Mildly Elevated Preoperative Inflammatory Markers and Postoperative Complications Following Total Knee Arthroplasty: A Propensity-Matched Cohort Analysis Authors: Josef W Kubofcik, Stefan W Fleps, Christopher J Drinkwater Published: 2026 ABSTRACT AbstractBackgroundThere is increasing use of preoperative biomarker assessment in patients undergoing total knee arthroplasty. Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) are the commonly used measures of ongoing inflammation. The impact of discordance between these values on postoperative outcomes remains poorly understood.MethodsThree cohorts were identified: one with a mildly elevated ESR and normal CRP, the other with a mildly elevated CRP and normal ESR, and the third with a mildly elevated ESR and CRP, with laboratory values obtained within 1 month of primary total knee arthroplasty. Groups were compared to patients with normal markers.ResultsAt 90 days postoperatively, patients with mildly elevated CRP and ESR had a significantly higher odds of periprosthetic joint infection (PJI) compared to patients with normal inflammatory markers (odds ratio [OR]: 3.90; 95% confidence interval [CI]: 2.28-6.68; P < .001). At 3 years postoperatively, patients with mildly elevated inflammatory markers had significantly higher odds of PJI (OR: 3.80; 95% CI: 2.41-6.00; P < .001), complications (OR: 1.34, 95% CI: 1.00-1.79; P = .047), and revision knee arthroplasty (OR: 3.34; 95% CI: 1.68-6.64; P < .001). At 90 days, patients with mildly elevated preoperative CRP (5.1-15 mg/L) and normal ESR demonstrated significantly increased odds of PJI compared to patients with normal markers (OR: 2.00, 95% CI: 1.37-2.87; P < .001).ConclusionsMildly elevated CRP and ESR were significantly associated with an increased risk of PJI, revision arthroplasty, and composite complications. Elevated ESR with normal CRP was associated with a higher risk of composite complications at 3 years postoperatively.Keywords: Total knee arthroplasty, C-reactive protein, Erythrocyte sedimentation rate, Periprosthetic joint infection, Propensity score matching Introduction Total knee arthroplasty (TKA) is among the most frequently performed elective orthopaedic procedures in the United States, with an annual volume of 1,013,533 procedures performed in 2022 [1]. Global forecasting predicts that by 2050, the volume of TKA is expected to rise 130% across all countries [2]. Importantly, due to the relative safety and efficacy of the procedure, revision TKA rates have increased steadily and are projected to continue to rise [3,4]. Common indications for revision include periprosthetic joint infection (PJI), aseptic implant loosening, and instability, with PJI being the primary cause of revision [5,6]. Complications and revisions in TKA place a significant financial burden on healthcare systems and, more importantly, patients [7]. Understanding which patient populations are at risk for complications and revision procedures is essential to optimizing patient outcomes and reducing the burden of revision surgery. An increasingly common method to assess preoperative risk in various surgical fields has been the adoption of preoperative biomarker screening [8,9]. Serologic inflammatory markers, specifically C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), are the 2 most routinely obtained markers in preoperative evaluations [10]. These tests are commonly used in orthopaedics to assess for PJI postoperatively but are not yet universally accepted as preoperative screening tools [11]. Additionally, there are no universally accepted preoperative thresholds defining what constitutes a mildly elevated CRP or ESR to guide surgical decision-making [12]. In TKA patients, preoperative elevations in CRP and ESR have been correlated with an increased risk of complications and revision procedures in some studies, but this finding is not universal [[12], [13], [14]]. A common clinical challenge arises when patients present with a mildly elevated CRP or ESR, which can reflect normal variation associated with aging, obesity, or low-grade chronic inflammation [15]. Furthermore, CRP and ESR are commonly tested in tandem, and discordance between values, where one falls within normal limits while the other is mildly elevated, occurs frequently in clinical practice [16]. This patient group has yet to be thoroughly studied in TKA, as most prior studies examine large elevations in either marker or do not rigorously account for discordance between the 2 [13,14,17]. We hypothesized that the most complications would be seen in the group with elevations in CRP and ESR. Additionally, we hypothesized that the mildly elevated CRP and ESR group would demonstrate increased odds of PJI. To address this gap, we utilized the TriNetX research network to perform a propensity score-matched analysis comparing postoperative outcomes following primary TKA in patients with mild isolated preoperative CRP elevation vs mild isolated ESR elevation as well as mildly elevated CRP and ESR, each compared to patients with entirely normal inflammatory markers. We examined whether mild isolated elevation of either marker was independently associated with increased rates of PJI, revision surgery, manipulation under anesthesia (MUA), or other adverse events at 90 days, 1 year, and 3 years postoperatively. Material and methods Study population This was a retrospective cohort study using data collected from the TriNetX research network on May 3, 2026, and included data from 109 healthcare organizations. This study utilized deidentified patient data from the real-time TriNetX platform and received a determination of Not Human Subjects Research by the institutional review board on February 2, 2026 (STUDY00011359). Informed consent was waived due to the use of deidentified data. The index event was defined as primary TKA (Current Procedural Terminology: 27447). Three separate 1:1 propensity-matched analyses were performed: (1) mildly elevated CRP and mildly elevated ESR vs normal markers, (2) mildly elevated CRP with normal ESR vs normal markers, and (3) mildly elevated ESR with normal CRP vs normal markers. The mildly elevated CRP with normal ESR group was defined by having a measured CRP of 5.1-15.0 mg/L (TNX:9063) and an ESR of 0-25.0 mm/h (TNX:9066) within 1 month of primary TKA. The mildly elevated ESR with normal CRP group was defined by having a measured ESR of 25.1-40.0 mm/h and a CRP of 0-5.0 mg/L within 1 month of primary TKA. The control group for each analysis was defined by having a measured CRP of 0-5.0 mg/L and an ESR of 0-25.0 mm/h within 1 month of primary TKA. The cohorts were matched 1:1 using the built-in TriNetX propensity score matching algorithm based on the following baseline characteristics: age at index, sex, race, ethnicity, body mass index, hypertensive diseases, diabetes mellitus, atrial fibrillation and flutter, heart failure, acute kidney failure and chronic kidney disease, other chronic obstructive pulmonary disease, rheumatoid arthritis, ankylosing spondylitis, other inflammatory spondylopathies, human immunodeficiency virus, long term use of systemic steroids, history of malignancy, nicotine dependence, and alcohol-related disorders (Supplementary Table 1). Comorbidities were identified based on diagnoses present any time up to 1 day before the index event. After matching, each analysis yielded 534 matched pairs in the concordantly elevated ESR and CRP group, 1126 matched pairs in the mildly elevated CRP group and 592 matched pairs in the mildly elevated ESR group. Study outcomes The following study outcomes were assessed at 90 days, 1 year, and 3 years postoperatively: MUA, PJI, arthrocentesis, and a composite complications endpoint including pulmonary embolism, acute myocardial infarction, deep vein thrombosis, sepsis, pneumonia, stroke, and acute kidney failure (Supplementary Table 2). Revision arthroplasty was analyzed at 1 and 3 years postoperatively. Patients with the outcome of interest prior to index were not excluded from the analysis. Statistical analysis All statistical analyses were performed within the TriNetX platform. Propensity score matching was performed using logistic regression. Differences in outcomes between matched cohorts were analyzed using risk difference and odds ratios (ORs) with associated 95% confidence intervals (CIs). Propensity score matching was selected as the primary analytic approach as it allows for cohort balance analogous to multivariable regression and allows for reporting of standardized mean differences. A P value of less than .05 was considered statistically significant throughout the study. Results After propensity score matching, the concordantly elevated CRP and ESR group had 534 pairs of matched patients. The elevated CRP analysis yielded 1126 matched pairs, and the elevated ESR analysis yielded 592 matched pairs. All baseline demographic and comorbidity characteristics were balanced between cohorts after matching (Table 1, Table 2, Table 3). The propensity matching for the mildly elevated ESR and CRP group is shown in Table 1. The propensity matching results for the mildly elevated ESR group with normal CRP are shown in Table 2. The propensity matching results for the mildly elevated CRP group with normal ESR are shown in Table 3. Table 1Propensity score matching for concordantly elevated CRP and ESR group.CharacteristicBefore matchingAfter matchingNormal (N = 2876) N (%)Both elevated (N = 543) N (%)SMDNormal (N = 534) N (%)Both elevated (N = 534) N (%)SMDDemographics Age (y), mean ± SD66.1 ± 10.264.7 ± 10.20.13964.6 ± 10.264.8 ± 10.20.016 Sex, N (%) Female1580 (54.9%)386 (71.1%)0.339384 (71.9%)377 (70.6%)0.029 Male1282 (44.6%)157 (28.9%)0.329150 (28.1%)157 (29.4%)0.029 Race, N (%) White2430 (84.5%)400 (73.7%)0.269398 (74.5%)400 (74.9%)0.009 Black or African American272 (9.5%)99 (18.2%)0.25690 (16.9%)92 (17.2%)0.010 Hispanic or Latino75 (2.6%)30 (5.5%)0.14828 (5.2%)28 (5.2%)<0.001 Asian35 (1.2%)10 (1.8%)0.05110 (1.9%)10 (1.9%)<0.001 American Indian or Alaska Native12 (0.4%)10 (1.8%)0.1350 (0%)10 (1.9%)0.195 Native Hawaiian or Pacific Islander10 (0.3%)10 (1.8%)0.14410 (1.9%)10 (1.9%)<0.001 Other race55 (1.9%)15 (2.8%)0.05618 (3.4%)14 (2.6%)0.044 Unknown race66 (2.3%)18 (3.3%)0.06216 (3.0%)17 (3.2%)0.011 Ethnicity, N (%) Not Hispanic or Latino2240 (77.9%)400 (73.7%)0.099394 (73.8%)395 (74.0%)0.004 Hispanic or Latino75 (2.6%)30 (5.5%)0.14828 (5.2%)28 (5.2%)<0.001 Unknown ethnicity561 (19.5%)113 (20.8%)0.033112 (21.0%)111 (20.8%)0.005 Comorbidities, N (%) Body mass index (Z68)623 (21.7%)189 (34.8%)0.295177 (33.1%)182 (34.1%)0.020 Diabetes mellitus558 (19.4%)168 (30.9%)0.268144 (27.0%)161 (30.1%)0.071 Essential primary hypertension1675 (58.2%)407 (75.0%)0.360409 (76.6%)398 (74.5%)0.048 Atrial fibrillation and flutter262 (9.1%)45 (8.3%)0.02946 (8.6%)45 (8.4%)0.007 Heart failure156 (5.4%)49 (9.0%)0.13947 (8.8%)48 (9.0%)0.007 Acute kidney failure/chronic kidney disease338 (11.8%)119 (21.9%)0.274103 (19.3%)110 (20.6%)0.033 Chronic obstructive pulmonary disease155 (5.4%)58 (10.7%)0.19653 (9.9%)54 (10.1%)0.006 Rheumatoid arthritis with rheumatoid factor (M05)78 (2.7%)33 (6.1%)0.16531 (5.8%)30 (5.6%)0.008 Other rheumatoid arthritis (M06)270 (9.4%)85 (15.7%)0.19076 (14.2%)80 (15.0%)0.021 Ankylosing spondylitis (M45)24 (0.8%)10 (1.8%)0.08810 (1.9%)10 (1.9%)<0.001 Other inflammatory spondylopathies (M46)136 (4.7%)28 (5.2%)0.02025 (4.7%)28 (5.2%)0.026 HIV (B20)10 (0.3%)10 (1.8%)0.14410 (1.9%)10 (1.9%)<0.001 Long-term systemic steroid use (Z79.52)105 (3.7%)34 (6.3%)0.12132 (6.0%)34 (6.4%)0.016 Personal history of malignant neoplasm (Z85)326 (11.3%)48 (8.8%)0.08348 (9.0%)46 (8.6%)0.013 Alcohol-related disorders (F10)105 (3.7%)30 (5.5%)0.09027 (5.1%)24 (4.5%)0.026 Nicotine dependence (F17)227 (7.9%)68 (12.5%)0.15361 (11.4%)56 (10.5%)0.030SMD, standardized mean difference.SMD <0.10 indicates adequate balance between cohorts. Cohort 1: TKA patients with normal inflammatory markers (CRP 0-5 mg/L and ESR 0-25 mm/h). Cohort 2: TKA patients with concordantly mildly elevated CRP (5.1-15 mg/L) and ESR (25.1-40 mm/h). Cohorts were balanced using 1:1 propensity score matching on the listed demographic and comorbidity covariates. Propensity score matching for concordantly elevated CRP and ESR group. SMD, standardized mean difference. SMD <0.10 indicates adequate balance between cohorts. Cohort 1: TKA patients with normal inflammatory markers (CRP 0-5 mg/L and ESR 0-25 mm/h). Cohort 2: TKA patients with concordantly mildly elevated CRP (5.1-15 mg/L) and ESR (25.1-40 mm/h). Cohorts were balanced using 1:1 propensity score matching on the listed demographic and comorbidity covariates. Table 2Propensity score matching for mildly elevated CRP with normal ESR group.CharacteristicBefore matchingAfter matchingNormal (N = 2637) N (%)Elevated CRP (N = 1147) N (%)SMDNormal (N = 1126) N (%)Elevated CRP (N = 1126) N (%)SMDDemographics Age (y), mean ± SD65.6 ± 10.162.5 ± 10.90.29562.6 ± 10.562.7 ± 10.90.005 Sex, N (%) Female1455 (55.2%)669 (58.3%)0.064658 (58.4%)653 (58.0%)0.009 Male1168 (44.3%)469 (40.9%)0.069463 (41.1%)466 (41.4%)0.005 Race, N (%) White2269 (86.0%)953 (83.1%)0.082927 (82.3%)940 (83.5%)0.031 Black or African American198 (7.5%)122 (10.6%)0.109129 (11.5%)119 (10.6%)0.028 Hispanic or Latino76 (2.9%)41 (3.6%)0.03935 (3.1%)38 (3.4%)0.015 Asian32 (1.2%)10 (0.9%)0.03410 (0.9%)10 (0.9%)<0.001 American Indian or Alaska Native12 (0.5%)10 (0.9%)0.05110 (0.9%)10 (0.9%)<0.001 Native Hawaiian or Pacific Islander10 (0.4%)10 (0.9%)0.06310 (0.9%)10 (0.9%)<0.001 Other race55 (2.1%)31 (2.7%)0.04036 (3.2%)30 (2.7%)0.032 Unknown race65 (2.5%)32 (2.8%)0.02024 (2.1%)28 (2.5%)0.024 Ethnicity, N (%) Not Hispanic or Latino2394 (90.8%)1018 (88.8%)0.0671008 (89.5%)1004 (89.2%)0.012 Hispanic or Latino76 (2.9%)41 (3.6%)0.03935 (3.1%)38 (3.4%)0.015 Unknown ethnicity167 (6.3%)88 (7.7%)0.05283 (7.4%)84 (7.5%)0.003 Comorbidities, N (%) Body mass index (Z68)697 (26.4%)433 (37.8%)0.244405 (36.0%)417 (37.0%)0.022 Diabetes mellitus509 (19.3%)259 (22.6%)0.081229 (20.3%)251 (22.3%)0.048 Hypertensive diseases1511 (57.3%)753 (65.6%)0.172756 (67.1%)734 (65.2%)0.041 Atrial fibrillation and flutter234 (8.9%)88 (7.7%)0.04488 (7.8%)87 (7.7%)0.003 Heart failure136 (5.2%)71 (6.2%)0.04568 (6.0%)67 (6.0%)0.004 Acute kidney failure/chronic kidney disease286 (10.8%)171 (14.9%)0.122159 (14.1%)159 (14.1%)<0.001 Chronic obstructive pulmonary disease156 (5.9%)93 (8.1%)0.08678 (6.9%)88 (7.8%)0.034 Rheumatoid arthritis with rheumatoid factor (M05)78 (3.0%)54 (4.7%)0.09142 (3.7%)47 (4.2%)0.023 Other rheumatoid arthritis (M06)263 (10.0%)139 (12.1%)0.068130 (11.5%)133 (11.8%)0.008 Ankylosing spondylitis (M45)24 (0.9%)14 (1.2%)0.03019 (1.7%)14 (1.2%)0.037 Other inflammatory spondylopathies (M46)106 (4.0%)68 (5.9%)0.08856 (5.0%)67 (6.0%)0.043 HIV (B20)10 (0.4%)10 (0.9%)0.06310 (0.9%)10 (0.9%)<0.001 Long-term systemic steroid use (Z79.52)96 (3.6%)66 (5.8%)0.10060 (5.3%)58 (5.2%)0.008 Personal history of malignant neoplasm (Z85)280 (10.6%)129 (11.2%)0.020126 (11.2%)124 (11.0%)0.006 Alcohol-related disorders (F10)104 (3.9%)71 (6.2%)0.10365 (5.8%)65 (5.8%)<0.001 Nicotine dependence (F17)215 (8.2%)175 (15.3%)0.222163 (14.5%)161 (14.3%)0.005SMD, standardized mean difference.SMD <0.10 indicates adequate balance between cohorts. Cohort 1: TKA patients with normal inflammatory markers (CRP 0-5 mg/L and ESR 0-25 mm/h). Cohort 2: TKA patients with mildly elevated CRP (5.1-15 mg/L) and normal ESR (0-25 mm/h). Cohorts were balanced using 1:1 propensity score matching on the listed demographic and comorbidity covariates. Propensity score matching for mildly elevated CRP with normal ESR group. SMD, standardized mean difference. SMD <0.10 indicates adequate balance between cohorts. Cohort 1: TKA patients with normal inflammatory markers (CRP 0-5 mg/L and ESR 0-25 mm/h). Cohort 2: TKA patients with mildly elevated CRP (5.1-15 mg/L) and normal ESR (0-25 mm/h). Cohorts were balanced using 1:1 propensity score matching on the listed demographic and comorbidity covariates. Table 3Propensity score matching for mildly elevated ESR with normal CRP group.CharacteristicBefore matchingAfter matchingNormal (N = 3050) N (%)Elevated ESR (N = 596) N (%)SMDNormal (N = 592) N (%)Elevated ESR (N = 592) N (%)SMDDemographics Age (y), mean ± SD66.2 ± 10.167.3 ± 9.80.11367.4 ± 10.167.3 ± 9.80.017 Sex, N (%) Female1676 (55.0%)431 (72.3%)0.367440 (74.3%)427 (72.1%)0.050 Male1360 (44.6%)161 (27.0%)0.373147 (24.8%)161 (27.2%)0.054 Race, N (%) White2595 (85.1%)440 (73.8%)0.281442 (74.7%)440 (74.3%)0.008 Black or African American277 (9.1%)109 (18.3%)0.270105 (17.7%)106 (17.9%)0.004 Hispanic or Latino76 (2.5%)17 (2.9%)0.02215 (2.5%)16 (2.7%)0.011 Asian35 (1.1%)13 (2.2%)0.08110 (1.7%)13 (2.2%)0.037 American Indian or Alaska Native12 (0.4%)10 (1.7%)0.12710 (1.7%)10 (1.7%)<0.001 Native Hawaiian or Pacific Islander10 (0.3%)10 (1.7%)0.13610 (1.7%)10 (1.7%)<0.001 Other race57 (1.9%)11 (1.8%)0.00212 (2.0%)11 (1.9%)0.012 Unknown race68 (2.2%)20 (3.4%)0.06820 (3.4%)19 (3.2%)0.009 Ethnicity, N (%) Not Hispanic or Latino2408 (79.0%)453 (76.0%)0.071441 (74.5%)451 (76.2%)0.039 Hispanic or Latino76 (2.5%)17 (2.9%)0.02215 (2.5%)16 (2.7%)0.011 Unknown ethnicity566 (18.6%)126 (21.1%)0.065136 (23.0%)125 (21.1%)0.045 Comorbidities, N (%) Body mass index (Z68)784 (25.7%)187 (31.4%)0.126182 (30.7%)186 (31.4%)0.015 Diabetes mellitus600 (19.7%)186 (31.2%)0.267164 (27.7%)182 (30.7%)0.067 Hypertensive diseases1811 (59.4%)430 (72.1%)0.272413 (69.8%)426 (72.0%)0.048 Atrial fibrillation and flutter276 (9.0%)58 (9.7%)0.02365 (11.0%)58 (9.8%)0.039 Heart failure164 (5.4%)69 (11.6%)0.22469 (11.7%)68 (11.5%)0.005 Acute kidney failure/chronic kidney disease350 (11.5%)111 (18.6%)0.201109 (18.4%)110 (18.6%)0.004 Chronic obstructive pulmonary disease170 (5.6%)51 (8.6%)0.11742 (7.1%)50 (8.4%)0.050 Rheumatoid arthritis with rheumatoid factor (M05)83 (2.7%)19 (3.2%)0.02819 (3.2%)19 (3.2%)<0.001 Other rheumatoid arthritis (M06)285 (9.3%)64 (10.7%)0.04657 (9.6%)64 (10.8%)0.039 Ankylosing spondylitis (M45)25 (0.8%)10 (1.7%)0.07710 (1.7%)10 (1.7%)<0.001 Other inflammatory spondylopathies (M46)142 (4.7%)26 (4.4%)0.01423 (3.9%)26 (4.4%)0.025 HIV (B20)10 (0.3%)10 (1.7%)0.13610 (1.7%)10 (1.7%)<0.001 Long-term systemic steroid use (Z79.52)108 (3.5%)27 (4.5%)0.05023 (3.9%)27 (4.6%)0.034 Personal history of malignant neoplasm (Z85)341 (11.2%)67 (11.2%)0.00260 (10.1%)67 (11.3%)0.038 Nicotine dependence (F17)239 (7.8%)71 (11.9%)0.13767 (11.3%)70 (11.8%)0.016 Alcohol-related disorders (F10)113 (3.7%)18 (3.0%)0.03820 (3.4%)18 (3.0%)0.019SMD, standardized mean difference.SMD <0.10 indicates adequate balance between cohorts. Cohort 1: TKA patients with normal inflammatory markers (CRP 0-5 mg/L and ESR 0-25 mm/h). Cohort 2: TKA patients with mildly elevated ESR (25.1-40 mm/h) and normal CRP (0-5 mg/L). Cohorts were balanced using 1:1 propensity score matching on the listed demographic and comorbidity covariates. Propensity score matching for mildly elevated ESR with normal CRP group. SMD, standardized mean difference. SMD <0.10 indicates adequate balance between cohorts. Cohort 1: TKA patients with normal inflammatory markers (CRP 0-5 mg/L and ESR 0-25 mm/h). Cohort 2: TKA patients with mildly elevated ESR (25.1-40 mm/h) and normal CRP (0-5 mg/L). Cohorts were balanced using 1:1 propensity score matching on the listed demographic and comorbidity covariates. Ninety-day outcomes At 90 days, patients with mildly elevated preoperative CRP and mildly elevated ESR demonstrated significantly increased odds of PJI compared to patients with entirely normal inflammatory markers (12.0% vs 3.4%; OR: 3.90; 95% CI: 2.28-6.69; P < .001) (Table 4). Elevated CRP with normal ESR was associated with significantly higher odds of PJI at 90 days compared to patients with normal preoperative inflammatory markers (7.6% vs 4.0%; OR: 1.99; 95% CI: 1.37-2.87; P < .001) (Table 4). No significant differences were observed for MUA, arthrocentesis, or composite complications in the CRP group at this time point. In the elevated ESR group, no statistically significant differences were observed for any outcome at 90 days (Table 4). MUA could not be assessed in the ESR group at 90 days due to insufficient sample size. Table 4Ninety-day postoperative outcomes after primary TKA in patients with mildly elevated preoperative inflammatory markers vs normal markers.OutcomeNormal n (%)Elevated n (%)Risk difference (95% CI)Odds ratio (95% CI)PElevated CRP and elevated ESR (CRP 5.1-15 mg/L, ESR 25.1-40 mm/h) |n = 534 vs 534 MUA————— PJI18/534 (3.4%)64/534 (12.0%)0.086 (0.055, 0.118)3.904 (2.280, 6.683)<.001 Arthrocentesis30/534 (5.6%)38/534 (7.1%)0.015 (−0.014, 0.044)1.287 (0.785, 2.110).316 Complications45/534 (8.4%)43/534 (8.1%)−0.004 (−0.037, 0.029)0.952 (0.615, 1.472).824Elevated CRP/normal ESR (CRP 5.1-15 mg/L, ESR 0-25 mm/h) |n = 1126 vs 1126 MUA33/1126 (2.9%)31/1126 (2.8%)−0.002 (−0.016, 0.012)0.938 (0.570, 1.541).800 PJI45/1126 (4.0%)86/1126 (7.6%)0.036 (0.017, 0.056)1.988 (1.372, 2.874)<.001 Arthrocentesis63/1126 (5.6%)69/1126 (6.1%)0.005 (−0.014, 0.025)1.101 (0.775, 1.565).590 Complications78/1126 (6.9%)71/1126 (6.3%)−0.006 (−0.027, 0.014)0.904 (0.648, 1.261).553Elevated ESR/normal CRP (ESR 25.1-40 mm/h, CRP 0-5 mg/L) |n = 592 vs 592 MUA————— PJI28/592 (4.7%)38/592 (6.4%)0.017 (−0.009, 0.043)1.381 (0.836, 2.282).205 Arthrocentesis30/592 (5.1%)31/592 (5.2%)0.002 (−0.023, 0.027)1.035 (0.619, 1.733).895 Complications38/592 (6.4%)47/592 (7.9%)0.015 (−0.045, 0.014)1.26 (0.807-1.96).311Bold P values indicate statistical significance (P < .05). Normal markers cohort: CRP 0-5 mg/L and ESR 0-25 mm/h. Odds ratios are expressed as elevated markers relative to normal markers (OR > 1 indicates increased risk with elevated markers). — Indicates insufficient sample size for analysis. Ninety-day postoperative outcomes after primary TKA in patients with mildly elevated preoperative inflammatory markers vs normal markers. Bold P values indicate statistical significance (P < .05). Normal markers cohort: CRP 0-5 mg/L and ESR 0-25 mm/h. Odds ratios are expressed as elevated markers relative to normal markers (OR > 1 indicates increased risk with elevated markers). — Indicates insufficient sample size for analysis. One-year outcomes At 1 year, elevated preoperative CRP with concomitant elevated ESR remained significantly associated with increased odds of PJI (14.6% vs 4.1%; OR: 3.98; 95% CI: 2.44-6.50; P < .001) (Table 5). Elevated CRP with normal ESR was associated with an increased odds of revision arthroplasty (3.7% vs 2.3%; OR: 1.64; 95% CI: 1.00-2.70; P = .049) and PJI (9.6% vs 5.6%; OR: 1.79; 95% CI: 1.30-2.47; P < .001) compared to patients with normal markers (Table 5). In the elevated ESR group, no outcome reached statistical significance at 1 year, although PJI again was similarly higher in the elevated ESR cohort (7.3% vs 5.4%; OR: 1.37; 95% CI: 0.86-2.20; P = .19) (Table 5). Table 5One-year postoperative outcomes after primary TKA in patients with mildly elevated preoperative inflammatory markers vs normal markers.OutcomeNormal n (%)Elevated n (%)Risk difference (95% CI)Odds ratio (95% CI)PElevated CRP and elevated ESR (CRP 5.1-15 mg/L, ESR 25.1-40 mm/h)|n = 534 vs 534 Revision————— MUA16/534 (3.0%)13/534 (2.4%)−0.006 (−0.025, 0.014)0.808 (0.385, 1.696).572 PJI22/534 (4.1%)78/534 (14.6%)0.105 (0.070, 0.139)3.981 (2.440, 6.496)<.001 Arthrocentesis69/534 (12.9%)86/534 (16.1%)0.032 (−0.010, 0.074)1.294 (0.919, 1.822).140 Complications71/534 (13.3%)77/534 (14.4%)0.011 (−0.030, 0.053)1.099 (0.776, 1.555).595Elevated CRP/normal ESR (CRP 5.1-15 mg/L, ESR 0-25 mm/h) |n = 1126 vs 1126 Revision26/1126 (2.3%)42/1126 (3.7%)0.014 (0.000, 0.028)1.640 (0.998, 2.696).049 MUA48/1126 (4.3%)40/1126 (3.6%)−0.007 (−0.023, 0.009)0.827 (0.539, 1.269).384 PJI63/1126 (5.6%)108/1126 (9.6%)0.040 (0.018, 0.062)1.788 (1.297, 2.466)<.001 Arthrocentesis147/1126 (13.1%)172/1126 (15.3%)0.022 (−0.007, 0.051)1.201 (0.946, 1.524).131 Complications132/1126 (11.7%)132/1126 (11.7%)0 (−0.027, 0.027)1.000 (0.773, 1.292)1.000Elevated ESR/normal CRP (ESR 25.1-40 mm/h, CRP 0-5 mg/L) |n = 592 vs 592 Revision13/592 (2.2%)18/592 (3.0%)0.008 (−0.010, 0.027)1.397 (0.677, 2.879).363 MUA17/592 (2.9%)14/592 (2.4%)−0.005 (−0.023, 0.013)0.819 (0.400, 1.677).585 PJI32/592 (5.4%)43/592 (7.3%)0.019 (−0.009, 0.046)1.370 (0.855, 2.195).189 Arthrocentesis70/592 (11.8%)77/592 (13.0%)0.012 (−0.026, 0.049)1.115 (0.789, 1.575).537 Complications64/592 (10.8%)78/592 (13.2%)0.024 (−0.014, 0.061)1.260 (0.890, 1.785).190Bold P values indicate statistical significance (P < .05). Normal markers cohort: CRP 0-5 mg/L and ESR 0-25 mm/h. Odds ratios are expressed as elevated markers relative to normal markers (OR > 1 indicates increased risk with elevated markers). — Indicates insufficient sample size for analysis. One-year postoperative outcomes after primary TKA in patients with mildly elevated preoperative inflammatory markers vs normal markers. Bold P values indicate statistical significance (P < .05). Normal markers cohort: CRP 0-5 mg/L and ESR 0-25 mm/h. Odds ratios are expressed as elevated markers relative to normal markers (OR > 1 indicates increased risk with elevated markers). — Indicates insufficient sample size for analysis. Three-year outcomes Concomitantly elevated preoperative CRP and ESR was associated with a significantly increased odds of revision arthroplasty (6.6% vs 2.1%; OR: 3.33; 95% CI: 1.68-6.64; P < .001), PJI (16.3% vs 4.9%; OR: 3.80; 95% CI: 2.41-6.00; P < .001), and composite complications (24.7% vs 19.7%; OR: 1.34; 95% CI: 1.00-1.79; P = .047) (Table 6). At 3 years, elevated preoperative CRP continued to be significantly associated with PJI (11.2% vs 6.4%; OR: 1.85; 95% CI: 1.37-2.49; P < .001) (Table 6). In the elevated ESR group, there was a significantly higher odds of arthrocentesis (22.8% vs 17.7%; OR: 1.37; 95% CI: 1.03-1.82; P = .030) and composite complications (22.1% vs 16.7%; OR: 1.41; 95% CI: 1.06-1.89; P = .019) (Table 6). Table 6Three-year postoperative outcomes after primary TKA in patients with mildly elevated preoperative inflammatory markers vs normal markers.OutcomeNormal n (%)Elevated n (%)Risk difference (95% CI)Odds ratio (95% CI)PElevated CRP and elevated ESR (CRP 5.1-15 mg/L, ESR 25.1-40 mm/h) |n = 534 vs 534 Revision11/534 (2.1%)35/534 (6.6%)0.045 (0.021, 0.069)3.335 (1.675, 6.639)<.001 MUA17/534 (3.2%)13/534 (2.4%)−0.007 (−0.027, 0.012)0.759 (0.365, 1.578).459 PJI26/534 (4.9%)87/534 (16.3%)0.114 (0.078, 0.150)3.803 (2.410, 6.001)<.001 Arthrocentesis118/534 (22.1%)123/534 (23.0%)0.009 (−0.041, 0.060)1.055 (0.792, 1.406).714 Complications105/534 (19.7%)132/534 (24.7%)0.051 (0.001, 0.100)1.342 (1.004, 1.793).047Elevated CRP/normal ESR (CRP 5.1-15 mg/L, ESR 0-25 mm/h) |n = 1126 vs 1126 Revision46/1126 (4.1%)64/1126 (5.7%)0.016 (−0.002, 0.034)1.414 (0.960, 2.087).078 MUA49/1126 (4.4%)43/1126 (3.8%)−0.005 (−0.022, 0.011)0.872 (0.575, 1.326).523 PJI72/1126 (6.4%)126/1126 (11.2%)0.048 (0.025, 0.071)1.846 (1.366, 2.494)<.001 Arthrocentesis229/1126 (20.3%)265/1126 (23.5%)0.032 (−0.002, 0.066)1.207 (0.987, 1.476).067 Complications202/1126 (17.9%)225/1126 (20.0%)0.020 (−0.012, 0.053)1.143 (0.926, 1.410).216Elevated ESR/normal CRP (ESR 25.1-40 mm/h, CRP 0-5 mg/L) |n = 592 vs 592 Revision21/592 (3.5%)26/592 (4.4%)0.008 (−0.014, 0.031)1.249 (0.694, 2.247).457 MUA17/592 (2.9%)16/592 (2.7%)−0.002 (−0.020, 0.017)0.940 (0.470, 1.882).860 PJI34/592 (5.7%)48/592 (8.1%)0.024 (−0.005, 0.053)1.447 (0.919, 2.281).109 Arthrocentesis105/592 (17.7%)135/592 (22.8%)0.051 (0.005, 0.096)1.370 (1.030, 1.823).030 Complications99/592 (16.7%)131/592 (22.1%)0.054 (0.009, 0.099)1.414 (1.058, 1.890).019Bold P values indicate statistical significance (P < .05). Normal markers cohort: CRP 0–5 mg/L and ESR 0–25 mm/h. Odds ratios are expressed as elevated markers relative to normal markers (OR > 1 indicates increased risk with elevated markers). — Indicates insufficient sample size for analysis. Three-year postoperative outcomes after primary TKA in patients with mildly elevated preoperative inflammatory markers vs normal markers. Bold P values indicate statistical significance (P < .05). Normal markers cohort: CRP 0–5 mg/L and ESR 0–25 mm/h. Odds ratios are expressed as elevated markers relative to normal markers (OR > 1 indicates increased risk with elevated markers). — Indicates insufficient sample size for analysis. Discussion This study examined whether mild, preoperative elevations in CRP or ESR are associated with adverse outcomes following primary TKA. Elevations in both CRP and ESR were associated with increased odds of PJI across all time points and increased odds of revision arthroplasty and composite complications at 3 years postoperatively. A mildly elevated preoperative CRP in the setting of a normal ESR was consistently and significantly associated with increased odds of PJI across all time points. By contrast, a mildly elevated preoperative ESR with a normal CRP was only associated with adverse outcomes at 3 years postoperatively. These findings suggest that CRP and ESR may carry different prognostic weight in the preoperative assessment of TKA candidates and that CRP and ESR elevation in tandem may warrant greater clinical attention than isolated mild CRP of ESR. The strength of the association between preoperative CRP elevation and PJI in our study is notable and aligns with prior work by Xu et al., who demonstrated that preoperatively elevated ESR and CRP in combination increased PJI risk following TKA in osteoarthritis patients [13]. Their study characterized elevated markers as any value >30 mm/h (ESR) or >10 mg/L (CRP), they then compared patients with isolated elevations and patients with both markers elevated [13]. They found that patients with both markers elevated had significantly higher odds of PJI compared to patients with normal values, which is in alignment with our findings. However, they found that patients with an isolated increase in either marker did not show significantly higher odds of PJI compared to patients with normal inflammatory markers. Our study did not corroborate this finding, which may be due to the large discrepancy in sample size. In the Xu et al. study, there were 108 patients with an isolated inflammatory marker elevation and 32 patients with both markers elevated; compared to 534 matched pairs in the ESR and CRP group, 1126 matched pairs in the mildly elevated CRP group, and 592 matched pairs in the mildly elevated ESR group. Additionally, we intentionally capped our cohort’s lab values as to not include any patient with markedly elevated values that would likely be excluded from surgery. The Xu et al. study included all patients with values above their threshold which may skew the data based on outliers. However, our study extends this finding by demonstrating that CRP elevation alone, even at mild levels and in the absence of concurrent ESR elevation, is associated with an elevated odds of PJI. This is consistent with Erdogan et al., who found that asymptomatic preoperative inflammatory marker elevations were associated with increased infection and revision rates following primary TKA, though their analysis did not specifically examine discordant CRP and ESR values [17]. Godoy et al. examined preoperative CRP as a predictor of postoperative complications in primary TKA but did not find a statistically significant association, likely due to their smaller sample size of 351 patients and a higher CRP threshold of 8.2 mg/L [12]. It is worth noting that our rates of PJI were significantly higher than rates that are typically reported in the literature. This may be due to the fact the patients who got inflammatory markers tested in the perioperative period represent a higher risk subgroup than the general elective TKA patient. Another explanation is our reliance on International Classification of Diseases, 10th Revision (ICD-10) coding to define PJI may over include patients who do not meet criteria for true PJI as defined by validated criteria. Our PJI findings should be interpreted cautiously and future studies investigated inflammatory markers in TKA that rigorously define PJI are warranted. We also demonstrated a dose-response relationship regarding inflammatory markers and PJI risk. In our study, patients with both markers elevated had an OR of 3.90 for PJI at 90 days, compared to an OR of 1.99 for PJI at 90 days in the isolated mildly elevated CRP group. Additionally, we found significantly increased odds of revision arthroplasty in the cohort with both markers mildly elevated at 3 years postoperatively which was not seen in the other cohorts at 3 years postoperatively. These findings indicate that assessment of both values should be considered by surgeons and patients with increases in both may warrant increased surveillance. Increased perioperative inflammatory markers have been shown to correlate with increased PJI and complication risk in a variety of orthopaedic operations [[18], [19], [20], [21]]. The current literature often defines elevations above a certain threshold and without a ceiling, making outliers a potential issue. However, there is a sparsity of evidence on if mild elevations in inflammatory markers correlate with increased risk, which our study aimed to investigate. CRP and ESR have been shown to rise and fall transiently in response to a number of events including systemic inflammation, obesity, injury, and anemia [[22], [23], [24], [25]]. The aim of this study was to demonstrate an association between inflammatory markers and odds of adverse events following TKA, not to establish a causal relationship. As CRP and ESR are downstream markers of underlying inflammatory and infectious process, our findings should be interpreted as identifying higher-risk patient phenotypes rather than establishing the markers themselves as causal mediators of poor outcomes. This principle is consistent with how preoperative screening tools are commonly used in clinical practice. Given the increased awareness surrounding preoperative screening in TKA, this study adds to the understanding of important and actionable risk factors in TKA [26,27]. Future data are needed to assess whether preoperative screening with CRP and ESR is necessary in all patients undergoing TKA. The broader literature currently points to increase PJI risk in patients with elevated inflammatory markers. However, isolated rises have not been fully studied and “borderline” elevations are not well understood. If patients do have elevated markers in the preoperative period, serial assessment of markers and potential delay of surgery may be indicated if future research demonstrates clear risk with mild elevations. Based on the results of this study, an isolated, mildly elevated CRP should prompt investigation for an identifiable cause rather than reflexive cancellation of surgery. Patients with concordantly elevated markers may warrant surgical delay until a source is identified. This study has several limitations. First, as a retrospective database study, it is subject to the inherent limitations of administrative data, including the accuracy of ICD-10 and Current Procedural Terminology coding across institutions. The TriNetX platform relies on deidentified aggregate data, which precludes individual chart review to confirm diagnoses such as PJI. Second, the CRP and ESR thresholds used to define mild elevations were selected based our institution’s upper value of normal of 5 mg/L for CRP and 25 mm/h for ESR, with the upper bound of mild elevation 15 mg/L for CRP and 40 mm/h for ESR selected as a reasonable threshold above which patients would typically be excluded from surgery. However, there are no universally accepted cutoffs for these values in the preoperative arthroplasty setting [12]. Third, while propensity score matching accounts for measured confounders, unmeasured confounders such as surgical technique, implant type, perioperative antibiotic protocols, and surgeon volume could not be controlled. Fourth, the elevated ESR cohort and the elevated ESR and CRP cohort were smaller which may have limited our ability to detect significant differences in this group. Fifth, the TriNetX network predominantly includes large, academic healthcare organizations, which may limit generalizability to community practice settings [28]. Finally, we utilized ICD-10 coding to define PJI which is less robust than validated criteria in defining PJI [29,30]; however, prior validation of administrative PJI coding in similar networks has been reported previously in the literature [[31], [32], [33]]. Conclusions In this propensity score-matched analysis using the TriNetX federated research network, concordantly elevated preoperative CRP and ESR were significantly and consistently associated with increased odds of PJI following primary TKA at 90 days, 1 year, and 3 years. Higher odds of PJI were seen in the cohort with both markers elevated, rather than an isolated increase in either CRP or ESR. Future prospective studies are warranted to validate these findings and to establish evidence-based preoperative CRP thresholds for TKA risk stratification. These findings are hypothesis-generating and require future prospective validation before informing clinical protocols. Surgeons should be aware of preoperative laboratory values, but further validation is needed prior to recommending routine use. Ethical approval This study was IRB approved and received a determination of Not Human Subjects Research University of Rochester Office for Human Subject Protection (Study ID: STUDY000111359). Data availability The data of this study are available in the article itself. Data may be requested from the authors. Conflicts of interest The authors declare there are no conflicts of interest. For full disclosure statements refer to https://doi.org/10.1016/j.artd.2026.102072. CRediT authorship contribution statement Josef W. Kubofcik: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Conceptualization. Stefan W. Fleps: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Conceptualization. Christopher J. Drinkwater: Writing – review & editing, Writing – original draft, Supervision, Resources, Methodology, Conceptualization. Supplementary data Conflict of Interest Statement for Drinkwater Conflict of Interest Statement for Kubofcik Conflict of Interest Statement for Fleps Supplementary Data Supplementary Table 1Codes used for propensity matching cohorts.Index event ProcedureCPT: 27447Arthroplasty, knee, condyle and plateau; medial and lateral compartments with or without patella resurfacing (total knee arthroplasty)Demographics AgeAge at IndexAge at index event SexFFemaleMMale Race2106-3White2054-5Black or African American2028-9Asian1002-5American Indian or Alaska Native2076-8Native Hawaiian or other Pacific Islander2131-1Other raceUNKUnknown race Ethnicity2186-5Not Hispanic or Latino2135-2Hispanic or LatinoUNUnknown ethnicity Comorbidities BMIICD-10-CM: Z68Body mass index (BMI) Diabetes mellitusICD-10-CM: E08–E13Diabetes mellitus Hypertensive diseasesICD-10-CM: I10–I1AHypertensive diseasesICD-10-CM: I10–I55∗Hypertensive diseases (deprecated 2018)∗ Atrial fibrillation and flutterICD-10-CM: I48Atrial fibrillation and flutter Heart failureICD-10-CM: I50Heart failure Kidney diseaseICD-10-CM: N17–N19Acute kidney failure and chronic kidney disease COPDICD-10-CM: J44Chronic obstructive pulmonary disease Rheumatoid arthritisICD-10-CM: M05ICD-10-CM: M06Rheumatoid arthritis with rheumatoid factorOther rheumatoid arthritis Ankylosing spondylitisICD-10-CM: M45Ankylosing spondylitis SpondylopathiesICD-10-CM: M46Other inflammatory spondylopathies HIVICD-10-CM: B20Human immunodeficiency virus Chronic steroid useICD-10-CM: Z79.52Long-term use of systemic steroids MalignancyICD-10-CM: Z85Personal history of malignant neoplasm Nicotine dependenceICD-10-CM: F17Nicotine dependence Alcohol disordersICD-10-CM: F10Alcohol-related disordersCOPD, chronic obstructive pulmonary disease; ICD-10-CM, International Classification of Diseases, 10th Revision. Codes used for propensity matching cohorts. COPD, chronic obstructive pulmonary disease; ICD-10-CM, International Classification of Diseases, 10th Revision. Supplementary Table 2Codes used to define study outcomes.OutcomeCodeDescriptionProcedural outcomes Revision arthroplastyCPT: 27486Revision of total knee arthroplasty, with or without allograft; 1 componentCPT: 27487Revision of total knee arthroplasty, with or without allograft; femoral and entire tibial component Manipulation under anesthesia (MUA)CPT: 27570Manipulation of knee joint under general anesthesia (includes application of traction or other fixation devices) ArthrocentesisCPT: 20610Arthrocentesis, aspiration and/or injection, major joint or bursa; without ultrasound guidanceCPT: 20611Arthrocentesis, aspiration and/or injection, major joint or bursa; with ultrasound guidance, with permanent recording and reportingInfection Periprosthetic joint infection (PJI)ICD-10-CM: T84.50Infection and inflammatory reaction due to unspecified internal joint prosthesisICD-10-CM: T84.53Infection and inflammatory reaction due to internal right knee prosthesisICD-10-CM: T84.54Infection and inflammatory reaction due to internal left knee prosthesisICD-10-CM: T84.59Infection and inflammatory reaction due to other internal joint prosthesisComplications Composite complicationsICD-10-CM: I26Pulmonary embolismICD-10-CM: I21Acute myocardial infarctionICD-10-CM: I82.4Acute embolism and thrombosis of deep veins of lower extremity (DVT)ICD-10-CM: A40Streptococcal sepsisICD-10-CM: A41Other sepsisICD-10-CM: J18Pneumonia, unspecified organismICD-10-CM: I63Cerebral infarction (stroke)ICD-10-CM: N17Acute kidney failureDVT, deep vein thrombosis. Codes used to define study outcomes. DVT, deep vein thrombosis. References Associated Data Supplementary Materials Conflict of Interest Statement for Drinkwater Conflict of Interest Statement for Kubofcik Conflict of Interest Statement for Fleps Data Availability Statement The data of this study are available in the article itself. Data may be requested from the authors.