Scandinavian Journal of Surgery 99: 162 166, 2010 Incidence and risk factors of surgical wound infection in children: a prospective study K. Varik, Ü. Kirsimägi, E.-A.Värimäe, M. Eller, R. Lõivukene, V. Kübarsepp Tartu University Department of Surgery, Tartu, Estonia Abstract Background and Aims: To establish the incidence and risk factors of surgical site infection (SSI) in children operated in the Department of Paediatric Surgery of the Clinic of Surgery of Tartu University Hospital. Material and Methods: The data of wound healing were obtained for 589 children operated between 15 March 2003 and 31 March 2005. The operations were divided into general surgical (451), orthopaedic (70) and urological (68). The surgical wounds were classified as clean (442), clean-contaminated (96), contaminated (36) and dirty-infected (15). Univariate and multivariate analyses were performed to identify risk factors. Results: There were 7 SSI cases, overall rate being 1.2%. Superficial wound infection occurred in 5 cases and deep wound infection occurred in 2 cases. There was no organ/ space infection. SSI was significantly more frequent in the case of contaminated and dirty-infected compared with clean or clean contaminated operations, 7.8% and 0.6%, respectively (p = 0.0008). Wound infection endangered more children who had operation related complications compared with non-complicated cases, 11.1% and 0.4%, respectively (p < 0.0001). Key words: Surgery; surgical site infection; wound infection; clean operations; clean-contaminated operations; contaminated operations; dirty operations; SSI risk factors; operation related complications; children Introduction The incidence of Surgical Site Infection (SSI) is an important indicator of surgical outcome, particularly in paediatric surgery. Wound infection and wound related complications are common causes of lengthening of hospital stay. According to literature data, the incidence of postoperative wound infection in children varies significantly, from 1.6 to 18.7% (1 5), and even up to 27% for contaminated operations (6) and up to 30% 40% for dirty-infected operations (7, 8). The incidence of SSI depends on wound class (3). The risk for development of wound infection is higher in emergency operations and operations lasting lon- Correspondence: Karin Varik, M.D. Tartu University Department of Surgery L. Puusepa 8, 51014 Tartu, Estonia Email: karin.varik@kliinikum.ee ger than 1 hour (8), as well as in operations in newborns and in operations where open drains were used (6). At the same time, available literature offers no specific analysis of the incidence of wound infection depending on operation related complications (5, 9). The aim of this study was to establish the incidence and risk factors of SSI in children operated in the Department of Paediatric Surgery of the Clinic of Surgery of Tartu University Hospital. No similar study has been conducted earlier in Estonia or in the Baltic countries. Material and Methods In cooperation with the surgeons (4 paediatric surgeons and 1 resident) of the Department of Paediatric Surgery of the Clinic of Surgery of Tartu University Hospital and the nurses of the paediatric operating theatre of the Surgical Service, we conducted a prospective study on the incidence of wound infection in children operated between 15.03.2003 and 31.03.2005.
Surgical wound infection in children 163 Data collection The study included paediatric inpatients requiring skin incision, who were operated in the operating theatre of paediatric surgery under general or epidural anaesthesia. The data of the patients, operation and postoperative course were recorded. Main focus was placed on operation urgency, wound class, antibiotic use (for prophylaxis and for treatment), operation related complications (surgical site haematoma, skin damage due to diathermia, skin necrosis, etc). Also child age and gender, diagnosis, length of hospital stay, type of operation, drain usage, etc. were recorded. Wound healing was followed during the whole hospital stay. Wound data were obtained from 90 patients during their hospital visit on day 7 and from 499 previously discharged patients on days 7 and 30 during an outpatient visit or using a phone interview, according to the protocol, conducted by one author (KV). In case there remained suspicion about wound infection after the telephone interview the wound was checked in the clinic by KV. The protocol was started for 613 operated children; the data of wound healing were obtained from 589 (96%) children enrolled in the study. Follow-up contact failed in 24 patients. There were no deaths during the study period. Methods The operations were divided, according to the specialities in our department, into general surgical, orthopaedic and urological. General surgical operations were grouped as operations of the inguinal region (inguinal hernia repair, hydrocelectomy, orchidopexy), abdominal operations (appendectomies, bowel resections, removal of abdominal cavity tumours, operations for congenital malformations of the intestinal tract) and other operations (umbilical hernia repair, removal of benign tumours of the skin and subcutaneous tissue, plastic operations). Among orthopaedic operations, more frequent were different osteotomies and removals of various benign tumours; among urological operations prevailing were hypospadia operations, pyeloplasties and antireflux operations. Regarding concomitant diseases, asthmatic bronchitis occurred in 12 cases and allergy to different agents occurred in 11 cases; there were no high-risk patients. Surgical wounds were classified as class I clean, class II clean-contaminated, class III contaminated and class IV dirty-infected according to the Centres for Disease Control (CDC). For classifying wound infections occurring within 30 days after operation, we used the CDC revised definition of nosocomial SSI: superficial incisional SSI infection involves only the skin or subcutaneous tissue of the incision; deep incisional SSI-infection involves the deep soft tissues (e.g., fascial and muscle layers) of the incision; organ/space SSI-infection involves any part of the anatomy (e.g., organs or spaces) (10, 11). Statistical analysis The mean values were compared with the Student t-test or the Mann-Whitney U test as appropriate. The association between the categorical variables were analysed by the chisquare test. The odds ratios (OR) with 95% confidence intervals (95% CI) were calculated to assess the relationship between the potential risk factors and wound infection. Logistic regression analysis was used to find independently associated risk factors for SSI. The level of significance was set at p < 0.05, all p-values were two-sided. Statistical analysis was done using the statistical package Statistica 7.0. Results In 589 prospectively studied children, wound infection occurred in 7 cases, overall SSI rate being 1.2% (95% CI 0.5 2.4). Superficial SSI occurred in 5/7 (71.4%) cases and deep SSI occurred in 2/7 (28.6%) cases; there was no organ/space SSI. All cases of wound infection were detected during hospital stay. All studied SSI cases are presented in Table 1. Wound infection appeared on average on day 3.3 (range 2 5). The mean hospital stay for patients with wound infection was 12.6 days longer than for other patients, 6.8 and 3.1 days, respectively (p < 0.0001). Suppurent secretion from the wound occurred in 3 children, among whom E. coli was detected in microbiological cultures in two cases and Staphylococcus aureus in one case. Both children with deep wound infection had an increase in body temperature and CRP, and one of them required incision of subfascial abscess following appendectomy. For treatment of wound infection, antibiotics were used according to sensitivity. The results of the study are presented in Table 2. The number of clean and clean-contaminated operations was 538 (91.3%). SSI was significantly more frequent in contaminated or dirty-infected operations compared with clean or clean-contaminated operations, 7.8% and 0.6%, respectively (p = 0.0008). Among the 98 (16.6%) children who underwent an emergency operation, the incidence of wound infection was significantly higher compared with the 491 children who underwent an elective operation, 4.1% Table 1 Cases of wound infection. Operation Emerg Wound Complic SSI Microbiology Signs Hospital class of infection stay Separation of fingers I yes S Day 3 32 Scar excision I yes S Day 3 17 Appendectomy yes II D E. coli Day 5 10 Saliva fistula closure III yes S Day 1 15 Wound toilet yes III yes D Staph.aureus Day 3 10 Appendectomy yes IV S E. coli Day 4 17 Appendectomy yes IV yes S Day 3 09
164 K. Varik, Ü. Kirsimägi, E.-A.Värimäe, M. Eller, R. Lõivukene, V. Kübarsepp Table 2 Results of the study. No of operations Wound infection (%) (%) n = 589 n = 7 Gender Boys 396 (67.2) 3 (10.0) Girls 193 (32.8) 4 (1.6) Age < 1 year 057 (9.7) 1 (1.8) 1 4 years 208 (35.3) 2 (1.0) 5 13 years 276 (46.9) 3 (1.1) 14 years 048 (8.1) 1 (2.1) Urgency Elective 491 (83.4) 3 (0.6) Emergency 098 (16.6) 4 (4.1) Wound class Clean 442 (75.0) 2 (0.5) Clean-contaminated 096 (16.3) 1 (1.0) Contaminated 036 (6.1) 2 (5.3) Dirty 015 (2.6) 2 (13.3) Duration of operation 30 min. 389 (66.1) 0 31 60 min 135 (22.9) 4 (3.0) 61 120 min 055 (9.3) 1 (1.8) > 120 min 010 (1.7) 2 (20.0) Clothing Textile 494 (83.9) 6 (1.1) Disposable 095 (16.1) 1 (1.2) Antibiotic usage Yes 117 (19.9) 4 (3.4) No 472 (80.1) 3 (0.6) Operation related complications Yes 045 (7.6) 5 (11.1) No 544 (92.4) 2 (0.4) and 0.6%, respectively (p = 0.009). The duration of operation was significantly longer for patients with SSI compared with the other patients, 82.9 and 32.8 min, respectively (p < 0.0001). Antibiotics were used in 117 patients and wound infection occurred in 4 (3.4%) of these patients. Antibiotics were not used in 472 children among whom wound infection occurred in 0.6% of all the cases. The overall number of operation related complications was 45 (7.6%). The most frequent complications were haematoma at the wound site, with 14 cases; hyperaemia at the wound site, caused by reaction to the suture material, or skin lesion due to diathermia, with 13 cases; secondary wound bleeding, with four cases; skin necrosis following plastic surgery, with two cases. The rate of SSI was higher among patients who had operation related complications compared with non-complicated cases, 11.1% and 0.4%, respectively (p < 0.0001). The incidence of wound infections, operation related complications and urgency according to the type of operation are presented in Table 3. It is evident that SSI are most frequent in abdominal operations versus all other operations, 3.8% vs 0.8%, respectively (p =0.031). At the same time, abdominal operations are significantly different from other operations in terms of urgency (89.9% vs 5.3%; p < 0.0001) and surgical wound class (classes III and IV occurred in 55.7% of the cases [p < 0.0001]). According to univariate analysis, the factors shown to be associated with SSI were the following: wound class, urgency, duration of operation, drain usage, antibiotic usage, operation related complications (Table 4). Among the 10 cases when the operation was longer than 120 min superficial SSI occurred in two cases: in one case it was a correction of an atypical position of the appendix and in the other case it was a procedure of separation of fingers. Patient age, gender, surgical clothing and surgeon versus trainee status did not affect the incidence of SSI. In multivariate logistic regression analysis, only wound class and operation related complications were independently associated with SSI: for contaminated and dirty-infected wounds OR = 11.0 (95% CI = 2.12 56.5; p = 0.0042) and for occurrence of operation related complications OR = 26.9 (95% CI = 4.8 150.7; p = 0.0002). The other factors that proved significant in univariate analysis (urgency, duration of operation, drain usage, antibiotic usage) remained nonsignificant as they are strongly associated with wound class. It should be noted that operations in the case of contaminated or dirty-infected wounds are characterised by high urgency (84.3%), high antibiotic usage (82.4%), high drain usage (33.3%) and longer duration (15.7% lasting > 1 h). Table 3 SSI rate and complications for different operations. Operation Number of Emergency Duration of Wound class Drain No of No of operations (%) op (median) III IV (%) usage complications infections (%) (%) General surgical abdominal* 079 71 (89.9) 35 44 (55.7) 15 05 (6.3) 3 (3.8) inguinal 230 05 (2.2) 20 00 00 12 (5.2) 0 other 142 05 (3.5) 15 04 (4.2) 03 13 (9.2) 3 (2.1) Orthopaedic 070 07 (10.0) 30 00 06 04 (5.7) 1 (1.4) Urological 068 10 (14.7) 60 01 (1.5) 37 11 (16.2) 0 Total (%) 589 98 (16.6) 49 (8.7) 61 45 (7.6) 7 (1.2) * 62 appendectomies, incl 6 perforative appendectomies
Surgical wound infection in children 165 Discussion Table 4 Association of variables with SSI in univariate analysis. Variable or 95% CI p value Gender (girls) 1.6 0.3 7.6 0.5800 Age (< 1 year) 1.6 00.2 10.8 0.2130 Operating trainee 1.0 0.0 7.0 > 0.9990 Clothing (textile) 1.2 00.2 27.0 > 0.9990 Emergency 6.9 01.4 37.4 0.0190 Wound class (III, IV) 15.0 03.0 82.6 0.0020 Duration of operation (> 60 min) 6.3 01.1 30.9 0.0360 Antibiotic usage 5.5 01.1 29.9 0.0360 Drain usage 6.8 01.2 33.4 0.0300 Complications 33.3 006.4 254.6 < 0.0001 The present study revealed one of the lowest reported incidences of SSI in children in all wound classes. Our study showed that the overall rate of SSI was 1.2%; for the clean and clean-contaminated operations it was 0.7% which is lower than in previously reported paediatric patients. The low overall rate of SSI and the rate of SSI in clean operations are obviously related to the facts that 83.4% of the patients in the present series underwent an elective operation and that 91.3% of the operations were clean or cleancontaminated. In our study SSI were most frequent for dirty-infected operations (13.3%), which is consistent with literature data according to which the rate of SSI for dirty-infected operations is 10 40.7% (5, 8, 12, 13). However, van Griethuysen et al. and Horwitz et al. reported that infection rate was lower for dirty operations than for contaminated operations (10, 14). These studies did not find any reason for this, but Bhattacharyya et al. explained the higher infection rate for contaminated operations with the fact that they did not close one-third of surgical wounds primarily in the case of dirty-infected operations (1). In presence of SSI, the hospital stay of patients increases significantly, in our study by 12.6 days, which is 2.5 days longer than the hospital stay in a study by Cruse (9). Although most studies report longer hospital stay in the case of SSI, the cause of such lengthening is not specified (13, 15). According to many studies, the duration of operation increases the rate of SSI (1, 3, 8, 12). Direct association between the duration of operation and higher rate of SSI was pointed out by Cruse et al., who showed that infection rate increases almost two times by every hour of operation (9). In the present study, the risk of wound infection was 6.3 times higher for operations lasting more than 1 hour. However, the duration of operation did not prove significant in multivariate analysis, where only wound class and operation related complications remained significant risk factors for SSI. Our findings are in agreement with those of Bhattacharyya et al. that SSI occurs most commonly in patients who have undergone surgery of the gastrointestinal tract; in their study wound infection occurred in 4.1% of the operations of the abdominal cavity, which was attributed to wound contamination by gut content (1). In our study the largest number of wound infections (3.8%) occurred in patients who underwent surgery of the abdominal cavity. Most of these operations involved opening of the gastrointestinal tract, as a result of which 55.7% of them were contaminated or dirty-infected, mostly emergency operations (89.9%). Data about wound drains and SSI in the literature are scarce. Only Porras-Hernandez et al. showed in their study, using multiple regression analysis, that the use of open drains was associated with SSI (5). In our study univariate analysis showed that drain usage increases the risk of SSI; however, in multivariate analysis this factor proved nonsignificant, which is related to the higher use of drains in contaminated or dirty-infected cases. Opinions about antibiotic use are discordant. Porras-Hernandez et al. and Fletcher et al. showed that antibiotic prophylaxis, although marginally significant, served as a protective factor against SSI (5, 12). Uludag et al. pointed out that despite the use of antibiotic prophylaxis, the rate of wound infection for dirty-infected operations was still 30% (8). We agree with Horwitz et al. who demonstrated the absence of association between preoperative antibiotic administration and SSI (10). In the present study, the rate of wound infection was higher when antibiotics were used. This is due to the fact that antibiotics were used more in contaminated and dirty-infected cases where the risk of SSI is higher. No data are available about the association of operation related complications (haematoma, skin damage due to diathermia, reaction to suture material) and wound infections. In the present study, the rate of SSI was 7.64% for complicated cases. In conclusion, according to our data the risk factors for wound infection were wound class and operation related complications. Acknowledgements We are grateful to Risto Rintala, Professor of Pediatric Surgery of the Hospital for Children and Adolescents of the University of Helsinki, for assistance in completing this paper. References 01. Bhattacharyya N, Kosloske AM: Postoperative wound infection in pediatric surgical patients: a study of 676 infants and children. J Pediatr Surg 1990;25:125 129 02. Brown SM, Eremin SR, Shlyapnikov SA, Petrova EA et al: Prospective surveillance for surgical site infection in St. Petersburg, Russian Federation. Infect Control Hosp Epidemiol 2007;28:319 325 03. Davis SD, Sobocinski K, Hoffmann RG, Mohr B, Nelson DB: Postoperative wound infections in a children s hospital. Pediatr Infect Dis 1984;3:114 116 04. Merei JM: Pediatric clean surgical wounds: is dressing necessary? J Pediatr Surg 2004;39:1871 1873 05. Porras-Hernandez JD, Vilar-Compte D, Cashat-Cruz M, et al: A prospective study of surgical site infections in a pediatric hospital in Mexico City. Am J Infect Control 2003;31:302 308
166 K. Varik, Ü. Kirsimägi, E.-A.Värimäe, M. Eller, R. Lõivukene, V. Kübarsepp 06. Sharma LK, Sharma PK: Postoperative wound infection in a pediatric surgical service. J Pediatr Surg 1986;21:889 891 07. Sangrasi AK, Leghari AA, Memon A et al: Surgical site infection rate and associated risk factors in elective general surgery at public sector medical university in Pakistan. Int Wound J 2008;5:74 78 08. Uludag Ö, Rieu P, Niessen M, Voss A: Incidence of surgical site infections in pediatric patients: a 3-month prospective study in an academic pediatric surgical unit. Pediatr Surg Int 2000;16: 417 420 09. Cruse PJE, Foord R: The epidemiology of wound infection: a 10-year prospective study of 62.939 wounds. Surg Clin North Am 1980; 60:27 40 10. Horwitz JR, Chwals WJ, Doski JJ et al: Pediatric wound infections: a prospective multicenter study. Ann Surg 1998;227:553 558 11. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR: Guideline for prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol 1999;20:250 278 12. Fletcher N, Sofianos D, Berkes MB, Obremskey WT: Prevention of perioperative infection. J Bone Joint Surg Am 2007; 89:1605 1618 13. Mishriki SF, Law DJ, Jeffery PJ: Factors affecting the incidence of postoperative wound infection. J Hosp Infect 1990;16:223 230 14. van Griethuysen AJ, Spies-van Rooijen NH, Hoogenboom- Verdegaal AM: Surveillance of wound infections and a new theatre: unexpected lack of improvement. J Hosp Infect 1996; 34:99 106 15. Sparling KW, Ryckman FC, Schoettker PJ, et al: Financial impact of failing to prevent surgical site infections. Qual Manag Health Care 2007;16:219 225. Received: December 22, 2009 Accepted: February 11, 2010