Efficacy of Intrawound Vancomycin in Prevention of Periprosthetic Joint Infection After Primary Total Knee Arthroplasty

44 45

Peri-prosthetic Joint Infection (PJI) after total knee arthroplasty (TKA) is a 46 devastating complication. Intra-wound vancomycin powder has been shown to reduce 47 infection rates in spine surgery. Previous studies on the efficacy of local vancomycin powder 48 in hip or knee arthroplasty are mostly retrospective case series. The aim of this prospective 49 RCT was to evaluate the efficacy and safety of intrawound vancomycin in preventing PJI 50 after primary TKA. 51

This study was a National Trial Registry-approved RCT of patients undergoing 52 primary TKA. 1022 patients were randomized to the study group (n=507, received 2g 53 intrawound vancomycin powder before arthrotomy closure) and control groups (n=515, no 54 local vancomycin). The minimum follow-up was 12-months. The primary outcome was PJI 55 rate. Secondary outcomes included surgical site infection (SSI) rates, incidence of revision 56 for PJI/SSI, and incidence of wound complications. High-risk groups (Obesity and Diabetes) 57 in both cohorts were also evaluated. 58

The overall infection rate in 1022 patients was 0.66%. There was no significant 59 difference in PJI rate in the study group (0.2%) versus the control group (0.58%), p=0.264. 60 Reoperation rates in the study group (N=4;0.78%) and Control (N=5;0.97%) and SSI rates in 61 the study (N=1;0.2%) and control groups (N=2;0.38%) were comparable. The Vancomycin 62 cohort however demonstrated a significantly higher number of minor wound complications 63 (n=67;13.9%) compared to the control group (n=39;8.4%, p<0.05). There was no difference 64 in PJI/SSI rates or minor surgical complications among high-risk groups and no cases of 65 nephrotoxicity were reported in the study. 66

Intra-wound vancomycin powder does not appear to reduce PJI/SSI rate in 67 primary total knee arthroplasties, including high-risk groups. Although safe from a renal 68 perspective, intra-wound vancomycin was associated with an increase in postoperative 69 aseptic wound complications such as persistent wound drainage. Intra-wound vancomycin 70 may not be effective in reducing the rate of PJI in primary TKA. 71 72

Infection; Arthroplasty; Knee; Prosthetic Joint Infection; Surgical site infection 73 74 75 76 77 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 3

78 There is an ever-increasing global burden of primary total knee and total hip arthroplasty 79 procedures, with a significant increase in surgery numbers projected by the year 2030 (1–3). 80 With increasing numbers of primary total joint replacement surgeries, healthcare systems 81 across the world are poised to handle an increased burden of revision arthroplasty cases. 82 Worldwide, the reported incidence of PJI after total joint replacement is around 0.8-1.5%, 83 with a 1-2% PJI risk after TKA (4–6). 84 85 Periprosthetic joint infection (PJI) is one of the most common indications for revision after 86 TKA (7–9) posing a significant burden to the healthcare system (10,11). There has been 87 increasing interest in the use of local antibiotic powder in the joint before closure to prevent 88 local contamination and biofilm formation. The use of intrawound vancomycin in preventing 89 surgical site infections is well documented in spine surgery (12–14). There are some 90 published reports of the benefits of local vancomycin powder used in surgeries of the elbow, 91 foot and ankle surgery (15,16). However, its use and clinical benefit in total knee arthroplasty 92 is debatable. 93 94 The main postulated advantages of using local vancomycin powder are its relatively cheap 95 cost, favourable bacterial spectrum (MRSA, coagulase-negative staphylococci), high local 96 concentrations without systemic adverse effects (17,18). Systematic reviews and meta-97 analyses on studies evaluating the use of vancomycin in primary total knee or hip arthroplasty 98 reported low-quality evidence with a high risk of study bias (19,20). No randomized control 99 trial (RCT) has evaluated the efficacy of local vancomycin powder in primary TKA. 100 101 The primary objective of this study was to evaluate the efficacy of intra-wound vancomycin 102 in reducing SSI or PJI rates in patients undergoing primary total knee arthroplasty. Secondary 103

include the evaluation of wound-healing-related complications, incidence of 104 nephrotoxicity, and the influence of diabetes and obesity on outcomes. 105 106 107

108 This is an Institutional Review Board (IRB) approved prospective randomized, controlled 109 trial (RCT), and was prospectively registered with the National Central Trial Registry 110 (CTRI/2021/02/031310). This is a single-centre RCT of patients from a high-volume tertiary 111 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 4 care institute, who underwent primary total knee arthroplasty (TKA) for primary 112 osteoarthritis of the knee, between January 2021 and February 2022. The RCT was a parallel-113 arm trial with a 1:1 allocation of patients into study and control groups. Study group patients 114 receive intra-articular 2g Vancomycin antibiotic powder before arthrotomy closure, control 115 group patients did not receive local vancomycin powder. The trial was conducted following 116 the guidelines of the Declaration of Helsinki on scientific studies involving human subjects. 117 118 All adult patients with primary osteoarthritis of the knee, consenting to primary manual jig-119 based TKA were eligible for recruitment in this trial. This includes patients with primary OA, 120 with either varus or valgus deformity of the knee. Patients were excluded from the trial if 121 they met any of the following exclusion criteria- refusal to participate in the trial, intra-122 operative findings suggestive of inflammatory arthropathy or non-specific synovitis, previous 123 knee surgeries, history of intra-articular knee injections within 3 years before surgery, known 124 allergy to Vancomycin, and if they were known cases of chronic immunosuppression 125 (secondary to human immuno-deficiency virus-HIV, malignancy or post-solid organ 126 transplant). Patients who used antibiotics for any cause, within 1-month leading up to surgery 127 were also excluded from the trial. 128 129 There were a total of 1208 eligible patients who underwent primary TKA during the study 130 period (January 2021 to January 2022). 56 patients declined to participate in the trial and a 131 further 24 patients were excluded due to exclusion criteria (12 patients used antibiotics for 132 Urinary tract infections, 8 patients with recent intra-articular knee visco-supplementation 133 injections and 4 patients with chronic immunosuppression). After exclusions, 1128 patients 134 were randomized to the study and control groups. 34 patients were excluded from the final 135 data analysis after randomization (16 deaths in the follow-up period, 18 cases with intra-136 operative findings suggestive of inflammatory arthropathy or non-specific synovitis). 137 138 The final study cohorts consisted of 549 patients in the study group and 545 patients in the 139 Control group. There was a loss to follow-up of 42 patients (8%) in the study group and 30 140 patients (5.76%) in the control group, Final analysis included 507 subjects in the study group 141 and 515 subjects in the control group. 142 143 Eligible trial participants undergoing primary total knee arthroplasty were randomized (1:1 144 allocation by computer randomization) to receive either normal saline lavage with (study 145 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 5 group) or without (control group) intra-articular 2g Vancomycin powder prior to arthrotomy 146 closure. All patients were operated on at the same institute, in laminar flow operating rooms. 147 Except for the use of vancomycin powder in the surgical wound, the operating procedures 148 and peri-operative protocols were common to both groups. The patients are blinded and 149 unaware of the assigned groups and follow-up evaluation for infection related complications 150 was performed by a blinded observer. 151 152 All cases in this trial received a standard antibiotic prophylaxis protocol, which includes 3 153 peri-operative intra-venous doses of 1.5g of Cefuroxime. The first dose was administered 1 154 hour before the skin incision and 2 doses were administered 8 hours and 24 hours after 155 surgery. None of the cases in this trial received prolonged Oral antibiotic use after surgery. 156 Skin preparation at the time of surgery was done with Chlorhexidine skin scrub (3M, 157 Avagard 4% Chlorhexidine Scrub). TKA was performed through a medial para-patellar 158 approach in all cases, under a tourniquet. Tourniquet was inflated from the time of the 159 incision till the cement was fully-set after final implantation. The knee joint was irrigated 160 with a total of 3 L of normal saline with pulse lavage before and after cementing of the 161 implant. All cases received bone cement without antibiotic impregnation (Palacos/ Stryker 162 Simplex low-viscosity cement). Patella was not resurfaced in any case, in either group. After 163 deflation of the tourniquet and haemostasis of bleeding vessels, the study group received 2g 164 vancomycin powder, which was placed into the medial and lateral joint gutters and 165 arthrotomy closed with No.2 Vicryl in an interrupted fashion followed by a continuous 166 STRATAFIX (Barbed-PDS) running stitch. Suction drains were not used in any case. 167 168 Post-operative blood investigations (Renal parameters, CBP) were obtained 24 hours after the 169 surgery. In all cases, wound dressing was changed 24 hours later in the PACU, with 170 MEPILEX surgical dressing (Mepilex Border Post-op, Mölnlycke, Göteborg, Sweden). 171 Patients in both groups received uniform rehabilitation with assisted walking and range of 172 motion exercises on the first postoperative day. Deep vein thrombosis prophylaxis was 173 common with low-Molecular weight heparin (LMWH) in the immediate postoperative period 174 and Oral Apixaban 2.5mg twice daily for 14 days following surgery. Surgical staple removal 175 is done between Day 14 and Day 21 after surgery. Staple removal after 21 days is considered 176 to be “Delayed staple/clip removal”. Patients were routinely followed up at 2 weeks, 4 weeks, 177 3 months and 6 months and 12 months after surgery. 178 179 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 6 Based on the criteria identified by Parvizi et al (21), early peri-prosthetic joint infection was 180 defined as an infection diagnosed within 90 days of surgery. Infection was diagnosed based 181 on the MSIS criteria and the 2018 Definition of Periprosthetic Hip and Knee Infection 182 consisting of 2 positive cultures (from an aspirate and/or at the time of debridement for PJI or 183 SSI), elevated serum ESR, CRP and elevated white cell counts in synovial fluid aspirates 184 (>10,000 cells/mm3), PMN percentage greater than 90% in the synovial aspirate(21,22). 185 Cases of possible infections based on the 2018 ICM (International Consensus Meeting) 186 scoring system were confirmed based on the findings of pus in aspirate and intra-operative 187 findings of purulence. 188 Surgical site infections and deep peri-prosthetic joint infections were considered major 189 surgical complications, necessitating re-operation. Minor complications included delayed 190 wound healing with or without dehiscence, delayed surgical staple removal, and stitch/suture 191 abscess (based on CDC Guidelines) necessitating oral antibiotic use. Persistent wound 192 drainage was defined as, wound drainage beyond 72-hours necessitating surgical dressing 193 change, based on previous definitions of persistent wound drainage (21,23). 194 Patients were also monitored in the peri-operative period for medical complications of 195 Myocardial infarction, DVT/PE, CVA and nephrotoxicity secondary to the use of 196 vancomycin. 197 198 Statistical analysis 199 Xu et al conducted a systematic review and meta-analysis of intra-wound vancomycin 200 powder used in primary total knee and hip arthroplasty(24). This meta-analysis concluded 201 that a minimum sample of 1000 patients was required for analysis, to conform to a decrease 202 of PJI rate from 2.74% to 1% in control versus study groups, with a power of 80% and 5% 203 significance level. Assuming a drop-out rate of 10%, we recruited over 1100 eligible subjects 204 for trial participation and randomization. Statistical analysis was performed using a 2-tailed 205 or independent samples t-test for continuous data parameters. The Chi-squared test or 206 Fisher’s Exact test was used for categorical data parameters. Statistical analysis was 207 performed using SPSS Version 24 (International Business Machines- IBM, Armonk, NY). 208 Assuming a power of 80% and 95% confidence intervals, a p-value less than 0.05 was 209 considered significant. 210 211 212 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 7

213 A total of 1022 patients were included in the final statistical analysis, with 507 patients in the 214 study group and 515 patients in the control group, after exclusions, loss to follow-up and 215 deaths. Demographic variables and baseline characteristics such as comorbidities, BMI, ASA 216 grading, and CCI grading were comparable between both groups and summarised in Table.1. 217 The mean tourniquet times and mean haemoglobin drop after surgery were also comparable 218 between both groups. The CONSORT Flowchart of participant recruitment is shown in 219 Figure.1 220 221 Infection rates 222 The overall infection rate in the study population was 0.66%. Periprosthetic joint infection 223 (PJI) was seen in 1 patient (0.19%) in the study group and 3 patients (0.58%) in the control 224 group, and the difference was not statistically significant. Surgical site infection (SSI) was 225 reported in 1 (0.19%) patient in the study group and 2 patients (0.38%) in the control group. 2 226 patients in the study group had periprosthetic fractures of the distal femur during the follow-227 up period and underwent ORIF. All complications in the trial are summarized in Table 2. 228 229 Peri-prosthetic Joint Infections (PJI) 230 In the study group, one patient (0.19%) underwent DAIR for a culture-negative PJI, 4 weeks 231 after surgery. This patient had elevated serum ESR, CRP and White cell counts and frank pus 232 aspirated from the joint before DAIR. In the control group, 3 patients (0.58%) underwent 233 DAIR, of which one patient had Staphylococcus aureus isolated, and later went on to have a 234 2-stage revision due to the persistence of infection. The other two patients had culture-235 negative PJI and underwent DAIR. 236 237 Surgical Site Infections (SSI) 238 One patient in the study group presented with an early SSI, with Staphylococcus epidermidis 239 the pathogenic organism isolated. This patient presented with persistent wound discharge 240 which was managed with superficial wound debridement and secondary closure and iv 241 antibiotics. 2 patients in the control group developed SSI, for which one patient underwent 242 debridement and secondary closure for local wound necrosis and the other patient developed 243 septic shock and Multi-Organ Dysfunction (MODS) following wound debridement at an 244 outside hospital, requiring hospitalization and intravenous antibiotics. The major 245 complications of PJI and SSI are summarized in Table 3. 246 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 8 247 Re-operation Rates 248 Re-operation rates between both groups were found to be statistically comparable, with 4 249 0.78%) patients in the study group (1-PJI, 1-SSI, 2-Periprosthetic fracture) versus 5 cases 250 (0.97%) in the control group (3-PJI, 2-SSI). (Table 2) 251 252 Wound Complications 253 The use of vancomycin was associated with a significantly higher number of minor wound 254 complications in the study group (n=67; 13.9%) compared to the control group (n=39; 8.4%, 255 p<0.05). Aseptic wound complications such as wound soakage, maceration and stitch abscess 256 were managed with regular dressings without any additional oral or intravenous antibiotic 257 use. Delayed suture removal (>3 weeks) was found to be similar in both groups and not 258 statistically significant. (Table 2) 259 260 Sub-group Analysis 261 We evaluated differences in complication rates amongst patients of both groups based on 262 BMI and compared diabetics with non-diabetics. There was no difference in PJI/SSI rates or 263 minor surgical complications among high-risk groups, Diabetics versus non-diabetics, (Table 264 4) and high BMI vs normal BMI (Tables 5 & 6). 265 266 Systemic Complications and Mortality 267 The use of intra-wound vancomycin was not associated with acute kidney injury (defined as 268 an increase in serum creatinine levels by more than 0.3mg/dl within 48 hours)(25). 269 A total of 16 patients expired in the follow up period (10 patients in the study group versus 6 270 patients in the control group. One patient in the control group developed DVT with 271 Pulmonary embolism 45 days post-surgery and expired. Other patients expired secondary to 272 non-surgical factors (1 patient secondary to complications of CKD in the control group, 1 273 patient with Pleural effusion, 4 patients expired after treatment for trauma unrelated to the 274 prior knee arthroplasty). No cases of CVA, AMI, Nephrotoxicity, Ototoxicity or anaphylactic 275 reactions were observed or documented in the post-operative 90-day follow-up period. 276 277 278 279 280 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 9

281 This randomised control trial did not demonstrate any additional clinical benefit of topical or 282 intra-wound vancomycin powder in preventing SSI/PJI in patients undergoing primary total 283 knee arthroplasty. On the contrary, there was a significant increase in wound healing 284 complications in the study group, with a significantly higher incidence of persistent wound 285 discharge and stitch abscess compared to the control group. 286 287 PJI after TKA bears a heavy toll on the patient and contributes to the economic burden on the 288 healthcare systems (6,11). The use of local antibiotic powder in orthopaedic surgery is not 289 new, with several reports of benefits in spine surgery (12–14,19,26). However, most of these 290 studies were retrospective and non-randomised. Tubaki et al published the only RCT on the 291 use of vancomycin in spine surgery and concluded that there was no difference in infection 292 rates with the use of vancomycin (27). 293 294 The role of topical or intra-wound vancomycin in primary total knee arthroplasty is 295 undecided, with studies both supporting and refuting the efficacy of vancomycin in reducing 296 post-operative PJI rates. 297 298 Heckman et al (28) published a meta-analysis and systematic review of intra-wound 299 vancomycin in total hip and knee arthroplasty with evidence demonstrating lower infection 300 rates with the use of vancomycin. However, all 6 studies included in the meta-analysis were 301 retrospective case series with Level-III evidence. Xu H et al published their meta-analysis of 302 4607 patients reporting a reduced PJI rate with the use of vancomycin but a higher incidence 303 of local wound complications and superficial surgical site infections (24). Our trial had 304 similar findings of a higher incidence of minor complications such as persistent wound 305 drainage and stitch abscesses in the study group. 306 307 Although previous studies attempted to evaluate the efficacy of intra-wound vancomycin, 308 most were either underpowered due to the relatively low incidence of PJI, or suffered from 309 low quality evidence secondary to retrospective study designs or lack of randomization. To 310 the best of our knowledge, this study is the first prospective randomized control trial on the 311 use of intra-wound vancomycin in primary TKA. 312 313 314 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 10 Effect of intra-wound vancomycin on PJI/SSI 315 In this trial, the use of intra-wound vancomycin powder in primary TKA failed to show a 316 reduction in the incidence of PJI/SSI. Previously published studies also showed no decrease 317 in the incidence of PJI with local vancomycin use (29,30)(31). On the contrary, Winkler et al 318 showed a statistically significant decrease in PJI in primary TKA. In the same study, 319 however, there was only a trend towards decreased PJI with local antibiotic administration in 320 THA and revision TKA/THA without statistical significance (32). The main limitation of this 321 report was the retrospective study design and heterogeneity of the study population which 322 included both primary and revision knee or hip arthroplasty, with varying intra-operative 323 protocols. 324 325 Most studies reporting reduced infection rates with the use of intra-wound vancomycin were 326 retrospective (33,34). Otte et al in their study of 1640 patients showed that 1g of intra-wound 327 vancomycin significantly reduced the incidence of PJI in both primary and revision scenarios 328 for both hips and knees (35). Patel et al also reported that intra-wound vancomycin was both 329 safe and effective in reducing rates of early PJI in both primary hip and knee arthroplasties 330 (36). 331 332 A meta-analysis by Peng et al, which included 4512 patients in 9 studies, also recommended 333 the use of vancomycin powder to reduce the incidence of PJI without modifying the bacterial 334 spectrum (37). A recent systematic review of 3371 patients did not demonstrate a significant 335 reduction in PJI in patients receiving vancomycin (38). This was similar to the findings of the 336 current trial. 337 338 This trial also showed a statistically significant increase in minor wound complications such 339 as persistent wound drainage and stitch abscesses with the use of vancomycin. However, this 340 should be interpreted with caution, as the trial may be underpowered to evaluate minor 341 complications. Similar wound complications were also reported in previously published 342 studies (20,24,26,33). The exact cause of this has not been determined but has been attributed 343 to the crystalline nature of the vancomycin salt, low pH of the vancomycin solution or as a 344 part of the body’s inflammatory response to the vancomycin which leads to seroma formation 345 and subsequent wound complications (39). The aseptic wound complications in our study 346 were managed using repeated sterile wound dressings without the routine use of antibiotics. 347 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 11 Management of persistent wound drainage and breakdown can vary with some institutes 348 opting for wound debridement and secondary closure in the operation theatre (20,33). 349 350 Use of intra-wound vancomycin in diabetics and other high-risk groups 351 The use of vancomycin did not appear to significantly influence SSI/PJI rates in diabetics 352 versus non-diabetics in the study population. There was no difference in outcomes based on 353 the BMI classification of patients. 354 355 The strengths of this study include the prospective randomised study design, the trial 356 performed at a single high-volume arthroplasty institute with a standardised protocol 357 followed in all cases and a very low attrition during the follow-up period. 358 359 This study has some limitations. Firstly, this study is powered to detect a 1.7% difference in 360 the PJI rates between the study and control groups. The trial is not adequately powered to 361 detect smaller differences in PJI rates. The study may be underpowered for the minor 362 complications such as prolonged wound drainage and delayed wound healing. This calls for 363 large-volume multi-centre RCTs, with uniform protocols in peri-operative management. 364 Registry-based studies will be high-powered but subject to bias due to varying institutional 365 protocols in the asepsis procedures and antibiotic policies. Secondly, the results in this trial 366 are based on data from a single institute. Although the surgical technique, aseptic precautions 367 and antibiotic policy have been standardised across the study population, the findings may 368 not be generalizable to other institutions dissimilar to ours. Other limitations include sub-369 analysis of other factors potentially contributing to PJI rate such as smoking or individual 370 comorbidities, which were not independently evaluated. 371 372

373 Intra-wound vancomycin powder application does not appear to reduce PJI/SSI rate in 374 primary total knee arthroplasties, including in high-risk groups. The use of intra-wound 375 vancomycin was associated with an increase in postoperative aseptic wound complications 376 such as persistent wound drainage. Intra-wound vancomycin may not be effective in reducing 377 the rate of PJI in primary TKA. 378 379 380 381 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 12

382 1. Ackerman IN, Bohensky MA, Zomer E, Tacey M, Gorelik A, Brand CA, et al. The 383 projected burden of primary total knee and hip replacement for osteoarthritis in 384 Australia to the year 2030. BMC Musculoskelet Disord. 2019 Feb;20(1):90. 385 2. Kim TW, Kang S-B, Chang CB, Moon S-Y, Lee Y-K, Koo K-H. Current Trends and 386 Projected Burden of Primary and Revision Total Knee Arthroplasty in Korea Between 387 2010 and 2030. J Arthroplasty. 2021;36(1):93–101. 388 3. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip 389 and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am. 390 2007 Apr;89(4):780–5. 391 4. Springer BD, Cahue S, Etkin CD, Lewallen DG, McGrory BJ. Infection burden in total 392 hip and knee arthroplasties: an international registry-based perspective. Arthroplast 393 today. 2017 Jun;3(2):137–40. 394 5. Kapadia BH, Berg RA, Daley JA, Fritz J, Bhave A, Mont MA. Periprosthetic joint 395 infection. Lancet (London, England). 2016 Jan;387(10016):386–94. 396 6. Kurtz SM, Lau E, Schmier J, Ong KL, Zhao K, Parvizi J. Infection burden for hip and 397 knee arthroplasty in the United States. J Arthroplasty. 2008 Oct;23(7):984–91. 398 7. Clohisy JC, Calvert G, Tull F, McDonald D, Maloney WJ. Reasons for revision hip 399 surgery: a retrospective review. Clin Orthop Relat Res. 2004 Dec;(429):188–92. 400 8. Fehring TK, Odum S, Griffin WL, Mason JB, Nadaud M. Early failures in total knee 401 arthroplasty. Clin Orthop Relat Res. 2001 Nov;(392):315–8. 402 9. Vessely MB, Whaley AL, Harmsen WS, Schleck CD, Berry DJ. The Chitranjan 403 Ranawat Award: Long-term survivorship and failure modes of 1000 cemented 404 condylar total knee arthroplasties. Clin Orthop Relat Res. 2006 Nov;452:28–34. 405 10. Premkumar A, Kolin DA, Farley KX, Wilson JM, McLawhorn AS, Cross MB, et al. 406 Projected Economic Burden of Periprosthetic Joint Infection of the Hip and Knee in 407 the United States. J Arthroplasty. 2021 May;36(5):1484-1489.e3. 408 11. Kurtz SM, Lau E, Watson H, Schmier JK, Parvizi J. Economic burden of 409 periprosthetic joint infection in the United States. J Arthroplasty. 2012 Sep;27(8 410 Suppl):61-5.e1. 411 12. Dodson V, Majmundar N, Swantic V, Assina R. The effect of prophylactic 412 vancomycin powder on infections following spinal surgeries: a systematic review. 413 Neurosurg Focus. 2019 Jan;46(1):E11. 414 13. Caroom C, Tullar JM, Benton EGJ, Jones JR, Chaput CD. Intrawound vancomycin 415 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 13 powder reduces surgical site infections in posterior cervical fusion. Spine (Phila Pa 416 1976). 2013 Jun;38(14):1183–7. 417 14. Bakhsheshian J, Dahdaleh NS, Lam SK, Savage JW, Smith ZA. The use of 418 vancomycin powder in modern spine surgery: systematic review and meta-analysis of 419 the clinical evidence. World Neurosurg. 2015 May;83(5):816–23. 420 15. Yan H, He J, Chen S, Yu S, Fan C. Intrawound application of vancomycin reduces 421 wound infection after open release of post-traumatic stiff elbows: a retrospective 422 comparative study. J shoulder Elb Surg. 2014 May;23(5):686–92. 423 16. Wukich DK, Dikis JW, Monaco SJ, Strannigan K, Suder NC, Rosario BL. Topically 424 Applied Vancomycin Powder Reduces the Rate of Surgical Site Infection in Diabetic 425 Patients Undergoing Foot and Ankle Surgery. Foot ankle Int. 2015 Sep;36(9):1017–24. 426 17. Johnson JD, Nessler JM, Horazdovsky RD, Vang S, Thomas AJ, Marston SB. Serum 427 and Wound Vancomycin Levels After Intrawound Administration in Primary Total 428 Joint Arthroplasty. J Arthroplasty. 2017 Mar;32(3):924–8. 429 18. Burdon DW. Principles of antimicrobial prophylaxis. World J Surg. 1982 430 May;6(3):262–7. 431 19. Fleischman AN, Austin MS. Local Intra-wound Administration of Powdered 432 Antibiotics in Orthopaedic Surgery. J bone Jt Infect. 2017;2(1):23–8. 433 20. Hanada M, Nishikino S, Hotta K, Furuhashi H, Hoshino H, Matsuyama Y. Intrawound 434 vancomycin powder increases post-operative wound complications and does not 435 decrease periprosthetic joint infection in primary total and unicompartmental knee 436 arthroplasties. Knee Surg Sports Traumatol Arthrosc. 2019 Jul;27(7):2322–7. 437 21. Parvizi J, Gehrke T, Chen AF. Proceedings of the International Consensus on 438 Periprosthetic Joint Infection. Bone Joint J. 2013 Nov;95-B(11):1450–2. 439 22. Parvizi J, Tan TL, Goswami K, Higuera C, Della Valle C, Chen AF, et al. The 2018 440 Definition of Periprosthetic Hip and Knee Infection: An Evidence-Based and 441 Validated Criteria. J Arthroplasty. 2018 May;33(5):1309-1314.e2. 442 23. Ghanem E, Heppert V, Spangehl M, Abraham J, Azzam K, Barnes L, et al. Wound 443 management. J Orthop Res Off Publ Orthop Res Soc. 2014 Jan;32 Suppl 1:S108-19. 444 24. Xu H, Yang J, Xie J, Huang Z, Huang Q, Cao G, et al. Efficacy and safety of 445 intrawound vancomycin in primary hip and knee arthroplasty. Bone Joint Res. 2020 446 Nov;9(11):778–88. 447 25. Makris K, Spanou L. Acute Kidney Injury: Definition, Pathophysiology and Clinical 448 Phenotypes. Clin Biochem Rev. 2016 May;37(2):85–98. 449 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 14 26. Kang DG, Holekamp TF, Wagner SC, Lehman RAJ. Intrasite vancomycin powder for 450 the prevention of surgical site infection in spine surgery: a systematic literature 451 review. Spine J. 2015 Apr;15(4):762–70. 452 27. Tubaki VR, Rajasekaran S, Shetty AP. Effects of using intravenous antibiotic only 453 versus local intrawound vancomycin antibiotic powder application in addition to 454 intravenous antibiotics on postoperative infection in spine surgery in 907 patients. 455 Spine (Phila Pa 1976). 2013 Dec;38(25):2149–55. 456 28. Heckmann ND, Mayfield CK, Culvern CN, Oakes DA, Lieberman JR, Della Valle CJ. 457 Systematic Review and Meta-Analysis of Intrawound Vancomycin in Total Hip and 458 Total Knee Arthroplasty: A Call for a Prospective Randomized Trial. J Arthroplasty. 459 2019 Aug;34(8):1815–22. 460 29. Koutalos AA, Drakos A, Fyllos A, Doxariotis N, Varitimidis S, Malizos KN. Does 461 Intra-Wound Vancomycin Powder Affect the Action of Intra-Articular Tranexamic 462 Acid in Total Joint Replacement? Microorganisms. 2020 May;8(5). 463 30. Yavuz IA, Oken OF, Yildirim AO, Inci F, Ceyhan E, Gurhan U. No effect of 464 vancomycin powder to prevent infection in primary total knee arthroplasty: a 465 retrospective review of 976 cases. Knee Surg Sports Traumatol Arthrosc. 2020 466 Sep;28(9):3055–60. 467 31. Khatri K, Bansal D, Singla R, Sri S. Prophylactic intrawound application of 468 vancomycin in total knee arthroplasty. J Arthrosc Jt Surg [Internet]. 2017;4(2):61–4. 469 Available from: 470 https://www.sciencedirect.com/science/article/pii/S2214963517300330 471 32. Winkler C, Dennison J, Wooldridge A, Larumbe E, Caroom C, Jenkins M, et al. Do 472 local antibiotics reduce periprosthetic joint infections? A retrospective review of 744 473 cases. J Clin Orthop trauma. 2018 Mar;9(Suppl 1):S34–9. 474 33. Dial BL, Lampley AJ, Green CL, Hallows R. Intrawound Vancomycin Powder in 475 Primary Total Hip Arthroplasty Increases Rate of Sterile Wound Complications. Hip 476 pelvis. 2018 Mar;30(1):37–44. 477 34. Xu X, Zhang X, Zhang Y, Chen C, Yu H, Xue E. Role of intra-wound powdered 478 vancomycin in primary total knee arthroplasty. Orthop Traumatol Surg Res. 2020 479 May;106(3):417–20. 480 35. Otte JE, Politi JR, Chambers B, Smith CA. Intrawound Vancomycin Powder Reduces 481 Early Prosthetic Joint Infections in Revision Hip and Knee Arthroplasty. Surg 482 Technol Int. 2017 Feb;30:284–9. 483 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 15 36. Patel NN, Guild GN 3rd, Kumar AR. Intrawound vancomycin in primary hip and knee 484 arthroplasty: a safe and cost-effective means to decrease early periprosthetic joint 485 infection. Arthroplast today. 2018 Dec;4(4):479–83. 486 37. Peng Z, Lin X, Kuang X, Teng Z, Lu S. The application of topical vancomycin powder 487 for the prevention of surgical site infections in primary total hip and knee arthroplasty: 488 A meta-analysis. Orthop Traumatol Surg Res. 2021 Jun;107(4):102741. 489 38. Wong MT, Sridharan SS, Davison EM, Ng R, Desy NM. Can Topical Vancomycin 490 Prevent Periprosthetic Joint Infection in Hip and Knee Arthroplasty? A Systematic 491 Review. Clin Orthop Relat Res. 2021 Aug;479(8):1655–64. 492 39. Hoelen DWM, Tjan DHT, van Vugt R, van der Meer YG, van Zanten ARH. Severe 493 local vancomycin induced skin necrosis. Vol. 64, British journal of clinical 494 pharmacology. England; 2007. p. 553–4. 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 16 List of Tables 518 Characteristics Study Group N=507 Control Group N=515 p-value Mean Age (SD) 61.7 (7.52) 61.4 (7.39) 0.520 * Mean BMI (SD) 28.5 (4.49) 28.4 (4.52) 0.722 * Mean Pre-op Haemoglobin (SD) 12.2 (1.41) 12.3 (1.33) 0.243 * Mean Post-op Haemoglobin (SD) 10.3 (1.34) 10.4 (1.24) 0.215 * Mean drop in Haemoglobin (SD) -1.96 (1.143) -1.91 (1.111) 0.478 * Mean ASA (SD) 2.02 (0.143) 2.01 (0.186) 0.336 * Mean Tourniquet Time (min) (SD) 70.2 (16.35) 71.1 (21.91) 0.457 * Smoking N (%) 58 (11.44) 60 (11.65) 0.916 * CCI Risk Classification Mild risk Moderate risk High Risk 324 (63.9) 156 (30.76) 27 (5.34) 330 (64.07) 164 (31.84) 21 (4.09) 0.278 † Mean CCI (SD) 2.3 (1.15) 2.2 (1.14) 0.163 * Gender Female N (%) 360 (71) 358 (69.51) 0.586 † Male N (%) 147 (29) 157 (30.49) Diabetics N (%) 170 (33.53) 161 (31.26) 0.438 † Hypothyroidism N (%) 85 (16.76) 92 (17.86) 0.642 † CKD N (%) 2 (0.39) 4 (0.77) 0.424 † ASA N (%) ASA I 6 (1.2) 10 (1.94) 0.220 † N (%) ASA II 491 (96.84) 492 (95.53) N (%) ASA III 10 (1.96) 13 (2.53) * independent samples t-test, † Chi-square test 519 Table.1. Demographic and Baseline Characteristics of the study population 520 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 17 521 Study Group N(%) Control Group N(%) p-value Major Complications SSI N (%) 1 2 0.264 † PJI (DAIR) 1 3 Periprosthetic fracture 2 0 Minor Complications Persistent Wound Drainage 43 (8.48) 26 (5.04) 0.010 † Stitch Abscess 20 (3.94) 13 (2.52) Delayed stitch removal (>3 weeks) 36 (7.1) 36 (6.99) 0.945 † Medical Complications Septic shock and MODS 0 1 NS † DVT 0 1 CVA 0 0 Acute kidney injury/ Nephrotoxicity 0 0 Cardiac complications 0 0 †Chi-square test 522 Table 2: Summary of complications in the study population 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 18 539 Table 3: Clinical summary of PJI and SSI cases 540 541 542 543 544 545 546 Patient Age Range CCI Time to Second surgery (weeks) Culture report Diagnosis Treatment Study 1 51-60 1 5 No growth PJI DAIR 2 51-60 1 6 No growth SSI Debridement and secondary suturing Control 1 61-70 4 13 S. aureus PJI • DAIR • Two-stage revision 2 61-70 2 20 No growth PJI DAIR 3 61-70 2 6 No growth PJI DAIR 4 61-70 2 12 S. epidermidis SSI Debridement and secondary suturing 5 61-70 3 16 No growth SSI Debridement and secondary suturing Septic shock and MODS All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 19 547 548 549 Study Group n (%) Control Group n (%) p-value Number of Diabetics 170 (33.5) 161 (31.26) 0.444 † Mean HbA1C (SD) 6.6 (2.04) 6.8 (1.95) 0.109 * DM-CCI grading Mild 61 (35.9) 61 (37.9) 0.439 † Moderate 88 (51.8) 87 (54.0) Severe 21 (12.4) 13 (8.1) Mean CCI (SD) 3.0 (1.03) 2.9 (1.14) 0.141 * Infection rates Major Complications SSI 0 0 0.160 † PJI (DAIR) 0 2 Minor Complications Maceration, n (%) 9 (5.3) 6 (3.5) 0.329 † Stitch Abscess, n (%) 15 (9.3) 4 (2.5) Delayed staple removal (>3 weeks) n (%) 36 (7.5) 36 (7.3) 0.942 † * independent samples t-test, † Chi-square test 550 Table 4: Sub-group analysis of outcomes in the high-risk diabetic patients in both groups 551 552 553 554 555 556 557 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint 20 558 Study Group Minor complications n (%) Control Group Minor complications n (%) p-value BMI grading 0.759 † Normal 16 (26.2) 11 (29.7) Overweight 24 (39.3) 16 (43.3) Class I Obese 15 (24.6) 5 (13.5) Class II Obese 5 (8.2) 4 (10.8) Class III Obese 1 (1.6) 1 (2.7) Total 61 (100.0) 37 (100.0) † Chi-square test 559 Table 5: Sub-group analysis of minor complications based on the classification of obesity 560 561 562 Study Group Major complications n (%) Control Group Major complications n (%) p-value BMI grading 0.507 † Normal 1 (25.0) 0 (0.0) Overweight 2 (50.0) 3 (60.0) Class I Obese 1 (25.0) 2 (40.0) Class II Obese 0 (0.0) 0 Class III Obese 0 (0.0) 0 † Chi-square test 563 Table 6: Sub-group analysis of major complications based on the classification of obesity 564 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted May 8, 2023. ; https://doi.org/10.1101/2023.05.05.23289368doi: medRxiv preprint