Abstract
Background
Identifying preoperative risk factors for complications following total shoulder arthroplasty (TSA) has both clinical and financial implications. The purpose of this study was to determine the influence of different degrees of preoperative INR elevation on complications following TSA.
Methods
Patients undergoing primary TSA from 2007 to 2018 were identified in a national database. Patients were stratified into 4 cohorts: INR of <1.0, INR of >1.0 to 1.25, INR of >1.25 to 1.5, and INR of >1.5. Postoperative complications were assessed. Multivariate logistic regressions were performed to adjust for differences in demographics and comorbidities among the INR groups.
Results
Following adjustment and relative to patients with an INR of <1.0, those with INR of >1.0–1.25, >1.25–1.5, and >1.5 had 1.6-times, 2.4-times, and 2.8-times higher odds of having postoperative bleeding requiring transfusion, respectively (p < 0.05 for all). Relative to patients with INR <1.0, those with INR of > 1.25–1.5 and INR of >1.5 had 7.8-times and 7.0-times higher odds of having pulmonary complications, respectively (p < 0.05 for both).
Discussion
With increasing INR levels, there is an independent and step-wise increase in odd ratios for postoperative complications. Current guidelines for preoperative INR thresholds may need to be adjusted for more predictive risk-stratification for TSA.
Level of Evidence
III
Keywords: total shoulder arthroplasty, TSA, international normalized ratio, INR, complications, risk stratification
Introduction
Perioperative bleeding is one of the most common complications seen in total shoulder arthroplasty (TSA), occurring from 4% to 43% of the time and requiring transfusion in approximately 7% to 25% of cases.1–4 It is known that abnormal preoperative coagulation tests such as international normalized ratio (INR) increases risk of bleeding following various types of procedures, but much of the existing literature on associations between preoperative INR and perioperative bleeding and transfusion risk has been assessed in non-orthopaedic procedures such as cardiac procedures and neurosurgical procedures. 5 It is important to note that INR is only one component to determine one's risk of bleeding, and a normal INR does not necessarily indicate that there is a normal clotting response. Nevertheless, recent studies investigating the effect of preoperative INR on complications following total knee arthroplasty (TKA) and total hip arthroplasty (THA) reported significant, stepwise relationships between degree of INR elevation and postoperative bleeding, infection, transfusion, and mortality rates.6,7
With the increasing focus on cost-containment within the United States healthcare system, risk stratification for common postoperative complications such as blood loss and transfusion carry immense value both clinically and financially. Further, joint replacement surgeries such as TKA, THA, and TSA are among the most common elective surgeries performed in the country, and so risk stratification of patients at risk preventable complications following these procedures is particularly important on a national scale.8–10 Although INR is a routinely collected preoperative laboratory value, there is currently a paucity of data on risks for postpreparative complications following TSA performed in patients with various levels of preoperative INR. 11 With further investigation, standardized screening guidelines for coagulation tests such as INR can be implemented to more accurately risk-stratify patients undergoing TSA.11–13
The purpose of this study was to investigate the association between different degrees of preoperative INR elevation from baseline and postoperative complications in TSA patients. We hypothesized that there would be a step-wise progression of complications rates, including postoperative bleeding and transfusion requirement, with increasing preoperative INR values.
Materials and methods
This was a retrospective cohort study of adult patients who underwent primary TSA recorded in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database from the years 2007 to 2018. The ACS-NSQIP is a national database that includes data from more than 700 participating institutions, which is collected by trained clinical reviewers. It collects preoperative, demographic, and comorbidity data and documents perioperative events and postoperative complications occurring within 30 days for patients undergoing surgical procedures. 14 Previous studies have demonstrated high inter-rater reliability for data collection using this database, leading to its widespread using in orthopaedic investigations.15–18
Patient selection
Patients who underwent primary TSA were identified using Current Procedural Terminology (CPT) code 23472 (arthroplasty, glenohumeral joint; total shoulder). Since separate CPT codes are not available for the varying types of TSA, CPT code 23472 includes anatomic TSA, reverse TSA, and total shoulder resurfacing. Patients were excluded from this study if their INR was not recorded within 1 day before the TSA or if they were < 18-years-old (Figure 1). Based on previous studies’ stratification of INR, four subgroups were used to assess the impact of preoperative INR on complications following surgery: <1.0, >1.0 to 1.25, >1.25 to 1.5, and >1.5.6,7
Figure 1.
Exclusion criteria for patients undergoing total shoulder arthroplasty.
Baseline patient characteristics
Patients’ baseline demographics data and clinical characteristics included age, sex, race, Body Mass Index (BMI), American Society of Anesthesiologists (ASA) class, length of stay, smoking status, and functional status. Patient comorbidities and intraoperative variables included chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), hypertension, dialysis, renal failure, steroid use, bleeding disorder, disseminated cancer, prior operation within 30 days, diabetes, dyspnea, anesthesia type, and operative time.
Postoperative complications
The primary outcome was bleeding events requiring transfusion up to 72 h postoperatively. Secondary thirty-day outcomes were classified into clinically relevant groups based on the type of complication. These groups included wound (superficial surgical site infection, deep surgical site infection, organ or space infections, or wound disruption), cardiac (cardiac arrest or myocardial infarction), pulmonary (pneumonia, reintubation, or failure to wean off ventilator for more than 48 h), renal (renal failure or renal insufficiency), thromboembolic (deep vein thrombosis, pulmonary embolism, or stroke), and sepsis (sepsis or septic shock). Mortality, urinary tract infection, extended length of hospital stay, readmission, and reoperation were also recorded. Extended length of stay was defined as > 6 days, or one standard deviation above the mean length of stay of 2.6 days for the patients in this study.
Statistical analysis
Bivariate and multivariate analyses were conducted using the Statistical Package for the Social Sciences (SPSS; Version 26; Armonk, NY) software. Data on patient demographics, clinical characteristics, comorbidities, and postoperative complications were analyzed with bivariate analysis using Pearson's Chi-squared test and analysis of variance where appropriate to analyze differences by INR class. All binary comparisons were compared with the cohort of patients who had an INR of <1.0. Demographics, clinical characteristics, and comorbidities were included in the multivariable analysis for p-values < 0.20, consistent with previous methodologies. 19 Postoperative complication variables with a p-value < 0.05 were selected for multivariable analyses. Only one INR range that was statistically significant on bivariate analysis was required for the complication variable to be included on multivariable analysis. Multivariable analysis results were reported as odds ratios with 95% confidence intervals. On multivariable analysis, a p-value of < 0.05 was the cut-off value for statistical significance.
Results
Demographics and comorbidities
In total, 1712 primary TSA patients were included in the analysis after the exclusion criteria were applied. There were 830 patients (48.5%) in the INR <1.0 group, 742 (43.3%) in the INR >1.0–1.25 group, 119 (7.0%) in the INR >1.25–1.5 group, and 21 (1.2%) in the INR >1.5 group (Table 1). Compared to patients with an INR of <1.0, those with an INR of >1.0–1.25 were more likely to be older (73.1 vs. 70.2 years, p < 0.001), male (48.1% vs. 37.2%, p < 0.001), have an ASA class of III (70.7% vs.60.4%, p < 0.001), be partially dependent on others in terms of functional status (7.2% vs. 3.3%, p = 0.001), and have a longer length of hospital stay (2.8 vs. 2.5 days, p = 0.006). Relative to patients with an INR of <1.0, those with an INR of >1.25–1.5 were more likely to be older (75.3 vs. 70.2 years, p < 0.001) and have an ASA class of III (77.3% vs. 60.4%, p < 0.001). Compared to patients with an INR of <1.0, those with an INR of >1.5 were more likely to have an ASA class of IV (23.8% vs. 4.2%, p < 0.001) and experience a longer length of hospital stay (5.1 vs. 2.5 days, p < 0.001) (Table 1).
Table 1.
Demographics and nclinical characteristics for patients undergoing total shoulder arthroplasty.
| Demographics | INR <1.0 | INR >1.0-1.25 | INR >1.25-1.5 | INR >1.5 | p-value: INR >1.0-1.25 vs INR <1.0 | p-value: INR >1.25-1.5 vs INR <1.0 | p-value: INR >1.5 vs INR <1.0 |
|---|---|---|---|---|---|---|---|
| Total patients, n | 830 | 742 | 119 | 21 | |||
| Sex, n (%) | < 0.001¶ | 0.633¶ | 0.331¶ | ||||
| Female | 521 (62.8) | 385 (51.9) | 72 (60.5) | 11 (52.4) | |||
| Male | 309 (37.2) | 357 (48.1) | 47 (39.5) | 10 (47.6) | |||
| Ethnicity, n (%) | 0.421¶ | 0.885¶ | 0.758¶ | ||||
| Caucasian | 639 (87.3) | 603 (88.8) | 97 (89.0) | 16 (80.0) | |||
| Black or African American | 37 (5.1) | 33 (4.9) | 5 (4.6) | 2 (10.0) | |||
| Hispanic | 33 (4.5) | 33 (4.9) | 5 (4.6) | 1 (5.0) | |||
| American Indian or Alaska Native | 12 (1.6) | 4 (0.6) | 2 (1.8) | 1 (5.0) | |||
| Asian | 10 (1.4) | 5 (0.7) | 0 (0.0) | 0 (0.0) | |||
| Native Hawaiian or Pacific Islander | 1 (0.1) | 1 (0.1) | 0 (0.0) | 0 (0.0) | |||
| ASA, n (%) | < 0.001¶ | < 0.001¶ | < 0.001¶ | ||||
| I | 5 (0.6) | 4 (0.5) | 0 (0.0) | 0 (0.0) | |||
| II | 289 (34.8) | 154 (20.8) | 16 (13.4) | 3 (14.3) | |||
| III | 501 (60.4) | 524 (70.7) | 92 (77.3) | 13 (61.9) | |||
| IV | 35 (4.2) | 58 (7.8) | 11 (9.2) | 5 (23.8) | |||
| Smoker, n (%) | 92 (11.1) | 65 (8.8) | 7 (5.9) | 5 (23.8) | 0.125¶ | 0.083¶ | 0.070 |
| Functional status preoperative, n (%) | 0.001¶ | 0.132¶ | 0.673 | ||||
| Independent | 796 (96.4) | 681 (92.7) | 109 (93.2) | 21 (100.0) | |||
| Partially dependent | 27 (3.3) | 53 (7.2) | 8 (6.8) | 0 (0.0) | |||
| Totally dependent | 3 (0.4) | 1 (0.1) | 0 (0.0) | 0 (0.0) | |||
| Mean age, yrs (SD) | 70.17 (9.91) | 73.09 (9.67) | 75.31 (8.90) | 72.71 (8.70) | < 0.001** | < 0.001** | 0.244** |
| Mean BMI, kg/m2 (SD) | 30.84 (6.86) | 31.04 (7.20) | 31.25 (7.44) | 28.28 (5.69) | 0.558** | 0.543** | 0.090** |
| Mean Length of Stay, days (SD) | 2.47 (2.15) | 2.82 (2.82) | 2.18 (9.73) | 5.10 (6.69) | 0.006** | 0.463** | < 0.001** |
Pearson's chi-squared test.
**Analysis of variance.
Bolding equals significance p < 0.05.
INR, International Normalized Ratio; ASA, American Society of Anesthesiologists; SD, standard deviation; BMI, body mass index.
Compared to patients with an INR of <1.0, those with an INR of >1.0–1.25 were more likely to have CHF (p = 0.012), hypertension (p = 0.012), bleeding disorders (p < 0.001), and prior operation within 30 days (p < 0.001) (Table 2). Relative to patients with an INR of <1.0, those with an INR of >1.25–1.5 were more likely to have CHF (p = 0.004), hypertension (p = 0.028), bleeding disorders (p < 0.001), disseminated cancer (p = 0.032), prior operation within 30 days (p < 0.001), but were less likely to have diabetes (p = 0.043). Compared to patients with an INR of <1.0, those with an INR of >1.5 were more likely to have CHF, dialysis requirement, renal failure, and bleeding disorders (p < 0.001 for all) (Table 2).
Table 2.
Comorbidities and intraoperative variables Among patients undergoing total shoulder arthroplasty.
| Comorbidities | INR <1.0 | INR >1.0-1.25 | INR >1.25-1.5 | INR >1.5 | p-value: INR >1.0-1.25 vs INR <1.0¶ | p-value: INR >1.25-1.5 vs INR <1.0¶ | p-value: INR >1.5 vs INR <1.0¶ |
|---|---|---|---|---|---|---|---|
| Total patients, n | 830 | 742 | 119 | 21 | |||
| COPD, n (%) | 66 (8.0) | 58 (7.8%) | 14 (11.8) | 3 (14.3) | 0.921 | 0.161 | 0.294 |
| CHF, n (%) | 5 (0.6) | 15 (2.0) | 4 (3.4) | 2 (9.5) | 0.012 | 0.004 | < 0.001 |
| Hypertension, n (%) | 582 (70.1) | 562 (75.7) | 95 (79.8) | 15 (71.4) | 0.012 | 0.028 | 0.897 |
| Dialysis, n (%) | 7 (0.8) | 10 (1.3) | 2 (1.7) | 2 (9.5) | 0.334 | 0.378 | < 0.001 |
| Renal failure, n (%) | 0 (0.0) | 2 (0.3) | 0 (0.0) | 1 (4.8) | 0.134 | - | < 0.001 |
| Steroid use, n (%) | 51 (6.1) | 34 (4.6) | 4 (3.4) | 1 (4.8) | 0.172 | 0.224 | 0.794 |
| Bleeding disorder, n (%) | 58 (7.0) | 94 (12.7) | 27 (22.7) | 9 (42.9) | < 0.001 | < 0.001 | < 0.001 |
| Disseminated cancer, n (%) | 5 (0.6) | 4 (0.5) | 3 (2.5) | 0 (0.0) | 0.868 | 0.032 | 0.721 |
| Prior operation within 30 days, n (%) | 0 (0.0) | 1 (0.1) | 1 (0.8) | 0 (0.0) | < 0.001 | < 0.001 | - |
| DM status, n (%) | 0.647 | 0.043 | 0.834 | ||||
| No DM | 669 (80.6) | 584 (78.7) | 100 (84.0) | 16 (76.2) | |||
| Noninsulin-dependent DM | 107 (12.9) | 105 (14.2) | 18 (15.1) | 3 (14.3) | |||
| Insulin-dependent DM | 54 (6.5) | 53 (7.1) | 1 (0.8) | 2 (9.5) | |||
| Dyspnea, n (%) | 0.614 | 0.211 | 0.772 | ||||
| No dyspnea | 754 (90.8) | 678 (91.4) | 102 (85.7) | 20 (95.2) | |||
| Moderate exertion | 71 (8.6) | 57 (7.7) | 16 (13.4) | 1 (4.8) | |||
| At rest | 5 (0.6) | 7 (0.9) | 1 (0.8) | 0 (0.0) | |||
| Anesthesia type, n (%) | 0.022 | 0.337 | 0.929 | ||||
| General | 776 (93.5) | 717 (96.6) | 116 (97.5) | 19 (90.5) | |||
| Neuraxial | 4 (0.5) | 0 (0.0) | 0 (0.0) | 0 (0.0) | |||
| Regional | 45 (5.4) | 20 (2.7) | 2 (1.7) | 2 (9.5) | |||
| MAC | 3 (0.4) | 2 (0.3) | 1 (0.8) | 0 (0.0) | |||
| OR Time > 3 h, n (%) | 83 (10.0) | 55 (7.4) | 14 (11.8) | 3 (14.3) | 0.070 | 0.552 | 0.520 |
Pearson's chi-squared test.
Bolding equals significance p < 0.05.
INR, International Normalized Ratio; COPD, chronic obstructive pulmonary disease; CHF, congestive heart failure; DM, diabetes mellitus; MAC, monitored anesthetic care; OR, operating room.
Complications
On bivariate analysis with patients who had an INR of <1.0 as the reference group, increased INR was associated with a step-wise increase in risk of any complications (8.4% for INR <1.0 vs. 11.6% for INR >1.0–1.25, 16.8% for INR >1.25–1.5, and 23.8% for INR >1.5; p < 0.05 for all; Table 3). Increasing INR was also significantly associated with a step-wise increase in risk of bleeding requiring transfusion (5.1% for INR <1.0 vs. 8.4% for INR >1.0–1.25, 13.4% for INR >1.25–1.5, and 23.8% for INR >1.5; p < 0.05 for all; Table 3). Increasing INR was associated with a significant and step-wise increase in risk of extended length of hospital stay greater than 6 days (3.9% for INR <1.0 vs. 6.1% for INR >1.0–1.25, 10.1% for INR >1.25–1.5, and 19.0% for INR >1.5; p < 0.05 for all; Table 3). Compared to patients with an INR of <1.0, patients with an INR of >1.25–1.5 were more likely to experience pulmonary complications (0.7% vs. 4.2%, p = 0.001), and patients with an INR >1.5 were more likely to have both pulmonary complications (0.7% vs. 4.8% p = 0.043) and cardiac complications (0.5% vs. 4.8%; p = 0.011) (Table 3). Out of the 24 pulmonary complications experienced by patients in the various INR groups, 15 were pneumonia (2 in the INR <1.0 group, 8 in the INR >1.0–1.25 group, 4 in the INR >1.25–1.5 group, 1 in the INR >1.5 group), 6 were reintubations (2 in the INR <1.0 group, 2 in the INR >1.0–1.25 group, 2 in the INR >1.25–1.5 group), and 3 were failure to wean off ventilator (1 in the INR <1.0 group, 1 in the INR >1.0–1.25 group, 1 in the INR >1.25–1.5 group). Pneumonia was defined as meeting the criteria for both radiology and signs/symptoms/laboratory sections. Reintubation was defined as the patient requiring placement of an endotracheal tube or other similar breathing tube and ventilator support following surgery which was not intended or planned. Failure to wean off ventilator was defined as the total duration of ventilator-assisted respirations during postoperative hospitalization was greater than 48 h.
Table 3.
Univariate analysis of postoperative complications of patients following total shoulder arthroplasty.
| Complications | INR <1.0 | INR >1.0-1.25 | p-value: INR >1.0-1.25 vs INR <1.0¶ | INR >1.25-1.5 | p-value: INR >1.25-1.5 vs INR <1.0¶ | INR >1.5 | p-value: INR >1.5 vs INR <1.0¶ |
|---|---|---|---|---|---|---|---|
| Total patients, n | 830 | 742 | 119 | 21 | |||
| Any complication, n (%) | 70 (8.4) | 86 (11.6) | 0.037 | 20 (16.8) | 0.004 | 5 (23.8) | 0.014 |
| Death, n (%) | 3 (0.4) | 4 (0.5) | 0.597 | 1 (0.8) | 0.451 | 0 (0.0) | 0.783 |
| Bleeding requiring transfusion, n (%) | 42 (5.1) | 62 (8.4) | 0.009 | 16 (13.4) | < 0.001 | 5 (23.8) | < 0.001 |
| Wound complication, n (%) | 4 (0.5) | 2 (0.3) | 0.495 | 0 (0.0) | 0.448 | 0 (0.0) | 0.750 |
| Cardiac complication, n (%) | 4 (0.5) | 3 (0.4) | 0.818 | 1 (0.8) | 0.614 | 1 (4.8) | 0.011 |
| Pulmonary complication, n (%) | 6 (0.7) | 12 (1.6) | 0.096 | 5 (4.2) | 0.001 | 1 (4.8) | 0.043 |
| Renal complication, n (%) | 2 (0.2) | 0 (0.0) | 0.181 | 0 (0.0) | 0.592 | 0 (0.0) | 0.822 |
| Thromboembolic complication, n (%) | 5 (0.6) | 5 (0.7) | 0.859 | 2 (1.7) | 0.199 | 0 (0.0) | 0.721 |
| Sepsis complication, n (%) | 7 (0.8) | 3 (0.4) | 0.274 | 0 (0.0) | 0.315 | 0 (0.0) | 0.673 |
| Urinary tract infection, n (%) | 12 (1.4) | 15 (2.0) | 0.380 | 1 (0.8) | 0.595 | 0 (0.0) | 0.579 |
| Extended length of stay (> 6 days), n (%) | 32 (3.9) | 45 (6.1) | 0.043 | 12 (10.1) | 0.003 | 4 (19.0) | 0.001 |
| Readmission, n (%) | 28 (5.4) | 37 (6.8) | 0.343 | 9 (10.3) | 0.077 | 0 (0.0) | 0.297 |
| Reoperation, n (%) | 13 (1.6) | 14 (1.9) | 0.625 | 1 (0.8) | 0.539 | 1 (4.8) | 0.256 |
Pearson's chi-squared test.
Bolding equals significance p < 0.05.
INR, International Normalized Ratio.
Demographic and comorbidity variables included in the multivariable analysis were as follows: age, gender, BMI, ASA classification, smoking and functional status, COPD, CHF, hypertension, dialysis, renal failure, steroid use, bleeding disorder, disseminated cancer, prior operation within 30 days, diabetes, and anesthesia type. Postoperative complications included any complication, bleeding requiring transfusion, cardiac or pulmonary complications, and extended length of stay. On multivariable analysis, increasing INR values was associated with a significant and step-wise increase in the risk of both bleeding requiring transfusion and any complications. Compared to patients with an INR of <1.0, those with INRs of >1.0–1.25, >1.25–1.5, and >1.5 had 1.6-times (95% CI 1.030 to 2.587; p = 0.037), 2.4-times (95% CI 1.163 to 4.783; p = 0.017), and 2.8-times (95% CI 1.428 to 5.357; p = 0.003) higher odds of having bleeding requiring transfusion, respectively (Table 4). Compared to patients with INR <1, those with INRs of >1.25–1.5 and INRs of > 1.5 had 1.9-times (95% CI 1.021 to 3.431; p = 0.043) and 2.1-times (95% CI 1.148 to 3.658; p = 0.015) higher odds of experience any postoperative complications, respectively (p < 0.05 for both; Table 4). Compared to patients with INR <1, those with INRs of > 1.25–1.5 and INRs of > 1.5 had 7.8-times (95% CI 1.835 to 33.230; p = 0.005) and 7.0-times (95% CI 1.729 to 28.446; p = 0.006) higher odds of having pulmonary complications, respectively (p < 0.05 for both; Table 4). There were no statistically significant differences among the various INR groups with regards to thromboembolic complications on multivariable analysis.
Table 4.
Multivariable analysis of postoperative complications of patients following total shoulder arthroplasty.
| Complications | INR >1.0-1.25 | INR >1.25-1.5 | INR >1.5 | |||
|---|---|---|---|---|---|---|
| p-value | Odds ratio (INR >1.0-1.25/ INR <1.0) (95% CI) | p-value | Odds ratio (INR >1.25-1.5/ INR <1.0) (95% CI) | p-value | Odds ratio (INR >1.5/ INR <1.0) (95% CI) | |
| Any complication | 0.270 | 1.236 (0.848 to 1.801) | 0.043 | 1.872 (1.021 to 3.431) | 0.015 | 2.050 (1.148 to 3.658) |
| Bleeding requiring transfusion | 0.037 | 1.633 (1.030 to 2.587) | 0.017 | 2.359 (1.163 to 4.783) | 0.003 | 2.766 (1.428 to 5.357) |
| Cardiac complication | 0.515 | 0.586 (0.117 to 2.937) | 0.506 | 2.462 (0.173 to 35.070) | 0.747 | 1.554 (0.107 to 22.571) |
| Pulmonary complication | 0.405 | 1.575 (0.540 to 4.596) | 0.005 | 7.808 (1.835 to 33.230) | 0.006 | 7.014 (1.729 to 28.446) |
| Extended length of stay (> 6 days) | 0.997 | 0.177 (0.000 to -) | 0.223 | 1.627 (0.744 to 3.557) | 0.999 | 1.542 (0.000 to -) |
Bolding equals significance p < 0.05.
INR, International Normalized Ratio; CI, confidence interval.
Discussion
Understanding the influence of preoperative INR on postoperative complications is important for risk-stratifying TSA patients prior to surgery. Physiologically, elevated INR values represent prolonged time needed for blood to clot, which can lead to higher risks of bleeding and other complications. We found that there was an independent and step-wise increase in risk for postoperative complications with increasing levels of INR relative to a baseline INR of <1.0. The risk of transfusion requirement is significantly increased with an INR > 1.0, and risk for pulmonary and other complications is increased for INR > 1.25.
To reduce perioperative bleeding risks, the American Academy of Orthopedic Surgeons (AAOS) recommends preoperative screening for coagulation status prior to joint replacement procedures. 20 Historically, general surgical guidelines indicate that patients’ target INR prior to surgery must be <1.5, particularly for patients taking anticoagulants.7,20–23 However, the validity of these general surgical guidelines for orthopaedic procedures like total joint arthroplasty continues to be debated. In a retrospective study of 636,231 patients, Tamim et al. showed that INR is a strong predictor of postoperative complications in various types of major surgeries, including orthopaedic surgery, vascular surgery, cardiac surgery, and neurosurgery. 23 In orthopaedic procedures specifically, recent studies have demonstrated increased risk of bleeding, transfusion, infection, and mortality in patients with elevated INR undergoing TKA and THA. Our study advances existing literature by assessing the risk profile of patients with varying degrees of INR elevation who undergo TSA, a procedure whose utilization has increased nearly two-fold from 2012–2017. 8
Rudasill et al. conducted a recent study assessing the influence of different degrees of INR elevation on 30-day complications following TKA in 21,239 patients using the NSQIP database. 6 Similar to our study on TSA, Rudasill et al. reported step-wise increases in bleeding and transfusion risks at an INR > 1.0 compared to patients with INR < 1.0. 6 They further demonstrated that, compared to patients with INR <1.0, an INR of > 1.25 was associated with significantly increased risks of infection and mortality following TKA. 6 Rudasill et al. subsequently conducted a similar analysis assessing INR thresholds in a cohort of 17,567 THA patients. In this study, patients undergoing THA were found to have bleeding risk at INR ≥ 1.25 and increased risk of mortality at INR ≥1.5, relative to a control group of patients with INR <1.0. 7 In contrast to Rudasill et al.'s studies on TKA and THA, we did not find increased INR to be associated with increased risk of infection or, more importantly, an increased risk in mortality following TSA. Even though we did find an association between increased INR and pulmonary complications, our results and Rudasill et al.'s findings collectively suggest that performing shoulder replacement for patients with INR > 1.25 may not carry the same severity of complications compared to performing knee or hip replacement for this patient population. In other words, elevated INR may not have as drastic of an effect on postoperative complications following TSA compared to TKA or THA. Notably, current general surgical guidelines recommend a target preoperative INR of <1.5, however, our study demonstrates that risk of bleeding requiring transfusion following TSA increases even with INR of 1.0 to 1.25. As such, current guidelines for INR risk-stratification may need to be reconsidered to reduce risk of postoperative transfusion requirement, or more consideration should be given to the use of other agents to promote local clotting within current INR guidelines.
Compared to baseline INR of 1.0, we found that increased INR > 1.25 was associated with significantly increased risk of pulmonary complications, including pneumonia, reintubation, or failure to wean off ventilator for more than 48 h. It is possible that the increased risk of pulmonary complications with higher INR values found in our study is driven by the increased risk of blood transfusions. Blood transfusions can cause pulmonary transfusion reactions such as transfusion-related acute lung injury, which is the leading cause of transfusion-related mortality in the United States.24–26 This association has been demonstrated in various general surgery procedures, where intraoperative blood transfusions have been shown to increase risk of pulmonary complications by up to 5.09-times.26–30 Previous studies have not reported pulmonary complications in patients with elevated INR who undergo joint replacement, so more studies are needed to assess the influence of increased INR on postoperative pulmonary complications.6,7
The association between INR and postoperative complications in orthopaedic procedures appears to be related to surgical invasiveness. In a prospective study of 107 patients with elevated preoperative INRs undergoing hand surgery, Edmunds et al. found that one patient suffered a postoperative hematoma, whereas none of the other patients had any postoperative complications. 31 More recent studies have similarly concluded minimal impact of preoperative INR on complications following hand surgery, which is hypothesized to be due to the minimally invasive nature of the majority of hand procedures.31–33 As such, preoperative INR should be more scrutinized prior to more invasive orthopaedic procedures, such as joint replacement and traumatic surgeries, compared to less invasive surgeries such as hand procedures.
Further, the progressively increasing risk of complications and length of stay with higher preoperative INR levels is important from a financial perspective. 34 Patients with unanticipated complications who require longer hospital stays pose substantial financial burdens to the healthcare system.35–37 Bundled payments are an increasingly popular way to adjust for the risk of postoperative complications and hospital stay, ultimately providing for a more effective mechanism of predicting and pricing surgical cases that may require increased resource utilization. 38 Our results suggest that patients with higher preoperative INR values are likely to have higher healthcare utilization and thus have more costly episodes of care that should be appropriately adjusted for in reimbursement plans to reduce inefficient resource allocation.
This study contains several limitations. First, the number of patients in the elevated INR cohorts was substantially less than the number of patients in the control cohort or minimally elevated INR cohorts. However, due to the large overall sample size of the NSQIP database, our total sample size of 1712 patients was sufficient to perform the analysis presented in this study. Additionally, since significant differences were found between INR groups, it is unlikely that a type II error could have occurred, as the sample size in this study was sufficient. This is also true concerning the excluded patients with missing data, as using only the sample of patients with complete data still demonstrated significant differences between groups which limits any possibility of type II error or insufficient sample size. Second, it is possible that patients in this study were on anticoagulants for coagulation disorders, but NSQIP does not record medication data and we were unable to determine patients who were on anticoagulants and if and when these medications were stopped preoperatively. It is important to note that INR may be normal for patients on Factor Xa inhibitors, direct thrombin inhibitors, and anti-platelets, therefore the results of this study should be interpreted in light of these limitations. We were also unable to determine who received Tranexamic Acid, which has increased in popularity over the study period. Even so, knowledge of how INR is associated with postoperative complications, regardless of anticoagulation regimen, remains useful. Another limitation of utilizing the NSQIP database is that it does not report patients who require postoperative venous thromboembolism prophylaxis. It is possible that the prophylaxis is the cause of bleeding rather than the preoperative INR. Since patients with higher INR are more likely to have comorbidities, they would be more likely to receive postoperative thromboprophylaxis. Additionally, the NSQIP database does not indicate whether patients have atrial fibrillation or preoperative venous thromboembolism which are common reasons as to why patients may be on anticoagulants. This can limit the analysis as these medical comorbidities cannot be controlled for during analysis, which may impact the results. A further limitation is that a sizable number of patients were excluded if they had an INR recorded prior to a day of surgery but not within a day of surgery. However, by only including patients with INRs recorded within a day of surgery, we were able to ensure that the preoperative INR documented had the least variation from INR on the day of surgery. Also, some complications in our study, such as pulmonary complications, were quite small in number. This can contribute to the wide confidence intervals observed in our study which can reduce the strength of our conclusions. Further, NSQIP only records bleeding events requiring transfusion up to 72 h postoperatively, therefore future studies should evaluate transfusion as part of the same admission, which can be past the 72 h, to provide more information on the topic. In this study, only 30-day postoperative complications were reported as this is what NSQIP records. However, future studies should examine complications past the 30-day period to elucidate longer-term complications. Another limitation is that INR is only one component to assess one's bleeding risk, and a normal INR does not necessarily signify that there is a normal clotting response. Future studies should examine other variables to determine a patient's risk of bleeding to further strengthen the current study. A further limitation of utilizing INR is that the results of INR measurements are dependent on patients’ compliance with medications and the time of the actual measurement. Consequently, the results of this study should be interpreted with consideration to this limitation. Finally, as with other administrative database investigations, our analyses are limited by potential coding errors within the NSQIP registry.
In conclusion, risk of transfusion requirement is significantly increased with an INR above 1, and risk for pulmonary and other complications is increased with an INR over 1.25. With increasing INR levels, there is an independent and step-wise increase in odd ratios for postoperative complications. Current guidelines for preoperative INR thresholds may need to be adjusted for more predictive risk-stratification for TSA.
Footnotes
Disclosure: Each author certifies that he or she has no commercial associations (e.g. consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article.
Ethics: The study was deemed exempt from institutional review board approval.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iDs: Theodore Quan https://orcid.org/0000-0001-8730-0804
Matthew J Best https://orcid.org/0000-0002-4401-2834
Uma Srikumaran https://orcid.org/0000-0002-2314-8855
References
- 1.Anthony CA, Westermann RW, Gao Yet al. et al. What are risk factors for 30-day morbidity and transfusion in total shoulder arthroplasty? A review of 1922 cases. Clin Orthop Relat Res 2015; 473: 2099–2105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Gruson KI, Accousti KJ, Parsons BOet al. et al. Transfusion after shoulder arthroplasty: an analysis of rates and risk factors. J Shoulder Elbow Surg 2009; 18: 225–230. [DOI] [PubMed] [Google Scholar]
- 3.Hardy JC, Hung M, Snow BJ, et al. Blood transfusion associated with shoulder arthroplasty. J Shoulder Elbow Surg 2013; 22: 233–239. [DOI] [PubMed] [Google Scholar]
- 4.Malcherczyk D, Hack J, Klasan A, et al. Differences in total blood loss and transfusion rate between different indications for shoulder arthroplasty. Int Orthop 2019; 43: 653–658. [DOI] [PubMed] [Google Scholar]
- 5.Mont MA, Jacobs JJ, Boggio LN, et al. Preventing venous thromboembolic disease in patients undergoing elective hip and knee arthroplasty. J Am Acad Orthop Surg 2011; 19: 768–776. [DOI] [PubMed] [Google Scholar]
- 6.Rudasill SE, Liu J, Kamath AF. Revisiting the international normalized ratio (INR) threshold for complications in primary total knee arthroplasty: an analysis of 21,239 cases. J Bone Joint Surg Am 2019; 101: 514–522. [DOI] [PubMed] [Google Scholar]
- 7.Rudasill SE, Liu J, Kamath AF. Revisiting the international normalized ratio threshold for bleeding risk and mortality in primary total hip arthroplasty: a national surgical quality improvement program analysis of 17,567 patients. J Bone Joint Surg Am 2020; 102: 52–59. [DOI] [PubMed] [Google Scholar]
- 8.Best MJ, Aziz KT, Wilckens JHet al. et al. Increasing incidence of primary reverse and anatomic total shoulder arthroplasty in the United States. J Shoulder Elbow Surg 2021; 30: 1159–1166. [DOI] [PubMed] [Google Scholar]
- 9.Kurtz SM, Lau E, Ong Ket al. et al. Future young patient demand for primary and revision joint replacement: national projections from 2010 to 2030. Clin Orthop Relat Res 2009; 467: 2606–2612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kremers H M, Larson DR, Crowson CS, et al. Prevalence of total hip and knee replacement in the United States. J Bone Joint Surg Am 2015; 97: 1386–1397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shikdar S, Vashisht R, Bhattacharya PT. International normalized ratio (INR). In: Statpearls. Treasure Island (FL): StatPearls Publishing, May 10, 2021. [PubMed] [Google Scholar]
- 12.Belli S, Aytac HO, Yabanoglu H, et al. Results of surgery in general surgical patients receiving warfarin: retrospective analysis of 61 patients. Int Surg 2015; 100: 225–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bernstein DN, Keswani A, Ring D. Perioperative risk adjustment for total shoulder arthroplasty: are simple clinically driven models sufficient? Clin Orthop Relat Res 2017; 475: 2867–2874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bovonratwet P, Malpani R, Ottesen TD, et al. Aseptic revision total hip arthroplasty in the elderly: quantifying the risks for patients over 80 years old. Bone Joint J 2018; 100-B: 143–151. [DOI] [PubMed] [Google Scholar]
- 15.Bohl DD, Ondeck NT, Darrith Bet al. et al. Impact of operative time on adverse events following primary total joint arthroplasty. J Arthroplasty 2018; 33: 2256–2262.e4. [DOI] [PubMed] [Google Scholar]
- 16.Surace P, Sultan AA, George J, et al. The association between operative time and short-term complications in total hip arthroplasty: an analysis of 89,802 surgeries. J Arthroplasty 2019; 34: 426–432. [DOI] [PubMed] [Google Scholar]
- 17.Tornero E, García-Oltra E, García-Ramiro S, et al. Prosthetic joint infections due to Staphylococcus aureus and coagulase-negative staphylococci. Int J Artif Organs 2012; 35: 884–892. [DOI] [PubMed] [Google Scholar]
- 18.Trickey AW, Wright JM, Donovan J, et al. Interrater reliability of hospital readmission evaluations for surgical patients. Am J Med Qual 2017; 32: 201–207. [DOI] [PubMed] [Google Scholar]
- 19.Tihista M, Gu A, Wei Cet al. et al. The impact of long-term corticosteroid use on acute postoperative complications following lumbar decompression surgery. J Clin Orthop Trauma 2020; 11: 921–927. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kandil A, Griffin JW, Novicoff WMet al. et al. Blood transfusion after total shoulder arthroplasty: which patients are at high risk? Int J Shoulder Surg 2016; 10: 72–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Blasdale C, Lawrence CM. Perioperative international normalized ratio level is a poor predictor of postoperative bleeding complications in dermatological surgery patients taking warfarin. Br J Dermatol 2008; 158: 522–526. [DOI] [PubMed] [Google Scholar]
- 22.Sølbeck S, Ostrowski SR, Johansson PI. A review of the clinical utility of INR to monitor and guide administration of prothrombin complex concentrate to orally anticoagulated patients. Thromb J 2012; 10:5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Tamim H, Habbal M, Saliba A, et al. Preoperative INR and postoperative major bleeding and mortality: a retrospective cohort study. J Thromb Thrombolysis 2016; 41: 301–311. [DOI] [PubMed] [Google Scholar]
- 24.Cho MS, Modi P, Sharma S. Transfusion-related acute lung injury. In: Statpearls. Treasure Island (FL): StatPearls Publishing, July 26, 2021. [PubMed] [Google Scholar]
- 25.Goldman M, Webert KE, Arnold DM, et al. Proceedings of a consensus conference: towards an understanding of TRALI. Transfus Med Rev 2005; 19: 2–31. [DOI] [PubMed] [Google Scholar]
- 26.Pessaux P, van den Broek MA, Wu T, et al. Identification and validation of risk factors for postoperative infectious complications following hepatectomy. J Gastrointest Surg 2013; 17: 1907–1916. [DOI] [PubMed] [Google Scholar]
- 27.Choudhuri AH, Chandra S, Aggarwal Get al. et al. Predictors of postoperative pulmonary complications after liver resection: results from a tertiary care intensive care unit. Indian J Crit Care Med 2014; 18: 358–362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Chughtai M, Gwam CU, Mohamed N, et al. The epidemiology and risk factors for postoperative pneumonia. J Clin Med Res 2017; 9: 466–475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Miki Y, Makuuchi R, Tokunaga M, et al. Risk factors for postoperative pneumonia after gastrectomy for gastric cancer. Surg Today 2016; 46: 552–556. [DOI] [PubMed] [Google Scholar]
- 30.Nobili C, Marzano E, Oussoultzoglou E, et al. Multivariate analysis of risk factors for pulmonary complications after hepatic resection. Ann Surg 2012; 255: 540–550. [DOI] [PubMed] [Google Scholar]
- 31.Edmunds I, Avakian Z. Hand surgery on anticoagulated patients: a prospective study of 121 operations. Hand Surg 2010; 15: 109–113. [DOI] [PubMed] [Google Scholar]
- 32.Twoon M, Hallam MJ. Anticoagulation therapy and hand surgery: do we worry too much? Indian J Plast Surg 2015; 48: 326–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wallace DL, Latimer MD, Belcher HJ. Stopping warfarin therapy is unnecessary for hand surgery. J Hand Surg Br 2004; 29: 201–203. [DOI] [PubMed] [Google Scholar]
- 34.Foley M, Kifaieh N, Mallon WK. Financial impact of emergency department crowding. West J Emerg Med 2011; 12: 192–197. [PMC free article] [PubMed] [Google Scholar]
- 35.Akinleye DD, McNutt LA, Lazariu Vet al. et al. Correlation between hospital finances and quality and safety of patient care. PLoS One 2019; 14: e0219124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Bayley MD, Schwartz JS, Shofer FS, et al. The financial burden of emergency department congestion and hospital crowding for chest pain patients awaiting admission. Ann Emerg Med 2005; 45: 110–117. [DOI] [PubMed] [Google Scholar]
- 37.Krochmal P, Riley TA. Increased health care costs associated with ED overcrowding. Am J Emerg Med 1994; 12: 265–266. [DOI] [PubMed] [Google Scholar]
- 38.Hardin L, Kilian A, Murphy E. Bundled payments for care improvement: preparing for the medical diagnosis-related groups. J Nurs Adm 2017; 47: 313–319. [DOI] [PMC free article] [PubMed] [Google Scholar]