Primary closure and prophylactic antibiotics for treatment of traumatic wounds caused by mammals, a systematic review and meta-analysis

Abstract

To compare primary closure (PC) with delayed/no closure (DC/NC), and compare prophylactic use of antibiotics (PUA) with no use of antibiotics (NUA) in the treatment of traumatic wounds caused by mammals by a systematic review and meta-analysis. PubMed and Embase databases were searched for eligible randomized clinical trials (RCTs) and observational studies. Qualities of RCTs were assessed according to Cochrane risk of bias tool, qualities of observational studies were assessed according to Newcastle–Ottawa Scale. Primary outcomes included the incidence of wound infection or poor wound healing and the rate of wound cosmesis satisfaction. The relative risks (RRs) of RCTs, odds ratios (ORs) of observational studies and their 95% confidence interval (CI) were extracted directly from included studies or calculated according to the 2 × 2 table obtained by the incidence. The sensitivity analysis, meta-regression and subgroup analysis were performed to identify clinical factors that caused the heterogeneity between studies. Of 26 included studies, 17 studies (8 RCTs and 9 observational studies, 8091 patients) compared PC with DC/NC and 14 studies (7 RCTs and 7 observational studies, 2508 patients) compared PUA with NUA. The pooled OR of all studies (PC versus DC/NC) for wound infection or poor wound healing was 0.79 (95%CI: 0.54, 1.17), the pooled RR of RCTs for wound infection was 0.73 (0.51, 1.06). The pooled OR for cosmesis satisfaction was 3.68 (1.27, 10.68) of 2 studies (PC versus DC) that did not use the negative pressure sealing drainage technique. Subgroup analysis demonstrated that the pooled OR was significant under specific clinical conditions: (1) comparison of PC and DC (pooled OR: 0.49 [0.27, 0.90]), (2) prophylactic use of antibiotics (0.56 [0.33, 0.94]), (3) no use of antibiotics (0.63 [0.41, 0.98]), (4) wounds located in limbs/trunk (0.41 [0.23, 0.73]), (5) time to the first medical presentation (TTP) ≤ 10 h (0.59 [0.39, 0.89]). While the pooled OR (PC versus NC) was not significant (0.84 [0.51, 1.37]). The pooled OR of all studies for wound infection (PUA versus NUA) was 0.73 (95%CI: 0.46, 1.17), the pooled RR of RCTs for wound infection was 0.81 (0.46, 1.44). No included studies (PUA versus NUA) reported the outcome of wound cosmesis. Subgroup analysis demonstrated that the pooled OR was significant under specific clinical conditions: (1) injury caused by other type of mammals other than dog (pooled OR: 0.24 [0.06–0.98]), (2) wounds located in face/head (0.13 [0.03, 0.52]). Regardless of whether prophylactic antibiotics are used or not, compared to delayed closure, primary closure should be given priority in treating traumatic wounds caused by mammals which can decrease the incidence of wound infection or poor wound healing and obtain the better wound cosmesis, but it does not show the superiority compared to no closure, unless under some specific clinical conditions. Prophylactic use of antibiotics may not benefit in prevention of wound infection unless under specific clinical conditions, such as wounds caused by mammals other than dogs or wounds located in face/head.