Timing of endoscopy for acute upper gastrointestinal bleeding in North West England: Results from a multicentre trainee-led network

We read with interest the recent paper by Siau et al. ¹ reporting results from a multicentre trainee-led gastrointestinal bleed (AUGIB) audit in the UK. Through our newly formed Gastroenterology Trainee Research and Improvement Network North West (GasTRIN NoW), we would like to share our results of AUGIB management within North West England.

A prospective multicentre AUGIB audit was undertaken across 11 hospitals in North West England between 30 October 2017 and 26 November 2017. All adult patients admitted with suspected AUGIB who underwent gastroscopy were included. The standards assessed were as defined by Joint Advisory Group on GI Endoscopy (JAG) and reported in Siau et al. ¹

A total of 133 patients were included, of which 84% (n = 112) were referred from Accident and Emergency while the remainder were from primary care. Median age was 65 years (range: 18–96), of which 65% (n = 87) were male. Of the patients, 51% (n = 68) were admitted during weekdays between 07:00 and 19:00. At admission, only 47% (n = 62) had a Glasgow-Blatchford bleeding score (GBS) documented.

Of the patients, 40% (n = 54) underwent an Oesophagogastroduodenoscopy (OGD) within 24 hours of admission (median: 28 hours, interquartile range (IQR): 18–50). Time to endoscopy did not differ significantly across centres (Kruskal p = 0.46; Figure 1). Delayed endoscopy occurred in 60% (n = 79) and this did not vary across centres (χ² p = 0.08). Outcomes of early and delayed endoscopy are summarised in Table 1.

Figure 1.

Time to endoscopy (hours) for each participating centre. Red dotted line denotes 24 hours.

Table 1.

Outcomes of early (<24 hours) and delayed (>24 hours) endoscopy groups.

	Time to endoscopy
Variable	<24 hours	>24 hours	p-value
Rebleed within 8 days of endoscopy	4	0	0.02
Death within 30 days of endoscopy	2	5	0.70
Length of stay (days)	5 (3–8)	4 (3–7)	0.43

Patients requiring endoscopic therapy had a significantly higher GBS score than those not requiring therapy (median GBS = 11, IQR: 8–13 versus 3, IQR: 2–4; p = 0.009). In total, 64% of patients who required therapy were endoscoped within 24 hours (n = 16). OGD was performed earlier in those who required intervention for bleeding compared with those who did not (median 19 hours, IQR 11–27 versus 32 hours, IQR: 21–54; p = 0.007). Length of stay of these patients was longer compared to those not requiring therapy (median 6 days, IQR: 5–7 versus 4 days, IQR: 2–7; p = 0.04). There were seven deaths at 30-day follow-up, but only one was directly attributable to AUGIB.

Our data report similar findings to Siau et al., that most centres failed to meet the JAG standards. Within our cohort, patients who required therapy for AUGIB had early endoscopy compared to those that did not require therapy; however, 37% of these cases did not meet the JAG guidelines for timing to endoscopy. These data provide impetus to improve the standards of AUGIB care across regions in England.