Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors
Search in posts
Search in pages
Filter by Categories
Case Report
Clinical research study
Current Issue
Editorial Board
Literature Review
Narrative review
Original Article
Research Article
Review Article
Short Report
Surgical techniques
Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors
Search in posts
Search in pages
Filter by Categories
Case Report
Clinical research study
Current Issue
Editorial Board
Literature Review
Narrative review
Original Article
Research Article
Review Article
Short Report
Surgical techniques
View/Download PDF

Translate this page into:

73 (); 128-136
doi:
10.1016/j.jor.2025.12.002

Serological risk factors associated with arthroplasty complications

, , , , , , , ,
a Department of Orthopaedic Surgery, College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
b College of Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

⁎Corresponding author: Johannes van der Merwe. jov777@mail.usask.ca

Received: , Accepted: ,
The Authors
Disclaimer:
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.

Abstract

Abstract

Post-operative complications can have a devastating impact on patients' quality of life. Peri-operative bloodwork is often routinely conducted and may predict post-operative complications. The purpose of this study is to understand if pre- or post-operative serology is associated with post-operative complications arising from hip and knee arthroplasty.

A retrospective chart review of 383 participants with post-operative total hip and knee arthroplasty complications was conducted. One hundred and forty-one participants with post-operative complications were included and compared to a control group of 120 participants. Patient, surgical, and pre- and post-operative serological data were collected. L2-regularized logistic regression was utilized to assess whether identified factors independently predicted post-operative complications.

The complications group had higher age, male sex, presence of osteoporosis, Charlson Comorbidity Index and bilateral arthroplasties compared to the control group (p < 0.05). Pre-operatively, the complications group had a lower hemoglobin value (p = 0.027; OR = 0.64) and higher Basophil Count (p = 0.012, OR = 2.23) and Monocyte–Lymphocyte ratio (p = 0.003, OR = 2.23), higher rates of undergoing a general anesthetic (p = 0.019, OR = 0.43) and psychiatric illness (p = 0.001, OR = 3.07). Post-operatively, the complications group had lower WBC (p = 0.040, OR = 0.55) and Neutrophil counts (p = 0.001, OR = 0.45), increased Eosinophil count (p = 0.018, OR = 2.36), increased age (p = 0.034, OR = 2.06) and BMI (p = 0.049, OR = 1.74), having postoperative anticoagulation other than ASA (p = 0.000, OR = 3.99).

We identified multiple peri-operative patient, surgical, and serological risk factors for developing post-operative complications following hip and knee arthroplasty. These markers could be prioritized for monitoring high-risk patients for possible intervention.

Keywords

Complications
Preoperative blood work
Postoperative blood work
Serological risk factors
1

1 Introduction

Total knee arthroplasty (TKA) and total hip arthroplasty (THA) are commonly performed procedures for osteoarthritis, with their frequency continuing to rise worldwide.1,2 Arthroplasty-specific complications include peri-prosthetic joint infections (PJI), aseptic loosening, and dislocation, in addition to broader post-operative complications which necessitate revision surgery, increase patient morbidity, mortality, and hospital stay, placing an increased economic burden on the healthcare system.1,3–6 The literature suggests that the rate of THA and TKA complications were 2.90 and 2.42 %, respectively, and can increase healthcare costs by up to threefold compared to cases without complications.6,7

PJIs are a particularly significant complication to joint replacements, occurring in 1–2 % of arthroplasties and remain the leading cause of revision arthroplasties.1,8 The etiology of PJIs is variable and can be caused by a diverse group of bacteria and fungi. Risk factors such as previous joint surgery, current infection, bacteremia, seasonal variation, and medical comorbidities can predispose patients to developing PJI.9,10 While the incidence of PJIs increases, they remain difficult to predict in clinical practice.1 Patients undergoing arthroplasty procedures are also at risk of medical complications such as myocardial infarction, pneumonia, and venous thromboembolism (VTE).6 These complications and the resultant hospital admissions increase the overall morbidity, mortality, and economic burden of arthroplasty procedures.6,11,12,13

Studies have suggested that peri-operative bloodwork may be useful in predicting post-operative complications, including creatinine, albumin, WBC, CRP, and hemoglobin.14,15 Additionally, the combination of serological tests has been suggested to have higher predictive value than single tests for PJIs.16 A recent meta-analysis has suggested that the combined use of CRP, WBC, neutrophils, and stool calprotectin provided the highest diagnostic accuracy for detecting post-operative complications in colorectal surgery.17 While there are studies investigating serological risk factors for detecting PJIs post THA and TKAs, there is limited literature assessing serological risk factors for predicting broader post-operative complications from THA and TKAs. Effective anticipation of arthroplasty complications may improve diagnosis and management. Furthermore, a shift towards more replacements being performed as day surgeries means identification of high-risk patients seems prudent.1 Investigating perioperative serological risk factors and their link to postoperative complications may offer valuable insights for improving patient outcomes.

The primary objective of this study was to evaluate the predictive value of individual and combined perioperative serology on postoperative THA and TKA associated complications. The secondary objective is to investigate other patient and surgical risk factors that are associated with postoperative complications.

2

2 Methods

Following Research Ethics Board approval (BIO-4954, July 5, 2024), we conducted a multicenter retrospective chart review evaluating patient serological risk factors that are associated with PJIs and other post-operative medical complications following THA and TKAs.

Three hundred and eighty-nine patient charts from April 2012 to March 2024 with post-operative complications were identified through diagnostic codes. Inclusion and exclusion criteria were applied resulting in a study group of 141 patients which were then compared against a control group of 120 patients and analyzed. The study group was subcategorized into a PJI complication group of 37 patients, and other post-operative complication group of 104 patients (See Fig. 1).

Participant flowchart on screening and exclusion process.
Fig. 1 Participant flowchart on screening and exclusion process.

Charts were abstracted for post-operative complications, peri-operative serology, and patient (age, sex, BMI, comorbidities) and surgical (arthroplasty type, laterality, post-op anticoagulation, anesthetic type) characteristics. Preoperative serology was defined as serology collected within 30 days before the date of operation. Collected preoperative serology data included White Blood Cell Count (WBC); Red Blood Cells (RBC); Hemoglobin (HB); Hematocrit; Mean Corpuscular Volume (MCV); Mean Corpuscular Hemoglobin (MCH); Mean Corpuscular Hemoglobulin Concentration (MCHC); Platelets; Mean Platelet Volume (MPV); Neutrophil-, Lymphocyte-, Monocyte-, Eosinophil-, and Basophil count; Monocyte: Lymphocyte Ratio (MLR); and Neutrophil: Lymphocyte Ratio (NLR); Platelet: Mean Platelet Volume ratio; and Platelet: Lymphocyte ratio. Postoperative serology was defined as serology collected within 30 days after the date of operation and included the same metrics as preoperative serology.

2.1

2.1 PJI group

Acute PJIs were defined with the Musculoskeletal Infection Society (MSIS) and European Bone and Joint Infection Society (EBJIS) criteria.18,19 PJI cases were only included if both the MSIS and EBJIS criteria confirmed infection. Two hundred patients who underwent THA or TKA and then went on to develop a PJI were identified through our Discharge Abstract Database using procedure codes for potential inclusion. All patients identified through the database did undergo either a debridement, antibiotic and implant retention (DAIR) procedure or a one-or two stage revision. Thirty-seven participants with PJI after a THA or TKA were included in the final analysis. One hundred and sixty-three participants were excluded. One hundred and thirty-four participants were excluded due to non-acute infection (>90 days after primary arthroplasty). Twenty-three participants were excluded due to missing pre-operative or post-operative serology. Five participants were excluded due to the PJI involving a unicompartmental knee replacement. One case was excluded due to not having a complete chart in the online medical record software.

2.2

2.2 Medical complications group

Participants were identified through our Discharge Abstract Database using procedure codes for medical complications post-operatively for potential inclusion (see Table 1). One hundred and eighty-nine participants who underwent a THA or TKA who then developed medical complications postoperatively were identified. One hundred and four participants were included. Eighty-five participants were excluded: 46 were excluded due to nonrelevant diagnosis code, 28 were excluded for participants undergoing a non-primary arthroplasty, 8 were excluded for missing serology, and 3 were excluded for other missing critical chart information.

Table 1 Descriptive analysis. PJI: Periprosthetic Joint Infections, BMI: Body Mass Index; eGFR: estimated glomerular filtration rate; HBA1C: Glycated hemoglobulin.
Variable Study Group (Combined PJI and Comorbidities) Control Group (Combined TKA and THA) P-value
Age (mean, Minimum – maximum; years) 72.31 (43–94) 68.20 (45–94) 0.0034
Sex (female percentage) 53.6 61.2 0.2591
BMI (mean, Minimum – Maximum, kg/L2) 35.49 (18.27–64.01) 33.11 (18.11–56.20) 0.1003
Laterality (left percentage) 58.8 50.8 0.2118
Hypertension (percentage) 28.1 22.1 0.7911
Severe Renal Impairment, eGFR <30 (percentage) 7.9 4.1 0.3011
Diabetes Melitis or HBA1C > 6.5 (percentage) 28.1 22.1 0.3183
Hyperlipidaemia (percentage) 43.9 55.7 0.0633
No pulmonary disease (percentage) 70 75.2 n/a
No Cardiac disease (percentage) 75.4 72.9 n/a
No liver disease (percentage) 94.8 94.3 n/a
Alcohol/drug usage (%) 7.9 2.5 0.058
Smoking (%) 33.1 23.8 0.1022
No Psychiatric illness (%) 75.4 88.5 n/a
No Inflammatory conditions (%) 95.7 97.5 n/a
Chronic steroid usage (> 3 months) (%) 4.3 0.8 0.1255
Osteoporosis (%) 15.2 3.3 0.0012
Neuromuscular Disorders (%) 3.6 2.5 0.7271
Connective tissue disease (%) 0.7 0.8 1.0
2.3

2.3 Control group

Three-hundred and two participants who underwent total hip and knee arthroplasty without perioperative joint infection or post-operative medical complications were screened. One hundred and eighty-two were excluded leaving a total of 120 participants with 60 undergoing TKA and 60 undergoing THA. One hundred and fifty-nine participants were excluded due to incomplete serology; 22 were excluded due to missing documentation; 1 was excluded for undergoing a non-primary arthroplasty.

Data collection occurred from August 2024 until December 2024. Chart abstractors were blinded to each other's findings. Data was not analyzed until all data was collected.

2.4

2.4 Statistical analysis

The study used L2-regularized logistic regression to assess whether pre- and post-operative lab values predict post-surgical complications. Comparisons were made between patients without complications and those in complication groups Covariates such as age, sex, BMI, and comorbidities were adjusted for, and statistical significance was evaluated using z-scores and bootstrapped confidence intervals. Descriptive statistics, including means, standard deviations, and frequency counts, were calculated. Group comparisons were conducted using the Mann-Whitney U test for continuous variables and Fisher's Exact Test for categorical data. Power was estimated using Monte Carlo simulations and t-test approximations. Predictive analysis involved ROC curve evaluation and AUC calculations for bloodwork variables. The optimal diagnostic thresholds were determined using the Youden Index, and Sensitivity, Specificity, PPV, and NPV were reported to assess diagnostic accuracy in predicting post-surgical complications. We also used composite predictor analysis using linear regression. In this analysis, we aimed to create composite predictors by linearly combining the top 4 pre-operative and post-operative features for each comparison group. Linear regression was used to generate a composite risk score. This score was then evaluated using ROC analysis to determine its diagnostic performance.

3

3 Results

We included a total of 261 patients in the study. The study group consisted of 141 patients and the control group consisted of 120 patients (See Fig. 1). The study group was further divided into 37 patients with acute postoperative PJI's and 104 patients who developed medical complications postoperatively. The mean age was 70 years (43–94 years of age) and 53.6 % females in the study group compared to 61 % in the control group (p = 0.25). The mean BMI was 34.29 kg/L2 (35.49 in the study group, 33.11 in the control group, p = 0.10). The study group consisted of 22.1 % of THA's, 29.3 % hip hemiarthroplasty, and 47.9 % total knee arthroplasty. The majority of the patients underwent a unilateral joint arthroplasty (97.1 % in the study group compared to 100 % in the control group) and 58.8 % were left sided in the study group compared to 50.8 % in the control group. The majority of patients received a spinal anesthetic prior to surgery (study group 65.4 %; control group 96.6 %). Postoperative anticoagulation was Aspirin in 78.7 % in the control group while a selective direct factor Xa inhibitor was used most commonly in the study group (21.5 %) (Please seeTable 1). We found a statistical significance between the study and control group's demographics in regards to age, Charlson Comorbidity Index (CCI) and presence of osteoporosis (See Table 1). All the other variables did not reach statistical significance.

3.1

3.1 Subgroup analysis: Demographics

Looking at the PJI sub-group in comparison with the control group the only statistical significance was determined with BMI (Mean BMI PJI subgroup: 37.85, Mean BMI control group: 33.11; p = 0.0068). Comparing the medical complications group (MCG) and the control group we found a statistical significance in age (MCG = 73; Control group = 68; p = 0.0002), presence of osteoporosis (MCG = 18.6 %; Control group = 3.3 %; p = 0.0002), CCI (MCG mean = 4.54, control group mean = 3.27; p = 0.00) and bilateral arthroplasties (MCG = 3.9 %; control group = 0 %; p = 0.0416). Comparing the medical complications group to the PJI only group there was a statistical significance found with age (MCG = 73; PJI = 68, p = 0.0137), BMI (MCG = 34.12, PJI = 37.85, 0.0378), CCI (MCG = 4.54, PJI = 3.64, p = 0.0046).

3.2

3.2 Pre - and - postoperative laboratory values

Group comparison analysis comparing the preoperative laboratory values between the study and control group found a statistically significance in relation to hemoglobin (HB); Red Blood Cells (RBC); Hematocrit; Mean Corpuscular Hemoglobin Concentration (MCHC); White Blood Cell Count (WBC); Lymphocyte -, Monocyte - and Neutrophil count; Monocyte-Lymphocyte Ratio (MLR); and Neutrophil-Lymphocyte Ratio (NLR) (See Table 2). Postoperative laboratory values between the control and study group reached statistical significance with MCHC, WBC, Neutrophil -, Eosinophil - and Basophil counts, platelets and NLR (See Table 3). Subgroup analysis was performed and Table 5 summarizes the statistical significant findings in pre - and - postoperative laboratory values between the MCG and PJI group compared to the control group.

Table 2 Group Comparison of Preoperative Laboratory results between study group (combined PJI and medical complications) and control group.
Variable Test Statistic P-Value Power
Age Mann-Whitney U 10331.0 0.0034 0.8727
Sex Fisher's Exact 1.3645 0.2591 0.246
BMI Mann-Whitney U 4724.5 0.1003 0.4203
Laterality Fisher's Exact 0.7233 0.2118 0.248
Bilateral Arthroplasty Fisher's Exact 0.0 0.1253 0.231
Hypertension Fisher's Exact 1.0934 0.7911 0.066
Hyperlipidaemia Fisher's Exact 1.6102 0.0633 0.486
Diabetes Melitis or HbA1c ≥ 6.5 % Fisher's Exact 0.7287 0.3183 0.172
Severe Renal Impairment; eGFR < 30 Fisher's Exact 0.4973 0.3011 0.188
Alcohol/Drug abuse Fisher's Exact 0.2934 0.058 0.477
Smoking Fisher's Exact 0.6304 0.1022 0.361
Chronic Steroid Use (> 3 months) Fisher's Exact 0.1832 0.1255 0.305
Osteoporosis Fisher's Exact 0.1889 0.0012 0.913
Neuromuscular Disorder Fisher's Exact 0.6756 0.7271 0.038
Connective Tissue Disease Fisher's Exact 1.1405 1.0 0.005
= Metal Allergy/Hypersensitivity Fisher's Exact 0.0 1.0 0.003
Charlson Comorbidity index Mann-Whitney U 10864.5 0.0001 0.9754
Hemoglobin Mann-Whitney U 6053.0 0.0001 0.9928
Red Blood Cells Mann-Whitney U 6785.5 0.009 0.9073
Hematocrit Mann-Whitney U 6535.5 0.0025 0.9528
Mean corpuscular volume (MCV) Mann-Whitney U 8267.0 0.8818 0.0819
Mean Corpuscular Hemoglobin Mann-Whitney U 7518.5 0.1637 0.1201
Mean corpuscular hemoglobin concentration (MCHC) Mann-Whitney U 5972.5 0.0001 0.9843
White Blood Cell Count Mann-Whitney U 10060.5 0.0047 0.7836
Lymphocyte count Mann-Whitney U 6553.0 0.0027 0.8219
Monocyte count Mann-Whitney U 9924.5 0.0092 0.628
Neutrophil count Mann-Whitney U 10226.0 0.0019 0.0957
Eosinophil count Mann-Whitney U 8164.0 0.7489 0.1462
Basophil count Mann-Whitney U 9157.5 0.1796 0.4192
Platelets Mann-Whitney U 7217.5 0.0584 0.1554
Mean platelet volume (MPV) Mann-Whitney U 7878.5 0.4267 0.1981
Monocyte - Lymphocyte ratio Mann-Whitney U 10651.0 0.0 0.9941
Neutrophil - Lymphocyte ratio Mann-Whitney U 10164.5 0.0004 0.0566
Platelet: Mean Platelet volume ratio Mann-Whitney U 7220.0 0.138 0.1336
Platelet - lymphocyte ratio Mann-Whitney U 8851.5 0.1965 0.4918
Table 3 Group Comparison of Postoperative Laboratory results between study group (combined PJI and medical complications) and control group.
Variable Test Statistic P-Value Power
Hemoglobin Mann-Whitney U 7288.0 0.0503 0.751
RBC Mann-Whitney U 7536.5 0.1216 0.496
Hematocrit Mann-Whitney U 7498.0 0.1071 0.5909
Mean corpuscular volume (MCV)0.1 Mann-Whitney U 8397.0 0.8934 0.1495
Mean Corpuscular hemoglobin Mann-Whitney U 7648.5 0.1724 0.2321
Mean corpuscular hemoglobin concentration (MCHC) Mann-Whitney U 6763.5 0.0048 0.0741
WBC Mann-Whitney U 5314.5 0.0 0.9986
Lymphocyte count Mann-Whitney U 7834.5 0.2899 0.1439
Monocyte count Mann-Whitney U 8393.0 0.8882 0.1072
Neutrophil count Mann-Whitney U 5142.5 0.0 0.9996
Eosinophil count Mann-Whitney U 12413.0 0.0 0.9982
Basophil count Mann-Whitney U 11129.5 0.0 0.0559
Platelets Mann-Whitney U 7274.0 0.0477 0.296
Mean platelet volume (MPV) Mann-Whitney U 7653.0 0.1746 0.4716
Monocyte - Lymphocyte ratio Mann-Whitney U 9376.0 0.0644 0.4398
Neutrophil - Lymphocyte ratio Mann-Whitney U 7008.5 0.0347 0.1914
Platelet - Mean Platelet volume ratio Mann-Whitney U 7625.0 0.2804 0.0804
Platelet - lymphocyte ratio Mann-Whitney U 8215.5 0.9273 0.131
3.3

3.3 Logistic regression analysis of preoperative laboratory values and demographics (Table 4)

Comparing the study group with the control group we found a statistical significance with the following variables. There was higher chance of complications with a lower hemoglobin value (optimal cut-off value (OCV) - 151; p = 0.027; OR = 0.64); higher Basophil Count (OCV - 0.07; p = 0.012, OR 2.23); higher Monocyte – Lymphocyte ratio (OCV - 0.34; p = 0.003; OR 2.23), being male (p = 0.012, OR 0.47), not using ASA for postoperative anticoagulation (p = 0.00, OR 3.99) undergoing a general anesthetic (p = 0.019, OR 0.43), or having a psychiatric illness (p = 0.001; OR = 3.07)(See Table 2). Statistically significant findings of the subgroup analysis involving the PJI only and medical complications only subset and comparing it with the control group can be found in Table 5.

Table 4 Logistic regression analysis of Preoperative bloodwork between study group and control group.
Variable Odds Ratio 95 % CI Lower 95 % CI Upper Z-Score P-Value Power VIF
Hemoglobin 0.6401 0.4428 0.9711 −2.2116 0.027 0.613 5.9057
Red Blood Cells 1.1628 0.6977 2.0038 0.5385 0.5902 0.081 111.5918
Hematocrit 0.704 0.467 1.1312 −1.5419 0.1231 0.357 103.6408
Mean corpuscular volume (MCV) 0.8628 0.4959 1.4665 −0.5403 0.589 0.088 23.8632
Mean Corpuscular Hemoglobin 0.9686 0.6707 1.344 −0.1912 0.8484 0.065 2.5701
MCHC 0.8532 0.4397 1.5481 −0.4952 0.6205 0.084 1.7701
WBC 0.9452 0.4844 2.0254 −0.1559 0.8761 0.052 3.489
Lymphocyte count 0.4868 0.2237 1.2882 −1.6163 0.106 0.377 6.5309
Monocyte count 1.2572 0.7292 2.1287 0.8292 0.407 0.133 18.5486
Neutrophil count 0.9944 0.7138 1.6211 −0.032 0.9745 0.075 1199.0428
Eosinophil count 0.9529 0.4645 2.0828 −0.1286 0.8976 0.054 2.3221
Basophil count 2.2252 1.164 4.2023 2.508 0.0121 0.713 1.9279
Platelets 0.7565 0.4084 1.3515 −0.9091 0.3633 0.156 8.3598
MPV 0.7942 0.4555 1.5835 −0.7913 0.4288 0.115 3.4913
Monocyte - Lymphocyte ratio 2.226 1.3276 3.8323 2.9845 0.0028 0.852 32.5655
Neutrophil - Lymphocyte ratio 1.0783 0.7361 1.9585 0.3295 0.7418 0.101 1210.0847
Platelet - MPV ratio 0.7885 0.4994 1.2915 −0.9924 0.321 0.153 2.1932
Platelet - lymphocyte ratio 0.6846 0.2756 1.4446 −0.8515 0.3945 0.152 13.8037
3.4

3.4 Logistic regression analysis of postoperative laboratory values

Comparing the combined study group (PJI and medical complications) with the control group we did find a statistically significant relationship between developing a complication and lower WBC count (OCV - 26.03; p = 0.040, OR 0.55), lower Neutrophil count (OCV - 3.34; p = 0.001, OR 0.45), increased Eosinophil count (OCV - 0.02; p = 0.018, OR 2.36), increasing age (p = 0.034, OR 2.06) and increased BMI (p = 0.049, OR 1.74), receiving a general anesthetic (p = 0.001, OR 0.34), not receiving ASA (p = 0.000, OR 4.49) and history of psychiatric illness (p = 0.006, OR 2.39). Table 5 highlights the statistically significant findings of the subgroup analysis comparing PJI only - and medical complications only group against the control group.

Table 5 Subgroup analysis only listing the statistical significance. PJI and medical complications only (MCG) were compared to the control group. n/s = cot significant.
Preop PJI Postop PJI Preop MCG Postop MCG
HB (lower) n/s n/s 0.026 (OR - 0.68) n/s
WBC count (lower) n/s 0.032 (OR 0.60) n/s 0.032 (OR - 0.59)
Neutrophil count (lower) n/s 0.034 (OR 0.62) n/s 0.002 (OR 0.49)
Eosinophil count (higher) n/s 0.023 (OR 2.26) n/s n/s
Basophil counts (high) n/s n/s 0.0007 (OR - 2.64) n/s
Lymphocyte count (low) 0.011 (OR - 0.45) n/s n/s n/s
MLR (high) 0.035 (OR - 1.87) 0.015 (OR 1.79) 0.0016 (OR - 2.24) n/s
Age (increased) n/s n/s 0.005 (OR 2.22) 0.019 (OR 2.14)
Sex (Male) 0.001 (OR 0.31) 0.039 (OR 0.49) n/s n/s
BMI (high) 0.004 (OR 2.61) 0.000 (OR 3.32) n/s n/s
Type of anesthesia (General anesthesia) n/s n/s 0.008 (OR 0.38) 0.001 (OR 0.32)
Postop anticoagulation (anticoagulation other than ASA) 0.000 (OR 3.82) 0.000 (OR 3.66) 0.000 (OR - 3.06) 0.000 (OR 4.04)
Hypertension (absence) n/s n/s 0.042 (OR 0.58) n/s
Liver disease (absence) n/s n/s 0.019 (OR 0.58) 0.039 (OR 0.64)
Psychiatric illness (present) n/s n/s 0.004 (OR 2.34) 0.009 (OR 2.25)
3.5

3.5 Composite predictor analysis using linear regression (Table 6)

By combining the top 4 predictors in the preoperative laboratory values we found that the combination of a lower value of MCHC (OCV 362.0),- HB (OCV 151),- Lymphocyte count (OCV 2.75)- and- Hematocrit (OCV 0.48) had an area under the curve (AUC) of 0.69 (Sensitivity 66.91 %; Specificity 67.23 %). Subgroup analysis comparing the MCG to the control group we found an AUC of 0.70 by combining Hemoglobin (OCV 151), MCHC (OCV 362), Lymphocyte count (OCV 2.75) and RBC count (OCV 5.37).

Table 6 Composite analysis evaluating pre - and - post operative.
Variables Group Pre - or - postoperative laboratory values AUC Sensitivity Specificity
MLR, NLR, Neutrophil count, WBC Combined (PJI and MCG) Pre - operative 0.65 0.70 0.55
MLR, NLR, Neutrophil count, WBC MCG Pre - operative 0.68 0.66 0.65
WBC, Monocyte count, Neutrophil count, Eosinophil Count PJI Pre - operative 0.63 0.86 0.42
Eosinophil, Basophil count, MLR, PLR Combined (MCG + PJI) Post - operative 0.76 0.76 0.73
Eosinophil count, Basophil count, MLR, PLR MCG Post - operative 0.74 0.76 0.63
Eosinophil count, Basophil count, Monocyte count, Lymphocyte count PJI Post - operative 0.84 0.89 0.70

Postoperative the top 4 predictors: Higher values of Basophil count (OCV 0.03),- MLR (OCV 0.83),- PLR (OCV 380) -and - Eosinophil count (OCV 0.02) had an AUC of 76.35 % (Sensitivity 76.09 %; Specificity 68.07 %) (See Fig. 2). Subgroup analysis comparing the MCG to the control group we found an AUC of 0.74 by combining variables Eosinophil count (OCV 0.02), Basophil Count (OCV 0.03), MLR (OCV 0.83), PLR (312) (See Fig. 3). Assessing the postoperative PJI only group to the control group we calculated an AUC of 0.84 combining variables Eosinophil count (OCV 0.02), Basophil count (OCV 0.03), Lymphocyte count (OCV 0.83) and Monocyte Count (OCV 0.98) (See Fig. 4).

ROC curve: Postoperative the top 4 predictors: Basophil count (OCV 0.03), MLR (OCV 0.83), PLR (OCV 380) and Eosinophil count (OCV 0.02) had an AUC of 76.35 % (Sensitivity 76.09 %; Specificity 68.07 %).
Fig. 2 ROC curve: Postoperative the top 4 predictors: Basophil count (OCV 0.03), MLR (OCV 0.83), PLR (OCV 380) and Eosinophil count (OCV 0.02) had an AUC of 76.35 % (Sensitivity 76.09 %; Specificity 68.07 %).
ROC curve: Subgroup analysis comparing the MCG to the control group we found an AUC of 0.74 by combining variables Eosinophil count (OCV 0.02), Basophil Count (OCV 0.03), MLR (OCV 0.83), PLR (312).
Fig. 3 ROC curve: Subgroup analysis comparing the MCG to the control group we found an AUC of 0.74 by combining variables Eosinophil count (OCV 0.02), Basophil Count (OCV 0.03), MLR (OCV 0.83), PLR (312).
Roc Curve: Assessing the postoperative PJI only group to the control group we calculated an AUC of 0.84 combining variables Eosinophil count (OCV 0.02), Basophil count (OCV 0.03), Lymphocyte count (OCV 0.83) and Monocyte Count (OCV 0.98).
Fig. 4 Roc Curve: Assessing the postoperative PJI only group to the control group we calculated an AUC of 0.84 combining variables Eosinophil count (OCV 0.02), Basophil count (OCV 0.03), Lymphocyte count (OCV 0.83) and Monocyte Count (OCV 0.98).
4

4 Discussion

Postoperative complications were associated with several factors, including increased age, male sex, higher body mass index (BMI), undergoing general anesthesia, the use of alternatives to acetylsalicylic acid (ASA) for venous thromboembolism prophylaxis, and a history of psychiatric illness. While certain preoperative laboratory values—specifically lower hemoglobin levels, elevated basophil counts, and increased monocyte-to-lymphocyte ratio (MLR)—were identified as independent predictors of postoperative complications, the combination of postoperative laboratory parameters, including elevated basophil count, MLR, platelet-to-lymphocyte ratio (PLR), and eosinophil count, demonstrated greater predictive value.

Increasing age, higher CCI, presence of osteoporosis and more frequent use of direct factor Xa inhibitors in the study group were the only demographic variables that reached significance (see Table 1) compared to the control group. Older age is independently associated with a higher risk of medical complications and mortality following a TKA or THA. The risk increases progressively with advancing age. In addition, comorbidities further amplify the risk of complications in older patients.20–22 Interestingly, surgical complications do not increase as markedly with age as medical complications. Studies show similar rates of surgical complications across age groups.23,24 Studies have shown an increased risk of medical and surgical complications with the usage of direct factor Xa inhibitors. In comparison to Aspirin, which currently is the gold-standard postoperative anticoagulation in low-risk patients, direct factor Xa inhibitors increase the risk of bleeding, leading to higher rates of hematoma, wound healing complications and transfusion requirements.25,26 Larger BMIs were both statistically significant comparing the PJI subgroup to the control and MCG respectively. This is not an unexpected finding, seeing that BMI is strongly and independently associated with an increased risk of PJI's, with the risk rising linearly with increases in BMI.27,28 We did see however that patients in the PJI subgroup were generally younger and had lower CCI compared to the MCG. This is in concordance with the medical literature that found that age alone may not be an independent risk factor for PJI when adjusted for comorbidities.29

When we look at the preoperative and postoperative demographic predictors for postoperative complications, the only clinically meaningful predictors were, usage of general anesthesia, and non-Aspirin anticoagulation. Being male, older age, and having a history of psychiatric illness are both non modifiable risk factors. Therefore, even though it is interesting to know, there is nothing that can be modified to decrease the risk. The literature demonstrated that GA is associated with a modest but clinically meaningful increase in medical complications after TKA and THA, particularly in higher risk patients.30,31 Therefore, spinal, regional or neuraxial anesthesia should be the method of choice in the higher risk group if possible. Similarly, Aspirin should be the anticoagulation of choice, if pre-existing medical conditions, (e.g. previous DVT, Pulmonary embolus, atrial fibrillation etc) do not preclude its usage. Elevated BMI was a postoperative predictor for complications. Patients should be encouraged to lose weight prior to surgery. This should be achieved preferably at least a year before surgery to offset higher rates of complications closer to surgery (weight loss within 9 months of surgery are associated with higher rates of complications).32

In the current study we found that preoperative laboratory predictors of complications included a lower HB value, higher basophil count and higher Monocyte-lymphocyte ratio (MLR). This is in accordance with the literature where multiple cohort studies and meta-analysis demonstrated that preoperative anaemia is an independent risk factor for adverse events (PJI, medical and surgical complications and mortality), even after controlling for comorbidities. They showed that the risk increases in a stepwise fashion as hemoglobulin decreases, aiming for levels around 13–14 g/dl.33,34 Similarly, MLR is associated with increased odds of aggregate postoperative complications and longer hospital stays. MLR is a marker for systemic inflammation and studies therefore recommend perioperative anti-inflammatory interventions, e.g. dexamethasone to reduce the risk of postoperative complications.35 Interestingly, there is limited evidence available associating higher basophil count and postoperative complications. This can be explained by low circulating numbers of basophils (<1 % of white blood cells) which leads to poor sensitivity and reproducibility with small fluctuations from baseline values.36

The postoperative laboratory predictors of complications included lower WBC and neutrophil count, and a higher Eosinophilic count. The current literature is insufficient to confirm using these markers as predictors for postoperative complications following TKA or THA. Usually, patients experience a surge in their WBC and neutrophil counts post-surgery as part of the normal surgical response, and lower values postoperatively might be an indication of an underlying condition (inadequate immune response, malnutrition or an overwhelming infection), which could theoretically increase the risk for postoperative complications.37 Eosinophils elevation postoperatively is not associated with the normal immune response following surgery, and an elevation can be an indication of immune imbalance.

4.1

4.1 Implications

The literature is clear that optimizing certain abnormal preoperative laboratory values (Hb, albumin, and poor glycaemic control) prior to surgery is associated with improved postoperative outcomes.38 The current study identified preoperative composite predictors with the highest AUC with the combination of MCHC, HB, Lymphocyte count and hematocrit. Optimizing preoperative Hb or hematocrit can be achieved by delaying surgery for treatments that include blood transfusions, iron transfusions or Erythropoietin treatments. MCHC may reflect underlying anaemia or iron deficiency and can be treated similarly as aforementioned. Lymphocyte counts are a marker of nutritional depletion and immunosuppression and can be corrected preoperatively by nutritional intervention, and management of immunosuppressive medications.39

Postoperatively the combination of Basophil count, MLR, PLR and Eosinophils led to an AUC of 0.76. The current literature suggests that postoperative optimization should focus on the management and monitoring of elevated MLR, PLR and NLR using immune modulating therapies such as dexamethasone to reduce complications 40. Currently there are no recommendations for treatment of basophil and eosinophilic counts.

5

5 Limitations

The control group consisted solely of elective TKA and THA cases, whereas the study group included patients undergoing hemiarthroplasty, the majority of whom were treated for hip fractures not amenable to fixation. These patients are generally older, more medically complex, and less thoroughly optimized preoperatively than elective arthroplasty patients, introducing a potential source of selection bias that must be considered when interpreting the findings. As a retrospective chart review, the study is also subject to the inherent limitations of this design, including variability in documentation and incomplete control of covariates. The relatively small number of PJI cases may have limited the statistical power of subgroup analyses. Additionally, differences in preoperative optimization between elective arthroplasty patients and the hip fracture cohort could influence both serologic values and complication rates. Finally, the possibility of unmeasured confounding cannot be excluded.

6

6 Conclusion

We identified multiple peri-operative patient, surgical, and serological risk factors for developing post-operative complications following hip and knee arthroplasty. These markers could be prioritized for monitoring high-risk patients for possible intervention. Ultimately, these results underscore the need for more rigorous, prospective research to better define which laboratory abnormalities truly influence postoperative outcomes and to determine which patient cohorts may safely forgo routine preoperative testing. Further studies with larger, more homogeneous samples are warranted to validate these observations and guide evidence-based perioperative evaluation strategies.

Ethical approval and consent to participate

Approved on the July 5, 2024, BIO-4954 Consent obtained from all patients to participate in the study.

Availability of supporting data

All the raw data and materials described in the manuscript is available upon requests to any scientist wishing to use them for non-commercial purposes.

Consent for publication

Not applicable.

Disclosures

The authors have no conflicts of interest to declare.

Author contributions

Dr Mars Zhao: Conceptualization, Methodology, Software, Formal analysis, Investigating, Writing original draft.

Dr Cole Elaschuk, MD: Formal analysis, Investigating, Writing original draft.

Dr. Janan Ashique, MD: Formal analysis, Investigating, Writing original draft.

Nathan Oster: Formal analysis, Investigating, Writing original draft.

Mikayla Rudniski: Formal analysis, Writing original draft.

Jenna England: Formal analysis, Investigating, Writing original draft.

Mason Beaulieu: Formal analysis, Investigating, Writing original draft.

Davidson Fadare: Formal analysis, Investigating, Writing original draft.

Dr Johannes M. van der Merwe, MBChB, FRCSC: Conceptualization, Methodology, Validation, Formal analysis, Investigating, Writing original draft.

Data available statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics statement

Description: Studies on patients or volunteers require ethics committee approval, which should be documented in the paper.

Funding

No funding for the research study.

References

  1. CJRR annual report: hip and knee replacements in Canada, 2021–2022.
    [Google Scholar]
  2. , , , , . Incidence rate of total knee arthroplasties in eleven European countries: do they reach a plateau? PLoS One. 2025;20(1)
    [Google Scholar]
  3. , , , et al . Causative factors for varus tibial baseplate subsidence in primary total knee arthroplasty: a retrospective analysis. J Orthop. 2025;67:59-67.
    [Google Scholar]
  4. , , , . Total hip arthroplasty techniques. StatPearls August 4, 2023
    [Google Scholar]
  5. , , , , . Total knee arthroplasty techniques. StatPearls May 6, 2024
    [Google Scholar]
  6. , , , , , . Incidence and timing of postoperative complications after total hip and knee arthroplasty. Can J Surg. 2024;67(6):E377-E382.
    [Google Scholar]
  7. , , , et al . Costs of joint replacement in osteoarthritis: a study using the national joint registry and clinical practice research datalink data sets. Arthritis Care Res. 2022;74(3):392-402.
    [Google Scholar]
  8. , , , , , , . Mortality during total knee periprosthetic joint infection. J Arthroplast. 2018;33(12):3783-3788.
    [Google Scholar]
  9. , , , , , , . Perioperative antibiotic prophylaxis: single and 24-Hour antibiotic dosages are equally effective at preventing periprosthetic joint infection in total joint arthroplasty. J Arthroplast. 2021;36(7S):S308-S313.
    [Google Scholar]
  10. , , , et al . Seasonal trends and risk factors in prosthetic joint infections: a retrospective analysis. J Orthopaedic Rep. 2025;4(2)
    [Google Scholar]
  11. , , , , , , . Has health care reform legislation reduced the economic burden of hospital readmissions following primary total joint arthroplasty? J Arthroplast. 2017;32(11):3274-3285.
    [Google Scholar]
  12. , , , , , , . Understanding readmission after primary total hip and knee arthroplasty: who's at risk? J Arthroplast. 2014;29(2):256-260.
    [Google Scholar]
  13. , , , et al . Impact of hospital readmissions reduction program penalties on hip and knee replacement readmissions: Comparison of hospitals at risk of varying penalty amounts. J Bone Joint Surg Am. 2020;102(1):60-67.
    [Google Scholar]
  14. , , , et al . The role of preoperative blood parameters to predict the risk of surgical site infection. Am J Surg. 2018;215(4):651-657.
    [Google Scholar]
  15. , , , , , . Predictors of surgical complications: a systematic review. Surgery (United States). 2015;158(1):58-65.
    [Google Scholar]
  16. , , , , , , . Comparative diagnostic value of serological and synovial tests for periprosthetic joint infections: a comprehensive analysis. JBJS Open Access. 2025;10(2)
    [Google Scholar]
  17. , . Optimizing early detection of post-operative complications in colorectal surgery: a meta-analysis of combined inflammatory markers. World J Colorectal Surg. 2024;13(4):105-116.
    [Google Scholar]
  18. , , , et al . The 2018 definition of periprosthetic hip and knee infection: an evidence-based and validated criteria. J Arthroplast. 2018;33(5):1309-1314.e2.
    [Google Scholar]
  19. , , , et al . The EBJIS definition of periprosthetic joint infection. Bone Joint Lett J. 2021;103-B(1):18-25.
    [Google Scholar]
  20. , , , et al . Trends in comorbidities and complications among octogenarians and Nonagenarians undergoing primary total joint arthroplasty in the United States. Anesth Analg. 2025 Jul;23
    [Google Scholar]
  21. , , , et al . Hip and knee arthroplasty outcomes for nonagenarian patients. J Am Acad Orthop Surg. 2022 Nov 15;30(22):1090-1097.
    [Google Scholar]
  22. , , , , , , . Morbidity and mortality analysis of primary total knee arthroplasty in patients over 90 years of age: insights from the German arthroplasty registry. J Arthroplast 2025 Nov 6
    [Google Scholar]
  23. , , , , . Older age increases short-term surgical complications after primary knee arthroplasty. Clin Orthop Relat Res. 2013 Aug;471(8):2611-2620.
    [Google Scholar]
  24. , , , , . The impact of older age on patient outcomes following primary total knee arthroplasty. Bone Joint Lett J. 2018 Nov;100-B(11):1463-1470.
    [Google Scholar]
  25. , , , , , . Direct factor Xa inhibitors versus low molecular weight heparins or vitamin K antagonists for prevention of venous thromboembolism in elective primary hip or knee replacement or hip fracture repair. Cochrane Database Syst Rev. 2025 Jan 27;1(1)
    [Google Scholar]
  26. , , , et al . Safety and efficacy of rivaroxaban in primary total hip and knee arthroplasty. J Arthroplast. 2023 Aug;38(8):1613-1620.e4.
    [Google Scholar]
  27. , , , et al . Correlation between body mass index and periprosthetic joint infection following total joint arthroplasty: a protocol for systematic review and meta analysis. Medicine (Baltim). 2020 Jun 12;99(24)
    [Google Scholar]
  28. , , , , , . Patient-related risk factors for periprosthetic joint infection after total joint arthroplasty: a systematic review and meta-analysis. PLoS One. 2016 Mar 3;11(3)
    [Google Scholar]
  29. , , , , , , . Higher age, female gender, osteoarthritis and blood transfusion protect against periprosthetic joint infection in total hip or knee arthroplasties: a systematic review and meta-analysis. Knee Surg Sports Traumatol Arthrosc. 2021 Jan;29(1):8-43.
    [Google Scholar]
  30. , , , et al . Comparison of pneumonia and major complications after total joint arthroplasty with spinal versus general anesthesia: a propensity-matched cohort analysis. J Am Acad Orthop Surg. 2024 Jan 1;32(1):33-40.
    [Google Scholar]
  31. , , , , . Does regional anesthesia reduce complications following total hip and knee replacement compared with general anesthesia? An analysis from the national joint registry for England, Wales, Northern Ireland and the Isle of Man. J Arthroplast. 2020 Jun;35(6):1521-1528.e5.
    [Google Scholar]
  32. , , , et al . The impact of preoperative weight loss timing on surgical outcomes in total hip arthroplasty. J Arthroplast. 2024 Jun;39(6):1424-1431.
    [Google Scholar]
  33. , , , et al . Increased severity of anaemia is associated with 30-day complications following total joint replacement. Bone Joint Lett J. 2020 Apr;102-B(4):485-494.
    [Google Scholar]
  34. , , , , , . Preoperative anemia and complications after total joint arthroplasty: a systematic review and meta-analysis. Eur Rev Med Pharmacol Sci. 2022 Oct;26(20):7420-7430.
    [Google Scholar]
  35. , , , et al . Poor performance of monocyte-to-lymphocyte ratio, neutrophil-to-lymphocyte ratio, and fibrinogen when screening for occult infection among patients with sequelae of suppurative hip arthritis before total hip arthroplasty. Int Orthop. 2022 Sep;46(9):1929-1935.
    [Google Scholar]
  36. , , , et al . Basophils from allergy to cancer. Front Immunol. 2022 Dec 12;13
    [Google Scholar]
  37. , , , , , . Leukocytosis is common after total hip and knee arthroplasty. Clin Orthop Relat Res. 2011 Nov;469(11):3031-3036.
    [Google Scholar]
  38. , , , , , , . Patient preoperative optimization: how to Do it and how to be paid for the work. J Arthroplast. 2025 Sep;40(9S1):S66-S71.
    [Google Scholar]
  39. , , , , , , . Preoperative risk factor screening protocols in total joint arthroplasty: a systematic review. J Arthroplast. 2020 Nov;35(11):3353-3363.
    [Google Scholar]

Fulltext Views
108

PDF downloads
17
View/Download PDF
Download Citations
BibTeX
RIS
Show Sections

Suggested read for related articles: