Skip to main content
. 2021 Nov 6;398(10312):1687–1699. doi: 10.1016/S0140-6736(21)01548-8

Table 4.

Secondary outcomes for skin preparation type by treatment group and wound classification stratum, in randomised patients who had abdominal surgery

Clean-contaminated stratum (n=3051)
Contaminated or dirty stratum (n=2662)
Chlorhexidine (n=1521) Povidone–iodine (n=1530) Treatment effect (RR, 95% CI) Chlorhexidine (n=1337) Povidone–iodine (n=1325) Treatment effect (RR, 95% CI)
SSI at discharge
Yes 93 (6·4%) 94 (6·4%) 0·99 (0·76–1·30) 229 (19·7%) 247 (21·4%) 0·93 (0·80–1·08)
No 1367 (93·6%) 1384 (93·6%) .. 936 (80·3%) 905 (78·6%) ..
Missing 61 52 .. 172 173 ..
Reoperation for SSI within 30 days of surgery
Yes 10 (0·7%) 6 (0·4%) 1·58 (0·58–4·28)* 38 (3·2%) 38 (3·3%) 0·97 (0·63–1·49)
No 1441 (99·3%) 1459 (99·6%) .. 1143 (96·8%) 1118 (96·7%) ..
Missing 70 65 .. 156 169 ..
Mortality within 30 days of surgery
Yes 40 (2·7%) 28 (1·9%) 1·44 (0·90–2·31) 121 (9·3%) 119 (9·3%) 1·00 (0·79–1·27)
No 1449 (97·3%) 1467 (98·1%) .. 1178 (90·7%) 1155 (90·7%) ..
Missing 32 35 .. 38 51 ..
Unplanned wound opening within 30 days of surgery
Yes 211 (14·5%) 216 (14·7%) 0·98 (0·84–1·15) 261 (21·9%) 288 (24·7%) 0·90 (0·79–1·03)
No 1246 (85·5%) 1252 (85·3%) .. 929 (78·1%) 876 (75·8%) ..
Missing 64 62 .. 147 161 ..
Length of hospital stay for index admission (days)
Median (IQR) 5·0 (4·0–8·0) 5·0 (4·0–8·0) 1·01 (0·94–1·09) 8·0 (5·0–12·0) 8·0 (5·0–13·0) 1·08 (1·00–1·18)
Missing 64 51 .. 172 174 ..
Readmission within 30 days of surgery
Yes 29 (2·0%) 40 (2·7%) 0·72 (0·45–1·15) 53 (4·5%) 60 (5·2%) 0·88 (0·62–1·26)
No 1423 (98·0%) 1427 (97·3%) .. 1129 (95·5%) 1098 (94·8%) ..
Missing 64 52 .. 155 167 ..
Return to normal activities within 30 days of surgery
Yes 594 (40·9%) 597 (40·7%) 1·01 (0·94–1·07) 364 (30·8%) 349 (30·2%) 1·00 (0·91–1·10)
No 859 (59·1%) 870 (59·3%) .. 818 (69·2%) 808 (69·8%) ..
Missing 68 63 .. 155 168 ..

Data are number (%), unless otherwise indicated. Percentages presented ignore any missing data. RR=relative risk. SSI=surgical site infection.

*

Age removed from model as model did not converge due to multicollinearity.

Data are hazard ratio (95% CI).

Yes=completely returned to normal activities.