Accepted Manuscript

Title: Randomized, Controlled Trial of Povidone-Iodine to

Reduce Simple Traumatic Wound Infections in the Emergency
Department

Author: Hamed Basir Ghafouri Morteza Zare Azam

Bazrafshan NiloofarAbazarian Tayeb Ramim

PII: S0020-1383(16)30203-0
DOI: http://dx.doi.org/doi:10.1016/j.injury.2016.05.031
Reference: JINJ 6744

To appear in: Injury, Int. J. Care Injured

Received date: 16-11-2015

Revised date: 1-5-2016
Accepted date: 20-5-2016

Please cite this article as: Ghafouri Hamed Basir, Zare Morteza, Bazrafshan Azam,
NiloofarAbazarian , Ramim Tayeb.Randomized, Controlled Trial of Povidone-Iodine
to Reduce Simple Traumatic Wound Infections in the Emergency Department.Injury
http://dx.doi.org/10.1016/j.injury.2016.05.031

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Randomized, Controlled Trial of Povidone-Iodine to Reduce Simple Traumatic Wound
Infections in the Emergency Department

Authors: Hamed Basir Ghafouri1 , Morteza Zare2 , Azam Bazrafshan3 , NiloofarAbazarian5* , Tayeb
Ramim6
1
MD, Assistant Professor of Emergency Medicine , Emergency Medicine Management Research
Center, Iran University of Medical Sciences, Tehran, Iran.
2
MSc, Neuroscience Research Center, Institute of Neuropharmacology, Kerman University of
Medical Sciences, Kerman, Iran
3
MSc, Neuroscience Research Center, Institute of Neuropharmacology, Kerman University of
Medical Sciences, Kerman, Iran
4
MD, Sina Trauma and Surgery Research Center, Tehran University of Medical Science, Tehran, Iran
(CORRESPENDING AUTHOR) abzarian.niloofar@gmail.com, Telephone: +983432263855
5
MD, Sina Trauma and Surgery Research Center, Tehran University of Medical Science, Tehran, Iran

Abstract

Introduction

Povidone Iodine (PVI) has been used to prevent wound infection for a long time, yet the merits and
effectiveness of this agent in reducing the rates of infection in simple traumatic wounds have been
debated. The aim of this study is determine the effect of PVI as skin disinfectant in preventing simple
traumatic wound infection after repair in emergency departments.

Methods

This study is a single-blind, randomized, controlled trial, conducted at the emergency departments
of two university hospitals. In this trial, those collecting and analyzing data were unaware of the
assigned treatment. Participants included ED patients with simple traumatic wounds, randomly
assigned to either experiment or control groups. Wounds were similarly irrigated with normal saline
in both groups. But the skin around the wounds of patients in the experiment group was cleansed by
PVI solution. The patients were followed till their next ED visit for stitch removal. The presence or
absence of wound infection was considered as the primary outcome and assessed by an ED
physician.

Results

In total, 444 patients were screened and included in the study. The patients were statistically similar
according to their baseline characteristics. Men constituted 85% of the study participants. The
highest numbers of reported wound sites belonged to head (30%), lower limbs (24%) and upper
limbs (19%) accordingly. The rates of infections in the experiment and control groups were similar
(p=0.231) but a significant association was observed between the rate of infections and wound
location. Lower limbs (OR=9.23, p<0.0001) and upper limbs (OR=5.47, p=0.011) indicated the highest
risks of wound infections among other locations.

Conclusion

Our study showed that using PVI in the management of traumatic wounds did not reduced rate of
infections.

Key Words: Povidone-Iodine, Wound Healing, Wounds and Injuries, Emergency Service

Introduction

Wounds and lacerations are common problems in the emergency departments (EDs). Restoring
integrity and function with minimal scar are therefore the most important goals in emergency
wound management .To achieve the above goals, it is important to prevent wound infection which
can be achieved by proper wound preparation before repair(1, 2). According to previous studies
wound infection occurs in approximately 3-5% of traumatic wounds that are repaired in EDs;
however the infection rate is varied based on some factors such as mechanism of injury, wound
location and patients health status (3). As we know wound irrigation and debridement are most
important validated steps in wound preparation (4, 5); but about skin disinfection there are some
controversies.

With the global increasing rates of antibiotic resistance, topical antiseptics have been suggested as
alternative means of reducing wound infection in primary care (6-8). These agents inhibit action or
reduce the number of microorganisms on intact skin or within a wound (9) and are recognized to be
essential for wounds infection control (6, 10). Wound cleansing by topical antiseptic agents seemed
to reduce wound infections effectively but, there are some evidence suggesting that using these
agents for all wounds might be unnecessary (6) and in some cases, may have a negative effect on
wound healing process (11-16). It is therefore rational to use appropriate solutions and avoid
unnecessary cleansing to optimize the healing process and decrease potential adverse effects on
normal tissue (6).

Topical antiseptic agents have commonly been used on skin around traumatic wounds to prevent
infection for a long time, yet the merits and effectiveness of antiseptic fluid irrigation have been
debated and the choice of antiseptic agents is controversial (6, 7). Besides, there is still limited
information on the effectiveness of different methods currently used for the irrigation and cleansing
traumatic wounds. In an attempt to address these challenges, the aim of this study is to provide a
detailed analysis of the rate of infection with presence or absence of Povidone-iodine (PVI) as a skin
disinfection agent, in simple traumatic wounds with identifying potential risk factors. Povidone
Iodine solution10% is an aqueous solution containing Povidone-iodine, 10% (1% available iodine) as
active ingredient and purified water and sodium hydroxide as inactive ingredients.

Methods
Study Design and Setting

This study is a single-blind, randomized controlled trial, conducted at the emergency departments of
two university hospitals of Iran University of Medical Sciences during April 2013-April 2014. In this
trial, those collecting and analyzing clinical data were unaware of the assigned treatment. Trial was
registered in Iranian Registry of Clinical Trials (IRCT) (IRCT2013070610017N5)

Written consent was obtained from all patients or their legal guardians and assessments were
conducted by an ED physician. Ethical approval has been granted by the Iran University of Medical
Sciences Research Ethics Committee and carried out in accordance with Declaration of Helsinki
(1989).

Selection of participants

Potential participants included all ED patients over 16 years old with simple traumatic wounds who
did not meet the exclusion criteria of recent history or current use of antibiotics or diabetes mellitus,
cirrhosis, renal failure, chronic substance or alcohol abuse, splenectomy or immunodeficiency
(congenital, acquired, or receiving immunosuppressive therapy), delayed presentation (> 12 hours),
allergy to iodine and any wound with indication of antibiotic prophylaxis based on literature or just
based on physician judgement. Simple traumatic wounds are defined as wounds that are not caused
by animal or human bite, not involving deeper structures (eg. nerves, tendons, vascular, bones or
joints), and are not crushed or penetrating and have not visible contamination (17). Using
convenience sampling method, included patients were assigned to either experiment or control
group.

Intervention

After enrolment, participants were randomly assigned to either experiment or control groups, using
random numbers table. At the beginning of wound management adequate local anaesthesia by
injecting lidocaine 1% in wound edges was made in both groups ; Then in the control group, wounds
were first irrigated by normal saline solution and either in the experiment group, wounds were
primarily irrigated by normal saline solution similar to control group and then PVI 10%(Toliddaru.co,
Iran) was rubbed on the intact skin5 cm apart from the wound edges by means of sterile cotton pads
letting it to dry on room air using sterile gloves; the wound itself was not cleansed by anything but
normal saline solution in both groups and so in experiment group povidone iodine was only used on
intact skin not the wound. then, using sterile method, debridement of dead and contaminated
tissues were done if needed based on physician judgement and wounds were closed by no
absorbable nylon sutures with cutting needle suture size was chosen based on physician judgement.
After suturing the wound was covered by light dressing composed of sterile dry cotton gauzes.
Preparation irrigation suturing dressing all were done by emergency medicine resident physician
who is by that time in charge of the shift all of physicians who participated at the study were
informed of the study protocol. Before discharge changing the dressing was thought to all patients
and they were asked to change their dressing every 12 hours till 48 hours after suturing. All patients
were alerted about wound infection signs and were asked to come to emergency department as
soon as they inspected these signs. 10 according to literature for simple traumatic wounds with
characteristics we defined there is no indication for prophylactic systemic antibiotics; (18) topical
antibiotics were not used on any wounds either. At the wound closure step it was not possible to
blind the physician who in suturing the wound because of the colour and smell of povidone iodine
solution, but all patients were asked to take shower and wash the skin around the wound after 48
hours so at the time of suture removal it was not possible for the wound investigator physician to
detect which wound was prepared using povidone iodine and which is not.

All of the patients were required to visit ED physician48 hours after repair and also at the time of
suture removal within 7-10 days depending on the wound location. Patients required referring to ED
physician or call the researcher in case of any wound infection during their wound healing period
(the day after receiving treatment until sutures removal). Outcome assessments were made by a
blinded ED physician. All of the participants received standard wound care, and were followed
during 1 month after their first ED visit by a telephone call.

Outcome Measures

Primary outcome measure was the difference in wound infection rates between two studied groups.
Based on CDC criteria for superficial incisional surgical site infection wound, infection is defined as
presence of at least one of the following condition ,Purulent drainage, with or without laboratory
confirmation, from the superficial incision; Organisms isolated from an aseptically obtained culture
of fluid or tissue from the superficial incision; At least one of the following signs or symptoms of
infection: pain or tenderness, localized swelling, redness, or heat and superficial incision is
deliberately opened by surgeon, unless incision is culture-negative; or diagnosis of superficial
incisional surgical site infection by the surgeon or attending physician with exception of the isolation
of organism from fluid or tissue, in fact diagnosis of infection was made based on clinical criteria (3,
19). A wound infection was considered to be significant if the follow-up physician’s impression was
that there was a wound infection requiring antibiotics, intervention, admission or referral.
Secondary outcome measures were the occurrence of skin irritation and wound dehiscence.

Primary Data Analysis

. All data were entered in a file and analyzed with Stata software (version 12.0). Demographic and
clinical data were presented descriptively as frequencies and means with SDs where appropriate.
Univariate logistic regression models were used to predict the odds of wound infection in the
experiment and control groups. For eliminating potential confounders, those variables with
significancy level < 0.03 were entered into multivariate regression model. The significance level was
0.05 for two sided tests.

Results

Characteristics of Study Subjects

During the study period, 443 patients were screened and finally enrolled. Of those patients, 37 were
excluded. (Figure 1) There were no differences in the baseline characteristics of the experiment
(with PVI) and control (without PVI) groups (Table 1). Men constituted 85% of the study participants.
The sites of wounds were head in 44% of patients, neck or trunk in 10% of patients, upper limbs in
22% of patients and lower limbs in 24% of patients. Wounds were typically treated on average 2
hours after the event. The average depth and length of traumatic wounds were similar in both
groups (p=0.317, p=0.971 accordingly).

The observed infection rate was 8.7% (19 patients, odds ratio=1.5,95% confidence interval, 0.76 to
2.97) in the experiment group. The difference in infection rates was not statistically significant
between these groups (odds ratio1.5; 95% CI 0.74 to 3.23; P=0.236). The difference in infection rates
were statistically significant according to the sites of wounds in which lower limbs and upper limbs
found to have the highest infection rates (p<0.05). Despite the higher rate of infection at limbs we
did not observe any statistically significant difference in experiment and control group for limb
infection incidence. (p=0.058 for upper limb and p=0.0001 for lower limb). There was no significant
different in the infection rates of between injury type subgroups (Table 2). Skin irritation and wound
dehiscence did not occur in any case.

Table 1- Demographic characteristics of the study patients and baseline characteristics of the
wounds Characteristics with PVI (N=192) without PVI (N=214)

Table 2- Univariate and multivariate logistic model for odds of traumatic wound infections

Discussion

In this study we tried to clarify the role of povidone iodine as a typical antiseptic solution in wound
infection prevention. We designed the study so that we can control or eliminate risk factors for
infection to reach this goal we excluded patient who had clinical circumstances that prone their
wound to infection such as diabetes, tissue ischemia and other conditions mentioned at exclusion
criteria and by selecting only simple traumatic wounds we tried to eliminate confounding factors like
presence of foreign body, delayed closure, and deep structures involvement. To control other risk
factors for infection that we may not be able to eliminate we used proper randomization that based
on comparing characteristics of two groups, which showed no statistically significant difference, we
were successful to reach this goal. Our findings indicated no significant association between the use
of PVI 10% solution and the rate of infections in the simple traumatic wounds. The results of the
regression analysis showed that the site of wound is a better predictor for wound infections than
using antiseptic agents In this regard lower indicated the highest rates of wound infections that
other sites.

(Our findings failed to support this argument that using topical antiseptic agents on intact skin
around the simple traumatic wounds prevents wound infections).(According to our findings, no
significant difference was reported between the rates of observed infections in the experiment and
the control groups either treated with normal saline or PVI. However there are evidences that using
topical antiseptics like PVI significantly reduces the overall infection rates and recommended for
prevention of wound infections (6, 20, 21). The association between PVI administration and the risk
of infection has been widely investigated in the complicated and surgical site wounds. Although the
results are mixed (22, 23) but several studies indicated that skin preparation with PVI may reduce
postoperative infection rates (24-26). Several studies have revealed that using PVI conferred no
benefits as compared to normal saline or chlorhexidine for cleaning wounds (7, 22, 23, 27).
However, information on the role of PVI in simple traumatic wound infection prevention is limited.
In this context, our results provide further evidence that routine skin cleansing by PVI is not well
associated with reduction in overall infection rates in simple traumatic wounds.

There is a great body of literature discouraging the use of routine antiseptics agents as a standard
step of preparation in all types of wounds. These arguments are supported with laboratory, human
and animal studies that examined their toxicity and the adverse effects of using topical antiseptics
on normal tissue healing (6, 11-13, 16, 28-32).These studies demonstrated the negative effects of
antiseptics on production of fibroblasts, collagen, epithelial migration and microcirculation during
tissue healing process (7). Besides, it is believed that antiseptics are inactivated while contacting
body fluids, blood or proteins (7) therefore; they need prolonged contact to reduce the bacterial
loads. While different antiseptic agents are used for wound cleansing and irrigation in primary care,
the choice of the best treatment, including elimination of skin disinfection by means of topical
antiseptics, is still controversial and further randomized clinical trials are needed to provide more
information about true effect of antiseptics on infection rate and healing process of traumatic
wounds. Previous studies showed that there is no significant difference between alcohol containing
skin antiseptic agents and aqueous povidone iodine solution in traumatic wound infection
prevention (33) although recent evidence on this issue is lacking. Our results also indicated
significant association between the wound location and the rate of infections, in which upper and
lower limbs indicated the highest rates of infections. Besides, previous studies suggested that
wounds on head and neck were less commonly associated with infection development than other
locations. In this regard, the incidence of infection estimated 1% to 2% in head and neck, whereas
legs and lower extremities were as high is 23% (34-37). Since these sites are considered to be more
infection-prone (20), it is therefore important to consider the site of wounds once deciding to
choose cleansing agents with better efficacy and lower adverse effects.

Our results also indicated no significant association among the length, depth, wound age (under 12
hours) and the rate of infections. This finding was in contrast to previous studies that proposed
wound age, and depth as the most predictive factors of wound infections (20, 34).Although it is
important to notice that we did not include wounds older than 12 hours.

We are aware that present study has some limitations one of them is that it is a single centre trial,
we could assess some secondary outcomes like pain and time to wound healing, maybe using
antiseptics on skin around the wound change the duration of time needed for complete healing, that
is the point we will sure consider to assess in further studies. In addition, other antiseptic agents
were not examined and because of lack of previous supporting evidence only simple traumatic
wounds were included it is obvious that rate of infection is not so high in these types of traumatic
wounds.

In summary, our findings support the argument that using PVI in the management of simple
traumatic wounds has no significant association with the rate of infections. Besides, the sites of
wound were proposed as the most important risk factor for wound infections in ED settings. Despite
our finding according to the fact that using povidone iodine on skin around the wounds is standard
of care, we believe that it is too soon to change wound management in emergency department
based on this single study of course further studies with larger sample size and improved method
are needed to clearly define the role of povidone iodine as skin antiseptic in wound infection
prevention. Besides, studies on other antiseptic agents specially alcohol containing agents is
necessary and will provide valuable information about exact role of skin disinfection in traumatic
wound management. We are planning to design further studies to eliminate our limitations and
provide stronger evidence on the issues mentioned above in near future.

Funding: No funding has been received by any authors of this work.

Disclosure of Conflict of Interest:

All authors declared no conflicts of interests regarding this work.

References

1. Capellan O, Hollander JE. Management of lacerations in the emergency department.

Emergency medicine clinics of North America. 2003;21(1):205-31.
2. Hollander JE, Singer AJ. Laceration management. Annals of emergency medicine.
1999;34(3):356-67.
3. Hollander JE, Singer AJ, Valentine SM, Shofer FS. Risk factors for infection in patients with
traumatic lacerations. Academic Emergency Medicine. 2001;8(7):716-20.
4. Fernandez RS, Griffiths RD, Ussia C. Wound cleansing: which solution, what technique? 2001.
5. Moscati RM, Mayrose J, Reardon RF, Janicke DM, Jehle DV. A multicenter comparison of tap
water versus sterile saline for wound irrigation. Academic Emergency Medicine. 2007;14(5):404-9.
6. Atiyeh BS, Dibo SA, Hayek SN. Wound cleansing, topical antiseptics and wound healing.
International wound journal. 2009;6(6):420-30.
7. Khan MN, Naqvi AH. Antiseptics, iodine, povidone iodine and traumatic wound cleansing.
Journal of tissue viability. 2006;16(4):6-10.
8. The Best Practice Advocacy Centre New Zealand. Should I prescribe a topical antiseptic
cream instead of a topical antibiotic for minor skin infections? Best Practice Journal 2015;68:18-20.
9. Swanson T, Asimus M, McGuiness B. Wound Management for the Advanced Practitioner.
Melbourne: IP Communications; 2014.
10. Knobler SL, Mahmoud AAF, Pray LA. Department Of Health And Human Services: Food and
Drug Administration [Docket No. 01N–0494]. 2002.
11. Cooper ML, Laxer JA, Hansbrough JF. The cytotoxic effects of commonly used topical
antimicrobial agents on human fibroblasts and keratinocytes. Journal of Trauma and Acute Care
Surgery. 1991;31(6):775-84.
12. Damour O, Hua SZ, Lasne F, Villain M, Rousselle P, Collombel C. Cytotoxicity evaluation of
antiseptics and antibiotics on cultured human fibroblasts and keratinocytes. Burns. 1992;18(6):479-
85.
13. McCauley RL, Linares HA, Pelligrini V, Herndon DN, Robson MC, Heggers JP. In vitro toxicity
of topical antimicrobial agents to human fibroblasts. Journal of Surgical Research. 1989;46(3):267-
74.
14. Thomas GW, Rael LT, Bar-Or R, Shimonkevitz R, Mains CW, Slone DS, et al. Mechanisms of
delayed wound healing by commonly used antiseptics. Journal of Trauma and Acute Care Surgery.
2009;66(1):82-91.
15. Welch JS, Bunting TG. The cytotoxic effects of commonly used topical antimicrobial agents
on human fibroblasts and keratinocytes. The Journal of trauma. 1993;35(3):491-2.
16. Wilson JR, Mills JG, Prather ID, Dimitrijevich SD. A toxicity index of skin and wound cleansers
used on in vitro fibroblasts and keratinocytes. Advances in skin & wound care. 2005;18(7):373-8.
17. Lee CK, Hansen SL. Management of acute wounds. Surgical Clinics of North America.
2009;89(3):659-76.
18. Tintinalli JE, Stapczynski JS, Ma OJ, Yealy DM, Meckler GD, Cline DM. Tintinalli's Emergency
Medicine: A Comprehensive Study Guide. 8th ed. Tintinalli JE, editor.: McGrawHill; 2015.
19. U.S. Centers for Disease Control and Prevention. Surgical Site Infection (SSI) Event. 2016
[updated 2016; cited 2016 01.25]; 1-29]. Available from:
http://www.cdc.gov/nhsn/PDFs/pscManual/9pscSSIcurrent.pdf.
20. Nakamura Y, Daya M. Use of appropriate antimicrobials in wound management. Emergency
medicine clinics of North America. 2007;25(1):159-76.
21. Roberts AHN, Roberts FEV, Hall RI, Thomas IH. A prospective trial of prophylactic povidone
iodine in lacerations of the hand. Journal of Hand Surgery (British and European Volume).
1985;10(3):370-4.
22. Darouiche RO, Wall Jr MJ, Itani KMF, Otterson MF, Webb AL, Carrick MM, et al.
Chlorhexidine & alcohol versus povidone-iodine for surgical-site antisepsis. New England Journal of
Medicine. 2010;362(1):18-26.
23. Noorani A, Rabey N, Walsh SR, Davies RJ. Systematic review and meta―analysis of
preoperative antisepsis with chlorhexidine versus povidone–iodine in clean―contaminated
surgery. British Journal of Surgery. 2010;97(11):1614-20.
24. Chundamala J, Wright JG. The efficacy and risks of using povidone-iodine irrigation to
prevent surgical site infection: an evidence-based review. Canadian Journal of Surgery.
2007;50(6):473.
25. Fournel I, Tiv M, Soulias M, Hua C, Astruc K, Aho Glele LS. Meta―analysis of intraoperative
povidone–iodine application to prevent surgical―site infection. British Journal of Surgery.
2010;97(11):1603-13.
26. Swenson BR, Hedrick TL, Metzger R, Bonatti H, Pruett TL, Sawyer RG. Effects of Preoperative
Skin Preparation on Postoperative Wound Infection Rates A Prospective Study of 3 Skin Preparation
Protocols. Infection Control. 2009;30(10):964-71.
27. Stringer MD, Lawrence JC, Lilly HA. Antiseptics and the casualty wound. Journal of Hospital
Infection. 1983;4(4):410-2.
28. Fernandez R, Griffiths R, Ussia C. Water for wound cleansing. Cochrane Database Syst Rev.
2008;1.
29. Ghafouri HB, Zoofaghari SJ, Kasnavieh MH, Ramim T, Modirian E. A pilot study on the repair
of contaminated traumatic wounds in the emergency department using sterile versus non-sterile
gloves. Hong Kong Journal of Emergency Medicine. 2014;21(3):148.
30. Main RC. Should chlorhexidine gluconate be used in wound cleansing? Journal of wound
care. 2008;17(3):112-4.
31. Tatnall FM, Leigh IM, Gibson JR. Comparative study of antiseptic toxicity on basal
keratinocytes, transformed human keratinocytes and fibroblasts. Skin Pharmacology and Physiology.
1990;3(3):157-63.
32. Zamora JL. Chemical and microbiologic characteristics and toxicity of povidone-iodine
solutions. The American journal of surgery. 1986;151(3):400-6.
33. Edlich RF, Custer J, Madden J, Dajani AS, Rogers W, Wangensteen OH. Studies in
management of the contaminated wound: III. Assessment of the effectiveness of irrigation with
antiseptic agents. The American journal of surgery. 1969;118(1):21-30.
34. Nicks BA, Ayello EA, Woo K, Nitzki-George D, Sibbald RG. Acute wound management:
revisiting the approach to assessment, irrigation, and closure considerations. International journal of
emergency medicine. 2010;3(4):399-407.
35. Pfaff JA, Moore GP. Reducing risk in emergency department wound management.
Emergency medicine clinics of North America. 2007;25(1):189-201.
36. Quinn JV, Polevoi SK, Kohn MA. Traumatic lacerations: what are the risks for infection and
has the ‘golden period’of laceration care disappeared? Emergency Medicine Journal.
2014;31(2):96-100.
37. Zehtabchi S, Tan A, Yadav K, Badawy A, Lucchesi M. The impact of wound age on the
infection rate of simple lacerations repaired in the emergency department. Injury.
2012;43(11):1793-8.
 Figure 1 CONSORT Flow Diagram

Assessed for eligibility (n= 443)

Enrollment

Excluded (n= 37)

 Not meeting inclusion criteria (n= 37 )
 Declined to participate (n= 0 )
 Other reasons (n= 0 )

Randomized (n= 406)

Allocation
Allocated to intervention (n= 192) Allocated to intervention (n=214)
 Received allocated intervention (n= 192 )  Received allocated intervention (n= 214)
 Did not receive allocated intervention (give  Did not receive allocated intervention (give
reasons) (n= 0 ) reasons) (n=0 )

Follow-Up
Lost to follow-up (give reasons) (n=0) Lost to follow-up (give reasons) (n= 0)

Discontinued intervention (give reasons) (n= 0) Discontinued intervention (give reasons) (n=0)

Analysis
Analysed (n=192) Analysed (n= 214)
 Excluded from analysis (give reasons) (n=0)  Excluded from analysis (give reasons) (n=0)
Table 1- Demographic characteristics of the study patients and baseline characteristics of the
wounds characteristics with PVI (N=192) without PVI (N=214)

Characteristics With PVD(N=192)(%) Without P Value

PVD(N=214)(%)
Age, y, mean ± SD 29.1 ± 14.2 31.6 ± 13.9 0.581

Sex
Men 169 (88) 175 (82) 0.081
Women 23 (12) 39 (18)

Site
Head 94 (49) 85 (39.7)
Neck or Trunk 15 (7.8) 26 (12.1)
Upper limbs 44 (22.9) 44 (20.6) 0.097
Lower limbs 39 (20.3) 59 (27.6)
Injury type
blunt 63 (32.8) 95 (44.4)
stab with glass 18 (9.4) 24 (11.2)
stab without glass 100 (52.1) 82 (38.3) 0.044
wood 11 (5.7) 13 (6.1)
Time passed, h, mean ± SD 1.9 ± 1.9 1.8 ± 1.2 0.731

Depth, mean ± SD 1.0 ± 0.6 1.1 ± 0.6 0.317

Length, mean ± SD 4.2 ± 3.7 3.7 ± 2.5 0.971

Contamination
Yes 16 (8.3) 13 (6.1)
No 176 (91.7) 201(93.9) 0.360
Table 2- Univariate and multivariate logistic model for odds of traumatic wound infections

Univariate model Multivariate model

95% CI 95% CI
variable Odds Lower Upper P value Odds Lower Upper P Value
ratio ratio
Intervention
With PVI Ref -- -- -- Ref -- -- --
Without PVI 1.55 0.74 3.23 0.231 1.51 0.67 3.36 0.313
Age 0.98 0.95 1.01 0.353
Sex
Male Ref -- -- --
Female 1.15 0.42 3.11 0.783
Location
Head Ref -- -- -- Ref -- -- --
Neck or Trunk 0.01 0.15 11.78 0.998 0.01 0.15 10.79 0.863
Upper limbs 3.13 0.96 10.18 0.058 5.47 1.47 20.28 0.011
Lower limbs 7.09 2.56 19.63 <0.0001 9.23 2.98 28.57 <0.0001
Others 1.21 0.13 10.78 2.53 0.26 24.50
Injury type 0.862 0.422
Blunt Ref -- -- Ref -- --
stab with glass 0.22 0.02 1.76 -- 0.16 0.02 1.39 --
stab without glass 0.70 0.32 1.52 0.157 0.57 0.24 1.32 0.099
wood 0.01 0.17 6.78 0.371 0.02 0.02 1.76 0.192
Wound age 1.08 0.80 1.25 0.999 0.999
Depth 0.67 0.31 1.45 0.945
Length 0.99 0.88 1.12 0.317
Contamination 0.971
Yes Ref -- --
No 0.55 0.15 1.96 --0.360