Standardised survival and excess Life Years Lost in patients with type 3 intestinal failure

Long term outcomes have been reported in home parenteral nutrition (HPN)-dependent patients with type 3 intestinal failure (IF), but there are limited survival data standardised to the general population that would help provide a meaningful prognosis for patients and clinicians. The primary aim of this study was therefore to investigate the survival of HPN-dependent patients and to evaluate the specific impact of type 3 IF on their life expectancy standardised to that of the general population.

This was a cohort study of adult patients initiated on HPN between 1978 and 2018 at a national UK IF reference centre and followed up until death or censoring date of 31st December 2020. The standardised mortality ratio (SMR) was calculated as observed deaths divided by expected deaths using UK Office for National Statistics database. Excess Life Years Lost (LYL) were calculated separately for each sex as the differences in average life expectancy between patients with type 3 IF and the general population. Survival data were evaluated using cox regression models adjusting for confounding.

In total, 1046 patients were identified, with a total observation time of 7344.1 patient-years. Patients with malignancy (n = 206) were excluded from the survival analysis. Of the remaining 840 patients, 398 were alive by the end of follow-up. The probability of survival was 91.8% at 1 year, 69.3% at 5 years, 54.3% at 10 years, 29.8% at 20 years and 16.7% at 30 years. Patients who did not achieve nutritional autonomy had an increased likelihood of death compared to patients who ceased HPN. In total, 40 (9.0%) deaths were HPN or IF-related, while underlying disease leading to IF accounted for 98 (22.2%) deaths. There were 270 (61.1%) deaths not related to IF, with the majority of these patients dying from infections unrelated to HPN. Overall mortality rates were higher among patients with a diagnosis of type 3 IF compared with the general UK population with a SMR of 7.48 (95% CI 6.80 to 8.21) and an excess mortality rate of 54.0 per 1000 person-years. All mechanisms of IF were associated with excess mortality, with SMR ranging from 6.82 (95% CI 5.98 to 7.72) for short bowel syndrome to 15.51 (95% CI 11.73 to 20.03) for dysmotility. On average, the excess LYL was 17.45 years for males and 17.39 years for females compared with the general population of the same age.

This the largest single-centre series reporting survival outcomes in patients with type 3 IF over more than a four-decade period and the first to report LYL in this patient cohort. Type 3 IF was associated with more than seven-fold higher mortality rates than for the general UK population and shorter life expectancies of more than 17 years. Survival, however, was better in those able to achieve nutritional autonomy. Since the majority of deaths were due to non-HPN or non-IF causes, there is clearly a need now to further explore these causes of death in order to improve our understanding of excessive mortality in type 3 IF and develop ways to prevent it.