Authors: Kenny Ling, Richelle P. Fassler, Andrew J. Nicholson, David E. Komatsu, Edward D. Wang
Categories: Shoulder, Age, Mortality, Nonagenarian, Octogenarian, Readmission, Septuagenarian, Total shoulder arthroplasty
Source: JSES International
Authors: Kenny Ling, Richelle P. Fassler, Andrew J. Nicholson, David E. Komatsu, Edward D. Wang
Increased age is a well-known risk factor for development of osteoarthritis. Total shoulder arthroplasty (TSA) is a common treatment option for patients with severe glenohumeral osteoarthritis. The purpose of this study was to investigate the association between the septuagenarian, octogenarian, and nonagenarian populations and postoperative outcomes following TSA.
The American College of Surgeons National Surgical Quality Improvement Program database was queried for all patients who underwent TSA between 2015 and 2020. Patients were divided into cohorts based on sexagenarians (60-69), septuagenarians (70-79), octogenarians (80-89), and nonagenarians (90+). Multivariate logistic regression was used to identify associations between age and postoperative complications.
On bivariate analysis, compared to sexagenarians, septuagenarians were significantly associated with higher rates of myocardial infarction (P = .038), blood transfusion (P < .001), organ/space surgical site infection (P = .048), readmission (P = .005), and nonhome discharge (P < .001. Compared to septuagenarians, octogenarians were significantly associated with higher rates of urinary tract infection (P < .001), blood transfusion (P < .001), readmission (P = .002), non-home discharge (P < .001), and mortality (P = .027). Compared to octogenarians, nonagenarians were significantly associated with higher rates of sepsis (P = .013), pneumonia (P = .003), reintubation (P = .009), myocardial infarction (P < .001), blood transfusion (P < .001), readmission (P = .026), nonhome discharge (P < .001), and mortality (P < .001).
From age 60, each decade of age was identified to be an increasingly significant predictor for blood transfusion, readmission, and nonhome discharge following TSA. From age 70, each decade of age was additionally identified to be an increasingly significant predictor for mortality.
Glenohumeral osteoarthritis affects up to 17% of patients with shoulder pain, greatly hindering quality of life and impairing activities of daily living.^15^^,^^25^ Its prevalence increases with age, with radiographic evidence noted in 32.8% of individuals aged older than 60.^16^ The aging population is on the rise, with the proportion of Americans over age 65 expected to increase from 16% to 23% by 2060.^26^ As the US population ages and osteoarthritis becomes more prevalent, it is important to optimize the management of individuals who choose to undergo surgical treatment.
Total shoulder arthroplasty (TSA), including both anatomic and reverse TSA, is a common treatment option for patients with severe glenohumeral osteoarthritis.^15^ TSA also may be considered to treat patients with massive rotator cuff tears or proximal humerus fractures, although primary osteoarthritis remains the most common indication for TSA.^12^^,^^18^^,^^39^ Specifically, reverse TSA has proven to be particularly beneficial to treat osteoarthritis in the elderly, due to severe bone loss and joint deformation.^2^ Overall, joint replacement has shown to significantly decrease pain and improve function of the shoulder joint in patients with a history of osteoarthritis, especially among older individuals.^25^
Increased age is a well-known risk factor for development of osteoarthritis, along with female sex, genetics, past trauma, and obesity.^31^ Since arthritis is common in the aging population, the safety and efficacy of TSA as a treatment option is valuable to investigate. Prior studies on joint replacement surgery among older patients compared to younger patients have reported increased risk of multiple medical complications, longer hospital stays, readmission, and mortality.^4^^,^9, 10, 11^,^^19^^,^^21^ However, understanding how each increased decade of life affects adverse outcomes has not been explored. Further investigation to stratify postoperative outcomes within older age groups may help to better understand as to which adverse outcomes certain patient groups are at higher risk.
The purpose of this study was to investigate the association between the septuagenarian, octogenarian, and nonagenarian populations and postoperative outcomes following TSA. We hypothesized that increased age will be associated with increased rate of postoperative adverse outcomes.
The American College of Surgeons National Surgical Quality Improvement Program database was queried for all patients who underwent TSA between 2015 and 2020. The NSQIP database is fully deidentified, therefore rendering this study exempt from approval by our University’s institutional review board. Data in the NSQIP database are obtained from over 600 hospitals in the United States and are collected by trained surgical clinical reviewers. The data are periodically audited to maintain high fidelity.
Current Procedural Terminology code 23472 was used to identify patients who underwent TSA, both anatomic and reverse, from 2015 to 2020. Cases for patients younger than 18 years of age or TSA performed for trauma were automatically excluded from the database. Cases were excluded if any of the following variables had missing age, height, weight, functional status, discharge destination, American Society of Anesthesiologists (ASA) classification. Cases were also excluded for age <60.
Variables collected in this study included patient demographics, comorbidities, surgical characteristics, preoperative laboratory values, and 30-day postoperative complication data. Patient demographics included age, body mass index (BMI), gender, functional status, ASA classification, and smoking status. Preoperative comorbidities included insulin-dependent and noninsulin dependent diabetes, severe chronic obstructive pulmonary disease (COPD), hypertension requiring medication, bleeding disorders, open wound/wound infection, disseminated cancer, and congestive heart failure (CHF). Surgical characteristics included operative duration in minutes. Preoperative laboratory values included hematocrit to assess for preoperative anemia. Postoperative complications within 30 days included sepsis, septic shock, pneumonia, reintubation, urinary tract infection (UTI), stroke, cardiac arrest, myocardial infarction, blood transfusion, deep vein thrombosis, pulmonary embolism, failure to wean off ventilator, deep incisional surgical-site infection (SSI), superficial incisional SSI, organ/space SSI, wound dehiscence, readmission, reoperation, nonhome discharge, and mortality.
The initial pool of patients was divided into cohorts based on sexagenarians (60-69), septuagenarians (70-79), octogenarians (80-89), nonagenarians (90+). Of note, the NSQIP database codes for all patients over the age of 90 as “90+.” Therefore, it was possible that the nonagenarian cohort included patients who were older than the nonagenarian range of 90-99. Three sets of analyses were performed, such that sexagenarians, septuagenarians, and octogenarians each served as the reference cohort for patients older than the reference cohort.
A total of 27,050 patients who underwent primary TSA were identified in NSQIP from 2015 to 2020. Cases were excluded as 152 for missing height/weight, 11 for missing discharge destination, 29 for missing ASA classification, 2 for missing readmission status, 227 for missing functional health status prior to surgery, 3736 for age <60. Of the 26,629 patients remaining after exclusion criteria, 9085 (39.7%) patients were included in the sexagenarian cohort, 10,307 (45.0%) in the septuagenarian cohort, 3335 (14.6%) in the octogenarian cohort, and 166 (0.7%) in the nonagenarian cohort.
All statistical analyses were conducted using SPSS Software version 29.0 (IBM Corp., Armonk, NY, USA). Patient demographics and comorbidities were compared between cohorts using bivariate logistic regression. Postoperative complications were also compared between cohorts using bivariate logistic regression.
Multivariate logistic regression, adjusted for all significantly associated patient demographics and comorbidities, was used to identify associations between age and postoperative complications. Odds ratios (OR) were reported with 95% confidence intervals (CI). The level of statistical significance was set at P < .05.
Compared to sexagenarians, the patient demographics and comorbidities significantly associated with septuagenarians were BMI 18.5-29.9 (P < .001), female gender (P < .001), dependent functional status (P < .001), ASA ≥3 (P < .001), nonsmoker status (P < .001), hypertension (P < .001), COPD (P = .010), bleeding disorder (P < .001), preoperative anemia (P < .001), and operative duration 0-79 minutes (P < .001) (Table I). The patient demographics and comorbidities significantly associated with octogenarians were BMI 18.5-29.9 (P < .001), female gender (P < .001), dependent functional status (P < .001), ASA ≥3 (P < .001), nonsmoker status (P < .001), hypertension (P < .001), COPD (P = .039), bleeding disorder (P < .001), no chronic steroid use (P = .039), CHF (P = .007), preoperative anemia (P < .001), and operative duration 0-79 minutes (P < .001). The patient demographics and comorbidities significantly associated with nonagenarians were BMI 18.5-29.9 (P < .001), female gender (P < .001), dependent functional status (P < .001), ASA ≥3 (P < .001), non-smoker status (P < .001), hypertension (P < .001), CHF (P < .001), preoperative anemia (P < .001), and operative duration 0-79 minutes (P = .008).Table IPatient demographics/comorbidities based on age for patients who underwent total shoulder arthroplasty between 2015 and 2020, with age 60-69 as the reference group.CharacteristicAge 60-69Age 70-79Age 80-89Age 90+NumberPercentNumberPercentP valueNumberPercentP-valueNumberPercentP valueTotal9085100.010,307100.03335100.0166100.0Body mass index (kg/m^2^)<.001****<.001****<.001 <18.5660.7690.7331.063.6 18.5-29.9386342.5512249.7210063.012675.9 30-34.9242026.6278127.081024.32414.5 35-39.9152516.8139713.62567.763.6 ≥40120113.29218.91303.942.4Gender**<.001****<.001****<.001** Female481052.9606158.8224567.313279.5 Male427547.1424641.2109032.73420.5Functional status**<.001****<.001****<.001** Independent895398.510,09397.9318595.514486.7 Dependent1321.52142.11504.52213.3ASA classification**<.001****<.001****<.001** 1-2425146.8395438.4%100730.23822.9 ≥3483453.2635361.6232869.812877.1Smoker**<.001****<.001****<.001** No790387.0977894.9324897.416498.8 Yes118213.05295.1872.621.2Diabetes mellitus.090.056.120 No diabetes739681.4826680.2280684.114386.1 Non-insulin dependent118713.1145914.240012.0159.0 Insulin dependent5025.55825.61293.984.8Hypertension**<.001****<.001****<.001** No336537.0284127.676823.03923.5 Yes572063.0746672.4256777.012776.5COPD**.010****.039**.146 No851793.7956792.8309292.715191.0 Yes5686.37407.22437.3159.0Bleeding disorders**<.001****<.001**.056 No890098.010,01397.1319795.915995.8 Yes1852.02942.91384.174.2Chronic steroid use.808**.015**.425 No863395.0980295.1320496.116096.4 Yes4525.05054.91313.963.6Open wound/wound infection.525.197.056 No905899.710,27199.7332099.616498.8 Yes270.3360.3150.421.2Disseminated cancer.907.697.998 No906499.810,28499.8332699.7166100.0 Yes210.2230.290.300.0Congestive heart failure.087**.007****<.001** No903699.510,23199.3330299.015895.2 Yes490.5760.7331.084.8Preoperative anemia**<.001****<.001****<.001** No686975.6746772.4218465.59456.6 Yes114712.6177817.387226.16740.4Operative duration (minutes)<.001****<.001****.008 0-79216423.8284127.6103731.15432.5 80-128453950.0520450.5165749.78350.0 ≥129238226.2226221.964119.22917.5ASA, American Society of Anesthesiologists; COPD, chronic obstructive pulmonary disease.Bold P values indicate statistical significance with P < .05.
Bivariate analysis identified postoperative complications significantly associated with each age cohort, with reference to the sexagenarian cohort (Table II). Septuagenarians had significantly higher rates of myocardial infarction (P = .038), blood transfusion (P < .001), organ/space SSI (P = .048), readmission (P = .005), and nonhome discharge (P < .001). Octogenarians had significantly higher rates of septic shock (P = .033), pneumonia (P = .023), UTI (P < .001), stroke (P = .050), blood transfusion (P < .001), deep vein thrombosis (P = .038), failure to wean off ventilator (P = .013), readmission (P < .001), nonhome discharge (P < .001), and mortality (P = .016). Nonagenarians had significantly higher rates of sepsis (P = .008), pneumonia (P < .001), reintubation (P < .001), UTI (P < .001), myocardial infarction (P < .001), blood transfusion (P < .001), pulmonary embolism (P = .044), failure to wean off ventilator (P = .019), readmission (P < .001), nonhome discharge (P < .001), and mortality (P < .001).Table IIBivariate analysis of 30-day postoperative complications based on age group, with age 60-69 as the reference group.Postoperative complicationAge 60-69Age 70-79Age 80-89Age 90+NumberPercentNumberPercentP valueNumberPercentP-valueNumberPercentP valueSepsis150.17100.10.19350.15.85221.20**.008Septic shock10.0170.07.08940.12.03300.001.000Pneumonia360.40530.51.227240.72.02353.01<.001Reintubation130.14140.14.984100.30.05921.20<.001Urinary tract infection510.56710.69.263441.32<.00153.01<.001Stroke50.0690.09.40860.18.05000.00.999Cardiac arrest40.0440.04.85840.12.15600.00.999Myocardial infarction180.20370.36.03880.24.65242.41<.001Blood transfusions880.972001.94<.0011253.75<.0011810.84<.001Deep vein thrombosis240.26300.29.723170.51.03800.00.998Pulmonary embolism250.28280.27.963120.36.44421.20.044Failure to wean off ventilator40.04120.12.09270.21.01310.60.019Deep incisional SSI90.1050.05.20000.00.99900.00.999Superficial incisional SSI250.28190.18.188100.30.81810.60.442Organ/space SSI240.26140.14.04840.12.14300.00.998Wound dehiscence60.0730.03.24620.06.90600.00.999Readmission2182.403153.06.0051394.17<.001137.83<.001Reoperation1291.421231.19.165501.50.74210.60.390Nonhome discharge4164.589118.84<.00176122.82<.0018048.19<.001Mortality90.10120.12.714100.30.01642.41<.001**SSI, surgical site infection.Bold P values indicate statistical significance with P < .05.
After adjusting for the patient variables significantly associated with each age cohort, multivariate logistic regression identified the complications independently associated with each cohort, with reference to the sexagenarian cohort (Table III). Septuagenarians were independently associated with higher rates of blood transfusion (OR 1.67, 95% CI 1.27-2.19; P < .001), readmission (OR 1.24, 95% CI 1.02-1.50; P = .028), and non-home discharge (OR 1.82, 95% CI 1.59-2.08; P < .001). Octogenarians were independently associated with higher rates of UTI (OR 1.73, 95% CI 1.09-2.76; P = .021), blood transfusion (OR 2.13, 95% 1.53-2.97; P < .001), failure to wean off ventilator (OR 7.67, 95% CI 1.70-34.72; P = .008), readmission (OR 1.39, 95% CI 1.07-1.81; P = .012), and nonhome discharge (OR 5.47, 95% CI 4.67-6.40; P < .001). Nonagenarians were independently associated with higher rates of pneumonia (OR 5.05, 95% CI 1.72-14.81; P = .003), reintubation (OR 4.53, 95% CI 1.02-20.17; P = .048), UTI (OR 3.26, 95% CI 1.12-9.42; P = .030), myocardial infarction (OR 6.30, 95% CI 1.59-25.01; P = .009), blood transfusion (OR 3.07, 95% CI 1.52-6.18; P = .002), nonhome discharge (OR 10.94, 95% CI 7.33-16.32; P < .001), and mortality (OR 5.80, 95% CI 1.31-25.59; P = .020).Table IIIMultivariate analysis of 30-day postoperative complications based on age group, adjusted for significantly associated patient demographics/comorbidities, with age 60-69 as the reference group.Postoperative complicationAge 70-79Age 80-89Age 90+OR95% CIP valueOR95% CIP-valueOR95% CIP valueSepsis------3.950.74-21.02.107Septic shock---7.840.83-73.86.072---Pneumonia---1.600.89-2.90.1185.051.72-14.81**.003Reintubation------4.531.02-20.17.048Urinary tract infection---1.731.09-2.76.0213.261.12-9.42.030Stroke---1.20.33-4.39.780---Myocardial infarction1.510.83-2.77.179---6.301.59-25.01.009Blood transfusions1.671.27-2.19<.0012.131.53-2.97<.0013.071.52-6.18.002Deep vein thrombosis---1.820.86-3.86.12---Pulmonary embolism------3.960.73-21.43.110Failure to wean off ventilator---7.671.70-34.72.0088.560.43-172.07.161Organ/space SSI0.540.26-1.13.101------Readmission1.241.02-1.50.0281.391.07-1.81.0121.900.97-3.74.063Non-home discharge1.821.59-2.08<.0015.474.67-6.40<.00110.947.33-16.32<.001Mortality---1.610.58-4.49.3625.801.31-25.59.020**OR, odds ratio; CI, confidence interval; SSI, surgical site infection.Bold P values indicate statistical significance with P < .05.
Compared to septuagenarians, the patient demographics and comorbidities significantly associated with octogenarians were BMI 18.5-29.9 (P < .001), female gender (P < .001), dependent functional status (P < .001), ASA ≥3 (P < .001), nonsmoker status (P < .001), no diabetes (P < .001), hypertension (P < .001), bleeding disorder (P < .001), no chronic steroid use (P = .021), preoperative anemia (P < .001), and operative duration 0-79 minutes (P < .001) (Table IV). The patient demographics and comorbidities significantly associated with nonagenarians were BMI 18.5-29.9 (P < .001), female gender (P < .001), dependent functional status (P < .001), ASA ≥3 (P < .001), nonsmoker status (P = .037), no CHF (P < .001), and preoperative anemia (P < .001).Table IVPatient demographics/comorbidities based on age for patients who underwent total shoulder arthroplasty between 2015 and 2020, with age 70-79 as the reference group.CharacteristicAge 70-79Age 80-89Age 90+NumberPercentNumberPercentP valueNumberPercentP valueTotal10,307100.03335100.0166100.0Body mass index (kg/m^2^)<.001****<.001 <18.5690.7331.063.6 18.5-29.9512249.7210063.012675.9 30-34.9278127.081024.32414.5 35-39.9139713.62567.763.6 ≥409218.91303.942.4Gender**<.001****<.001** Female606158.8224567.313279.5 Male424641.2109032.73420.5Functional status**<.001****<.001** Independent10,09397.9318595.514486.7 Dependent2142.11504.52213.3ASA classification**<.001****<.001** 1-2395438.4100730.23822.9 ≥3635361.6232869.812877.1Smoker**<.001****.037** No977894.9324897.416498.8 Yes5295.1872.621.2Diabetes mellitus**<.001**.144 No diabetes826680.2280684.114386.1 Non-insulin dependent145914.240012.0159.0 Insulin dependent5825.61293.984.8Hypertension**<.001**.245 No284127.676823.03923.5 Yes746672.4256777.012776.5COPD.836.360 No956792.8309292.715191.0 Yes7407.22437.3159.0Bleeding disorders**<.001**.300 No10,01397.1319795.915995.8 Yes2942.91384.174.2Chronic steroid use**.021**.448 No980295.1320496.116096.4 Yes5054.91313.963.6Open wound/wound infection.410.088 No10,27199.7332099.616498.8 Yes360.3150.421.2Disseminated cancer.628.998 No10,28499.8332699.7166100.0 Yes230.290.300.0Congestive heart failure.157**<.001** No10,23199.3330299.015895.2 Yes760.7331.084.8Preoperative anemia**<.001****<.001** No746772.4218465.59456.6 Yes177817.387226.16740.4Operative duration (minutes)<.001.230 0-79284127.6103731.15432.5 80-128520450.5165749.78350.0 ≥129226221.964119.22917.5ASA, American Society of Anesthesiologists; COPD, chronic obstructive pulmonary disease.Bold P values indicate statistical significance with P < .05.
Bivariate analysis identified postoperative complications significantly associated with each age cohort, with reference to the septuagenarian cohort (Table V). Octogenarians had significantly higher rates of UTI (P < .001), blood transfusion (P < .001), readmission (P = .002), nonhome discharge (P < .001), and mortality (P = .027). Nonagenarians had significantly higher rates of sepsis (P < .001), pneumonia (P < .001), reintubation (P < .001), UTI (P = .001), myocardial infarction (P < .001), blood transfusion (P < .001), pulmonary embolism (P = .042), readmission (P < .001), nonhome discharge (P < .001), and mortality (P < .001).Table VBivariate analysis of 30-day postoperative complications based on age group, with age 70-79 as the reference group.Postoperative complicationAge 70-79Age 80-89Age 90+NumberPercentNumberPercentP valueNumberPercentP valueSepsis100.1050.15.42721.20**.001Septic shock70.0740.12.36400.00.999Pneumonia530.51240.72.17153.01<.001Reintubation140.14100.30.05421.20<.001Urinary tract infection710.69441.32<.00153.01.001Stroke90.0960.18.17000.00.999Cardiac arrest40.0440.12.11000.00.999Myocardial infarction370.3680.24.30042.41<.001Blood transfusions2001.941253.75<.0011810.84<.001Deep vein thrombosis300.29170.51.06400.00.998Pulmonary embolism280.27120.36.41521.20.042Failure to wean off ventilator120.1270.21.21510.60.114Deep incisional SSI50.0500.00.99900.00.999Superficial incisional SSI190.18100.30.21310.60.248Organ/space SSI140.1440.12.82600.00.999Wound dehiscence30.0320.06.42800.00.999Readmission3153.061394.17.002137.83<.001Reoperation1231.19501.50.17110.60.494Non-home discharge9118.8476122.82<.0018048.19<.001Mortality120.12100.30.02742.41<.001**SSI, surgical site infection.Bold P values indicate statistical significance with P < .05.
After adjusting for the patient variables significantly associated with each age cohort, multivariate logistic regression identified the complications independently associated with each cohort, with reference to the septuagenarian cohort (Table VI). Octogenarians were independently associated with higher rates of UTI (OR 1.73, 95% CI 1.15-2.61; P = .009), readmission (OR 1.28, 95% CI 1.03-1.59; P = .029), and nonhome discharge (OR 2.81, 95% CI 2.50-3.17; P < .001). Nonagenarians were independently associated with higher rates of pneumonia (OR 5.02, 95% CI 1.85-13.63; P = .002), reintubation (OR 4.91, 95% CI 1.08-22.28; P = .039), UTI (OR 2.94, 95% CI 1.07-8.08; P = .037), myocardial infarction (OR 3.83, 95% CI 1.17-12.56; P = .027), blood transfusion (OR 1.86, 95% CI 1.03-3.35; P = .038), readmission (OR 2.17, 95% CI 1.17-4.01; P = .014), nonhome discharge (OR 12.80, 95% CI 1.28-128.40; P = .030), and mortality (OR 9.87, 95% CI 2.62-37.16; P < .001).Table VIMultivariate analysis of 30-day postoperative complications based on age group, adjusted for significantly associated patient demographics/comorbidities, with age 70-79 as the reference group.Postoperative complicationAge 80-89Age 90+OR95 CIP valueOR95% CIP valueSepsis---6.820.94-49.29.057Pneumonia---5.021.85-13.63**.002Reintubation---4.911.08-22.28.039Urinary tract infection1.731.15-2.61.0092.941.07-8.08.037Myocardial infarction---3.831.17-12.56.027Blood transfusions1.220.95-1.57.1151.861.03-3.35.038Pulmonary embolism---4.160.89-19.39.069Readmission1.281.03-1.59.0292.171.17-4.01.014Reoperation------Non-home discharge2.812.50-3.17<.00112.801.28-128.40.030Mortality2.400.97-5.89.0579.872.62-37.16<.001**OR, odds ratio; CI, confidence interval.Bold P-values indicate statistical significance with P < .05.
Compared to octogenarians, the patient demographics and comorbidities significantly associated with nonagenarians were BMI 18.5-29.9 (P < .001), female gender (P < .001), dependent functional status (P < .001), ASA ≥3 (P = .046), CHF (P < .001), and preoperative anemia (P < .001) (Table VII).Table VIIPatient demographics/comorbidities based on age for patients who underwent total shoulder arthroplasty between 2015 and 2020, with age 80-89 as the reference group.CharacteristicAge 80-89Age 90+NumberPercentNumberPercentP valueTotal3335100.0166100.0Body mass index (kg/m^2^)<.001 <18.5331.063.6 18.5-29.9210063.012675.9 30-34.981024.32414.5 35-39.92567.763.6 ≥401303.942.4Gender**<.001** Female224567.313279.5 Male109032.73420.5Functional status**<.001** Independent318595.514486.7 Dependent1504.52213.3ASA classification**.046** 1-2100730.23822.9 ≥3232869.812877.1Smoker.274 No324897.416498.8 Yes872.621.2Diabetes mellitus.785 No diabetes280684.114386.1 Non-insulin dependent40012.0159.0 Insulin dependent1293.984.8Hypertension.889 No76823.03923.5 Yes256777.012776.5COPD.401 No309292.715191.0 Yes2437.3159.0Bleeding disorders.960 No319795.915995.8 Yes1384.174.2Chronic steroid use.839 No320496.116096.4 Yes1313.963.6Open wound/wound infection.190 No332099.616498.8 Yes150.421.2Disseminated cancer.999 No332699.7166100.0 Yes90.300.0Congestive heart failure**<.001** No330299.015895.2 Yes331.084.8Preoperative anemia**<.001** No218465.59456.6 Yes87226.16740.4Operative duration (min).835 0-79103731.15432.5 80-128165749.78350.0 ≥12964119.22917.5ASA, American Society of Anesthesiologists; COPD, chronic obstructive pulmonary disease.Bold P values indicate statistical significance with P < .05.
Bivariate analysis identified postoperative complications significantly associated with nonagenarians, with reference to the octagenarian cohort (Table VIII). Nonagenarians had significantly higher rates of sepsis (P = .013), pneumonia (P = .003), reintubation (P = .009), myocardial infarction (P < .001), blood transfusion (P < .001), readmission (P = .026), nonhome discharge (P < .001), and mortality (P < .001).Table VIIIBivariate analysis of 30-day postoperative complications based on age group, with age 80-89 as the reference group.Postoperative complicationAge 80-89Age 90+NumberPercentNumberPercentP valueSepsis50.1521.20**.013Septic shock40.1200.00.999Pneumonia240.7253.01.003Reintubation100.3021.20.009Urinary tract infection441.3253.01.078Stroke60.1800.00.999Cardiac arrest40.1200.00.999Myocardial infarction80.2442.41<.001Blood transfusions1253.751810.84<.001Deep vein thrombosis170.5100.00.999Pulmonary embolism120.3621.20.113Failure to wean off ventilator70.2110.60.324Deep incisional SSI00.0000.00-Superficial incisional SSI100.3010.60.505Organ/space SSI40.1200.00.999Wound dehiscence20.0600.00.999Readmission1394.17137.83.026Reoperation501.5010.60.363Non-home discharge76122.828048.19<.001Mortality100.3042.41<.001**SSI, surgical site infection.Bold P values indicate statistical significance with P < .05.
After adjusting for the patient variables significantly associated with the nonagenarian cohort, multivariate logistic regression identified the complications independently associated with the nonagenarian cohort, with reference to octogenarians (Table IX). Nonagenarians were independently associated with higher rates of sepsis (OR 8.37, 95% CI 1.42-49.29; P = .019), pneumonia (OR 3.92, 95% CI 1.41-10.91; P = .009), reintubation (OR 5.48, 95% CI 1.45-20.77; P = .012), myocardial infarction (OR 10.24, 95% CI 2.76-38.03; P < .001), blood transfusion (OR 1.84, 95% CI 1.04-3.27; P = .037), nonhome discharge (OR 2.36, 95% CI 1.68-3.32; P < .001), and mortality (OR 4.90, 95% CI 1.38-17.36; P = .014).Table IXMultivariate analysis of 30-day postoperative complications based on age group, adjusted for significantly associated patient demographics/comorbidities, with age 80-89 as the reference group.Postoperative complicationAge 90+OR95% CIP valueSepsis8.371.42-49.29**.019Pneumonia3.921.41-10.91.009Reintubation5.481.45-20.77.012Myocardial infarction10.242.76-38.03<.001Blood transfusions1.841.04-3.27.037Readmission1.640.89-3.05.115Nonhome discharge2.361.68-3.32<.001Mortality4.901.38-17.36.014**OR, odds ratio; CI, confidence interval.Bold P values indicate statistical significance with P < .05.
In this study, we used a large national database to investigate the association between an aging population and 30-day postoperative outcomes in patients undergoing TSA between 2015 and 2020. By comparing each decade of life to an increasingly older age group, we were able to analyze the strength of significance between each age group and adverse postoperative outcomes. We identified age to be an increasingly significant predictor for blood transfusions, readmission, and nonhome discharge from age 60 and older. From the septuagenarian cohort, each increasing decade of age was found to be a significant predictor of mortality.
Arthritis is a major cause of disability in the older population due to a multitude of risk factors along with biological changes in joint structure with age.^14^ Due to the functional importance of the upper extremity, glenohumeral osteoarthritis threatens the independence of many elderly patients. The degeneration of cartilage in osteoarthritis leads to unequal load distribution within the joint cavity, leading to inflammation, osteophyte formation, subchondral bone changes, and synovial proliferation.^15^^,^^25^ These structural changes often lead to pain and limit range of motion.^2^ TSA for osteoarthritis has successfully provided favorable outcomes, notable pain relief, and improved range of motion.^1^^,^^2^^,^^12^^,^^30^ Surgical treatment can therefore drastically improve both physical and emotional well-being.^15^^,^^23^ A study by Cho et al reported improved psychological status in patients following TSA for the treatment of osteoarthritis.^7^
As the US population ages, a larger number of older individuals are seeking orthopedic surgery. However, the literature has consistently shown that older individuals are at higher risk of postoperative medical complications, including deep vein thrombosis, UTI, acute renal failure, and pneumonia.^35^ Therefore, as a greater number of older individuals undergo TSA, it is important to understand potential adverse outcomes of these different age groups.
Our results align with those of previous studies in orthopedic literature, including studies on total hip and total knee arthroplasty (THA and TKA, respectively). For example, prolonged length of stay, non-home discharge, and postoperative mortality have been more commonly reported among elderly patient cohorts following total joint arthroplasty.^9^^,^^28^^,^^29^ Furthermore, both octogenarian and nonagenarian patients have been found to be at increased postoperative risk of transfusions following TKA and THA compared to younger patients.^10^^,^^28^ Multiple orthopedic studies have also reported associations between increased age and unplanned readmission.^4^^,^^6^^,^^13^^,^^28^^,^^32^^,^^40^ Additionally, a study by Bovonratwet et al found that both older age and bleeding disorders increased risk of readmission following THA.^4^^,^^6^ On the other hand, when compared to younger cohorts, older patients have been reported to be at decreased risk for dislocation, perioperative fracture, implant related complications, and surgical site infections.^35^^,^^38^
In our study, we identified the older-aged cohorts to be increasingly significant predictors of blood transfusions following TSA. Prior studies on shoulder arthroplasty have found female sex, low preoperative hemoglobin, and traumatic surgical indication to be significant predictors for postoperative transfusion.^33^ These findings are consistent with clinical practice, as women are more likely to be anemic and their lower baseline hemoglobin may increase the need for transfusion. Similar to our results, a study on revision TSA cases also found older age to be a risk factor for postoperative transfusion.^1^
As one ages, the likelihood of developing anemia increases due to a number of different etiologies including anemia of chronic disease, chronic kidney disease, and iron deficiency anemia.^3^^,^^22^^,^^34^ Anemia has been reported in 10% of individuals over age 65 and in over 50% of those over age 80.^3^ The increasing incidence of anemia in older populations may increase the likelihood of low preoperative hemoglobin and thus increase the probability of needing a postoperative transfusion. Additionally, preoperative anemia has been reported to be a risk factor for postoperative complications and mortality following TJA.^36^ Therefore, preoperative evaluation and management of anemia in older populations may help to mitigate adverse outcomes.
We also identified increasingly significant rates of readmission as our patient cohorts aged. Although readmission rates following TSA are overall low, they have been reported to be as high as 5.5% among the elderly patient population.^17^ Interestingly, prior research on THA and TKA have reported a majority of readmissions due to falls, which may be less likely following upper extremity surgery. Consistent with our results, other studies have reported increased age as an independent risk factor for readmission following TSA, with medical causes accounting for up to 50% of readmissions.^24^^,^^37^^,^^38^ A study on primary TSA found that old age leads to higher rates of readmission, most commonly due to pneumonia, dislocation, pulmonary embolism, and surgical site infections.^8^ Similarly, a study by White et al found older age was more likely to be associated with readmission, as well as multiple medical complications such as pulmonary embolism, infection, and respiratory complications.^38^
Across all age groups, we found increasing age to be a significant predictor of nonhome discharge. This is in line with prior research, as a study on octogenarian outcomes following TSA found increased risk of non-home discharge compared to nonoctogenarian patients.^5^ Nonhome discharge following TSA may negatively impact postoperative outcomes and increase the likelihood of readmission, which raises important considerations for postoperative management of medically complex patients.^20^ From age 70, our study also identified age to be a significant predictor of mortality. Although overall mortality and complication rates are low, patients older than age 80 have been reported to have higher rates of early mortality following TSA.^35^ This again could be related to the medical complexity that often goes along with aging patients. However, a study by McCormick et al suggests that the mortality rate following TSA is still lower than that of THA and TKA.^27^
Our study is limited to the data that can be analyzed through the American College of Surgeons National Surgical Quality Improvement Program database. This database is limited to a 30-day postoperative outcomes period, and therefore is unable to identify complications that occur outside of this 30-day window. Potential long-term postoperative complications, such as implant failure or revision surgery, are unable to be accounted for. Additionally, operative factors such as hospital location and surgeon skill level were unable to be identified. Despite these limitations, we used a large national database to better understand the impact of our aging population on postoperative outcomes following TSA. Furthermore, this study allowed us to identify transfusion, nonhome discharge, readmission, and mortality to be increasingly statistically significant outcomes as a patient ages. Future research needs to be done to further understand which comorbidities are associated with adverse outcomes in the older population to better preoperatively manage these patients.
From age 60, each decade of age was identified to be an increasingly significant predictor for blood transfusion, readmission, and nonhome discharge following TSA. From age 70, each decade of age was additionally identified to be an increasingly significant predictor for mortality. Increasing age was consistently an independent predictor for non-home discharge. As the patient population continues to age, understanding the complications associated with increasing age may help to improve outcomes.
Funding: No funding was disclosed by the authors.
Conflicts of The authors, their immediate families, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.