Factors Associated With Elevated Inflammatory Markers Prior to Shoulder Arthroplasty

Abstract

Background: Preoperative erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) ranges for several shoulder arthroplasty indications are not well understood. Purpose: We sought to compare preoperative ESR and CRP values for a variety of shoulder arthroplasty indications and evaluate risk factors for elevated preoperative ESR and CRP values. Methods: We conducted a retrospective cohort study of shoulder arthroplasty cases performed at a single academic medical institution from 2013 to 2018. Preoperative ESR and CRP values for 235 shoulder arthroplasties with various indications were recorded. Independent risk factors for elevated values (CRP > 10.0 mg/L and ESR > 30.0 mm/h) were determined via multiple variable logistic regression. Results: Patients undergoing shoulder arthroplasty for osteoarthritis had an ESR (mean ± SD) of 22.6 ± 17.8, with 29.8% of patients elevated, and a CRP of 6.5 ± 6.4, with 25.5% of patients elevated. Arthroplasty for acute fracture and prosthetic joint infection (PJI) had higher preoperative ESR and CRP values. Multivariate analysis identified several predictors of elevated ESR, including infection, acute fracture, diabetes, and female sex. It also identified predictors of elevated CRP, including infection, acute fracture, and younger age. Conclusions: Preoperative ESR and CRP values may be elevated in 25% to 30% of patients undergoing primary shoulder arthroplasty. Arthroplasty for both acute fracture and PJI, along with several other patient factors, was associated with elevated preoperative ESR and CRP. Thus, routine collection of ESR and CRP preoperatively may not be of benefit, as elevated values are common. Further study is warranted.

Introduction

Glenohumeral osteoarthritis (OA) is one of the most common indications for shoulder arthroplasty to reduce pain and improve function [3–5,30]. Other pathologies, such as rotator cuff arthropathy, proximal humeral fracture, glenohumeral dislocation, rotator cuff tear, and avascular necrosis, may be treated with various arthroplasty techniques including total shoulder arthroplasty (TSA), reverse total shoulder arthroplasty (rTSA), and hemiarthroplasty (HA). Favorable outcomes, evolving implant designs, and improvements in surgical techniques have resulted in an increasing use of shoulder arthroplasty over the past few decades [16,31,36].

Although rare (<1% prevalence), the consequences of periprosthetic shoulder infections can be devastating, including need for further surgery, systemic infection, and inferior outcomes [7,8]. One common method used to diagnose prosthetic joint infection (PJI) is determining whether serum inflammatory markers, such as erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP), are elevated [7]. Evidence suggests that markers such as ESR and CRP may not be accurate as diagnostic tools for low-grade and indolent infections that are commonly seen in the shoulder [6,11,27]. Furthermore, a study investigating multiple anatomic sites showed that ESR and CRP are more sensitive/specific in detecting infections in the hip, knee, and spine than in the shoulder [28]. Pathogens such as Cutibacterium acnes (formerly Propionibacterium acnes) constitute a large portion (38.9%) of shoulder PJIs [20] but are indolent and slow-growing anaerobic bacteria [13,14]. The presence of organisms such as C acnes has been associated with reduced sensitivity of ESR or CRP [6,15,26,29] in detecting shoulder PJI. Moreover, elevated values of ESR or CRP are not highly specific for infection and may occur in a number of different scenarios, in response to injury or illness, or even in the absence of infection [11,28]. Baseline ESR values for various shoulder arthroplasty indications have not been well described to date in the literature, and baseline CRP values have only been described for a few indications [35]. Establishment of baseline ESR and CRP values for different shoulder arthroplasty indications could add to existing literature by providing surgeons appropriate context to interpret these inflammatory markers for various patient scenarios in the pre- and postoperative periods.

The primary purpose of the current study was to investigate preoperative values for ESR and CRP for surgical indications such as OA, fracture, rotator cuff arthropathy, revision arthroplasty without infection, and active PJI. Specifically, our primary hypothesis was that patients undergoing primary arthroplasty for OA would have a lower rate of elevated preoperative ESR and CRP than would patients with PJI or fracture. The secondary purpose was to identify patient risk factors for elevated preoperative ESR and CRP.

Methods

After obtaining Institutional Review Board approval, a retrospective cohort study was conducted by reviewing the electronic medical records of shoulder arthroplasty cases (n = 1211) performed at a single academic medical institution from 2013 to 2018. This time range was selected due to the initiation of routine collection of ESR and CRP by our institution in 2013, according to a change in institutional protocol for arthroplasty patients. Surgeries were conducted by 1 of 4 shoulder and elbow fellowship trained orthopedic surgeons. Patients undergoing partial or TSA for the following indications were included: OA with an intact rotator cuff, rotator cuff arthropathy, fracture, revision arthroplasty without an active infection, and active PJI. An acute fracture was defined as a fracture that occurred less than 4 weeks from the date of injury to the date of surgery. Revisions were defined as operations performed on a previous TSA, rTSA, or HA without infection. An active PJI was defined as treatment of a suspected or confirmed infection by debridement, component exchange, and/or explantation of components with or without antibiotic therapy [8]. Exclusion criteria included missing ESR or CRP data, known concurrent systemic infection, or a surgical indication other than those previously mentioned (such as avascular necrosis, shoulder instability, or malignancy). Patients with native glenohumeral septic arthritis (n = 2) were excluded from the analysis due to the small sample size.

For each patient, indications for surgery were determined by reviewing the preoperative diagnosis in the surgeon’s operative report. The following additional characteristics were noted for each patient: age, sex, preoperative ESR and CRP values and their collection dates, presence of diabetes mellitus, presence of rheumatoid arthritis or autoimmune disease, tobacco use status, and prior surgical history of the operative shoulder. Data gathered regarding revision operations included preoperative implant type, postoperative implant type, and whether there was a history of infection.

Only patients with available preoperative ESR (mm/h) and CRP (mg/L) results were included in the study. If multiple lab samples were available preoperatively, the value closest in time to the surgical date was used. No ESR or CRP values were used that were obtained more than 30 days prior to their operations. Thresholds for elevated ESR and CRP values were 30 mm/h and 10 mg/L, respectively, which was based on proceedings from the international PJI consensus meeting, similar studies, and the reference ranges for adults in our institution [12,25,28].

Statistical Analysis

After performing an a priori power analysis, consecutive cases of OA with an intact rotator cuff, rotator cuff arthropathy, fracture, revision arthroplasty without an active infection, and active PJI were included until a goal of 45 cases were identified for each surgical indication. The power analysis indicated that this was adequate to detect a mean 6.5 point difference in ESR and CRP values by indication with 80% power. Fracture (n = 44) and PJI (n = 33) did not meet the a priori goal of 45 as all available cases were collected during the study period. A post hoc sample size estimation indicated that n = 33 per group was sufficient to detect a mean 7.5 point difference in mean ESR and CRP between groups with 80% power.

Statistical analysis was performed with a standard software package (JMP 14.2, SAS Institute, Cary, NC). Descriptive statistics were used to summarize the findings of our sample population. The distributions of ESR and CRP values were determined according to surgical indication. Chi-square tests were performed comparing each indication to OA with an intact rotator cuff as a baseline. To determine independent predictors of ESR and CRP values, multiple variable linear and logistic regression models were created, respectively. Linear regression models were created for ESR and CRP as continuous variables, and logistic regression models were created for elevated CRP or elevated ESR. A backwards selection method was used with an exit criterion of P > .05. Variables assessed for potential inclusion in the multivariate models were surgical indication, age, sex, history of prior shoulder surgery, history of prior shoulder infection, tobacco use status, history of rheumatoid arthritis or autoimmune disorder, and history of diabetes mellitus. The effect of fracture on ESR and CRP was time-dependent, and receiver operating characteristic curve (ROC) analysis was used to determine the ideal timepoint from injury to arthroplasty as a predictor of ESR or CRP; this was defined as ≤8 vs >8 weeks for CRP and ≤4 vs >4 weeks for ESR.

Results

Of the included patients (n = 235), 103 (43.8%) were male and 132 (56.2%) were female. The mean age was 64.0 ± 9.6 years (Table 1). Most of the patients underwent reverse TSA (52.7%); the next most common procedure was anatomic TSA (22.1%), followed by HA (11.1%), and revision procedures (13.6%). A total of 47 (20.0%) patients underwent arthroplasty for OA, 51 (21.7%) for rotator cuff arthropathy, 44 (18.7%) for fracture, 51 (21.3%) for revision arthroplasty without infection, and 33 (14.0%) for an active PJI. The remaining 9 patients had a revision surgery with a resolved infection (3.8%).

Table 1.

Candidate predictors analyzed in univariate analysis.

Variable	All study patients (N = 235 patients)
Demographics and medical history
Age, y	64.0 ± 9.6
Sex (number of females)	132 (56.2%)
Current tobacco smoker	50 (21.3%)
RA diagnosis	13 (5.5%)
Non-RA autoimmune disease diagnosis	43 (18.3%)
Diabetes mellitus	66 (28.1%)
Surgical history
Prior arthroplasty	91 (38.7%)
Prior rotator cuff repair	30 (12.8%)
Prior fracture fixation	20 (8.5%)
Surgical indication
Osteoarthritis, rotator cuff intact	47 (20.0%)
Rotator cuff arthropathy	51 (21.7%)
Fracture occurred ≤4 wk preoperatively	19 (8.1%)
Fracture occurred ≤8 wk preoperatively	26 (11.1%)
Fracture occurred >8 wk preoperatively	18 (7.7%)
Time from fracture to arthroplasty, wk	17.8 ± 20.4
Revision arthroplasty for PJI	33 (14%)
Revision arthroplasty, without infection	51 (21.7%)
Procedure
Hemiarthroplasty	26 (11.1 %)
Anatomic TSA	52 (22.1%)
Reverse TSA	124 (52.7%)
Antibiotic spacer	21 (8.9%)
Polyethylene exchange	4 (1.7%)
Resection arthroplasty	7 (3.0%)

Candidate predictor variables for elevated inflammatory markers studied in the univariate analysis. Mean ± SD are provided for normal variables and number and proportion are given for categorical variables. RA rheumatoid arthritis, PJI prosthetic joint infection, TSA total shoulder arthroplasty.

Using patients undergoing TSA for glenohumeral OA as a baseline (ESR: 22.6 ± 17.8; CRP: 6.5 ± 6.4), patients undergoing arthroplasty for a fracture less than 4 weeks had elevated values with a mean preoperative ESR of 41.1 ± 25.1 (P = .013) and CRP of 36.0 ± 51.8 (P < .001) (Table 2). Patients with active PJI also had elevated values compared with the baseline group, with a mean preoperative ESR of 48.1 ± 37.7 (P < .001) and CRP of 40.2 ± 53.3 (P < .001). The remaining indications showed normal average baseline ESR and CRP values without intergroup differences.

Table 2.

Comparison of ESR and CRP values by surgical indication.

Indication	ESR, mm/h (Mean ± SD); range	P value	CRP, mg/L (mean ± SD); range	P value
OA, rotator cuff intact	22.6 ± 17.8; r: 1–75	Referent	6.5 ± 6.4; r : 1–22.7	Referent
Rotator cuff arthropathy	30.6 ± 31.8; r: 1–130	.123	4.6 ± 4.33; r : 1–20.7	.714
Fracture ≤4 wk preop	41.1 ± 25.1; r : 1–99	.013	36.0 ± 51.8; r : 1.9–193.5	<.001
Fracture ≤8 wk preop	38.2 ± 25.9; r : 1–99	.018	30.9 ± 45.5; r : 1.9–193.5	<.001
Fracture >8 wk preop	22.2 ± 12.1; r : 1–37	.957	5.2 ± 6.1; r : 1–26.2	.848
PJI	48.1 ± 37.7; r : 1–130	<.001	40.2 ± 53.3; r: 1–209.5	<.001
Revision arthroplasty, without infection	25.2 ± 20.3; r : 1–119	.545	6.2 ± 6.5; r : 1–31	.988

Comparison of ESR and CRP by each surgical indication, with patients with OA and an intact rotator cuff (“OA, Rotator Cuff Intact”) used as the baseline comparison group. Mean ± SD, as well as range of minimum and maximum values (“r”) are provided for each group. Here, PJI and fractures are stratified by how many weeks elapsed from the date of injury for the fracture to the date of surgery. ESR erythrocyte sedimentation rate, CRP C-reactive protein, OA osteoarthritis, PJI prosthetic joint infection.

Bold indicates statistically significant values.

Despite a normal average value, the preoperative ESR was elevated more than 30 mm/h in 29.8% (14/47) of patients undergoing anatomic TSA for glenohumeral OA. The percentage of patients with baseline elevated ESR was higher in patients with fracture less than 4 weeks (63.2%, P = .012) and PJI (60.6%, P = .006) compared with patients with OA (Fig. 1). Rotator cuff arthropathy (33.3%, P = .706), fracture greater than 4 weeks (40.0%, P = .381), and revision arthroplasty without an infection (39.2%, P = .327) showed no difference in rates of baseline ESR elevation compared with patients with OA.

Fig. 1.

Proportion (percentage) of patients with elevated preoperative erythrocyte sedimentation rate (ESR), categorized by surgical indication for arthroplasty. An elevated ESR level was defined as a laboratory measurement of serum ESR greater than 30 mm/h. ^†Significant difference between acute fracture (defined as a fracture occurring within 4 weeks of the arthroplasty procedure) and the baseline comparison group of osteoarthritis with intact rotator cuff (“OA, cuff intact”) patients (P = .012). ^‡Significant difference between active periprosthetic joint infection (PJI) and the baseline comparison group of osteoarthritis with intact rotator cuff (“OA, cuff intact”) patients (P = .006).

Despite a normal average value, the preoperative CRP was elevated more than 10 mg/L in 25.5% (12/47) of patients undergoing anatomic TSA for glenohumeral OA. The proportion of patients with preoperative elevated CRP was higher in the subgroup with fractures less than 8 weeks (57.7%, P = .006) and PJI (54.6%, P = .008) compared with patients with OA (Fig. 2). Fracture greater than 8 weeks (11.1%, P = .206) and revision arthroplasty without an infection (20.0%, P = .483) showed no difference in rates of baseline CRP elevation compared with patients with OA. The percentage with elevated CRP was lower in patients with rotator cuff arthropathy (9.8%, P = .040).

Fig. 2.

Proportion (percentage) of patients with elevated preoperative C-reactive protein (CRP), categorized by surgical indication for arthroplasty. An elevated CRP level was defined as a laboratory measurement of serum CRP greater than 10 mg/L. *Significant difference between rotator cuff deficient (“rotator cuff arthropathy”) and the baseline comparison group of osteoarthritis with intact rotator cuff (“OA, cuff intact”) patients (P = .040). ^†Significant difference between acute fracture (defined as a fracture occurring within 4 weeks of the arthroplasty procedure) and the baseline comparison group of osteoarthritis with intact rotator cuff (“OA, cuff intact”) patients (P = .006). ^‡Significant difference between active periprosthetic joint infection (“PJI”) and the baseline comparison group of osteoarthritis with intact rotator cuff (“OA, cuff intact”) patients (P = .008).

In multivariate analysis, independent predictors of elevated CRP were arthroplasty for fracture that occurred within 8 weeks of surgery and active PJI, with older age decreasing the likelihood of elevated CRP. The independent predictors in combination had good ability to predict an elevated CRP value (whole model area under the receiver operating characteristic curve [AUROC] = 0.72; P < .001) (Table 3). Surgery for indications other than fracture 8 or less weeks or active PJI did not affect the likelihood of a preoperative CRP value 10.0 mg/L or greater (P > .25).

Table 3.

Independent predictors of elevated preoperative CRP (as a binary variable).

Predictor	Adjusted odds ratio	95% confidence interval	P value
Fracture occurred ≤8 wk preoperatively	5.95	(2.35–15.1)	<.001
Active periprosthetic joint infection	4.88	(2.19–10.9)	<.001
Age, per 5-yr increase	0.83	(0.70–0.995)	.03
Whole model area under receiver operating characteristic curve: 0.71			<.001

Independent predictors of elevated preoperative CRP determined by multivariate analysis, with CRP treated as a binary outcome variable. An elevated CRP level was defined as a laboratory measurement of serum CRP greater than 10 mg/L. Variables assessed and ultimately excluded from the multivariate model due to lack of significance were surgical indication (other than fracture or periprosthetic infection), sex, history of prior shoulder surgery, history of prior shoulder infection, tobacco use status, history of rheumatoid arthritis or autoimmune disorder, and history of diabetes mellitus. CRP C-reactive protein.

Bold indicates statistically significant values.

Independent predictors of elevated CRP (as a continuous variable) included surgery for active PJI, arthroplasty for fracture that occurred within 8 weeks of surgery, and age on day of surgery, which together explained 27% of the variance in CRP values (R² = .27, P < .001) (Table 4). Surgery for indications other than fracture 8 or less weeks or active PJI did not independently affect preoperative CRP values (P > .25).

Table 4.

Independent predictors of preoperative CRP (as a continuous variable).

Predictor	Effect size (Mean change in CRP)	Standard error	P value
Fracture occurred ≤8 wk preoperatively	13.44	2.71	<.001
Active periprosthetic joint infection	15.48	2.37	<.001
Age on day of surgery, per 5-yr increase	−3.66	0.88	<.001
Whole model R2 value: .27			<.001

Independent predictors of preoperative CRP determined by multivariate analysis, with CRP treated as a continuous outcome variable. Variables assessed and ultimately excluded from the multivariate model due to lack of significance were surgical indication (other than fracture or periprosthetic infection), sex, history of prior shoulder surgery, history of prior shoulder infection, tobacco use status, history of rheumatoid arthritis or autoimmune disorder, and history of diabetes mellitus. CRP C-reactive protein.

Bold indicates statistically significant values.

Independent predictors of elevated ESR in multivariate analysis were arthroplasty for fracture 4 weeks or less before surgery, surgery for active PJI, prior rotator cuff repair, and female sex which together had fair ability to predict an elevated ESR value (whole model AUROC = 0.69; P < .001) (Table 5). Surgery for indications other than fracture 4 or less weeks or active PJI did not independently affect the likelihood of a preoperative ESR value 30.0 mm/h or greater (P > .25).

Table 5.

Independent predictors of elevated preoperative ESR (as a binary variable).

Predictor	Adjusted odds ratio	95% confidence interval	P value
Fracture occurred ≤4 wk preoperatively	3.50	(1.13–10.8)	.02
Active periprosthetic joint infection	3.45	(1.48–8.03)	.003
Female sex	2.37	(1.31–4.26)	.003
Prior rotator cuff repair	0.26	(0.09–0.72)	.004
Whole model area under receiver operating characteristic curve: 0.69			<.001

Independent predictors of elevated preoperative ESR determined by multivariate analysis, with ESR treated as a binary outcome variable. An elevated ESR level was defined as a laboratory measurement of serum ESR greater than 30 mm/h. Variables assessed and ultimately excluded from the multivariate model due to lack of significance were surgical indication (other than fracture or periprosthetic infection), age, history of prior shoulder surgery (other than rotator cuff repair), history of prior shoulder infection, tobacco use status, history of rheumatoid arthritis or autoimmune disorder, and history of diabetes mellitus. ESR erythrocyte sedimentation rate.

Bold indicates statistically significant values.

Independent predictors of elevated ESR (as a continuous variable) included surgery for active PJI, history of prior rotator cuff repair, history of diabetes, and female sex, which together explained 17% of the variance in ESR values (R² = .16, P < .001) (Table 6). Surgery for indications other than active PJI did not independently affect preoperative ESR values (P > .25).

Table 6.

Independent predictors of preoperative ESR (as a continuous variable).

Predictor	Effect size (Mean change in ESR)	Standard error	P value
Female sex	4.79	3.20	.005
Active periprosthetic joint infection	12.58	2.45	<.001
Diabetes mellitus	4.24	1.83	.02
Prior rotator cuff repair	−6.30	2.44	.01
Whole model R2 value: .16			<.001

Independent predictors of preoperative ESR determined by multivariate analysis, with ESR treated as a continuous outcome variable. Variables assessed and ultimately excluded from the multivariate model due to lack of significance were surgical indication (other than fracture or periprosthetic infection), age, history of prior shoulder surgery (other than rotator cuff repair), history of prior shoulder infection, tobacco use status, history of rheumatoid arthritis or autoimmune disorder. ESR erythrocyte sedimentation rate.

Bold indicates statistically significant values.

Discussion

Our results demonstrate that more than 1 out of 4 patients undergoing primary shoulder arthroplasty for OA have elevated preoperative ESR (29.8%) and CRP (25.5%) values. Regarding our primary hypotheses, we established that differences in preoperative ESR existed between (1) patients with active PJI and primary OA (elevated in 60.6% vs 29.8%, P = .006) and (2) patients with acute fracture and primary OA (elevated in 63.2% vs 29.8%, P = .012). Also, differences existed in preoperative CRP between (1) patients with active PJI and primary OA (elevated in 54.6% vs 25.5%, P = .008) and (2) patients with acute fracture and primary OA (elevated in 57.7%, P = .006). These results imply that elevated preoperative ESR and CRP values may not be a relative contraindication to surgery and it may be acceptable to proceed with arthroplasty despite having these lab values.

The use of inflammatory markers such as ESR and CRP in shoulder surgery has been primary for the screening of infections, particularly PJI [13,14,20]. Analyses of PJIs and noninfected cases following hip and knee arthroplasties have elucidated ideal thresholds for ESR and CRP to enhance sensitivity (up to 100%) and specificity (54.7%) for PJIs [2]. However, similar efforts to effectively use ESR and CRP to diagnose shoulder PJI are complicated by a high proportion of indolent-growing organisms such as C acnes, with sensitivity and specificity of approximately 66% [14]. Values of ESR <30 and CRP <3 are seen even with confirmed infections [19]. Torrens et al [35] expanded knowledge of preoperative CRP values for diagnoses other than PJI, such as primary glenohumeral OA, rotator cuff arthropathy, and acute fracture. However, Torrens and associates did not compare these CRP values with other diagnoses, nor did they evaluate ESR for the indications studied [35]. Others have characterized normal perioperative ESR and CRP values for arthroplasty in other joints [22–24,33,38] and reported normal baseline preoperative ESR and CRP in uncomplicated patients. Currently, the associations between elevated preoperative ESR and CRP values in the shoulder and the risk of postoperative complications (such as PJI) are unknown.

Elevated ESR and CRP values prior to arthroplasty for acute fracture are consistent with the elevated preoperative CRP values (35 ± 38 mg/L) reported by Torrens et al [35]. Also, the elevation of preoperative ESR and CRP in more than 1 out of 4 patients with primary OA is congruent with findings by Shih et al [33] of elevated preoperative ESR, but normal CRP, in patients undergoing uncomplicated primary total hip arthroplasty. For shoulder arthroplasty, the elevated inflammatory markers may be due to a number of factors, such as recent upper respiratory tract infection, reduced clearance of ESR and CRP from the serum, or an occult autoimmune condition. Additional collection of historical and laboratory data, ideally in a prospective manner, would be helpful in elucidating if these are potential contributory factors.

Patient factors apart from surgical indication also affect preoperative ESR and CRP values. Previous investigations identified that female sex and older age are associated with higher ESR levels in the general population [18,21,32,34]. Our results identify female sex as a risk factor for higher ESR, but not older age. For CRP, age actually had the opposite effect; younger age independently predicted higher preoperative CRP values. This is in contrast to previous reports that younger and more physically active patients have a lower baseline CRP [9,37]. One possible explanation could be that younger patients undergoing shoulder arthroplasty may be suffering from additional chronic disease and poorer overall health, although we attempted to control for this with multivariate regression incorporating these factors. A wide variety of comorbid conditions are known to elevate levels of both ESR and CRP [1,17]. Of note, the patient factors assessed in this study explained 17% and 27% of the variance for ESR and CRP, respectively, which indicates that there is wide individual variation for each test not well explained by the factors assessed. Thus, future studies are needed to determine how possible comorbid conditions contribute to the fluctuation of ESR and CRP values both before and after shoulder arthroplasty.

This study has several limitations. First, ESR and CRP are nonspecific markers of inflammation, and patient characteristics such as systemic illness, malignancy, and other parameters that were not assessed in the study may contribute to the results [10]. Second, our database contained patients at a single academic medical center, which may limit its external validity. Not all indications for arthroplasty were assessed, as certain surgical indications such as malignancy and avascular necrosis were excluded from analysis due to the small number of patients with these conditions. Because this was a retrospective study, it can only imply associations, not causal relationships, which would require a prospective investigation. Finally, postoperative values were not analyzed, so the implications of elevated baseline values could not be determined.

In conclusion, preoperative ESR and CRP are elevated for patients undergoing shoulder arthroplasty for acute fracture and PJI compared with arthroplasty for glenohumeral OA, with patient factors including diabetes, sex, and age also contributing. Nevertheless, preoperative values may be elevated in 25% to 30% of patients undergoing primary arthroplasty for glenohumeral OA. Thus, routine collection of ESR and CRP preoperatively may not be of benefit as elevated values are common. Future investigations, ideally with a prospective design, would be helpful in determining other potential contributory factors to elevated preoperative ESR and CRP prior to elective shoulder arthroplasty and to assess if patients with these conditions are at increased risk of postoperative complications.