Table 5.

Hospital-level variables used in the hierarchical regression model to determine which structural factors affect outcomes

		7-day mortality OR (95% CI)	30-day mortality OR (95% CI)	30-day readmission OR (95% CI)	Long length of stay OR (95% CI)
Transfer (ref no)	Transfer in	1.62 (0.82 to 3.22)	1.49 (0.94 to 2.38)	1.08 (0.82 to 1.41)	1.42 (1.25 to 1.61)*
Volume (ref <3000)	3000–4000	0.87 (0.66 to 1.13)	0.82 (0.68 to 0.99)	0.98 (0.88 to 1.09)	1.04 (0.95 to 1.14)
	>4000	1.07 (0.85 to 1.35)	0.87 (0.74 to 1.04)	0.96 (0.87 to 1.07)	0.96 (0.89 to 1.05)
Consultant onsite 24 hours (ref not)	Onsite	0.74 (0.48 to 1.12)	0.67 (0.47 to 0.98)	0.94 (0.77 to 1.15)	0.91 (0.78 to 1.07)
Consultant cleared of elective commitments (ref no)	Cleared	1.35 (1.09 to 1.67)*	1.04 (0.83 to 1.31)	1.01 (0.88 to 1.16)	0.80 (0.72 to 0.89)*
Primary assessment (ref trainee)	Assessment by consultant	0.89 (0.65 to 1.72)	1.08 (0.80 to 1.45)	1.11 (0.93 to 1.32)	1.11 (0.96 to 1.27)
Handovers (ref no)	Handovers	0.79 (0.57 to 1.10)	1.01 (0.78 to 1.31)	0.99 (0.86 to 1.15)	1.19 (1.06 to 1.34)*
Dedicated EGS operating theatre (ref no)	Dedicated EGS operating theatre	0.92 (0.70 to 1.22)	0.95 (0.78 to 1.17)	1.05 (0.93 to 1.19)	0.89 (0.80 to 0.98)
Surgical assessment unit (ref no)	Surgical assessment unit	1.26 (0.99 to 1.62)	1.29 (1.09 to 1.54)*	1.09 (1.00 to 1.20)	1.02 (0.95 to 1.10)
ICU beds		0.94 (0.91 to 0.98)*	0.98 (0.95 to 1.02)	1.00 (0.99 to 1.02)	1.01 (1.00 to 1.02)

The data presented are for the subgroup of patients who underwent surgery (cohort of 19 082 patients).

The four funnel plots (figures 1–4) were created by using patient-level variables in a multivariate logistic regression model and demonstrate differences in geographical trends and outcome between the 23 GC centres included in this study. (n=69 490 in all funnel plots.)

----=99.8% control limits.

*p<0.01.