Skip to main content
NCBI home page
Frontiers in Surgery logoLink to Frontiers in Surgery
. 2026 Jan 8;12:1715571. doi: 10.3389/fsurg.2025.1715571

Risk factors for periprosthetic joint infection following primary total knee arthroplasty: a systematic review and meta-analysis

Jiashun Li 1,†, Zhen Liu 1,†, Zheng Wang 2,†, Deyu Li 1,†, Heng Zhang 2, Shuo Hao 3, Qingxue Zhu 4, Maihemuti Yakufu 1,*, Ainikaer Abulaiti 1,*, Abuduwupuer Haibier 1,*
PMCID: PMC12823960  PMID: 41583845

Abstract

Objective

Periprosthetic joint infection (PJI) represents a severe and devastating complication following primary total knee arthroplasty (TKA), leading to higher morbidity and implant failure. However, existing evidence regarding its risk factors remains inconsistent.

Method

PubMed, Cochrane Library, Embase, China National Knowledge Infrastructure (CNKI), Wanfang Database, VIP Chinese Journal Database, and the Chinese Biomedical Literature Service System (SinoMed) were searched for primary studies from the database inception to August 2025. Data were managed with EndNote X9 software, meta-analyses were performed using Review Manager 5.4, infection risk was analyzed using STATA 18, and the comprehensive incidence rate of PJI was analyzed using R 4.5.1.

Results

A total of 1,569 studies were identified, of which 20 were included in this systematic review. The meta-analysis including 20 studies analyzed eight risk factors: longer operation time (OR = 9.10, 95% CI: 7.66–10.80), obesity (OR = 13.95, 95% CI: 12.06–16.14), male gender (OR = 2.67, 95% CI: 1.80–3.95), diabetes (OR = 2.98, 95% CI: 2.27–3.92), longer hospital stay (OR = 1.73, 95% CI: 1.38–2.16), use of immunosuppressants (OR = 5.76, 95% CI: 2.77–11.98), hypoalbuminemia (OR = 6.24, 95% CI: 4.00–9.73), and underlying systemic inflammatory disease (OR = 3.47, 95% CI: 1.92–6.27). All identified risk factors were associated with an increased risk of developing PJI after primary TKA.

Conclusion

The meta-analysis confirmed that longer operative time, obesity, male gender, diabetes, longer hospital stay, use of immunosuppressants, hypoalbuminemia, and underlying systemic inflammatory disease are risk factors for PJI after primary TKA. This systematic review and meta-analysis provide Level II evidence (according to the Oxford Centre for Evidence-Based Medicine criteria) for the identified risk factors.

Keywords: meta-analysis, periprosthetic joint infection, PJI, risk factors, TKA, total knee arthroplasty

1. Introduction

Total knee arthroplasty (TKA) is a highly successful and cost-effective surgical intervention for end-stage knee osteoarthritis, significantly alleviating pain and restoring function in millions of patients worldwide (1, 2). With the aging global population and rising prevalence of osteoarthritis, the annual volume of TKA procedures continues to increase. However, this success is tempered by the risk of postoperative complications, among which periprosthetic joint infection (PJI) remains one of the most devastating and challenging. PJI is the leading cause of early revision TKA, accounting for approximately 25%–30% of all revision cases (3–6), and it poses a tremendous burden on healthcare systems, with treatment costs often exceeding those of the primary procedure by severalfold (7).

The reported incidence of PJI following primary TKA varies in the literature but is generally estimated to be between 1% and 2% (8–10). Although this percentage may seem modest, it corresponds to a substantial absolute number of affected patients due to the high procedural volume, making PJI a significant public health concern. The consequences for affected patients are severe and can include prolonged antibiotic therapies, multiple complex surgical procedures (including debridement, antibiotics, and implant retention, one- or two-stage revision), extended rehabilitation, and often a permanent functional deficit (11–13). Furthermore, PJI is associated with a significantly higher mortality rate compared with aseptic revision procedures (14).

Given the profound implications of PJI, identifying at-risk patients is paramount for implementing targeted preemptive strategies. Numerous studies have investigated potential risk factors, which can be broadly categorized as patient-specific modifiable (e.g., obesity, diabetes mellitus, and malnutrition) and non-modifiable factors (e.g., age, gender, and rheumatoid arthritis), as well as procedure-related and healthcare system-related factors. However, evidence from individual studies is often inconsistent or underpowered. Therefore, we conducted a systematic review and meta-analysis to synthesize the evidence and establish the risk factors for developing PJI after primary TKA.

2. Method

2.1. Study search strategy

The PubMed, Cochrane Library, Embase, China National Knowledge Infrastructure (CNKI), Wanfang Database, VIP Chinese Journal Database, and the Chinese Biomedical Literature Service System (SinoMed) from their inception to August 2025. A combined search strategy using both subject headings and free-text terms was employed. The search keywords included: total knee arthroplasty, TKA, prosthesis, periprosthetic infection, PJI, and risk factors. Gray literature (conference abstracts, doctoral theses, master's theses) obtained from the Chinese databases CNKI, Wanfang, and VIP was also considered. The complete search strategy is provided in Supplementary Text S1.

2.2. Diagnostic criteria for PJI

According to the International Musculoskeletal Infection Society (MSIS), the International Consensus Meeting (ICM), and the European Bone and Joint Infection Society (EBJIS), PJI is defined as follows:

Major criteria: Presence of a sinus tract communicating with the prosthesis, or isolation of the same pathogen from two or more separate samples of synovial fluid or intraoperative tissue cultures. Minor criteria: Serological markers [significantly elevated c-reactive protein (CRP) and erythrocyte sedimentation rate (ESR)], synovial fluid analysis [white blood cell count >3,000/µL or polymorphonuclear neutrophil (PMN) percentage >80%], intraoperative findings (purulent secretion or suppuration around the prosthetic tissue), histopathological examination (>5 polymorphonuclear leukocytes per high-power field), and emerging biomarkers (positive synovial α-defensin or D-dimer), meeting any single major criterion confirms the diagnosis (3).

2.3. Inclusion and exclusion criteria

Four authors (JL, ZL, ZW, DL) independently determined study eligibility. Any difference in opinion regarding eligibility was resolved through consensus.

Inclusion criteria: (Ⅰ) Study subjects were adult patients who underwent primary total knee arthroplasty at the same hospital. (Ⅱ) The exposure factor is one or more risk factors potentially associated with PJI after primary TKA, such as obesity and diabetes. (Ⅲ) The control group consists of patients who are not exposed to the risk factors under study. (Ⅳ) The outcome was the occurrence of periprosthetic joint infection after primary total knee arthroplasty. (Ⅴ) Patients were followed up for at least 6 months. (Ⅵ) The study design is a cohort study or case–control study, with the research focus on validating the risk of PJI infection following primary TKA.

Exclusion criteria: (Ⅰ) Patients not hospitalized during the whole study period; (Ⅱ) patients undergoing revision total knee arthroplasty; (Ⅲ) studies with missing medical data; (Ⅳ) duplicate publications, letters, or conference papers; (Ⅴ) animal studies, review articles, and systematic reviews; and (Ⅵ) studies not related to total knee arthroplasty.

2.4. Study selection

All the studies containing abstracts and titles were imported to EndNote X9. After removing duplicate papers, three investigators (JL, ZL, ZW) independently screened titles and abstracts according to the inclusion and exclusion criteria and then further screened the initially included literature by reading the full text to exclude those that do not meet the inclusion criteria. In cases of any disagreement in this step, a fourth investigator (DL) performed the same screening process as described above, and a consensus was reached through discussion.

2.5. Data abstraction and validity assessment

For studies meeting the inclusion criteria, the following information was extracted: first author, publication year, study design type, sample size, evaluated variables or risk factors, odds ratio, and 95% CI.

2.6. Quality assessment of studies

The methodological quality of the included literature was evaluated using the Newcastle–Ottawa scale (NOS) to assess the quality of all included studies.

2.7. Inclusion in the meta-analysis, data extraction, and statistical methods

Meta-analysis on the included studies was performed using Review Manager 5.4. The mean values and their standard deviations were extracted from each study to calculate the weighted mean difference and 95% confidence interval (CI), as well as the pooled odds ratio (OR) and 95% CI. For the heterogeneity test model selection, the Cochrane's Q test was employed to evaluate the degree of heterogeneity among studies. When heterogeneity was low (e.g., I2 < 50%) or P-value >0.10, indicating that the differences were mainly due to sampling error, a fixed-effects model was used. When heterogeneity was high (e.g., I2 ≥ 50%) or P-value ≤0.10, indicating the presence of genuine differences among studies, a random-effect model was used. The pooled odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for each risk factor for PJI following primary TKA, including longer operation time, obesity, male gender, diabetes, longer hospital stay, use of immunosuppressants, hypoalbuminemia, and underlying systemic inflammatory disease. A significance level of P < 0.05 was adopted to indicate statistical significance. Meta-analysis was performed using both fixed-effects and random-effects models.

3. Results

3.1. Search results

Figure 1 shows the study selection process. After excluding 254 duplicate studies, 1,037 were then excluded based on abstract and title screening, and an additional 278 studies were discarded after full-text review. Therefore, 20 studies reporting PJI following primary TKA were included.

Figure 1.

Flowchart showing the identification and screening of studies via databases and registers. Initially, 1,569 records were identified. After removing duplicates and non-conforming studies, 278 records were screened. Further screening reduced to 44 records, from which 24 were excluded due to issues like merging indicators and study design. Finally, 20 studies were included in the review.

Open in a new tab

Flow diagram of study selection (Supplementary Text S2).

3.2. Characteristics of the studies included

Table 1 provides a detailed description of the studies analyzed and lists the risk factors identified in each study. The 20 studies included were published from database inception to August 2025. Among them, 13 were cohort studies (15–17, 19–27, 34), and 7 were case–control studies (18, 28–33). Three studies were conducted in the USA (15, 23, 25), 2 in Australia (16, 17), 1 in Pakistan (19), 1 in Finland (20), 1 in Germany (21), 1 in Japan (24), and 11 in China (18, 22, 26–34).

Table 1.

Characteristics of the studies included.

Study inclusion Publication year Country Study type Number of cases Total sample size Incidence of PJI (%) Risk factors
Anis et al. (15) 2019 USA Cohort 79 11,840 0.7 1, 2
Armit et al. (16) 2018 Australia Cohort 29 1,058 2.7 3, 4, 5, 6
Dowsey and Choong (17) 2009 Australia Cohort 18 1,214 1.5 2, 3, 7
Guo et al. (18) 2020 China CC 54 162 33.33 1, 2, 14, 15, 16, 17, 18
Iqbal et al. (19) 2020 Pakistan Cohort 48 4,269 1.1 1, 2, 19
Jämsen et al. (20) 2012 Finland Cohort 31 3,915 0.79 2, 7
Claus et al. (21) 2006 German Cohort 144 17,641 0.8 3, 6, 27, 28
Siu et al. (22) 2018 China Cohort 34 2,543 1.34 20
Remily et al. (23) 2023 USA Cohort 32,093 13,80,536 2.32 2, 7, 29, 30, 31, 32
Suzuki et al. (24) 2011 Japan Cohort 17 2,022 0.84 2, 3, 15
Wang et al. (25) 2019 USA Cohort 70 7,861 0.89 1, 2, 31, 34
Cheng et al. (26) 2022 China Cohort 14 525 2.67 7, 10, 27
Ning et al. (27) 2017 China Cohort 12 1,249 0.96 1, 2, 7, 13, 17, 20
Baoyu (28) 2016 China CC 15 45 33.33 1, 2, 13, 14, 15, 16, 17, 18, 21
Di (29) 2020 China CC 14 42 33.33 1, 2, 7, 14, 16, 17
Tao et al. (30) 2017 China CC 22 402 5.47 3, 10, 24, 25
Zhe et al. (31) 2021 China CC 21 105 20 1, 16, 22, 23
Shaoqiang et al. (32) 2024 China CC 69 276 25 1, 7, 17, 20, 26
Zhongren (33) 2023 China CC 46 131 35.11 7, 22, 24
Du (34) 2022 China Cohort 36 868 4.15 2, 17, 24, 36, 37
Open in a new tab

1, longer operative time; 2, obesity; 3, male gender; 4, ambient humidity >60%; 5, apparent temperature >30°C; 6, ASA ≥ III; 7, diabetes mellitus; 8, history of post-traumatic arthritis; 9, discharge to rehabilitation/nursing facility; 10, advanced age; 11, history of patellar resurfacing; 12, lower income; 13, high postoperative drainage volume; 14, longer hospital stay; 15, previous surgical history of the affected site; 16, use of immunosuppressants; 17, hypoalbuminemia; 18, superficial infection; 19, bilateral TKA; 20, concomitant inflammatory disease (RA); 21, high drainage volume on postoperative day 1; 22, bacteriuria; 23, history of intra-articular injection; 24, anemia; 25, prolonged drainage tube retention; 26, chronic pulmonary disease; 27, intraoperative allogeneic blood transfusion; 28, antibiotic use >24 h; 29, Crohn's disease; 30, ulcerative colitis; 31, smoking; 32, alcohol abuse; 33, retained previous hardware; 34, CCI; 35, AVN; 36, lymphocytopenia; 37, low total cholesterol. ASA, American Society of Anesthesiologists Physical Status Classification System; CCI, Charlson comorbidity index; AVN, avascular necrosis of the femoral head; CC, case–control.

All the included studies describe the occurrence of PJI following primary TKA. Among them, five studies specifically explored the risk factors for PJI after both primary TKA and THA (20, 25, 28, 29, 33).

3.3. Quality of the studies included

The eligible articles were evaluated using the NOS, with scores used in place of the traditional star ratings for this analysis. One study scored 6 points (16), twelve studies scored 7 points (18–20, 22, 26–32, 34), five studies scored 8 points (15, 21, 23, 24, 33), and two studies scored 9 points (17, 25). The quality assessment of the included studies is presented in Table 2.

Table 2.

Quality of the studies included.

Study inclusion Case selection Comparability Outcome NOS
Anis et al. (15) 4 1 3 8
Armit et al. (16) 4 0 2 6
Dowsey and Choong (17) 4 2 3 9
Guo et al. (18) 4 1 2 7
Iqbal et al. (19) 4 1 2 7
Jämsen et al. (20) 4 1 2 7
Claus et al. (21) 4 1 3 8
Siu et al. (22) 4 1 2 7
Remily et al. (23) 4 1 3 8
Suzuki et al. (24) 4 1 3 8
Wang et al. (25) 4 2 3 9
Cheng et al. (26) 4 1 2 7
Ning et al. (27) 4 1 2 7
Baoyu (28) 4 1 2 7
Di (29) 4 1 2 7
Tao et al. (30) 4 1 2 7
Zhe et al. (31) 4 1 2 7
Shaoqiang et al. (32) 4 2 1 7
Zhongren (33) 4 2 2 8
Du (34) 4 1 2 7
Open in a new tab

Study quality was assessed using the NOS for cohort studies. The NOS awards a maximum of nine stars across three domains: Selection of the study groups (0–4 stars), comparability of the groups (0–2 stars), and assessment of the outcome (0–3 stars). A higher total score (out of nine) indicates a higher methodological quality of the study.

3.4. Description of the risk factors

3.4.1. The incidence of PJI after primary TKA

The incidence rates from the 20 studies included in the synthesis via R4.5.1 were analyzed using a random-effects model to calculate the corresponding 95% confidence intervals (95% CI), weight values, and I2 for each study. The results indicated substantial heterogeneity among the studies (I2 = 99.0%), suggesting that the pooled incidence does not accurately represent the true rate of PJI following primary TKA. This discrepancy may be related to factors such as surgical duration, approach, technique, and patient status, among other factors. Additionally, seven of the studies were case–control studies (18, 28–33), which may not adequately reflect the incidence of PJI. The forest plot is shown in Figure 2.

Figure 2.

Forest plot presenting the incidence and confidence intervals for multiple studies on a specific event. Each study is listed with its event count, total sample size, incidence rate, and ninety-five percent confidence interval. Weights range around five percent per study. A blue square represents a summary measure, and the red diamond indicates the overall random effects model result. Heterogeneity is high at ninety-nine percent.

Open in a new tab

Analysis results of incidence rates from 20 studies.

3.4.2. Longer operative time

A total of five studies were included (18, 19, 27, 31, 32). The pooled OR calculated using the fixed-effects model was 9.10 (95% CI: 7.66–10.80), with I2 = 0 and P = 0.63, indicating no heterogeneity among the studies. Therefore, it demonstrates that longer operative time is a risk factor for PJI following primary TKA. The forest plot is shown in Figure 3.

Figure 3.

Forest plot showing odds ratios for five studies comparing experimental and control groups. The plot highlights individual study weights and confidence intervals. The overall odds ratio is 9.10 with a 95% confidence interval of 7.66 to 10.80. The test for overall effect shows a significant result.

Open in a new tab

Forest plot of longer operative time and PJI following primary TKA.

3.4.3. Obesity

A total of four studies were included (17, 19, 20, 27). The pooled OR calculated using the fixed-effects model was 13.95 (95% CI: 12.06–16.14), with I2 = 0 and P = 0.50, indicating no heterogeneity among the studies. Therefore, it demonstrates that obesity is a risk factor for PJI following primary TKA. The forest plot is shown in Figure 4.

Figure 4.

Forest plot showing odds ratios for four studies on a fixed-effects model. Dowsey (2009) odds ratio: 8.96. Iqbal (2020) odds ratio: 14.23. Jämsen (2012) odds ratio: 6.40. Ning (2017) odds ratio: 7.57. Overall odds ratio: 13.95 with confidence interval [12.06, 16.14]. Heterogeneity is low with chi-squared value of 2.36 and I-squared value of zero percent. The test for overall effect has a Z-score of 35.52, with a p-value less than 0.00001. The plot indicates a higher favor for the experimental groups over the controls.

Open in a new tab

Forest plot of obesity and PJI following primary TKA.

3.4.4. Male gender

A total of five studies were included (16, 17, 21, 24, 30). The pooled OR calculated by the random-effects model was 2.67 (95% CI: 1.80–3.95). The I2 value was 54% with a P-value of 0.07, indicating that the heterogeneity among the studies was within an acceptable range. This demonstrates that male gender is a risk factor for PJI following primary TKA. The forest plot for the meta-analysis of the association between male gender and PJI risk following primary TKA is shown in Figure 5.

Figure 5.

Forest plot showing odds ratios with 95% confidence intervals for five studies: Armit 2018, Claus 2006, Dowsey 2009, Suzuki 2011, and Tao 2017. Each study's log odds ratio, standard error, weight, and interval are detailed. The combined effect size is marked by a diamond at approximately 2.67, with a confidence interval of 1.80 to 3.95. Heterogeneity statistics show Tau-squared equals 0.09 and I-squared equals 54%. The test for overall effect shows a Z-score of 4.91 with P equals 0.00001.

Open in a new tab

Forest plot of male gender and PJI following primary TKA.

3.4.5. Diabetes

A total of six studies were included (17, 20, 26, 27, 29, 33). The pooled OR calculated using the fixed-effects model was 2.98 (95% CI: 2.27–3.92), with I2 = 13% and P = 0.33. The results indicate low heterogeneity among the studies, which is within an acceptable range. This demonstrates that diabetes is a risk factor for PJI following primary TKA. The forest plot is shown in Figure 6.

Figure 6.

Forest plot from a meta-analysis shows the odds ratios for six studies, each indicated by a square with a horizontal line representing the confidence interval. The overall effect size is represented by a diamond at an odds ratio of 2.98. The heterogeneity among studies is low, with Chi² = 5.75, degrees of freedom = 5, and I² = 13%. The overall test effect is significant, Z = 7.85, p < 0.00001.

Open in a new tab

Forest plot of diabetes and PJI following primary TKA.

3.4.6. Longer hospital stay

A total of three studies were included (18, 28, 29). The pooled OR calculated using the fixed-effects model was 1.73 (95% CI: 1.38–2.16), with I2 = 18% and P = 0.30, indicating low heterogeneity among the studies, which is within an acceptable range. This demonstrates that a longer hospital stay is a risk factor for PJI following primary TKA. The forest plot is shown in Figure 7.

Figure 7.

Forest plot illustrating the odds ratio with 95% confidence intervals for three studies: Baoyu L 2016 (1.77), Di Q 2020 (1.65), and Guo H 2020 (5.77). Overall odds ratio is 1.73. Heterogeneity is low (Chi² = 2.44, I² = 18%).

Open in a new tab

Forest plot of longer hospital stay and PJI following primary TKA.

3.4.7. Use of immunosuppressants

A total of four studies were included (18, 28, 29, 31). The pooled OR calculated using the fixed-effects model was 5.76 (95% CI: 2.77–11.98), with I2 = 0 and P = 1.00, indicating no heterogeneity among the studies. Therefore, it demonstrates that the use of immunosuppressants is a risk factor for PJI following primary TKA. The forest plot is shown in Figure 8.

Figure 8.

Forest plot showing odds ratios for studies Di Q 2020, Guo H. 2020, Yubao L 2016, and Zhe S 2021. Odds ratios range from 5.02 to 6.20 with confidence intervals. The combined odds ratio is 5.76. No heterogeneity is present, with an I-squared of zero percent, and the overall effect test is significant with Z = 4.68 (p < 0.00001).

Open in a new tab

Forest plot of use of immunosuppressants and PJI following primary TKA.

3.4.8. Hypoalbuminemia

A total of three studies were included (18, 27, 32). Pooled analysis using a fixed-effects model yielded an OR of 6.24 (95% CI: 4.00–9.73), with I2 = 0 and P = 0.37, indicating no heterogeneity among studies and suggesting that only sampling error was present. This confirms that hypoalbuminemia is a risk factor for PJI following primary TKA. The forest plot is presented in Figure 9.

Figure 9.

Forest plot showing odds ratios from three studies: Guo H. 2020, Ning L. 2017, and Shaoqiang Y. 2024. Studies indicate higher odds ratios with size reflecting weight: Guo 5.20 [2.92, 9.25], Ning 22.61 [2.80, 182.34], Shaoqiang 7.17 [3.43, 14.98]. Total effect: 6.24 [4.00, 9.73]. No heterogeneity (I² = 0%). Overall effect significant (P < 0.00001). Red squares represent individual studies, black diamond shows overall effect.

Open in a new tab

Forest plot of hypoalbuminemia and PJI following primary TKA.

3.4.9. Underlying systemic inflammatory disease

A total of three studies were included (22, 27, 32). The pooled OR calculated using the fixed-effects model was 3.47 (95% CI: 1.92–6.27), with I2 = 0 and P = 0.92, indicating no heterogeneity among the studies. Therefore, it demonstrates that underlying systemic inflammatory disease is a risk factor for PJI following primary TKA. The forest plot is shown in Figure 10.

Figure 10.

Forest plot showing the odds ratios from three studies: Ning L 2017, Shaoqiang Y 2024, and Siu K.T. 2018. Each study's result is represented by a square with horizontal lines indicating the confidence interval. The combined effect size is shown as a diamond at 3.47 with a confidence interval of 1.92 to 6.27, favoring the experimental group. Heterogeneity is low with an I-squared of zero percent.

Open in a new tab

Forest plot of underlying systemic inflammatory disease and PJI following primary TKA.

3.5. Pooled results from the meta-analysis of risk factors

Meta-analysis pooling all risk factors revealed that each factor had a P-value of <0.05, indicating statistically significant differences. This demonstrates that all nine factors are risk factors for PJI following primary TKA, as shown in Table 3.

Table 3.

Results of the meta-analysis on risk factors associated with PJI following primary TKA.

Risk factor Effect model OR 95% CI I2 (%) Z P
Longer operative time (Figure 3) Fixed 9.10 7.66–10.80 0 25.21 <0.00001
Obesity (Figure 4) Fixed 13.95 12.06–16.14 0 35.52 <0.00001
Male gender (Figure 5) Random 2.67 1.80–3.95 54 4.91 <0.00001
Diabetes (Figure 6) Fixed 2.98 2.27–3.92 13 7.85 <0.00001
Longer hospital stay (Figure 7) Fixed 1.73 1.38–2.16 18 4.79 <0.00001
Use of immunosuppressants Fixed 5.76 2.77–11.98 0 4.68 <0.00001
Hypoalbuminemia (Figure 8) Fixed 6.24 4.00–9.73 0 8.09 <0.00001
Underlying systemic inflammatory disease (Figure 9) Fixed 3.47 1.92–6.27 0 4.12 <0.0001
Open in a new tab

This table presents the pooled associations between potential risk factors and PJI, expressed as odds ratios (ORs) with 95% confidence intervals (95% CIs). OR > 1 indicates higher risk. A fixed-effects model was used when heterogeneity was low (I2 ≤ 50%); otherwise, a random-effects model was applied. The I2 statistic quantifies the percentage of total variation across studies due to heterogeneity. The Z-test and corresponding P-value assess the statistical significance of the pooled OR.

3.6. Heterogeneity analysis

Heterogeneity was observed among studies in all three analyses (Figures 4–6), but it remained within acceptable limits. All meta-analytic results were generated only after studies with substantial heterogeneity were excluded through sensitivity analysis. For example, in Figure 2, the final pooled outcome was obtained after excluding four studies (15, 25, 28, 29). The final pooled analysis of the five studies (18, 19, 27, 31, 32) showed no heterogeneity.

This review has certain limitations, which are reflected in the meta-analyses, particularly for the risk factors of longer operative time, obesity, and hypoalbuminemia (Figures 2, 3, 8). In these three meta-analyses, four (15, 25, 28, 29), eight (15, 18, 23–25, 28, 29, 34), and three (28, 29, 34) studies, respectively, were excluded due to significant heterogeneity. This heterogeneity may be attributed to clinical diversity, sample size variations, and variability in outcomes. Furthermore, the risk factors longer hospital stay, hypoalbuminemia, and underlying systemic inflammatory disease each included only three studies, which may lead to incomplete data. Nevertheless, all meta-analyses showed acceptable levels of heterogeneity among the included studies, and the final results are considered reliable.

4. Discussion

The results of this meta-analysis show that obesity has a combined OR value of 13.95 (Figure 3), indicating that this factor, particularly pathological obesity, is the most significant risk factor for PJI following primary TKA. Patients with obesity, especially pathological obesity, often have multiple comorbidities (17, 20), such as diabetes and cardiovascular diseases, as well as unhealthy lifestyle habits (e.g., smoking and alcohol abuse), which collectively influence surgical outcomes. Multiple studies have confirmed that obesity is an independent risk factor for PJI after TKA (35–37). Some studies have also demonstrated that the combination of obesity and diabetes significantly increases the risk of PJI, particularly in patients with pathological obesity (BMI ≥ 40 kg/m2) and diabetes. These patients tend to experience more wound complications, poor soft tissue healing, and greater surgical difficulty, which can prolong operative time. The meta-analysis revealed that longer operative time has a combined OR value of 9.10 (Figure 6), making it the second strongest risk factor. Prolonged surgery increases soft tissue damage due to anatomical traction and extends the exposure of peri-wound tissues to the environment, thereby elevating the risk of bacterial colonization and infection. Therefore, clinicians should consider whether such patients are suitable candidates for TKA or whether alternative interventions, such as weight and glycemic control, should be implemented prior to surgery to reduce operative time, which is crucial for minimizing PJI risk.

This study confirmed that hypoalbuminemia (OR = 6.24) and the use of immunosuppressants (OR = 5.76) are important risk factors (Figures 8, 9). Hypoalbuminemia is a marker of malnutrition, which directly impairs tissue repair capabilities and immunoglobulin synthesis, weakening the patient's systemic immune defense (38, 39). Similarly, long-term use of immunosuppressants systemically suppresses the patient's immune response, making them more susceptible to opportunistic infections, especially those with underlying systemic inflammatory diseases. The analysis also identified systemic inflammatory diseases (OR = 3.47) as a risk factor for PJI (Figure 9). Patients with inherent immune system dysregulation who require long-term immunosuppressive therapy for their primary condition face a “double hit”: they experience both abnormal immune inflammation and iatrogenic immunosuppression, significantly compromising their ability to combat pathogens introduced during surgery (40, 41). This underscores the importance of preoperative multidisciplinary assessment, optimized medication management, and enhanced nutritional support for these patients.

Furthermore, diabetes (OR = 2.98), male gender (OR = 2.67), and prolonged hospital stay (OR = 1.76) were also identified as significant risk factors (Figures 4–6). However, it is noteworthy that no PJI cases were reported in diabetic patients with a BMI < 30 kg/m2, whereas PJI cases were reported in diabetic patients with a BMI > 30 kg/m2, particularly those with pathological obesity (20). This suggests that in diabetic patients undergoing TKA, the effects of obesity and diabetes synergize, a finding supported by multiple studies (8, 9, 17, 20, 35). The exact mechanism behind male gender as a risk factor for PJI remains unclear but may be related to hormonal levels, social behavioral factors, or differences in comorbidities. Although a longer hospital stay may increase the risk of colonization with drug-resistant bacteria and cross-infection, especially in suboptimal ward conditions, it is more likely a consequence rather than a cause of PJI.

The strength of this study lies in the use of meta-analysis, which integrates data from multiple studies to provide higher-level evidence-based conclusions. All included factors exhibited highly significant P-values (P < 0.00001), indicating strong statistical validity (Figure 10). Additionally, the low heterogeneity (I2) for most factors (e.g., obesity, operative time, hypoalbuminemia, immunosuppressant use, and systemic inflammatory diseases) suggests consistent results across studies, enhancing the reliability of the conclusions. However, this study has certain limitations. First, the nature of meta-analysis means that the conclusions are constrained by the quality and methodological heterogeneity of the included studies. Second, most risk factors were analyzed as binary variables, preventing dose-response analysis (e.g., levels of glycemic control). Future prospective studies and in-depth investigations targeting specific high-risk populations are needed to develop more precise risk stratification models and personalized intervention strategies.

In summary, this meta-analysis demonstrates that PJI results from the combined effects of multiple factors. The most significant modifiable risk factors include obesity, operative time, nutritional status, immune status, and comorbid systemic inflammatory diseases. Surgeons should use this evidence to conduct comprehensive preoperative assessments and multidisciplinary optimization of patients, adhere to meticulous surgical techniques and aseptic principles during operations, and enhance postoperative monitoring. This multilayered defense strategy will ultimately reduce the incidence of PJI following primary TKA.

5. Conclusion

Based on Level II evidence from this systematic review and meta-analysis, surgeons should assess the identified risk factors present in patients. Periprosthetic joint infection is the most serious complication of total joint arthroplasty in clinical practice, and inadequate management can impose a significant burden on both patients and healthcare institutions. Therefore, the evidence synthesized here supports the development of personalized, preoperative optimization plans aimed at mitigating these risks to improve surgical success and prevent PJI.

Acknowledgments

We wish to thank Maihemuti Yakufu for his research guidance.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This work was financially supported by the Health Care and Medical Research Special Project of Xinjiang Uygur Autonomous Region (BL202573, BL202460, BL202675), Youth Medical Science and Technology Talent Special Research Project (WJWY-202331), National Natural Science Foundation of China (82460420), Research and Innovation Team Project, Xinjiang Medical University (XYD2024C12), Tianshan talent Training Program (TSYC202301B039 and TSYC202301B106), Key Laboratory of High Incidence Disease Research in Xingjiang (Xinjiang Medical University), Ministry of Education (2023B04), Natural Science Foundation of Xinjiang Uygur Autonomous Region (2022D01C821), Research and Innovation Team Project, Sixth Affiliated Hospital of Xinjiang Medical University (LFYKYZXJJ2024001), Tianshan Innovation Team Program (2025D14019), Tianshan District Science & Technology development Program (WLMQTSQ-2025-KJXM05).

Edited by: Sujith Konan, University College London, United Kingdom

Reviewed by: Konstantinos Markatos, Salamina Medical Center, Greece

Aidos Konkayev, Astana Medical University, Kazakhstan

Abbreviations PJI, periprosthetic joint infection; CI, confidence interval; ASA, American Society of Anesthesiologists Physical Status Classification System; AVN, avascular necrosis of the femoral head; CCI, Charlson comorbidity index.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

JL: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. ZL: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. ZW: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. DL: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Visualization, Writing – original draft, Writing – review & editing. HZ: Data curation, Investigation, Project administration, Resources, Validation, Writing – review & editing. SH: Data curation, Investigation, Resources, Validation, Visualization, Writing – review & editing. QZ: Data curation, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – review & editing. MY: Conceptualization, Methodology, Project administration, Supervision, Validation, Visualization, Writing – review & editing. AA: Data curation, Formal analysis, Project administration, Validation, Visualization, Writing – review & editing. AH: Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence, and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsurg.2025.1715571/full#supplementary-material

Datasheet1.zip (348.6KB, zip)

References

  • 1.Gunaratne R, Pratt DN, Banda J, Fick DP, Khan RJK, Robertson BW. Patient dissatisfaction following total knee arthroplasty: a systematic review of the literature. J Arthroplasty. (2017) 32(12):3854–60. 10.1016/j.arth.2017.07.021 [DOI] [PubMed] [Google Scholar]
  • 2.Canovas F, Dagneaux L. Quality of life after total knee arthroplasty. Orthop Traumatol Surg Res. (2018) 104(1S):S41–6. 10.1016/j.otsr.2017.04.017 [DOI] [PubMed] [Google Scholar]
  • 3.Patel R. Periprosthetic joint infection. N Engl J Med. (2023) 388(3):251–62. 10.1056/NEJMra2203477 [DOI] [PubMed] [Google Scholar]
  • 4.Nelson SB, Pinkney JA, Chen AF, Tande AJ. Periprosthetic joint infection: current clinical challenges. Clin Infect Dis. (2023) 77(7):e34–45. 10.1093/cid/ciad360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Dalury DF, Pomeroy DL, Gorab RS, Adams MJ. Why are total knee arthroplasties being revised? J Arthroplasty. (2013) 28(8 Suppl):120–1. 10.1016/j.arth.2013.04.051 [DOI] [PubMed] [Google Scholar]
  • 6.Schwartz AM, Farley KX, Guild GN, Bradbury TL, Jr. Projections and epidemiology of revision hip and knee arthroplasty in the United States to 2030. J Arthroplasty. (2020) 35(6S):S79–85. 10.1016/j.arth.2020.02.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Premkumar A, Kolin DA, Farley KX, Wilson JM, McLawhorn AS, Cross MB, et al. Projected economic burden of periprosthetic joint infection of the hip and knee in the United States. J Arthroplasty. (2021) 36(5):1484–89.e3. 10.1016/j.arth.2020.12.005 [DOI] [PubMed] [Google Scholar]
  • 8.Pulido L, Ghanem E, Joshi A, Purtill JJ, Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res. (2008) 466(7):1710–5. 10.1007/s11999-008-0209-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Depypere M, Morgenstern M, Kuehl R, Senneville E, Moriarty TF, Obremskey WT, et al. Pathogenesis and management of fracture-related infection. Clin Microbiol Infect. (2020) 26(5):572–8. 10.1016/j.cmi.2019.08.006 [DOI] [PubMed] [Google Scholar]
  • 10.Cooper AM, Shope AJ, Javid M, Parsa A, Chinoy MA, Parvizi J. Musculoskeletal infection in pediatrics: assessment of the 2018 International Consensus Meeting on Musculoskeletal Infection. J Bone Joint Surg Am. (2019) 101(24):e133. 10.2106/JBJS.19.00572 [DOI] [PubMed] [Google Scholar]
  • 11.Henderson RA, Austin MS. Management of periprosthetic joint infection: the more we learn, the less we know. J Arthroplasty. (2017) 32(7):2056–9. 10.1016/j.arth.2017.02.023 [DOI] [PubMed] [Google Scholar]
  • 12.Longo UG, De Salvatore S, Bandini B, Lalli A, Barillà B, Budhiparama NC, et al. Debridement, antibiotics, and implant retention (DAIR) for the early prosthetic joint infection of total knee and hip arthroplasties: a systematic review. J ISAKOS. (2024) 9(1):62–70. 10.1016/j.jisako.2023.09.003 [DOI] [PubMed] [Google Scholar]
  • 13.Sharma RK, Lumban-Gaol I, Vinayak U, Budhiparama NC. Timing-specific debridement, antibiotics, and implant retention and 2-stage revision for megaprosthesis-related infection: optimizing the window period for improved outcomes. Arthroplast Today. (2025) 33:101688. 10.1016/j.artd.2025.101688 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Zmistowski B, Karam JA, Durinka JB, Casper DS, Parvizi J. Periprosthetic joint infection increases the risk of one-year mortality. J Bone Joint Surg Am. (2013) 95(24):2177–84. 10.2106/JBJS.L.00789 [DOI] [PubMed] [Google Scholar]
  • 15.Anis HK, Sodhi N, Klika AK, Mont MA, Barsoum WK, Higuera CA, et al. Is operative time a predictor for post-operative infection in primary total knee arthroplasty? J Arthroplasty. (2019) 34(7S):S331–6. 10.1016/j.arth.2018.11.022 [DOI] [PubMed] [Google Scholar]
  • 16.Armit D, Vickers M, Parr A, Van Rosendal S, Trott N, Gunasena R, et al. Humidity a potential risk factor for prosthetic joint infection in a tropical Australian hospital. ANZ J Surg. (2018) 88(12):1298–301. 10.1111/ans.14916 [DOI] [PubMed] [Google Scholar]
  • 17.Dowsey MM, Choong PF. Obese diabetic patients are at substantial risk for deep infection after primary TKA. Clin Orthop Relat Res. (2009) 467(6):1577–81. 10.1007/s11999-008-0551-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Guo H, Xu C, Chen J. Risk factors for periprosthetic joint infection after primary artificial hip and knee joint replacements. J Infect Dev Ctries. (2020) 14(6):565–71. 10.3855/jidc.11013 [DOI] [PubMed] [Google Scholar]
  • 19.Iqbal F, Shafiq B, Zamir M, Noor S, Memon N, Memon N, et al. Micro-organisms and risk factors associated with prosthetic joint infection following primary total knee replacement-our experience in Pakistan. Int Orthop. (2020) 44(2):283–9. 10.1007/s00264-019-04472-1 [DOI] [PubMed] [Google Scholar]
  • 20.Jämsen E, Nevalainen P, Eskelinen A, Huotari K, Kalliovalkama J, Moilanen T. Obesity, diabetes, and preoperative hyperglycemia as predictors of periprosthetic joint infection: a single-center analysis of 7181 primary hip and knee replacements for osteoarthritis. J Bone Joint Surg Am. (2012) 94(14):e101. 10.2106/JBJS.J.01935 [DOI] [PubMed] [Google Scholar]
  • 21.Claus A, Asche G, Brade J, Bosing-Schwenkglenks M, Horchler H, Müller-Färber J, et al. Identifizierung von risikofaktoren postoperativer komplikationen in der primären knieendoprothetik [Risk profiling of postoperative complications in 17,644 total knee replacements]. Unfallchirurg. (2006) 109(1):5–12. 10.1007/s00113-005-0992-2; German. [DOI] [PubMed] [Google Scholar]
  • 22.Siu KT, Ng FY, Chan PK, Fu HC, Yan CH, Chiu KY. Bacteriology and risk factors associated with periprosthetic joint infection after primary total knee arthroplasty: retrospective study of 2543 cases. Hong Kong Med J. (2018) 24(2):152–7. 10.12809/hkmj176885 [DOI] [PubMed] [Google Scholar]
  • 23.Remily EA, Sax OC, Douglas SJ, Salib CG, Salem HS, Monárrez RG, et al. Inflammatory bowel disease is associated with increased complications after total knee arthroplasty. Knee. (2023) 40:313–8. 10.1016/j.knee.2022.12.010 [DOI] [PubMed] [Google Scholar]
  • 24.Suzuki G, Saito S, Ishii T, Motojima S, Tokuhashi Y, Ryu J. Previous fracture surgery is a major risk factor of infection after total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. (2011) 19(12):2040–4. 10.1007/s00167-011-1525-x [DOI] [PubMed] [Google Scholar]
  • 25.Wang Q, Goswami K, Shohat N, Aalirezaie A, Manrique J, Parvizi J. Longer operative time results in a higher rate of subsequent periprosthetic joint infection in patients undergoing primary joint arthroplasty. J Arthroplasty. (2019) 34(5):947–53. 10.1016/j.arth.2019.01.027 [DOI] [PubMed] [Google Scholar]
  • 26.Cheng L, He Q, Haijiao W, Yuwei L. Analysis of related factors in periprosthetic joint infection following primary total knee arthroplasty. OJC. (2022) 30(13):1158–62. 10.3977/j.issn.1005-8478.2022.13.02 [DOI] [Google Scholar]
  • 27.Ning L, Yiqin Z, Shu C, Jinhui P, Wuxiang H, Shiqi C, et al. Risk factors and pathogen distribution analysis of periprosthetic joint infection after primary total knee arthroplasty. Chin J Joint Surg. (2017) 11(05):461–6. 10.3877/cma.j.issn.1674-134X.2017.05.004 [DOI] [Google Scholar]
  • 28.Baoyu L. Risk factors analysis for prosthetic joint infections after primary hip and knee arthroplasty (PhD thesis). Nanjing University; (2016). [Google Scholar]
  • 29.Di Q. Analysis and treatment of the risk factors of the prosthetic joint infection after the first hip and knee replacement (PhD thesis). Yangzhou University; (2016). [Google Scholar]
  • 30.Tao W, Junwei W, Qi P, Wei S, Tao M, Yumin Z. Analysis of risk factors for infection after total knee arthroplasty. Shaanxi J Tradit Chin Med. (2017) 46(2):204–5. 10.3969/j.issn.1000-7377.2017.02.027 [DOI] [Google Scholar]
  • 31.Zhe S, Chaojun S, Hong L, Zhifu H, Chang Z, Yue L, et al. Risk factors of periprosthetic infection after primary total knee arthroplasty. Chin J Nosocomiol. (2021) 21(05):546–51. 10.16718/j.1009-7708.2021.05.008 [DOI] [Google Scholar]
  • 32.Shaoqiang Y, Peng F, Dongsheng W, Yu Z, Tao Y, Jianning Z, et al. Analysis of risk factors for periprosthetic joint infection after primary total knee arthroplasty and construction of a nomogram prediction model. Chin J Orthop Trauma. (2024) 26(3):234–40. 10.3760/cma.j.cn115530-20240118-00033 [DOI] [Google Scholar]
  • 33.Zhongren Z. Analysis of risk factors and diagnostic indexes of periprosthetic joint infection after primary joint replacement (PhD thesis). Jining Medical University; (2023). [Google Scholar]
  • 34.Du L. Correlation between preoperative nutritional status and periprosthetic joint infection in patients undergoing total knee arthroplasty (PhD thesis). Henan University; (2022). [Google Scholar]
  • 35.Kapadia BH, Berg RA, Daley JA, Fritz J, Bhave A, Mont MA. Periprosthetic joint infection. Lancet. (2016) 387(10016):386–94. 10.1016/S0140-6736(14)61798-0 [DOI] [PubMed] [Google Scholar]
  • 36.Bhanushali A, Tran L, Nairne-Nagy J, Bereza S, Callary SA, Atkins GJ, et al. Patient-related predictors of treatment failure after two-stage total hip arthroplasty revision for periprosthetic joint infection: a systematic review and meta-analysis. J Arthroplasty. (2024) 39(9):2395–2402.e14. 10.1016/j.arth.2024.04.053 [DOI] [PubMed] [Google Scholar]
  • 37.Agarwal AR, Wang KY, Mikula JD, Berger P, Kreulen RT, Nayar SK, et al. Prior nonshoulder periprosthetic joint infection increases the risk of surgical site infection, sepsis, and all-cause revision after primary total shoulder arthroplasty. J Am Acad Orthop Surg. (2022) 30(3):133–9. 10.5435/JAAOS-D-21-00745 [DOI] [PubMed] [Google Scholar]
  • 38.Gatta A, Verardo A, Bolognesi M. Hypoalbuminemia. Intern Emerg Med. (2012) 7(Suppl 3):S193–9. 10.1007/s11739-012-0802-0 [DOI] [PubMed] [Google Scholar]
  • 39.Wiedermann CJ. Hypoalbuminemia as surrogate and culprit of infections. Int J Mol Sci. (2021) 22(9):4496. 10.3390/ijms22094496 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Meier-Schiesser B, French LE. Autoinflammatory syndromes. J Dtsch Dermatol Ges. (2021) 19(3):400–26. 10.1111/ddg.14332 [DOI] [PubMed] [Google Scholar]
  • 41.Antonini L, Le Mauff B, Marcelli C, Aouba A, de Boysson H. Rhupus: a systematic literature review. Autoimmun Rev. (2020) 19(9):102612. 10.1016/j.autrev.2020.102612 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Datasheet1.zip (348.6KB, zip)

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Articles from Frontiers in Surgery are provided here courtesy of Frontiers Media SA

RESOURCES