Table 4.

Logistic regression analysis to identify factors associated with in‐hospital mortality during training (48 trainees, 24 272 surgical training episodes)

		Died*		Univariable analysis		Multivariable analysis
	Alive		Missing	Odds ratio†	P	Odds ratio†	P
Student role
Observer	4397	116 (2·6)	2	1·00 (reference)		1·00 (reference)
Assistant	9075	225 (2·4)	11	0·94 (0·75, 1·18)	0·592	0·99 (0·78, 1·25)	0·915
Directly supervised	5633	89 (1·6)	2	0·60 (0·45, 0·79)	< 0·001	0·74 (0·55, 0·99)	0·045
Indirectly supervised	4678	36 (0·8)	1	0·29 (0·20, 0·42)	< 0·001	0·47 (0·32, 0·71)	< 0·001
Missing	7	0	0
Sex
M	9974	262 (2·6)	8	1·00 (reference)		1·00 (reference)
F	13 804	204 (1·5)	8	0·56 (0·47, 0·68)	< 0·001	0·43 (0·35, 0·54)	< 0·001
Missing	12	0	0
Urgency
Planned	12 907	120 (0·9)	4	1·00 (reference)		1·00 (reference)
Emergency	10 865	345 (3·1)	12	3·42 (2·77, 4·21)	< 0·001	4·05 (3·18, 5·15)	< 0·001
Missing	18	1	0
Surgical procedure
Caesarean section	6383	54 (0·8)	1	1·00 (reference)		1·00 (reference)
Hernia repair	6435	33 (0·5)	3	0·61 (0·39, 0·94)	0·024	0·68 (0·41, 1·13)	0·135
Laparotomy	1937	199 (9·3)	6	12·14 (8·95, 16·48)	< 0·001	7·14 (5·06, 10·08)	< 0·001
Appendicectomy	826	8 (1·0)	0	1·14 (0·54, 2·41)	0·722	0·93 (0·44, 2·00)	0·861
Dilatation and curettage	860	6 (0·7)	0	0·82 (0·35, 1·92)	0·655	0·97 (0·41, 2·23)	0·917
Hysterectomy	654	13 (1·9)	0	2·34 (1·28, 4·33)	0·006	3·62 (1·93, 6·79)	< 0·001
Other	6695	153 (2·2)	6	2·70 (1·98, 3·69)	< 0·001	2·78 (1·96, 3·95)	< 0·001
Hospital
Government	6510	67 (1·0)	0	1·00 (reference)		1·00 (reference)
Private non‐profit	17 228	399 (2·3)	16	2·25 (1·73, 2·92)	< 0·001	1·54 (1·16, 2·03)	0·003
Missing	52	0	0

Values in parentheses are

^*percentages and

^†95 per cent confidence intervals.