Continuous variables are reported as medians and proportions as percentages with 95% confidence intervals (CIs). Group comparisons were performed using Fisher’s exact test for categorical data and exact logistic regression for continuous variables. Significance level was set at 0.05 with an odds ratio and number needed to harm extrapolated from the data sets.
Over a 12-month period, 723 patients met inclusion criteria with 85 lost to follow-up with a remaining cohort of 638 patients, aged between 3 months and 95 years. The postoperative infection rate for the patients treated within 24 hours was 1.8% (8 of 429 patients) as compared with 5.3% (11 of 209) in those treated over 24 hours after presentation. Using chi-square analysis, results demonstrated that patients managed over 24 hours after presentation have a statistically significant increased rate of infection with a P value of 0.024, an odds ratio of 2.9 (95% CI, 1.2–7.1) and a NNH of 29. Using a logistic regression model, the time to operating theater was not a significant predictor for infection (z = 1.028, P = 0.304).
The median age of the patients was 34 years, and 79% were male (n = 502). The median age of patients who sustained a postoperative infection was 30 years. The shortest time to surgery was 1.7 hours, and the longest time to surgery was 95 hours with an average time to surgery of 21 hours. All patients sustained open hand injuries.
Of the 19 patients who developed an infection postoperatively, 2 had a significant risk factor for increased likelihood of developing an infection, that being type 2 diabetes mellitus with one from each group. Table 1 demonstrates the mechanisms of injury in the 19 patients who developed a postoperative infection. Their histories were reviewed, and no other contributing factors such as cardiovascular incident, comorbidities, or other life-threatening events were identified to have contributed to a delay to operative management or poor wound healing. Of these patients, 14 of these patients were treated successfully with oral antibiotics in the outpatient setting, 2 required readmission for intravenous antibiotics, and 3 patients requirement surgical debridement.
Analysis of 401 patients who sustained hand injuries requiring a washout with simple debridement, primary repair, and/or tendon repair, excluding any patient who had a neurovascular bundle repair, any form of graft, k-wire insertion, or plating/metalwork to repair a fracture demonstrated that 2 of the 277 patients treated within 24 hours developed an infection (0.7%) compared with 6 of the 118 patients treated over 24 hours after presentation (5.1%). This demonstrated an increased risk of infection with a P value of 0.012 and an odds ratio of 7.0 (95% CI, 1.4–35.1); however, logistic regression analysis did not demonstrate a significant risk of infection with a coefficient of 0.02195 (z = 1.219, Pr (>|z|) = 0.223).
Given these differences between chi-square and logistic regression models, an exact logistic regression model with 45,000 Monte Carlo iterations and time grouped into 1 hour intervals that gave a significant result, with an estimated co-efficient of 0.01194 (95% CI, -0.01988, 0.04132) and P value of 0.01564 (Monte Carlo n = 45,000), which was statistically significant.
This study reignites the debate of the existence of a “Golden Period.” Where the period within which open hand wounds should be managed was previously open ended (Angly et al., 2012, McLain et al., 1991; Zehtabchi et al., 2012), this study highlights that there may be a critical time within which hand trauma wounds should be surgically managed. Delayed treatment of open hand injuries may occur for reasons such as difficulty accessing centers with a hand surgery unit, bed allocation constraints, limitations of operating room availability, and other comorbidities or injuries that require stabilization prior surgery (Angly et al., 2012; Juon et al., 2014). Review of the literature since the introduction of antibiotic therapy suggests that the existence of a critical time period for repair is not proven; however, these conclusions are drawn from patients treated within a 24-hour period (Hollander et al., 2001; Zehtabchi et al., 2012). No study as yet has reliably determined a time limit for wound age beyond which the infection rate would increase significantly.
Despite thorough review of individual patient medical histories for factors which may contribute to the risk of infection, limitations such as incomplete medical records are a limitation of retrospective cohort analyses. There was no ability to randomize patients nor standardize preoperative and postoperative care; however, the unit policies for the management of hand trauma were followed for all patients. These included intravenous antibiotics before and on induction of surgery and a postoperative course of oral antibiotics, tetanus status was checked and updated if required, all wounds preoperatively were irrigated and redressed using sterile technique in the principle hospital where the operation was undertaken and all patients were followed up 7–14 days postoperatively in the principle hospital outpatient clinic. A small group of 12 patients were excluded as their care deviated from the standard unit management. Factors such as smoking status, completion of postoperative course of oral antibiotics, dressing type and redressing regime, level of wound irrigation, care rendered in the emergency department, or in other outside facilities could also be considered in future research. Larger cohort studies could further assess specific patient and treatment variables for incorporation into a multivariate analysis of patients sustaining a postoperative wound infection to further enhance these findings and analyze for contributing factors. Specific type of wound infection using microbial analysis, severity of infection, standardization of wound dressings, operative time duration, length of hospital stay, and hardware use could also be analyzed in future prospective studies.
A variety of hand wounds were considered in this research, including animal and human bites. Patients were triaged in order of presentation, with children under the age of 16 the only exception to this triaging rule. Patients with animal and human bite injuries were therefore not excluded from this analysis as they were not given operative theater time priority. Using 3 different hospital sites increased the population diversity of captured patients making the extrapolation of these results more applicable to the greater metropolitan area; however, further studies capturing patients from regional and rural settings would increase the strength of these results and the sample size.
Postoperative infections are a financial burden to the healthcare system with increased costs of readmission, prolonged hospital stay, potential return to theater, use of hospital in the home/community nursing resources, and increased outpatient follow-up time. Additionally, there is a financial burden on public funding with postoperative infections delaying return to work and the long-term effects of increased scarring, chronic pain, and subsequent impaired joint mobility.
It is noted that the logistic regression model differed from the chi-square results and therefore we examined different cutoffs for bucketing the “Time to OT” predictor variable and found an additional significant cutoff at 6 hours, in addition to 24 hours as detailed earlier. This raised the question of whether the underlying distribution of the predictor variable, in addition to the relative rarity of the outcome variable, was causing the standard logistic regression to give a nonsignificant result (Fig. 1). Given this, we used an exact logistic regression model with 45,000 Monte Carlo iterations and time grouped into 1 hour intervals that gave a significant result. Future larger cohort studies could further assess this eliminating the need for exact logistic regression models.
There appears to be some evidence of a higher risk of postoperative infection in hand trauma patients surgically managed more than 24 hours after hospital presentation; however, when time is analyzed as a continuous variable, there is limited statistical significance. With increasing demands on the public healthcare system, patients are facing longer than 24 hour delays to surgery and to date there is a lack of research studying this time delay effect. This research appears to have implications for surgical hand wound management and recognizes time as an important factor perhaps overlooked in the management of open hand wounds. Further evaluation of this outcome should be considered with a larger cohort and standardization of perioperative management. This is the first study to raise the possibility of a double peak for infection risk giving basis for a larger prospective study which accounts for confounding variables to further assess the contention that a critical time period exists in the management of open hand wounds.
1. Angly B, Constantinescu MA, Kreutziger J, et al. Early versus delayed surgical treatment in open hand injuries: a paradigm revisited. World J Surg. 2012;36:826–829.
2. Hollander JE, Singer AJ, Valentine SM, et al. Risk factors for infection in patients with traumatic lacerations. Acad Emerg Med. 2001;8:716–720.
3. Juon BH, Iseli M, Kreutziger J, et al. Treatment of open hand injuries: does timing of surgery matter? A single-centre prospective analysis. J Plast Surg Hand Surg. 2014;48:330–333.
4. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of surgical site infection, 1999. Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol. 1999;20:250–278; quiz 279.
5. McLain RF, Steyers C, Stoddard M. Infections in open fractures of the hand. J Hand Surg Am. 1991;16:108–112.
6. Morgan WJ, Hutchison D, Johnson HM. The delayed treatment of wounds of the hand and forearm under antibiotic cover. Br J Surg. 1980;67:140–141.
7. Murphy GR, Gardiner MD, Glass GE, et al. Meta-analysis of antibiotics for simple hand injuries requiring surgery. Br J Surg. 2016;103:487–492.
8. Quinn JV, Polevoi SK, Kohn MA. Traumatic lacerations: what are the risks for infection and has the ‘golden period’ of laceration care disappeared? Emerg Med J. 2014;31:96–100.
9. Srour M, Inaba K, Okoye O, et al. Prospective evaluation of treatment of open fractures: effect of time to irrigation and debridement. JAMA Surg. 2015;150:332–336.
11. van den Baar MT, van der Palen J, Vroon MI, et al. Is time to closure a factor in the occurrence of infection in traumatic wounds? A prospective cohort study in a Dutch level 1 trauma centre. Emerg Med J. 2010;27:540–543.
Copyright © 2018 The Authors. Published by Wolters Kluwer Health, Inc. on behalf of The American Society of Plastic Surgeons.
12. Zehtabchi S, Tan A, Yadav K, et al. The impact of wound age on the infection rate of simple lacerations repaired in the emergency department. Injury. 2012;43:1793–1798.