clinically translatable estimates of prognosis for the purposes of healthcare delivery and planning and communication to patients.