Short- and long-term outcomes for patients with variceal haemorrhage in a tertiary hospital

Suboptimal Implementation of Evidence-based Therapy for Acute Variceal Hemorrhage: A Systematic Review of Observational Studies

BACKGROUND & AIMS

Gastroenterologists should strive to improve the outcomes associated with acute variceal hemorrhage (AVH) by optimizing care delivery and clinician preparedness through training. Unfortunately, data regarding contemporary outcomes and management of AVH are limited.

METHODS

We performed a systematic review of cohort studies examining outcomes and management of AVH from January 1990 to September 2016. We pooled data on mortality and utilization of medical therapy, endoscopic interventions, balloon tamponade (BT), and salvage procedures (portal decompressive surgery or transjugular portosystemic shunt).

RESULTS

The pooled 6-week mortality rate after AVH was 17.7% (95% confidence interval [CI], 14.4%-21.7%). Sclerotherapy was used in 18.7% of cases (95% CI, 9.2%-34.3%), and no endoscopic intervention was provided in 14.3% of patients (95% CI, 9.7%-20.6%). The overall rate of BT use was 10.8% (95% CI, 8.2%-14.1%). Salvage procedures were used in 5.7% (95% CI, 3.7%-8.6%), falling to 1.7% (95% CI, 0.7%-3.8%) among patients treated after 2000. Although pooled rates of timely endoscopy and vasoactive medication use were high (85.8% and 83.0%, respectively), only half of patients received prophylactic antibiotics (52.8%; 95% CI, 40.0%-66.2%). In studies that enrolled patients exclusively after 2000, 13.8% of patients (95% CI, 7.0%-25.4%) received sclerotherapy, salvage transjugular portosystemic shunt fell to 1.7% (95% CI, 0.7%-3.8%), BT use fell slightly to 8.7% (95% CI, 6.4%-11.6%), and vasoactive medication and prophylactic antibiotic use rose to 91.4% (95% CI, 86.0%-94.9%) and 62.7% (95% CI, 48.9%-74.9%), respectively.

CONCLUSIONS

AVH is associated with high mortality rates and suboptimal implementation of evidence-based therapies including prophylactic antibiotics and endoscopic interventions, suggesting a need for quality improvement. In addition, the frequent need for BT and sclerotherapy suggests that specific attention to these procedures in gastroenterology training curricula may be warranted.

Liver collagen in cirrhosis correlates with portal hypertension and liver dysfunction

Liver collagen proportionate area (CPA) assessed by computer-assisted digital image analysis has been proposed as an accurate and objective histological variable for subclassifying cirrhosis. The study aimed to examine the relationship between CPA and relevant clinical parameters in cirrhotic patients and to evaluate the sampling variability for CPA. The study included 48 consecutive liver transplantation patients with established cirrhosis. Hepatic venous pressure gradient (HVPG) and serum markers of liver failure were determined prior to transplantation. CPA was assessed in the explanted livers. In 20 of the livers, CPA was measured in more than one tissue sample. CPA showed significant correlations with HVPG and with various surrogate markers of hepatic dysfunction including albumin, bilirubin, INR, MELD score and Child-Pugh score. CPA reliably discriminated HVPG ≥10 mmHg, termed 'clinically significant portal hypertension' (area under receiver operator curve: 0.923, p < 0.001; odds ratio: 1.209, p = 0.003). CPA measured on tissue blocks showed no significant sampling variability (p > 0.5). In conclusion, the study correlated portal hypertension and hepatic dysfunction with the amount of collagen in cirrhotic livers. The findings support the presumption of CPA as a useful histological marker for subclassifying cirrhosis and as a helpful supplement to the qualitative description of hepatic architectural changes in routine pathology.