Improving 30-day mortality after radiologically inserted gastrostomy tube from 2007 to 2019: A population-based study of 15,605 patients

Abstract

Background and aims

Radiologically inserted gastrostomy (RIG) allows long-term enteral nutrition when percutaneous endoscopic gastrostomy (PEG) tube insertion is not feasible either due to technical difficulty or a higher risk of complications. The aims of this study were to examine mortality associated with RIG insertion.

Methods

Adult patients with RIG insertion from 2007 to 2019 were identified in the Hospital Episode Statistics database. Indications and adverse events were identified using International Statistical Classification of Diseases and Related Health Problems, 10th Revision codes. Provider nutrition support data were available from the Getting It Right First Time in Gastroenterology Report 2021. Multivariable logistic regression analysis examined factors associated with 30-day mortality following a RIG.

Results

15,605 patients were studied (68.0 % male; age 64(Interquartile range (IQR) 56–73)). There was a steady increase in the number of RIGs inserted from 510 per year in 2007 to 1787 per year in 2019. 59.9 % of RIGs were inserted as an outpatient. 63.3 % of RIGs were inserted in head and neck cancer patients.

Of the patients who had a RIG insertion, 4.7 % had pneumonia within 7 days and 6.9 % died within 30 days of RIG insertion. Thirty-day mortality fell from 12.2 % in 2007 to 5.8 % in 2019. Higher 30-day mortality rates were observed in patients with Dementia (16.4 %) and in NHS providers without a nutrition support nurse (11.5 %).

Factors associated with 30 day mortality included: increasing age (>81 years odds ratio (OR) 13.67 (95 % confidence interval (CI) 4.99–37.48), p < 0.001); increasing NHS provider volume of RIG insertion >12 per year (OR 0.69 (95 % CI (0.55–0.88), p = 0.003); RIG insertion during an emergency admission (OR 2.53 (95 % CI 2.19–2.93), p < 0.001); increasing comorbidity Charlson score >5 (OR 1.38 (95 % CI 1.10–1.75), p = 0.006); NHS provider without a nutrition support nurse (OR 1.38 (95%CI 1.09–1.75), p = 0.007) and other neurological conditions than stroke as indication for RIG (OR 1.55 (95%CI 1.24–1.95), p < 0.001).

Conclusions

Despite an increase in RIG insertion over the study period, 30-day mortality has fallen by 52 %. Providers without a nutrition support nurse and providers with a lower volume of RIG insertions were associated with higher mortality.