Correlation between body mass index and two-stage revision failure of periprosthetic joint infection following total joint arthroplasty: A systematic review and meta-analysis

Abstract

Background: Although the correlation between body mass index (BMI) and two-stage revision failure of periprosthetic joint infection (PJI) following total joint arthroplasty (TJA) have been frequently reported, the results remain controversial. Therefore, the correlation between them was systematically evaluated and meta-classified in this study. Methods: Literature on the correlation between BMI and two-stage revision failure of PJI following TJA was retrieved in PubMed, Embase and Cochrane Library due May 2020. Stata 13.0 software and Cochrane Collaboration Review Manager software (RevMan version 5.3) were applied to data synthesis, subgroup analysis, analyses of publication bias, and sensitivity. Results: A total of 15 observational studies included 1267 patients, of which 15 studies were included in systematic review and 11 studies in meta-analysis. Eight studies found a correlation between BMI and two-stage revision failure of PJI following TJA, but seven other studies found no correlation. Meta-analysis found that the risk of two-stage revision failure of PJI following TJA significantly boosted by 3.53 times in patients with BMI ≥ 30 kg/m² (OR = 3.53; 95% CI = 1.63–7.64 for the BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²) and the risk of two-stage revision failure of PJI following TJA significantly increased by 2.92 times in patients with BMI ≥ 40 kg/m² (OR = 2.92; 95%CI = 1.06–8.03 for the BMI ≥ 40 kg/m² vs. BMI < 30 kg/m²). The subgroup analysis showed that significant association was observed among the studies performed in TKA (OR = 3.63; 95% CI = 2.27–5.82), but not among those conducted in THA (OR = 3.06; 95% CI = 0.42–22.19). A significant association remained consistent, as indicated by sensitivity analyses. Because there are too few studies that can be combined in the included studies, Egger’s and Begg’s tests were not performed. Conclusion: Meta-analysis suggests that the risk of two-stage revision failure of PJI following TJA significantly boosted in obese patients. However, because there may be publication bias of this study, combined overall systematically evaluated and meta-analysis results, we cannot yet conclude that BMI is associated with two-stage revision failure of PJI following TJA.

Keywords

body mass index periprosthetic joint infection two-stage revision systematic review meta-analysis

Introduction

Total joint arthroplasty is one of the most important surgical operations in the 21st century, it is widely used in advanced arthritis. Periprosthetic joint infection is one of the most severe complications that occur in patients who undergo total joint arthroplasty. At present, the two-stage revision is the main way to treat the periprosthetic joint infection following total joint arthroplasty, but there is still a risk of failure after the two-stage revision, reinfection rates reported in the literature range from 4 to 50%.^1–4

More than 500 million people are obese worldwide, including one-third of men and women in the United States.^5,6 Obesity increases the risk of osteoarthritis, and it is a key component in the sustained rise in the number of patients undergoing TJA. Obesity (BMI [body mass index], ≥ 30 kg/m²) is a well-established risk factor for PJI.^7–11 However, the role of obesity or BMI in two-stage revision failure of PJI following TJA is unclear.

The correlation between BMI and two-stage revision failure of PJI following TJA have been reported by many lately, but the results of studies are divergent and even controversial. In this study, we retrieved the published literature on two-stage revision failure of PJI following TJA, extracted high-relevant data for a systematic review and meta-analysis, and evaluated the correlation between BMI and two-stage revision failure of PJI following TJA, in a bid to provide a reference for the prevention of two-stage revision failure of PJI.

Methods

Search strategies

This study was executed in line with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)¹² and reported based on the guidelines developed by the Meta-Analysis of Observational Studies in Epidemiology group.¹³ Because all the analyses were performed on the basis of previous published studies, no ethical approval or patient consent was required. In the initial screening, two investigators conducted the main search in the electronic databases of PubMed, Embase, and Cochrane Library to retrieve eligible articles on the correlation between BMI and two-stage revision failure of PJI following TJA from the inception of the databases to May 2020, without restrictions to languages, publication types or regions. The combined terms of Medical Subject Headings (MeSH) and non-MeSH were searched as follows: “Prosthesis-Related Infections,” “Prosthesis-Related Infections,” “Prosthesis Related Infections,” “Infections, Prosthesis-Related,” “Prosthesis-Related Infection,” “periprosthetic joint infection,” “prosthetic joint infections,” “periprosthetic infections,” “infection of joint,” “joint infection,” “PJI,” “Reoperation,” “Surgical Revision,” “Surgery, Repeat,” “Revision, Surgical,” “Revision Surgery,” “Revision Surgeries,” “Surgery, Revision,” “Repeat Surgery,” “Revision, Joint,” “Joint Revision,” “two-stage exchange,” “two stage exchange,” “2 stage exchange,” “two-stage revision,” “two stage revision,” “2 stage revision,” “second stage revision,” “2 stage revision,” “TSR,” “Body Mass Index,” “Index, Body Mass,” “Quetelet Index,” “Index, Quetelet,” “Quetelet’s Index,” “Quetelets Index,” “BMI,” “Obesity,” “fat,” and “Obese.” A third investigator irrelevant to the initial procedure was consulted in case of any discrepancy. Taking the pubmed database as an example, the literature search strategy is shown in Table 1.

Table 1.

The Pubmed database literature search strategy.

#1	“Prosthesis-related Infections” (Mesh)
#2	Prosthesis-related infections
#3	Prosthesis related infections
#4	Infections, prosthesis-related
#5	Prosthesis-related infection
#6	Periprosthetic joint infection
#7	Prosthetic joint infections
#8	Periprosthetic infections
#9	Infection of joint
#10	Joint infection
#11	PJI
#12	#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11
#13	“Reoperation” (Mesh)
#14	Reoperation
#15	Surgical revision
#16	Surgery, repeat
#17	Revision, surgical
#18	Revision surgery
#19	Surgery, revision
#20	Repeat surgery
#21	Revision, joint
#22	Joint revision
#23	Two-stage exchange
#24	Two stage exchange
#25	2 stage exchange
#26	Two-stage revision
#27	Two stage revision
#28	Two stage revision
#29	second stage revision
#30	TSR
#31	#13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30
#32	“Body Mass Index” (Mesh)
#33	Body mass index
#34	Index, body mass
#35	Quetelet index
#36	Index, quetelet
#37	Quetelet’s index
#38	Quetelets index
#39	BMI
#40	Obesity
#41	Fat
#42	Obese
#43	#32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43
#44	#12 AND #31 AND #43

Study selection criteria

Two independent investigators analyzed the initially selected articles to verify their relevance with the topic of the correlation between BMI and two-stage revision failure of PJI following TJA. Studies had to fulfill the following criteria for inclusion: outcome was two-stage revision failure of PJI following TJA; study design included case-control, retrospective and prospective cohorts, and cross-sectional studies; participants were selected without limitations to regions, ages, or social status. Trials were excluded according to following identifications: duplicate or overlapping data, animal experiments, conference abstracts, letters, and review articles. In case of any disagreement the results were discussed and unified by senior authors.

Data extraction

Data from the included studies were extracted and independently categorized by two of the authors in a predefined data extraction form. All disagreements were resolved by discussion. Design information, baseline population characteristics (mean age, sample size, and country), surgical approach, risk factors from all included studies were stratified into a standardized evidence table. All the data were rechecked to ensure accuracy. Study selections were shown in a PRISMA flow diagram.

Methodological quality assessment

The methodological quality of the included studies was evaluated by two independent reviewers based on the items of modified Newcastle-Ottawa Scale (NOS),¹⁴ comprising patient selection, study group comparability, and outcome assessment. The observational studies scored 0 to 9. Divaricate opinions were discussed among the authors.

Statistical analysis

The meta-analysis and statistical analysis were performed using Stata 13.0 software (Stata Corp, USA) and Cochrane Collaboration Review Manager software (RevMan version 5.3, Nordic Cochrane center, Copenhagen, Denmark). The OR and 95% CIs were calculated. The I-square (I²) test was adopted to evaluate the influence of heterogeneity on the output of the meta-analysis. I² values of 0%, 25%, 50%, and 75% represented no, low, medium, and high heterogeneity, respectively. Heterogeneity was tested using Cochran’s Q statistic and the I² metric: a I² > 25% was the cutoff of significant heterogeneity, and a fixed-effect model was used when a I² < 25%; otherwise, a random effect model was preferred.¹⁵ A p value of less than 0.05 was accepted as statistically significant. A sensitivity analysis¹⁶ was conducted by excluding one study at a time to evaluate the quality and consistency of the results. Egger’s and Begg’s linear regression tests for publication bias were carried out. Subgroup analyses were performed according to different countries, operation methods, and effect type.

Results

Study selection process

As a result, 1231 references were initially retrieved, 539 were left after eliminating duplicate literature; and then 499 without high-relevant to our topic were discarded by reading titles and abstracts, and 40 studies remained. Finally, 25 full-text articles were abandoned because of the following reasons: 15 studies on irrelevant topics; five studies without full-text materials; three studies without sufficient data for extraction; one study was poster abstracts; one study was letters to the editor. Therefore, 15 observational studies with 1267 patients were included, of which 15 studies were included in systematic review and 11 studies in meta-analysis. The flow chart describing the selection process of the study was shown in Figure 1.

Figure 1.

The flow diagram of literature search and selection.

Study characteristics and methodological quality

15 observational studies about the correlation between BMI and two-stage revision failure of PJI following TJA were included in this study. The 15 included references encompassed retrospective cohort and retrospective case-control, with the publication years differing from 2013 to 2020. Four were conducted in Germany, five in the USA, four in China, Taiwan, and one in Italy, Canada, respectively. In the selected clinical trials, the sample size varied between 16 and 245. Eight studies found a correlation between BMI and two-stage revision failure of PJI following TJA, but seven other studies found no correlation. The failure rate of two-stage revision ranged from 6.8% to 42.86%. The basic characteristics of these studies are summarized in Table 2 and Table 3. In addition, all studies were evaluated as high methodological quality in accordance with the NOS scores.

Table 2.

Characteristics of the included studies.

Included studies	Study design	Country	Study characteristics	Study period	Adjustment factors	Operation type	Conclusion by author
Hoell S 2016¹⁷	Retrospective cohort study	Germany	32 males, 73 ± 9.7 years	2004–2008	Sinus present, diabetes, smoking, BMI, Periprosthetic fracture, wound healing problems, corticosteriods, immune suppression, postoperative hematoma, blood transfusion, tumor disease	TKA	Significant
Watts CD 2014¹⁸	Retrospective cohort study	USA	33 males, 60 ± 9 years (the morbidly obese group); 62 ± 8 years (the non-obese group.)	1987–2007	No be adjusted	TKA	Significant
Houdek TM 2015¹⁹	Case-control study	USA	42 males, 63 ± 10 years	1987–2007	No be adjusted	THA	Significant
Ahmad SS 2019²⁰	Case-control study	Germany	69 males, 66 ± 12 years	1999–2012	Alcoholism, smoking, BMI, primary bacteremia, other primary focus of infection, rheumatoid arthritis, gout, diabetes, malignancy, immune deficiency, congestive heart failure, renal failure, liver cirrhosis, corticosteroids, chemotherapy, IV drug abuse	THA	Significant
Ma CY 2018²¹	Retrospective cohort study	China Taiwan	29 males, 70.1 ± 8.4 years	2005–2015	BMI, liver cirrhosis, reimplantation operative time, the AORI (Anderson orthopaedic research institute) types of bone defect in the femur, the tibia, type of revision component, isolation of Enterococcus spp, presence of polymicrobial organisms, gout	TKA	Significant
Logroscino G 2019²²	Retrospective cohort study	Italy	9 males, 69.4 ± 11.63 years	2012–2016	No be adjusted	THA	Significant
Jhan SW 2017²³	Retrospective cohort study	China Taiwan	43 males, 57 ± 14 years	2005–2012	BMI, drug abuse, liver cirrhosis, gram-negative organism, presence of sinus tract, repeated debridement between stage, revision operative time	THA	Significant
Kurd MF 2010²⁴	Retrospective cohort study	USA	56 males, Mean 67 (17–88) years	1998–2005	No be adjusted	TKA	No significant
Petis SM 2019²⁵	Case-control study	USA	123 males, Mean 68 (33–91) years	1991–2006	Age, sex, age-adjusted CCI(), BMI, ASA, previous revision surgery, McPherson extremity grade, McPherson infection type, McPherson host grade, retained hardware (plates, screws,nails), additional re-debridement prior to reimplantation, dose of antibiotics in spacer, antibiotic cement used for reimplantation, sinus tract present, 2 cultures positive for same organism, cultured organism, McPherson infection type, use of antibiotic suppression after reimplantation, use of antibiotic suppression after reimplantation, adjusted using propensity score, staphylococcus species	TKA	Significant
Su YJ 2017²⁶	Case-control study	China Taiwan	18 males, Mean 47.3 (22–67) years	2001–2012	No be adjusted	THA	No significant
Chen MJW 2020²⁷	Retrospective cohort study	China Taiwan	30 males, Mean 69.5 (38–92) years	2003–2013	Obesity, pre-operative CRP levels, cirrhosis, virulent pathogens, anaerobic pathogens, polymicrobial infection	TKA	No significant
Spiegl U 2013²⁸	Retrospective cohort study	Germany	Unclear, Mean 61.7 (22–90) years	2006–2008	No be adjusted	THA	Significant
Mortazavi SMJ 2011²⁹	Retrospective cohort study	USA	62 males, Mean 67.5 (37–88) years	1997–2007	No be adjusted	TKA	No significant
Claassen L 2015³⁰	Retrospective cohort study	Germany	21 males, 65.4 ± 10.6 years	2011–2012	No be adjusted	TKA	No significant
Drexler M 2015³¹	Case-control study	Canada	43 males, Mean 68 (39–87) years	2006–2012	No be adjusted	TKA	No significant

THA = total hip arthroplasty, TKA = total knee arthroplasty, ASA = the American Society of Anesthesiologists, BMI = body mass index, TJA = total joint arthroplasty. CCI = Charlson Comorbidity Index.

Table 3.

Characteristics of the included studies.

Included studies	Treatment protocols	The interval between the first and second stage	Replantation to reinfection time	Follow-up time
Hoell S 2016¹⁷	The protocol consisted of explantation of the prosthesis with implantation of a fixed antibiotic-loaded bone–cement spacer (1 g gentamicin and 1 g clindamycin/40 g cement) and at least 14 days of intravenous antibiotic administration, followed by at least 4 weeks of antibiotics orally. If necessary, additional antibiotics were mixed into the spacer, depending on the microbiological results, as an off-label application. After an interval of 14 days without antibiotics, C-reactive protein (CRP) was measured in serum. The second-stage procedure was performed 9–12 weeks after explantation	9–12 weeks	Mean 2.3(0.6–3.7) years	Mean 4.1 ± 2.7 years
Watts CD 2014¹⁸	All patients placement of a high-dose antibiotic-impregnated spacer, organism-specific intravenous antibiotic therapy, and delayed reimplantation using antibiotic-impregnated cement. For all resections and reimplantations, vancomycin and/or an aminoglycoside were added to Simplex P bone cement (stryker orthopaedics, Mahwah, New Jersey) at the time of surgery. Antibiotic spacers contained a median of 3 g/batch of vancomycin (range, 0–4 g/batch) and 3.6 g/batch (range, 0–4.8 g/batch) of aminoglycoside for both the morbidly obese and non-obese groups. Reimplantations were performed using cement containing a median of 1 g/batch (range, 0–2 g/batch) of vancomycin and a median of 1.2 g/batch (range, 0–2.4 g/batch) of aminoglycoside for both the morbidly obese and non-obese groups	Unclear	52 ± 29 months (the morbidly obese group); 48 ± 65 months (the non-obese group).	Mean 6.9 (5.1–10.8, the morbidly obese group); Mean 7.9 (5.0–11.1, the non-obese group)
Houdek TM 2015¹⁹	All patients placement of a high-dose antibiotic-impregnated spacer, organism-specific intravenous antibiotic therapy, and delayed reimplantation. For all resections and reimplantations, vancomycin and/or an aminoglycoside (tobramycin or gentamicin) were added to Simplex cement (stryker, Mahwah, New Jersey) used in the surgery. The antibiotic spacers used in both the morbidly obese and non-obese groups contained a median of 3 g/batch (range, 0–4 g/batch) of vancomycin and 3.6 g/batch (range, 0–4.8 g/batch) of aminoglycoside. The reimplantations in both the morbidly obese and non-obese groups were performed with use of cement containing a median of 1 g/batch (range, 0–2 g/batch) of vancomycin and a median of 1.2 g/batch (range, 0–2.4 g/batch) of aminoglycoside	Mean 34.3 (6–292) weeks; Mean 20.9 (6–208) weeks	51 ± 47 months; 107 months	Mean 8.1 (5.1–15.2); Mean 10.3 (5.1–20.8)
Ahmad SS 2019²⁰	With or without implantation of a gentamicin-impregnated cement spacer as a first-stage procedure (n = 9 [9%] temporary Girdlestone situation). All patients underwent 2 weeks of intravenous antibiotics followed by 10 weeks of oral antibiotics. After a 2-week interval without antibiotic treatment, a diagnostic joint aspiration was performed. Upon negative microbiology, the second-stage procedure was performed, which involved the removal of the cement spacer and the implantation of a revision implant	Unclear	Unclear	Mean 60 (24–170) months
Ma CY 2018²¹	During the first stage, necrotic tissue was thoroughly debrided following removal of all implanted components. Antibiotic-loaded cement beads or a spacer, comprised of 40 g bone cement with 2–4 gm of vancomycin and 2−4 gm of piperacillin or ceftazidime, was inserted. After the first stage, intravenous organism-specific antibiotics were administered for at least 4 weeks followed by oral antibiotics for 2 weeks	All reimplantations were performed after a minimum 2-week antibiotic holiday	22 ± 19 months	5.6 ± 2.3 (2–10)
Logroscino G 2019²²	The spacer’s implant (It’s not clear exactly)	127.6 ± 90.1 days (in healed patients); 72.5 ± 33.6 days (in the failure ones)	Unclear	23.8 ± 16.05 months
Jhan SW 2017²³	Implantation of antibiotic-loaded cement beads. The regimen of antibiotics in the bone cement was determined according the culture results from pre-operative joint aspiration or previous culture report. If the infecting microorganism could not be known at the time of resection arthroplasty, empirical combination of 2–4 g vancomycin and 2–4 g piperacillin per 40 g package of bone cement was used. Further debridement or exchange to sensitive antibiotic-loaded cement beads may be needed if the infection could not be controlled during the interim stage. Culture-specific parenteral antibiotics were given postoperatively for 4 weeks, followed by oral antibiotics for 2 weeks	20 ± 15.8 weeks	Unclear	5.7 ± 2.4 years
Kurd MF 2010²⁴	Two-stage exchange arthroplasty consisted of removal of total knee implants, as well as bone cement, irrigation, and de’bridement of the joint, and insertion of a static antibiotic-laden cement spacer block for at least 6 weeks (mean time to reimplantation 15 weeks; range, 6–62 weeks). The cement was impregnated with 4.0 g vancomycin and 3.6 g tobramycin for each 40 g Palacos1 bone cement	Mean 15 (6–62) weeks	Unclear	Mean 35 (24–90) months
Petis SM 2019²⁵	Insertion of an antibiotic polymethylmethacrylate spacer, cement dowels were inserted into the femoral and tibial medullary canals for all spacers. All cement was Simplex P with combinations of vancomycin in 216 cases (mean,2.7 g [range, 1–7 g] per batch) and tobramycin in 85 cases (mean, 2.8 g [range, 1.2–4.8 g] per batch) or gentamicin in 97 cases (mean, 3.4 g [range, 1.2–4.8 g] per batch).The mean duration of intravenous (IV) antibiotic therapy was 6 weeks (range, 4–26 weeks)	Mean 11 (5–54) weeks	Mean 47 (1–247) months	Mean 14 (2–25) years
Su YJ 2017²⁶	Placement of an antibiotic-containing cement spacer. The spacer was static and mixed with 2.5 g vancomycin and 4 g meropenem per 40 g of cement. Intravenous antibiotic therapy for 6 weeks followed the resection. If the clinical physical examination and serology showed eradication of infection after 3 months without antibiotics treatment, reimplantation was performed. In the RHA, the prosthesis was fixed without cement after adequate debridement. The intraoperative pathology was arranged to confirm the eradication of infection before prosthesis implantation	>3 months	174 (42–270) days; 230 (90–300) days	3(2–11) years
Chen MJW 2020²⁷	Antibiotic-loaded bone cement mobile spacer implantation. Antibiotic-loaded bone cement was hand-mixed intraoperatively using four g of vancomycin powder with four g of ceftazidime powder per 40-g pack of polymethylmethacrylate (PMMA) polymer. Articulating mobile spacers were routinely used, except for cases when spacer dislocation was a major concern due to massive bone loss or lack of soft tissue integrity [13]. Pre-made silicone mold templates were utilized to construct mobile spacers. Culture-specific parenteral antibiotics were given postoperatively for 4 weeks, followed by 2 weeks of oral antibiotics. With a minimum interim period of 3 months and antibiotic holiday of 2 weeks	>3 months	13.1 ± 12.6 months	65.1 (25–139) months
Spiegl U 2013²⁸	Implantation of an antibiotic-loaded cement spacer and a vacuum-assisted closure. For the most parts, the cement was loaded with gentamicin. In cases of bacterial resistance, vancomycin was used, according to the resistogram. Revision surgeries including changing of vacuum sealing, removing tissue specimens, and if necessary local debridement in cases of conspicuous tissue situation were continued until intraoperative biopsies showed no signs of bacterial growth. Subsequently, the antibiotic therapy was stopped. A surgical exploration followed after 4–7 days including sampling of bacterial biopsies, exchange of the spacer, secondary wound closure, and postoperative reuptake of antibiotic therapy for another 6 weeks. Another operative exploration was performed after an interval of 10–12 weeks. In cases of negative bacterial biopsies, the reimplantation of the prosthesis was performed with accompanying antibiotic prophylaxis for 6 weeks	10–12 weeks	Unclear	32 ± 8 months
Mortazavi SMJ 2011²⁹	Placement of an antibiotic-loaded cement spacer. The spacer was static versus dynamic at the discretion of the treating surgeon. It is standard at this institution to use 3 g vancomycin and 3.6 g tobramycin per 40 g of cement. After resection, the patient was treated with 6 weeks of intravenous antibiotics with a subsequent antibiotic holiday (2–6 weeks). Reimplantation was performed when deemed appropriate by the treating surgeon. At the time of reimplantation, antibiotic-loaded cement (1.2 g tobramycin and 1 g vancomycin per 40 g of cement) was used for fixation of the prosthesis	>6 weeks	Unclear	Mean 3.8 (2–9.4) years
Claassen L 2015³⁰	Placement of an antibiotic-loaded cement spacer. After the first stage the patient was started on vancomycin and clindamycin antibiotics until cultures returned with sensitivities. Then culture-directed antibiotics were administered intravenously for 2 weeks followed by antibiotic treatment per os for a total of 6 weeks of treatment. Eight weeks after first-stage procedure, with at least a 2-week antibiotic-free interval, a diagnostic aspiration was performed. When the final results of the aspiration were negative and there were no other signs of ongoing infection the second-stage procedure was performed. If signs or symptoms of infection persisted, we performed a second look operation with spacer exchange and re-debridement. A second course of antibiotic therapy was initiated based on the intraoperative and culture results for at least another 6 weeks	11.3 ± 5.6 weeks	Unclear	>12 months
Drexler M 2015³¹	An appropriately sized new femoral component and thin polyethylene liner were implanted (Nexgen LPS, Zimmer, warsaw, IN), using three bags of Palacos cement (Heraeus, Wehrheim, Germany) that were mixed with 12 g of powdered vancomycin and 12 g powdered ceftazidime (4 g of each antibiotic per bag). The implants were then cemented into place with as much cement as possible packed between the bone surfaces and the implants allowing a 0–90 range of motion. The wound was closed in a standard fashion. Drains were not utilized to avoid unwarranted loss of antibiotic-loaded fluid. Antibiotics were administered intravenously for 6 weeks through a peripherally inserted central catheter (PICC line)	Mean 4.9 (1.5–18) months	Unclear	36.2 (24–85) months

Overall meta-analysis

BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²

Data from six studies^{17,20,21,23,25,27} on BMI ≥ 30 kg/m² vs. BMI < 30 kg/m² were available for the meta-analysis by the random effect model due to a significant statistical heterogeneity (I² = 83.4%, p = 0.000). It was found that the risk of two-stage revision failure of PJI following TJA significantly boosted by 3.53 times in patients with BMI ≥ 30 kg/m² (OR = 3.53; 95% CI = 1.63–7.64), with high heterogeneity (I² = 83.4%, p = 0.000; Figure 2). Thus, subgroup analyses were conducted to investigate the potential factors that could substantially affect the between-study heterogeneity.

Figure 2.

The meta-analysis of the correlation between BMI (BMI ≥30 kg/m² vs. BMI <30 kg/m²) and two-stage revision failure of PJI following TJA.

BMI ≥ 40 kg/m² vs. BMI < 30 kg/m²

Data from two studies^18,19 on BMI ≥ 40 kg/m² vs. BMI < 30 kg/m² were available for the meta-analysis by the random effect model due to a significant statistical heterogeneity (I² = 71.3%, p = 0.062). It was found that the risk of two-stage revision failure of PJI following TJA significantly increased by 2.92 times in patients with BMI ≥ 40 kg/m² (OR = 2.92; 95%CI = 1.06–8.03; Figure 3).

Figure 3.

The meta-analysis of the correlation between BMI (BMI ≥40 kg/m² vs. BMI <30 kg/m²) and two-stage revision failure of PJI following TJA.

Other body mass index comparisons

Data from three studies^22,24,28 on BMI (continuous variable) were available for the meta-analysis by the random effect model due to a significant statistical heterogeneity (I² = 56%, p = 0.10). The results show that there were nonsignificant differences in BMI between the failure group and the success group (SMD = −0.48; 95%CI = −1.20–0.23; Figure 4).

Figure 4.

The meta-analysis of the correlation between BMI (continuous variable) and two-stage revision failure of PJI following TJA.

Subgroup analyses (BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²)

Subgroup analysis of different effect type

Subgroup analyses of different effect type were conducted. The subgroup analysis showed that significant correlations were basically consistent, significant association were observed among the studies performed in odds ratio (OR = 1.22; 95% CI = 1.07–1.40) and hazard ratio (OR = 4.10; 95% CI = 2.64–6.36). When original studies used hazard ratio as the effect type, there was no heterogeneity (I² = 0.00%, p = 0.467; Figure 5).

Figure 5.

Subgroup analysis of different effect type.

Subgroup analysis of different countries

Subgroup analyses of different countries were conducted. The subgroup analysis showed that significant association were observed among the studies performed in China, Taiwan (OR = 5.41; 95% CI = 2.51–11.68) and USA (OR = 3.10; 95% CI = 1.61–5.95), but not among those conducted in Germany (OR = 2.191; 95% CI = 0.57–8.44). And the heterogeneity was low when studies performed in China, Taiwan (I² = 12.3%, p = 0.320; Figure 6).

Figure 6.

Subgroup analysis of different countries.

Subgroup analysis of different operation method

Subgroup analyses of different operation method were conducted. The subgroup analysis showed that significant association were observed among the studies performed in TKA (OR = 3.63; 95% CI = 2.27–5.82), but not among those conducted in THA (OR = 3.06; 95% CI = 0.42–22.19). And the heterogeneity was no when studies performed in TKA (I² = 0.0%, p = 0.653; Figure 7).

Figure 7.

Subgroup analysis of different operation method.

Sensitivity analyses (BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²)

The sensitivity analysis was performed to assess whether individual studies would affect the overall results. We evaluated the effect of each study on the methodological quality through the sequential exclusion of single studies. The results showed that there was a nonsignificant difference in the stability of the results (Figure 8), which validated the rationality and reliability of our analysis.

Figure 8.

The influence analysis of included studies.

Evaluation of publication bias

Due to the insufficient number of included studies in the meta-analysis, evaluation of publication bias was not conducted (Including visual inspection of funnel plots, Egger and Begg tests).

Discussion

In this study, we conducted 15 observational studies which included 1267 patients, of which 15 studies were included in systematic review and 11 studies in meta-analysis. To ensure a reliable conclusion, previous published studies on the correlation between BMI and two-stage revision failure of PJI following TJA were retrieved, reviewed, and summarized to achieve those with high compliance and high quality, so as to answer various clinical questions of this malady. Our study shows that eight original studies found a correlation between BMI and two-stage revision failure of PJI following TJA, but seven other original studies found no correlation. The failure rate of two-stage revision ranged from 6.8% to 42.86%. And the meta-analysis results suggest that the risk of two-stage revision failure of PJI following TJA significantly boosted in obese patients. The subgroup analysis showed that significant association were observed among the studies performed in TKA (OR = 3.63; 95% CI = 2.27–5.82), but not among those conducted in THA (OR = 3.06; 95% CI = 0.42–22.19). A significant association remained consistent, as indicated by sensitivity analyses.

The two-stage revision is the gold standard for the treatment of periprosthetic joint infection following total joint arthroplasty, but there was still a 6.8%–42.86% failure rate. Previous meta-analysis^32–34 showed that obesity was closely related to periprosthetic joint infection following total joint arthroplasty, however, the association between BMI and two-stage revision failure of PJI following TJA remains unclear. Though the correlation between BMI and two-stage revision failure of PJI following TJA has been rapidly reported, their results still remain divergent and even controversial.^17–31 Our meta-analysis found that the risk of two-stage revision failure of PJI following TJA significantly boosted by 3.53 times in patients with BMI ≥ 30 kg/m² (BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²) and 2.92 times in patients with BMI ≥ 40 kg/m² (BMI ≥ 40 kg/m² vs. BMI < 30 kg/m²), this shows a significant increase in the failure rate of two-stage revision in obese patients. When subgroup analysis was performed in different surgical methods, in the THA group, there was no statistically significant difference in failure rates between obese and non-obese patients, this may attribute to the insufficient inclusion of eligible studies. Systematic review found that nearly half of the original studies concluded that BMI was not associated with two-stage revision failure, therefore, we cannot yet conclude that BMI is associated with two-stage revision failure of PJI following TJA.

As the passages have exposed, two significant advantages of our study are clear. Firstly, as the correlation between BMI and two-stage revision failure of PJI following TJA were controversial, this meta-analysis assessed the potential correlation between BMI and two-stage revision failure of PJI following TJA through a thorough systematic study with rigorous analytical methods. Secondly, the rationality and reliability of our meta-analysis have been prudently and significantly improved in that the overall comprehensive estimation is based on a large sample size. In addition, sufficient sensitivity analysis has been carried out to ensure the reliability of this study.

The current systematically evaluated and meta-analysis has the following limitations and must be considered before our results are accepted. First, the selected studies in the systematically evaluated and meta-analysis were published between 2013 and 2020, the outcome evaluation indexes of the included studies were not completely consistent, and most of them did not adjust for confounding factors, meta-analysis evaluation data is insufficient. Second, the research included in this analysis is insufficient, and potential publication bias still exists. Third, this study only includes references in English. Therefore, we may have lost data from those in other languages.

Conclusion

In summary, our meta-analysis suggests that the risk of two-stage revision failure of PJI following TJA significantly boosted in obese patients. However, because there may be publication bias of this study, combined overall systematic review and meta-analysis results, we cannot yet conclude that BMI is associated with two-stage revision failure of PJI following TJA. This conclusion needs to be verified by more prospective studies.

Footnotes

Author contributions

(I) Conception and design: X Deng; (II) Administrative support: H Wang; (III) Provision of study materials or patients: J Guo, S Wu, W Chen; (IV) Collection and assembly of data: J Guo, H Wang; (V) Data analysis and interpretation: J Guo, S Wu; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Xiaoqiang Deng

References

Jämsen

Huhtala

Puolakka

, et al. Risk factors for infection after knee arthroplasty. A register-based analysis of 43,149 cases. J Bone Jt Surg Am 2009; 91(1): 38–47.

Jämsen

Varonen

Huhtala

, et al. Incidence of prosthetic joint infections after primary knee arthroplasty. J Arthroplasty 2010; 25(1): 87–92.

Höll

Schlomberg

Gosheger

, et al. Distal femur and proximal tibia replacement with megaprosthesis in revision knee arthroplasty: a limb-saving procedure. Knee Surg Sports Traumatol Arthrosc 2012; 20(12): 2513–2518.

Iorio

Williams

Marcantonio

, et al. Diabetes mellitus, hemoglobin A1C, and the incidence of total joint arthroplasty infection. J Arthroplasty 2012; 27(5): 726–729.

Finucane

Stevens

Cowan

, et al. National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9·1 million participants. Lancet 2011; 377(9765): 557–567.

Flegal

Carroll

Kit

, et al. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. JAMA 2012; 307(5): 491–497.

Wagner

Kamath

Fruth

, et al. Effect of body mass index on complications and reoperations after total hip arthroplasty. J Bone Jt Surg Am 2016; 98(3): 169–179.

Triantafyllopoulos

Soranoglou

Memtsoudis

, et al. Rate and risk factors for periprosthetic joint infection among 36,494 primary total hip arthroplasties. J Arthroplasty 2018; 33(4): 1166–1170.

Meller

Toossi

Johanson

, et al. Risk and cost of 90-day complications in morbidly and superobese patients after total knee arthroplasty. J Arthroplasty 2016; 31(10): 2091–2098.

10.

Lenguerrand

Whitehouse

Beswick

, et al. Risk factors associated with revision for prosthetic joint infection after hip replacement:a prospective observational cohort study. Lancet Infect Dis 2018; 18(9): 1004–1014.

11.

Lenguerrand

Whitehouse

Beswick

, et al. Risk factors associated with revision for prosthetic joint infection following knee replacement:an observational cohort study from England and Wales. Lancet Infect Dis 2019; 19(6): 589–600.

12.

Moher

Liberati

Tetzlaff

, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statemen. BMJ 2009; 339(15): b25–35.

13.

Stroup

Berlin

Morton

, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000; 283(15): 2008–2012.

14.

Wells

Shea

Connell

, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. 2011. [Internet]. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. [Google Scholar].

15.

Higgins

Thompson

Deeks

, et al. Measuring inconsistency in meta-analyses. BMJ 2003; 327(7414): 557–560.

16.

Fox

Piepgras

Bartleson

. Anterolateral decompression of the atlantoaxial vertebral artery for symptomatic positional occlusion of the vertebral artery:case report. Neurosurg 1995; 83(4): 737–740.

17.

Hoell

Sieweke

Gosheger

, et al. Eradication rates, risk factors, and implant selection in two-stage revision knee arthroplasty: a mid-term follow-up study. J Orthop Surg Res 2016; 11(1): 93.

18.

Watts

Wagner

Houdek

, et al. Morbid obesity: A significant risk factor for failure of two-stage revision total knee arthroplasty for infection. J Bone Jt Surg Am 2014; 96(18): e154.

19.

Houdek

Wagner

Watts

, et al. Morbid obesity: A significant risk factor for failure of two-stage revision total hip arthroplasty for infection. J Bone Jt Surg Am 2015; 97(4): 326–332.

20.

Ahmad

Orlik

Ahmad

SJS

, et al. Obesity and smoking predict the results of two-stage exchange in septic revision hip arthroplasty: A cohort study. Orthop Traumatol Surg Res 2019; 105(3): 467–471.

21.

Bell

. Predictors of treatment failure following two-stage reimplantation for infected total knee arthroplasty: a 2-10 years follow-up. J Arthroplasty 2018; 33(7): 2234–2239.

22.

Logroscino

Campana

Pagano

, et al. Risk factors for failure of two-stage revision arthroplasty for infected hip prosthesis: review of the literature and single centre cohort analysis. Eur Rev Med Pharmacol Sci 2019; 23(2 Suppl): 65–75.

23.

Jhan

Lee

, et al. The risk factors of failed reimplantation arthroplasty for periprosthetic hip infection. BMC Musculoskelet Disord 2017; 18(1): 255.

24.

Kurd

Ghanem

Steinbrecher

, et al.

Two-stage exchange knee arthroplasty: does resistance of the infecting organism influence the outcome?

Clin Orthop Relat Res 2010; 468(8): 2060–2066.

25.

Petis

Perry

Mabry

, et al. Two-stage exchange protocol for periprosthetic joint infection following total knee arthroplasty in 245 knees without prior treatment for infection. J Bone Jt Surg Am 2019; 101(3): 239–249.

26.

Lin

Huang

, et al. Intravenous drug abuse is a risk factor in the failure of two-stage treatment for infected total hip arthroplasty. Kaohsiung J Med Sci 2017; 33(12): 1–7.

27.

Chen

Hung

Chang

, et al. Periprosthetic knee infection reconstruction with a hinged prosthesis: Implant survival and risk factors for treatment failure. Knee 2020; 27(3): 1–8.

28.

Spiegl

Friederichs

Pätzold

, et al. Risk factors for failed two-stage procedure after chronic posttraumatic periprosthetic hip infections. Arch Orthop Trauma Surg 2013; 133(3): 421–428.

29.

Mortazavi

Vegari

, et al. Two-stage exchange arthroplasty for infected total knee arthroplasty: predictors of failure. Clin Orthop Relat Res 2011; 469(11): 3049–3054.

30.

Claassen

Plaass

Daniilidis

, et al. Two-stage revision total knee arthroplasty in cases of periprosthetic joint infection: an analysis of 50 cases. Open Orthop J 2015; 9: 49–56.

31.

Drexler

Dwyer

Kuzyk

, et al. The results of two-stage revision TKA using Ceftazidime-Vancomycin-impregnated cement articulating spacers in tsukayama type II periprosthetic joint infections. Knee Surg Sports Traumatol Arthrosc 2016; 24(10): 3122–3130.

32.

Resende

VAC

Neto

Nunes

, et al. Higher age, female gender, osteoarthritis and blood transfusionprotect against periprosthetic joint infection in total hip or knee arthroplasties: a systematic review and meta-analysis. Knee Surg Sports Traumatol Arthrosc 2018; 26: 1–36.

33.

Guo

, et al. Meta-analysis shows that obesity may be a significant risk factor for prosthetic joint infections. Int Orthop 2016; 40(4): 659–667.

34.

Liu

Wahafu

Cheng

, et al. The influence of obesity on primary total hip arthroplasty outcomes: A meta-analysis of prospective cohort studies. Orthop Traumatol Surg Res 2015; 101(3): 289–296.

Correlation between body mass index and two-stage revision failure of periprosthetic joint infection following total joint arthroplasty: A systematic review and meta-analysis

Abstract

Keywords

Introduction

Methods

Search strategies

Study selection criteria

Data extraction

Methodological quality assessment

Statistical analysis

Results

Study selection process

Study characteristics and methodological quality

Overall meta-analysis

BMI ≥ 30 kg/m2 vs. BMI < 30 kg/m2

BMI ≥ 40 kg/m2 vs. BMI < 30 kg/m2

Other body mass index comparisons

Subgroup analyses (BMI ≥ 30 kg/m2 vs. BMI < 30 kg/m2)

Subgroup analysis of different effect type

Subgroup analysis of different countries

Subgroup analysis of different operation method

Sensitivity analyses (BMI ≥ 30 kg/m2 vs. BMI < 30 kg/m2)

Evaluation of publication bias

Discussion

Conclusion

Footnotes

Author contributions

Declaration of conflicting interests

Funding

ORCID iD

References

BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²

BMI ≥ 40 kg/m² vs. BMI < 30 kg/m²

Subgroup analyses (BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²)

Sensitivity analyses (BMI ≥ 30 kg/m² vs. BMI < 30 kg/m²)