Effect of Skin Antisepsis After Skin Closure on Wound Infection After Abdominal Surgery for Sepsis: a Preliminary Report

The beneficial role of this pre-operative skin preparation has been fully established in several studies, persisting debate being essentially limited to which agent or combination of agents provides superior antiseptic benefit.

However, following skin closure after abdominal surgery, and cleaning the wound edges and surrounding skin of blood and other tissue fluid with a soapy antiseptic agent, some surgeons clean the apposed skin edge with an antiseptic solution, either isopropyl alcohol (more commonly) or povidone-iodine, before application of sterile dressings. The effects of this practice after skin closure in terms of post-operative wound outcome remains to be established. This was a prospective randomized controlled study conducted at the University College Hospital Ibadan, Nigeria. Study participants were recruited from adult patients presenting through the Surgery Out-Patient (SOP) department, Emergency departments and non-surgical wards of University College Hospital, Ibadan. Assuming a standard normal deviate at 5% type I error of 1.96, power of 80%, an SSI prevalence of 38.1% in the control group10 with 15% degree of precision, a total sample size of 84 patients were obtained, having considered an attrition rate of 5%. This preliminary report is based on 37 patients recruited so far.

. Study participants were randomized into 2 groups: povidone iodine (PI) \& Control (C) groups using blocked sequence randomization. Consecutive adult surgical patients (aged 18 years and above) booked to have abdominal surgery for sepsis were prospectively enrolled in turns into the two study groups.

Shaving of abdominal hair was done for all patients just before the surgery using a surgical blade. They were given a single dose each of intravenous ceftriaxone (1g; Zonon; Sanofi;) and metronidazole (500mg; Metrone; Aventra) at the induction of anaesthesia. After surgery, patients with contaminated wounds were given intravenous ceftriaxone and metronidazole, discontinued after 24hours post-operatively. Patients with dirty wounds received therapeutic doses of post-operative antibiotics intravenously (iv ceftriaxone 1 g 12hourly and iv metronidazole 500mg 8hourly) for 10-14days. This was converted to oral antibiotics (cefpodoxime 200mg 12hourly and metronidazole 500mg 8hourly) once oral intake was established.

Skin preparation was done as follows: scrubbing of the operation site twice over 3-5 minutes using savlon (0.3% chlorhexidine-gluconate in alcohol + 3% cetrimide) was followed by drying of the skin with a sterile gauze. The operation site was painted with 10% povidone iodine followed by 70% isopropyl alcohol before draping. A midline incision that provides adequate exposure was made using a scalpel and deepened using monopolar diathermy. Upon gaining peritoneal access, a wound swab was taken from the focus of peritoneal contamination/sepsis. The surgical procedures relevant for each case was carried out.

Having closed the skin and cleaned with savlon and 70% isopropyl alcohol, patients in the C group had the apposed skin edge covered with dry sterile gauze while in PI group, 10% povidone iodine-soaked gauze dressing was used to cover the apposed skin edge. The surgical site was assessed on the 3rd, 7th and 10th post-operative day for evidence of SSI, defined in this study as purulent drainage from the surgical wound or a drain inserted at surgery.

A swab was taken from the wound edge of all patients with a clinical feature of SSI using a sterile swab stick for microscopy, culture and sensitivity. The type of intervention, type and duration of post-operative antibiotics and other relevant information were documented in an identifier-free patient's proforma. The total duration of hospital stay, calculated from the first post-operative day to the day of discharge, was recorded.

Comparison of groups for surgical site infection was done using the Pearson's Chi-square (X2) test. Multivariate logistic regression analysis was used to test for significance of association between potentially confounding variables and the development of SSI. Statistical significance was set at a p-value of <0.05 and Confidence Interval of 95%. Version 23 of the Statistical Package for Social Sciences for Windows (SPSS Inc. Il, USA) was used to analyze all data obtained from the study