Differences in Infection Rates by Surgical Approach in Total Hip Arthroplasty and Patient Sex: A Systematic Review

Abstract

Background

There exists conflicting data that patient sex may influence complication and revision rates when undergoing total hip arthroplasty (THA), specifically when comparing different surgical approaches. Differences in body fat or muscular distribution are proposed mechanisms, but these are poorly understood and not well described in current literature.

Methods

A systematic review of the literature was conducted from PubMed, Embase, and Web of Science from inception of the database through September 15, 2020. Studies were included if they included patients undergoing primary elective unilateral THA, delineated infections by surgical approach, and delineated infections by patient sex. Basic science, cadaveric, and animal studies were excluded as were case reports. Two authors screened abstracts and then extracted data from the full text article.

Results

Three studies, including 1,694 patients undergoing 1,811 THA were included. 80 infections were included. No study reported a statistically significant difference in infection risk by patient sex or surgical approach, though there was substantial heterogeneity in study design, approach, and analysis.

Conclusion

Limited data suggests no relationship between sexes across surgical approaches for infection rates. However, poor reporting and small sample sizes preclude definitive conclusions from being drawn. Future studies should emphasize reporting differences in outcomes by patient sex to better elucidate differences, if any, in adverse outcomes between sexes following THA across surgical approaches.

Level of Evidence: IV

Introduction

The total hip arthroplasty (THA) is one of the most commonly performed orthopaedic surgeries in the United States, with over 400,000 individuals undergoing the procedure every year.¹ In 2010, it was estimated that a total of 2.5 million adults are living with a prosthetic hip.² The surgery is deemed highly successful, with a low risk of revision, a reduction in chronic pain, and improvement in quality of life markers.^3–5 The surgical infection rate is low with a surgical site infection (SSI) rate of 2.5% and a peri-prosthetic joint infection (PJI) rate of 0.9%, however infectious complications are devastating in their economic impact.⁶ The additional cost of a PJI after a THA costs an average of $30,000 per patient, and the total cost of PJI’s after both hip and knee replacements is projected to be $753.4 million in the United States by 2030, with the increase mainly driven by increasing volume.⁶ In addition to the monetary costs, the morbidity and mortality of these infections is devastating with reduced functional hip scores, increased risk of further complications, and extended hospital stays.^7–10

Given the success of the procedure and the staggering number of patients who will undergo THA in the future, it is necessary to understand the risks of infectious complications. In the attempt to find the most successful operation, numerous surgical approaches have been designed to minimize complications and optimize functional outcomes. These include the direct anterior, anterolateral, direct lateral, posterolateral, and direct posterior approaches, each with their own specific risks and benefits. Specifically, the direct anterior approach (DAA) has gained popularity among orthopaedic surgeons due to the minimization of muscular damage as well as the easily hidden post-surgical scar.¹¹ However, concerns have been raised due to the proximity of the incision to the inguinal region and genitals, and some studies suggest an increased risk of complications with this approach.¹²

There is conflicting evidence on the risk of post-operation infectious complications based on sex, with some evidence supporting an increased risk in women, with an increasing risk in obese women without corresponding increase in obese men.^13–15 However, there are several large studies which suggest an increased risk of revision due to infection in men as opposed to women.^16,17 Some of these discrepancies have been attributed to geographic or regional perioperative differences, with European men at a higher risk of revision, and American men at a lower risk.¹⁸ Despite these incongruencies, there is a lack of research into further specific risk factors between men and women. Moreover, there have been no systematic attempts to evaluate the available research on sex specific infection rates based on THA surgical approach. This study therefore seeks to evaluate the available evidence regarding the risk of infection by the various THA approaches and further sub-stratifying by patient sex.

Methods

In order to perform an objective and thorough examination of the available data, this study was performed under the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines (Figure 1).¹⁹

Figure 1.

A PRISMA diagram outlining the studies eligible for and included in the present review.

Literature Search

A systematic review was conducted using PubMed, Embase, and Web of Science databases in order to identify studies that reported on infection rates stratified by sex and surgical approach following a primary elective THA. The databases were searched from their inception through September 15, 2020. Using Boolean operators AND and OR, the following terms were searched: “Infection”, “surgical site infection” “superficial infection”, “deep infection”, “postoperative complication”, “postoperative infection”, “surgical approach”, “anterior approach”, “lateral approach”, “posterior approach”, “anterolateral approach”, “posterolateral approach”, “comparison of approach”, “arthroplasty, replacement, hip”, “hip prosthesis”, “THA”, ”total hip”, “total hip replacement”, and “hip” with associated wildcards and MeSH terms.

Abstracts were selected if they were primary observational studies or randomized controlled trials (RCTs) written in the English language that also met the following criteria: 1) Patients underwent a primary elective unilateral THA procedure; 2) The study reported on infections by surgical approach; and 3) The study reported on infections by patient sex. Complications included diagnosis of superficial wound infection, wound dehiscence, deep wound infection, periprosthetic joint infection (PJI), or total infection rates. To avoid bias from non-standard approaches or, non-total hip arthroplasty, studies evaluating experimental minimal incision variants, bilateral THA’s, hemiarthroplasties, or robot-assisted surgeries were excluded. In addition, publications were excluded if they were basic science studies, cadaveric studies, animal studies, case reports, letters to the editor, editorials, personal correspondences, or review articles.

Two of the investigators (DTW and NVS) reviewed all titles and abstracts from the identified articles. Following review of selected abstracts, the full articles were obtained for all but two studies. Despite best efforts, the full text of two studies were unavailable for access. The references contained in all of the articles were used to further identify studies that were not captured in the initial database search. The authors then reviewed the full- text articles for inclusion/exclusion criteria. If a disagreement or uncertainty arose among the articles included, another author aided in resolution. The final inclusion or exclusion was unanimously decided.

Results

Following review of the available full-texts, two studies met inclusion criteria. Five additional studies that would be eligible for inclusion if sex related infection data was reported were identified. The corresponding authors of these papers were contacted to request infection data stratified by patient sex. The specific data was obtained for one of the five additionally identified studies. Therefore, that resulted in a total of three publications from which data was gathered. Cumulatively this data included 1,694 patients, who underwent 1,811 THA operations (several studies evaluated consecutive THA operations on the same patient). In total 80 infective wound complications were reported.

Jahng et al.²⁰ performed a retrospective study evaluating risk factors for post-operative complications in patients undergoing THA via the DAA. They included a consecutive 651 DAA THA operations in 611 patients performed by two surgeons over the course of three years. Forty patients underwent consecutive THAs on the contralateral hip, none were performed simultaneously. The author’s primary outcome measure was wound complications, which include wound dehiscence and superficial infection among others, requiring additional treatment in the first 90 days post- operatively. A total of 364 females and 287 males were included. Of the 364 females that underwent surgery, 37 (10.1%) sustained complications of which five (1.3%) required reoperation. Of the 287 males undergoing THA, there were 38 (13.2%) wound complications of which eight (2.7%) required reoperation, the difference between groups was non-significant (p=0.217). Univariate risk factors identified included diabetes mellitus (OR 4.1; 95% confidence interval [CI] 2.1-8.0), smoking (OR 2.9; 95% CI 1.2-7.0), and obesity. Obesity severity increased risk with class I (OR 2.1; 95% CI 1.2-3.9), obesity class II (OR 2.8, 95% CI 1.2-6.4), and morbid obesity (OR 9.7; 95% CI 3.8-24.8). Multivariate analysis also identified obesity and diabetes mellitus as risk factors.

Tsai et al.²¹ performed a prospective analysis on 1,003 patients undergoing a consecutive 1,077 anterolateral THAs. They had a mean follow-up period of 59 months. They did not specify primary outcome measures, but recorded perioperative factors such as incision length, blood loss, type of implant etc, as well as postoperative complications, which included infections. Of these 1,003 patients, 560 were male and 443 were female. Four infectious complications occurred, three males (0.5%) and one female (0.2%) experienced surgical site infections. They did not report whether these differences between the two sexes was significant. No risk factors for infectious or noninfectious complications were identified in this study.

Fransen et al.²² performed a retrospective cohort study evaluating all patients who underwent DAA THA without a fluoroscopy table, or posterolateral approach (PLA) THA by a single surgeon within one year. This was performed in a large volume hospital in the attempt to find differences in outcomes between the two approaches while accounting for the possible bias of a fluoroscopic table used in DAA that was reported in prior studies. The primary outcome was Harris Hip Score results (HHS), with secondary outcomes including infections and other complications. A total of 80 patients undergoing 83 operations were identified. Forty-five patients underwent the DAA approach, of which 15 were males and 30 were females. Whereas 35 patients underwent the PLA, of which 13 were male and 22 were female. Three of the patients undergoing PLA underwent consecutive contralateral surgeries totaling 38 hips undergoing a THA using the PLA approach. The authors identified only a single patient who experienced a superficial wound infection, a female in the PLA group (0.4%) and statistical significance was not reported. No infections in the DAA group were reported. No infectious risk factors including sex were reported.

In summary, this study identified four studies which reported on infection rates specific to both patient sex and surgical approach. None of the identified studies attempted to find a statistical difference in outcomes by sex; and the groups were too varied for group analysis and the studies were widely ranging in their surgical approaches, reported outcomes, and number of patients.

Discussion

Total hip arthroplasties are an effective operation with a favorable safety profile and low perioperative infection rate.^3–6 However, when infections arise, they can significantly impact quality of life, and drastically increase healthcare system costs.^23–25 It is crucial to identify risk factors for these adverse outcomes to reduce the morbidity and mortality inherent to these complications. Furthermore, multiple surgical approaches for performing THAs exist, all with their own risks and benefits. Stratification of risk factors for each approach allows for improved clinical decision making and peri-operative planning.

Obesity is another known risk factor with conflicting results. Obesity may increase risk on women but not men.¹⁵ Other studies have found obesity to be a risk factor, without a difference between men and women.²⁶ Concerns have been raised that the proximity to the groin would increase infection risk in DAA operations, as would an overhanging pannus in obese patients, it is unknown if the differing fat distribution among men and women would influence this.²⁷ However, it has been reported that DAA approach may be associated with increased infection risk when compared to other approaches.^12,27,28 Further prospective research is warranted to determine the effect of obesity on THA outcomes and the impact that patient sex has on this relationship.

This systematic review sought to answer whether sex and surgical approach were important variables for postoperative infection risk. The study identified a total of 1,833 unilateral THA operations performed on a total of 1,694 patients; these included 1077 hips undergoing the anterolateral approach, 696 hips undergoing the DAA approach, and 38 undergoing the posterolateral approach.^20–22 Only the DAA and PL were directly compared within the included studies.²² Eighty infective wound complications were reported, of which 39 occurred in females versus 41 in males. Thirty-seven of the complications that occurred in females were from those undergoing the DAA approach compared to 38 of the infections in males. None of the studies reported a statistical difference between these rates or identified sex as a risk factor for postoperative complications.

Statz et al.²⁹ reported on a series of 1573 DAA THA performed at a single institution and found that 18 patients subsequently required an irrigation and debridement (I&D) for superficial wound dehiscence. The authors identified body mass index (BMI) greater than 30, 35, and 40 to all be independent predictors for the need for a superficial I&D. They further reported that female sex was independently associated with the need for I&D (HR: 5.5). Recently, Bendich et al.³⁰ compared the DAA and posterior approach for multiple adverse postoperative events including infection. They reported no difference in risk of infection (OR: 1.5, 95% CI: 0.5-5.2) between surgical approaches. While the authors did not explicitly compare infection risk by sex, they reported 6/1498 female patients and 3/850 male patients undergoing a DAA THA and 3/1498 female patients and 3/850 male patients undergoing THA by posterior approach developed an infection. This data further underscores the need for more prospective research that delineates postoperative infection by sex and surgical approach.

There are several limitations to this systematic review that should be considered. The primary limitation is the paucity of data on the outcomes of interest. The total sample size of this review was 1,716 patients, with the majority of this from just two studies, Jahng et al.²⁰ and Tsai et al.²¹ The other study included the final 80 patients.²² Within this the majority of the data was comprised of just two approaches, and these studies did not compare surgical approaches. The small sample size reduces the power of this study and ability to find statistical differences that may be present.

Additionally, one of the studies by Tsai et al.²¹ was prospective in nature, whereas the other three studies were retrospective, which reduces the generalizability of their findings. The disparate methods of the present studies, outcomes of interest, and surgical approaches used do not allow for the combination of the present data to increase sample size and generalizability of their findings. For example, there is no standardization between what is defined as a severe compared to a mild infectious complication. These limitations reduce the ability to draw meaningful conclusions with respect to the study outcomes of interest. What this review highlights is a lack of data concerning an important clinical question. This study identified 79 studies that did not report on infection rates by patient sex, and five studies that directly related to the question of interest but lack outcome reporting by sex. This emphasizes a need for a shift in the norms of what data is reported in studies that examine THA outcomes.

The lack of data, combined with the aforementioned conflicting data concerning the differences in outcomes between THA and patient sex points to the necessity for further research. The collection of patient sex is a standard aspect of study design, and it is an easy step for all authors to include complication rates by patient sex in all future studies. This will allow future meta- analyses to take place, which can effectively answer the question of whether surgical approach and postoperative infections vary between male and female patients, allowing potential factors driving these differences to be identified. Additionally, large scale database studies and prospective trials should examine these differences directly. The results of these will allow for a better personalization of care for patients, and a reduction in postoperative complications.