Letter to the Editor: Complications Associated With Anesthesia Services in Endoscopic Procedures Among Patients With Cirrhosis

TO THE EDITOR:

We read with interest the recent study by Lieber et al.⁽¹⁾ evaluating complications in patients with cirrhosis using anesthesia services for endoscopic procedures, and we agree with their findings that complications in patients with cirrhosis receiving anesthesia for endoscopic procedures are rare.

The authors emphasize the importance of assessment of disease severity including the American Society of Anesthesiologists (ASA) score as well as clinical features such as a history of portal hypertension and use these in their model of risk factors for adverse events. Notably, measures of disease severity such as the Child-Pugh (CP) and Model for End-Stage Liver Disease (MELD) scores have emerged as the two predominant prognostic markers for predicting survival in decompensated cirrhosis,⁽²⁾ but these were not used in the Lieber study because the study used an administrative dataset that did not include laboratory data. The inclusion of subjective variables, i.e., ascites and encephalopathy, have the potential to induce intraobserver variation, which could change scores.⁽²⁾ This is also true with the ASA score, which is highly subjective. A recent study showed that nonanesthesia providers had a 30%–40% chance of underrating ASA scores when compared with anesthesia colleagues.⁽³⁾ Given the potential for variability between ASA scores, we emphasize that clinical prediction models of procedural or sedation risk should include more objective measures of liver disease severity, in particular the MELD score and the CP score.

We have examined the relationship between type of sedation and complication rates in patients with cirrhosis.⁽⁴⁾ We found that moderate sedation and monitored anesthesia care (MAC) had similar adverse event rates and that procedure type did not influence complication rates. Our model found that biochemical variables such as higher albumin and lower CP scores were protective of complications, whereas increasing procedure times, and elevated INR were risk factors for complications. Interestingly, MELD was not predictive of complications, which is likely due to the fact that the presence of portal hypertension seems to better relate to procedural complications, whereas MELD tends to better predict overall prognosis and mortality.

With recent concerns about rising health care costs, it is important to better understand the most appropriate and cost effective approach(es) to sedation in patients with cirrhosis.⁽⁴⁾ An algorithmic approach (based on the authors’ personal opinions) to the utilization of anesthesia services could help decrease expensive anesthesia utilization (Fig. 1).⁽⁵⁾ Further studies to better develop prognostic scoring systems to better use anesthesia assistance for routine endoscopic procedures are required, which will have the potential to both decrease health care cost and improve patient safety.

FIG. 1.

Algorithm for sedation of the patient with cirrhosis.