The application of topical antibiotics for the prevention of infections in primary joint arthroplasty. An umbrella review of systematic reviews and meta‐analysis

Abstract

This umbrella review aim to explore the effect of topical antibiotics in infection prevention after primary joint arthroplasty, and provide a specific theoretical basis for clinical treatment. The review process was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines. We searched PubMed, EMBASE, Medline, and the Cochrane Library on infection prevention by topical antibiotics from inception to 10 April 2023. The two researchers individually and strictly screened the literature according to the inclusion and exclusion criteria, performed the literature quality evaluation and data extraction, and used Stata 17 for data analysis. This study included six studies with one systematic review and five meta‐analyses. The pooled analysis showed that topical antibiotic administration effectively reduced the incidence of overall infection and periprosthetic joint infection. However, it does not reduce the risk of superficial infection. Besides, the topic of antibiotics significantly increases the incidence of other sterile complications of the incision. According to the current evidence, topical application of antibiotics can reduce the incidence of overall infection and periprosthetic joint infection after primary joint arthroplasty. Although it increases the incidence of complications such as delayed healing of incisions, the pros and cons should be weighed in clinical decision making. However, they should not be discarded due to side effects.

Abbreviations

CI

confidence intervals

ES

effect size

OR

odds ratio

PJI

periprosthetic joint infection

1. INTRODUCTION

Joint arthroplasty is an effective treatment for end‐stage osteoarthritis. Implanting artificial joints can effectively reduce patients' pain, correct joint deformities and dysfunctions, restore joint motor function, and improve patient's quality of life. ¹ With the aging of the population, the number of joint arthroplasty surgeries has gradually increased, and the occurrence of postoperative complications has risen steadily. Periprosthetic joint infection (PJI) is one of the severe and catastrophic complications of arthroplasty, with an incidence of 1%–2.5%. ² , ³ Once infection occurs around the prosthesis, it can lead to joint pain, limited joint deformity, disability, and even death, which seriously prolongs the patient's hospital stay, increases the psychological and physical pain of the patient, and even accelerates the patient's death, increasing the economic burden of the medical system. ⁴ , ⁵

Rational and standardized prophylactic antibiotics effectively reduce joint infection clinically. ⁶ The most common pathogens responsible for PJI are Staphylococcus aureus and coagulase‐negative Staphylococcus. ⁷ Incorporating gentamicin in bone cement effectively reduces the risk of periprosthetic infection. ⁸ However, Oliveira et al. have explored that topical gentamicin in the wound is ineffective in lowering periprosthetic infection. ⁹ With the increasing number of methicillin‐resistant Staphylococcus aureus (MRSA) infections in joint replacement, the frequency of clinical prophylactic use of narrow‐spectrum powerful antibiotics such as vancomycin has gradually increased. ¹⁰ The systemic reaction of topical application of vancomycin powder in the incision is slight, which can make the local drug concentration at a high level for a long time, thereby killing bacteria. Its effectiveness has been demonstrated in spinal surgery. ¹¹ In recent years, vancomycin has also been studied in patients with joint arthroplasty and has been found to significantly reduce the risk of PJI. ¹ , ¹² , ¹³ , ¹⁴ Given the relatively small incidence of PJI, clinical studies on this topic have been ambiguous, with numerous previous systematic reviews and meta‐analyses of the use of topical antibiotics in TJA with varying participants and outcomes. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ To this end, we conducted an umbrella review of relevant systematic reviews and meta‐analyses, categorized the evidence levels of each systematic review, summarized current knowledge on this topic, comprehensively assessed the efficacy and safety of topical antibiotics in patients with joint replacement, and provided a basis for decision making for clinical treatment.

2. METHODS

This meta‐analysis was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) statement. ²¹

2.1. Search strategy

Two authors (ZZ and WFC) searched PubMed, EMBASE, Medline, and the Cochrane Central Register of Controlled Trials databases from inception to 10 April 2023, using the keywords ‘(local or topical or intrawound or powder* or tissue* or regional* or wound*) and (Meta‐Analysis OR metaanaly* OR meta‐analy* OR Systematic review OR systematic review) AND (gentamicin or bacteria* or antibiotic* or Vancomycin) AND (Arthroplasty OR Replacement) AND (knee OR hip OR shoulder) AND (infect* OR complication).’ No language restrictions were applied.

2.2. Inclusion criteria

The inclusion criteria were as follows: (1) meta‐analysis or systematic review study design; (2) study of effects of topical antibiotics on infection or wound complications in patients undergoing joint arthroplasty.

2.3. Exclusion criteria

The exclusion criteria were as follows: (1) literature review or animal trial design; (2) inadequate reporting of data; and (3) lack of a quantitative synthesis.

2.4. Data extraction

Two researchers (ZYC and QHP) individually screened the retrieved studies against the inclusion and exclusion criteria. After removing duplicates, studies that met the inclusion criteria were subjected to title and abstract review. In cases of disagreement, the senior researcher (HL) intervened. Eligible studies were thoroughly evaluated based on the level of comparison, random‐effects summary, and I² statistics. Data on relevant outcome measures were extracted from the included studies, including author year, country, outcome, number of studies, I ² statistic, and effect size (ES) reported with 95% confidence intervals (CI).

2.5. Assessment of methodological quality

The methodological quality of each included meta‐analysis was assessed using the Assessment of Multiple Systematic Reviews (AMSTAR‐2) tools by two independent investigators (QHP and HL). The AMSTAR‐2 ranks the quality of a meta‐analysis from low to high according to 16 predefined items.

2.6. Data synthesis and analysis

Results from meta‐analyses were pooled using a 95% CI and odds ratio (OR). For studies that reported risk ratios, data were converted to odds ratios based on the primary data included in the review. The meta‐analysis results were precisely the same as those reported in the manuscript for studies with reported odds ratios. Inter‐study heterogeneity was indicated by I ², but values greater than 50% indicated high heterogeneity, and those greater than 75% indicated very high heterogeneity. We calculated small‐sample effect sizes using the regression asymmetry test, considered the effect size of the most extensive dataset, and estimated the power of each constituent study using an algorithm of non‐central t‐distributions. All statistical analyses were performed using Stata version 17.0 (StataCorp, College Station, TX, USA).

3. RESULTS

A total of 225 relevant studies were identified, and 80 duplicates and 138 irrelevant studies were removed after initial screening by reading titles and abstracts. Please read the full text to rescreen the remaining seven articles, one of which was an update of previous studies, so we included only the most recent study. A total of six studies were included in our review. The flow diagram is shown in Figure 1. Wong et al. only conducted a systematic review and did not perform a meta‐analysis, ¹⁹ so this study was included in the systematic review.

FIGURE 1.

Flow diagram of the study search and selection processes.

3.1. Study characteristics

In the six included studies, sample sizes ranged from 3714 to 33 731. The survey by Saidahmed et al. included the topical use of gentamicin and vancomycin. ¹⁸ The rest of the studies only involved topical vancomycin. Three studies were on primary joint arthroplasty, ¹⁷ , ¹⁸ , ²⁰ and the remaining three were on primary and revision arthroplasty. ¹⁶ , ¹⁹ , ²² Outcomes included overall infection, PJI, and complications. The included studies were all non‐randomized controlled studies, of which Movassaghi et al. did not report specific quality assessment tools, ¹⁶ and Saidahmed et al. ¹⁸ Xu et al. ²⁰ used a methodological index for non‐randomized studies and risk of bias in non‐randomized studies of interventions individually, and three used NOS scales. ¹⁷ , ¹⁹ , ²² In terms of fund support, three studies mentioned the relevant information on fund support. ¹⁷ , ²⁰ , ²² The essential characteristics of the included studies are shown in Table 1.

TABLE 1.

Characteristics of the included studies.

Study	Country	IRange of years	No. of primary studies	Duration	No. of subject	Outcomes	I 2	Statistical significance	Summary effect size	Funding	AMSTAR score
Martin 2022	China	2017–2021	14	Minimum 3 month	22 753	PJI	53%	p < 0.00001	OR 0.39 (95% CI: 0.26, 0.58)	Yes	Critically low
						Aseptic wound complications	90%	p = 0.65	OR 0.78 (95% CI: 0.27, 2.28)
						Deep venous thromboembolism	0%	p = 0.73	OR 1.29 (95% CI: 0.31, 5.42)
Movassaghi 2022	USA	2010–2021	16	Minimum 3 month	33 731	Overall infection	0%	p < 0.05	OR 0.44 (95% CI: 0.32, 0.60)	None	Critically low
						Aseptic wound complications	‐	‐	‐
Peng 2021	China	2010–2020	9	Not limited	4512	Overall infection	0%	p < 0.0001	OR 0.39 (95% CI: 0.26, 0.61)	Yes	Critically low
						PJI	0%	p < 0.0001	OR 0.36 (95% CI: 0.22, 0.59)
Saidahmed 2020	Canada	2010–2019	9	Not limited	3714	PJI	0%	p = 0.002	OR 0.51 (95% CI: 0.33, 0.79)	None	Critically low
						Superficial infection	0%	p = 0.43	OR 0.60 (95% CI: 0.17, 2.12)
						Aseptic wound complications	16%	p = 0.03	OR 2.44 (95% CI: 1.12, 5.34)
Wong 2021	Canada	2017–2020	9	Minimum 3 month	6255	PJI	‐	‐	‐	None	Critically low
						Aseptic wound complications	‐	‐	‐
Xu 2020	China	2010–2019	9	Not limited	4607	PJI	0%	p < 0.05	OR 0.44 (95% CI: 0.28, 0.69)	Yes	Critically low
						Superficial infection	0%	p > 0.05	OR 0.60 (95% CI: 0.17, 2.12)
						Aseptic wound complications	16.2%	p < 0.05	OR 2.44 (95% CI: 1.12, 5.34)

Abbreviations: CI, confidence interval; OR, odds ratio; PJI, periprosthetic joint infection.

3.2. Assessment of methodological quality

Four of the included studies restricted the publication language to English ¹⁶ , ¹⁷ , ¹⁸ , ²² ; two studies did not claim to have a protocol, ¹⁹ , ²² and one study did not mention the risk of bias assessment tool. ¹⁶ None of the studies provided further methodological explanations for the risk of bias for allocation concealment, blinding, and other outcomes. In addition, in the calculation of pooled results, the results were not combined after the adjustment for confounders; instead, they were connected directly, ¹⁶ , ¹⁷ , ¹⁸ , ²⁰ , ²² resulting in non‐compliance with essential item 11. Three studies did not discuss the risk of bias at the time of outcome analysis, ¹⁸ , ¹⁹ , ²⁰ resulting in non‐compliance with essential item 13. Finally, all six included studies had critically low methodological quality (Table 1).

3.3. Effect of intrawound antibiotics on overall infection

The efficacy of intrawound antibiotics on overall infection was reported in two studies. ¹⁶ , ¹⁷ The pooled result showed that topical antibiotics effectively reduced the incidence of overall infection in primary joint arthroplasty (ES = 0.42; 95% CI: 0.31–0.53, p < 0.0001; Figure 2). No heterogeneity was detected between the included studies (I ² = 0%, p = 0.662).

FIGURE 2.

Intrawound antibiotic versus control and associations with overall infection.

3.4. Effect of intrawound antibiotics on PJI

A total of four studies were reported on PJI. ¹⁷ , ¹⁸ , ²⁰ , ²² Among them, Martin et al., ²² including primary and revision arthroplasty, eliminated the revision cases, recalculated the results to obtain the OR, and then combined them with other results. The pooled result showed that topical antibiotics effectively reduced the incidence of PJI (ES = 0.41; 95% CI: 0.32–0.51, p < 0.0001; Figure 3). No heterogeneity detected (I ² = 0%, p = 0.766).

FIGURE 3.

Intrawound antibiotic versus control and associations with periprosthetic joint infection.

3.5. Effect of intrawound antibiotics on superficial infection

Superficial infection was reported in two studies. ¹⁸ , ²⁰ Topical antibiotics do not prevent superficial infection with primary joint replacement (ES = 0.60; 95% CI: −0.09–1.29, p = 0.088; Figure 4).

FIGURE 4.

Intrawound antibiotic versus control and associations with superficial infection.

3.6. Effect of intrawound antibiotics on aseptic wound complication

Three meta‐analyses examined the impact of intrawound antibiotics on aseptic wound complications. ¹⁸ , ²⁰ , ²² The pooled effect size revealed that intrawound antibiotics increase the aseptic wound complication compared with control (ES = 1.57; 95% CI: 0.35–2.79, p = 0.012; Figure 5).

FIGURE 5.

Intrawound antibiotic versus control and associations with aseptic wound complication.

3.7. Publication bias

Considering our meta‐analysis's small sample size (< 10), funnel plot analysis was not applicable for determining publication bias.

4. DISCUSSION

This study aimed to systematically collect evidence from multiple meta‐analyses on the efficacy of topical antibiotics in patients with primary joint arthroplasty and to integrate it into one accessible and up‐to‐date source. This study investigated the results of six studies, demonstrating that intrawound antibiotics could significantly decrease overall infection and PJI. Nevertheless, no meaningful change was observed in superficial infection. Besides, the intrawound antibiotics also increase the aseptic wound complication.

In spinal surgery, topical antibiotics applied intrawound effectively reduce the incidence of infection with definite efficacy and few complications. ²³ Given the catastrophic and refractory nature of periprosthetic infections, topical antibiotics have also gradually begun to be promoted in patients with joint arthroplasty. Presently, gentamicin and vancomycin powder are mainly used, and the risk of MRSA infection has steadily increased in recent years, so the current clinical topical antibiotics are especially vancomycin. ¹ , ²⁰ , ²⁴ Gentamicin powder has only been reported in one study, ⁹ and has concluded that the risk of infection does not decrease. Several studies have reported the effectiveness of topical vancomycin in reducing postoperative joint infection, ¹⁴ , ²⁵ and our findings summarize previous studies that have shown that vancomycin is effective in reducing the risk of overall infection and PJI after primary joint arthroplasty. Although topical vancomycin is used chiefly clinically in the joint cavity and deep fascia, the penetration of locally acting vancomycin into the superficial surface may affect the healing of local epidermal tissue. In addition, locally high concentrations of vancomycin may lead to increased exudation of cellular and interstitial fluid, resulting in secondary hematoma formation. ²⁶ For superficial infection, mainly includes whether there is redness and swelling at the local incision, and whether the skin temperature is elevated. However, the presentation of these clinical symptoms cannot be attributed to vancomycin alone. Moreover, these diagnoses are mostly subjective by clinicians, and there are no relevant quantitative indicators to clarify the diagnosis, resulting in a bias in the results. Although vancomycin can penetrate the superficial tissue, the incision in the outer layer is bandaged with dressings such as gauze, and superficial wound infection may still not be avoided during clinical dressing changes and environmental factors. Hence, we should also pay attention to superficial infections; if they are not well controlled, more secondary PJI occurs. Therefore, when considering whether to use vancomycin, it is necessary to weigh the advantages and harms of reducing infection rates and increasing incision complications.

Over the past 3 years, numerous systematic reviews and meta‐analyses have been conducted to evaluate the effectiveness of topical antibiotics for preventing infection. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²² However, these reviews differ in that they include only patients with a primary joint replacement in some reviews, while patients with primary and revision are included in others. In addition, the primary outcome was different in different thoughts. Some studies reported overall infection rates, while others said only superficial and periprosthetic infections, which led to different results in terms of efficacy. As the aim of these reviews is to provide readers with easily accessible and high‐quality information, the quality of these reviews also needs to be assessed before due consideration is given to conclusions or recommendations. To this end, we carry out this umbrella meta‐analysis.

Periprosthetic total hip infection ranks as the third most prevalent reason for total hip revision in the United States, exhibiting an incidence of 0.7% to 2.2%. ²⁷ , ²⁸ , ²⁹ According to Kurtz et al., utilizing the Nationwide Inpatient Sample, there has been a noteworthy increase in the incidence of periprosthetic total hip infections—from 1.99% in 2001 to 2.18% in 2009, reflecting an average 5.3% relative annual increase, even when adjusting for patient demographic factors. ²⁷ This complication, with its substantial morbidity and mortality, necessitates multiple surgical interventions, prolonged hospital stays, extended courses of intravenous antibiotics, and extended periods of rehabilitation and work absence. These factors collectively impose significant financial strain on both the individual and the healthcare system. In 2009, the cost of periprosthetic total joint infections to the U.S. healthcare system was $566 million, and projections indicate an increase to $1.62 billion by 2020. ²⁷ Due to the low cost of only $12 per vial, the expected cost savings of locally using vancomycin in total hip arthroplasty amount to $904 per patient. This aligns with cost‐effectiveness studies in the spinal literature, which report potential savings of up to $4400 per case. ³⁰ , ³¹ Calculated based on the cost of $12 per vial of vancomycin, the break‐even point for the Cost‐Effectiveness Analysis is only 0.01% of the Absolute Risk Reduction (ARR). This suggests that it is a cost‐conscious strategy for reducing infections. Risk reduction is sensitive only to drug costs and complications, while being insensitive to the infection rate, making it meaningful in various scenarios. ³²

The consensus in the scientific community is that exposure to antibiotics heightens the risk of drug resistance emergence. ³³ Employing antibiotic stewardship, a practice aimed at minimizing antibiotic exposure in surgical patients, is widely regarded as a best practice. ³⁴ In the context of using topical powdered antibiotics, the risk of resistance increase is theoretical and challenging to quantify. A specific concern lies in sustained systemic exposure to sub‐inhibitory levels of vancomycin, which may favor the development of resistant strains. In an in vitro model, the development of vancomycin‐intermediate resistant Staphylococcus was demonstrated with continuous vancomycin exposure exceeding 10 mg/L. ³⁵ However, it is crucial to note that experimental exposures were 5‐fold higher than the peak systemic levels typically achieved with topical intra‐wound vancomycin.

A separate study assessed changes in the microbiologic properties of a methicillin‐resistant Staphylococcus aureus (MRSA) strain in a patient subjected to chronic vancomycin exposure over a 9‐month period. ³⁶ No resistance was detected, and only minimal alterations in vancomycin susceptibility were observed. Lastly, studies on the use of topical vancomycin in spine surgery have not reported the emergence of vancomycin resistance. ²³ Given the significant public health concern related to antibiotic resistance, surgeons must carefully evaluate the risks and benefits associated with the use of topical antibiotics. ³⁷

4.1. Strength and limitation

To our knowledge, this is the first umbrella of meta‐analysis to present the existing evidence regarding the effects of intrawound antibiotics on infection and aseptic wound complication in primary joint arthroplasty. This study includes all available primary evidence. As such, it represents a comprehensive overview of the topic and critically analyzes previously conducted meta‐analyses. Given the catastrophic consequences of PJI, how to scientifically prevent PJI and its complications deserves further exploration. In addition, this study will generate significant interest from the public.

Our study had the following limitations. First, the included studies included 19 non‐randomized controlled studies with a low level of evidence, while 15 studies overlapped and were included in multiple reviews, which may have resulted in double counting of results. In this review, studies with earlier publication dates were included in more meta‐analyses and may therefore be over‐represented which biased our outcomes. Second, three studies did not limit the length of follow‐up. ¹⁷ , ¹⁸ , ²⁰ Less than 3 months of follow‐up may result in partial impairment of outcome events, which increases the bias of the result. Third, the methodological quality of all included studies could have been much higher, and to some extent, our confidence in the results could have been higher. In evidence‐based medicine, high‐quality systematic reviews are an essential source of evidence to optimize clinical use. Further high‐quality meta‐analysis studies and RCT studies are needed to confirm these findings.

5. CONCLUSIONS

According to the current evidence, intrawound antibiotics could significantly decrease overall infection and periprosthetic joint infection. However, it also increases the aseptic wound complication.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflict of interest.

Luo H, Chen Z, Pan Q, Mei H, Chen W, Zhu Z. The application of topical antibiotics for the prevention of infections in primary joint arthroplasty. An umbrella review of systematic reviews and meta‐analysis. Int Wound J. 2024;21(3):e14726. doi: 10.1111/iwj.14726

DATA AVAILABILITY STATEMENT

No data are available.