Opioid Utility and Hospital Outcomes Among Inpatients Admitted With Osteoarthritis and Spine Disorders : American Journal of Physical Medicine & Rehabilitation

Secondary Logo

Journal Logo

Original Research Articles

Opioid Utility and Hospital Outcomes Among Inpatients Admitted With Osteoarthritis and Spine Disorders

Lee, Se Won MD; Werner, Bryan MD; Nguyen, Dan MD; Wang, Charles DO; Kang, Mingon PhD; Ayutyanont, Napatkamon PhD; Lee, Soohyoung PhD

Author Information
American Journal of Physical Medicine & Rehabilitation 102(4):p 353-359, April 2023. | DOI: 10.1097/PHM.0000000000002101


What Is Known

  • Opioid analgesics are frequently used by patients with osteoarthritis (OA) and spine disorders. However, the overall utilization of opioids decreased in recent years. Opioid analgesics are associated with worse outcomes after hospitalization.

What Is New

  • Among 12,747 patients admitted with OA or spine disorders, the total number of patients using opioids decreased. However, the daily morphine milligram equivalent (MME) use for patients on opioids increased from 2017 to 2020. In addition, results show that increased daily MME is associated with nonhome discharge among patients with OA.

Spine disorders and osteoarthritis (OA) are the most commonly encountered musculoskeletal disorders within medicine,1 impacting more than 500 million people worldwide,2 and are the leading cause of adult disability in the United States (US).3 Osteoarthritis and spondylopathies/spondyloarthropathy both ranked within the top 20 principal diagnoses for all inpatient stays in US Hospitals (3rd and 13th, respectively) and accounted for 13.5% of all hospital stays.4 A higher percentage of patients with painful OA and spine disorders are prescribed opioids in comparison with those with other diagnoses in the US.5 Opioid prescriptions in the US are more than double that of countries within the European Union and Australia/New Zealand,6,7 despite limited evidence to support their clinical efficacy and increased awareness of the risks associated with their use. The percentage of patients being prescribed opioid medication for OA pain relief within the US continues to be highest in the acute care hospital setting.8

After the recent trends to reduce opioid prescribing, there have been few studies looking at opioid prescribing practices within the acute hospital setting for patients with OA and spine disorders as the primary admitting diagnosis.9 Investigating the similarities and differences in the utilization of pain medications and their impact on the hospital outcomes between two of the most common disabling musculoskeletal disorders can enhance our understanding of the impact and burden of opioid prescribing within the US healthcare system.10

This study aims to evaluate the trends of opioid and nonopioid analgesic utilization among patients admitted to the acute care hospital setting with OA and spine disorders to reveal associations between demographic and clinical characteristics, including the use of opioid and nonopioid analgesics, hospital-acquired complications (HACs), patient safety indicators (PSIs), and discharge disposition.


Patient data for this study were drawn from six private community hospitals in six states within a private healthcare system in the US spanning the period January 1, 2017, to December 31, 2020. Inclusion criteria for this study were adults 18 yrs and older, admission to an acute inpatient unit, length of stay involving at least one overnight stay, and patients who received International Classification of Disease, Ninth and Tenth Revision, Clinical Modification, diagnoses of OA (M15–M19) and spine disorders (M40–M54) as their primary admitting diagnoses. Excluded from the study were patients who were discharged from the emergency department or who were discharged on the day of admission, had missing demographic variables, and those who were admitted to inpatient rehabilitation units.

This study was reviewed and approved as exempt by the intuitional review board at the author’s institution, and the patient consent was waived by the intuitional review board as this study used a deidentified data set. The study findings are in accordance with Strengthening the reporting of observational studies in epidemiology guidelines (see Supplementary Checklist, Supplemental Digital Content 1, https://links.lww.com/PHM/B822).


The demographic variables assessed included age, sex, race, ethnicity, medical insurance coverage, length of hospitalization, and discharge destination. Spine disorders and OA diagnoses were determined using International Classification of Disease, Ninth and Tenth Revision, Clinical Modification, diagnostic codes.

The Charlson Comorbidity Index (CCI), a validated, weighted scoring system was used to quantify medical comorbidities. The CCI was previously used to predict the severity of illness and 1-yr mortality risk.11,12 The CCI has been used widely to stratify risk following medical conditions, including physically disabling conditions, such as stroke,13 OA,14 and spine disorders.15

Hospital-acquired complication and PSI events used in the analysis include pulmonary complications, venous thromboembolic disease (deep venous thrombosis and pulmonary embolism), falls, cardiovascular disease, gastrointestinal complications, acute kidney injury, and intraoperative and postoperative complications.16

Medications received during the acute hospitalization were organized using the therapeutic classification system from the US Pharmacopeia Drug Classification System.17 The 15 most frequently used medications were analyzed among the group with OA and spine disorders.

Morphine milligram equivalents (MMEs) of oral, intravenous, and transdermal opioid analgesics were calculated for standardization and comparison using the conversion factors from the Centers for Disease Control and Prevention.18,19


Descriptive statistics were used to analyze demographic and clinical characteristics. A t test was used to validate the normal distribution of samples based on the demographic and clinical variables; then, χ2 was performed to examine the independence of demographic and clinical variables and the categorical variables. A stepwise logistic regression model was used to evaluate significant relationships between predictive variables (demographic and clinical variables including frequently used medications, and surgical vs. nonsurgical) and the outcome variables of home discharge (0: home discharge vs. 1: other than home discharge [facility-based postacute care]), HAC, and PSI event (1: the presence of either HAC or PSI). While performing stepwise selection, an attempt was made to remove any insignificant variables from the model before adding a significant variable to the model.20 A P value less than 0.05 was considered significant. Analyses were performed using SAS, version 9.4 (SAS Institute, Inc, Cary, NC) and Stata version 12 (StataCorp, College Station, TX).


There were 12,747 unique patient admissions that met the inclusion criteria between January 1, 2017, to December 31, 2020. Seven thousand one hundred ninety-five or 56% of the patients were diagnosed with OA, while 5552 or 44% were diagnosed with spine disorders. The mean age ± SD of all patients was 65.38 ± 11.98 yrs. Patients with spine disorders were slightly younger than those with OA (62.31 ± 13.93 vs. 67.74 ± 9.58, P < 0.0001). The length of stay for both diagnostic groups was 2.27 ± 2.35 days, while patients with spine disorders had a slightly longer duration than those with OA (2.92 ± 3.18 vs. 1.77 ± 1.20, P < 0.0001; Table 1).

TABLE 1 - Demographic and clinical information
Total (N = 12,747) Osteoarthritis Group (n = 7195) Spine Disorders Group (n = 5552) P
Age, mean ± SD 65.38 ± 11.98 67.74 ± 9.58 62.31 ± 13.93 <0.0001
Male sex 5526 (43.35%) 2943 (40.90%) 2583 (46.52%) <0.0001
Race <0.0001
 Black 1261 (9.89%) 640 (8.9%) 621 (11.19%)
 Other 691 (5.42%) 256 (3.56%) 435 (7.84%)
 White 10,795 (84.69%) 6299 (87.55%) 4496 (80.98%)
Insurance <0.0001
 Medicaid 534 (4.19%) 115 (1.60%) 419 (7.55%)
 Medicare 7735 (60.68%) 4620 (64.21%) 3115 (56.11%)
 Uninsured 195 (1.53%) 33 (0.46%) 162 (2.92%)
 Other 971 (7.62%) 374 (5.20%) 597 (10.75%)
 Private 3312 (25.98%) 2053 (28.53%) 1259 (22.68%)
Body mass index, mean ± SD 30.60 ± 6.28 31.15 ± 6.10 29.89 ± 6.43 <0.0001
Length of stay, mean ± SD, d 2.27 ± 2.35 1.77 ± 1.20 2.92 ± 3.18 <0.0001
Top 5 admitting diagnoses Unilateral primary knee OA: 3649 (50.72%) L. Spinal stenosis: 1726 (31.09%)
Unilateral primary hip OA: 2379 (33.06%)
Primary shoulder OA: 509 (7.07%)
Bilateral primary OA of knee:
L. Radiculopathy: 631 (11.37%)
C. Spinal stenosis: 619 (11.15%)
L. Spondylolisthesis:
389 (5.41%)
Primary OA of ankle, and foot: 142 (1.97%)
372 (6.7%)
C. Disc disease: 336 (6.05%)
CCI, mean ± SD 2.826 ± 1.759 2.983 ± 1.550 2.622 ± 1.978 <0.0001
Discharge destination <0.0001
 Home 11,144 (87.42%) 6293 (87.46%) 4851 (87.37%)
 SNF 1238 (9.781%) 807 (11.22%) 431 (7.76%)
 IRF 283 (2.22%) 86 (1.2%) 197 (3.55%)
Discharge AMA 37 (0.29%) 0 37 (0.67%)
 Other 35 (0.27%) 6 (0.08%) 29 (0.52%)
 Deceased 6 (0.05%) 3 (0.04%) 3 (0.05%)
 Hospice 4 (0.03%) 0 4 (0.07%)
AMA, against medical advice; C, cervical; IRF, inpatient rehabilitation facility; L, lumbar; SNF, skilled nursing facility.

Opioid analgesics were the single most used medication (n = 11,737, 92.94% of 12,692 patients without missing medication information, slightly higher in the group with spine disorders, 93.32% vs. 91.83%, P = 0.002), followed by laxatives (n = 10,347, 81.93%). Short-acting opioids with acetaminophen were prescribed at a higher rate for patients with chronic spine disorders (5592, 77.84% vs. 4385, 79.61%, P = 0.0160), while tramadol (with/without acetaminophen) was used at a significantly higher rate within the OA group (2447, 34.06% vs. 307, 5.57%, P ≤ 0.0001). From 2017 to 2020, the rate of opioid use among patients decreased in both groups (OA group: 93.10% in 2017 to 91.05% in 2020, P = 0.028, spine group: 94.89% to 92.06%, P = 0.008; Fig. 1). Opposite to the decrease in the rate of opioid use, daily MME increased over the same period (47.08 ± 42.83 vs. 50.74 ± 41.42, P = 0.0001; Fig. 1). For 1107 of the patients (8.7%) prescribed opioids during hospitalization, opioid analgesics were not prescribed within the first 24 hrs of admission but were initiated before discharge. This occurred at a slightly higher rate in the OA group compared with the spine disorder group (9.24% vs. 8.01%, P = 0.015). From 2017 to 2020, there was an increase in the number of patients who were started on an opioid after the first 24 hrs of admission in the OA group (7.11% in 2017 to 9.81% in 2020, P < 0.0001), while new opioid use among the spine disorder group showed a declining trend that was not statistically significant (9.04% in 2017 to 7.04% in 2020, P = 0.173).

Trends of the rate of overall and specific opioid analgesic utility among patients with OA (A) and spine disorders (B).

Other frequently used analgesics include acetaminophen (n = 6476, 51.28%), Meloxicam/other nonsteroidal anti-inflammatory drugs (NSAIDs, n = 5441, 43.09%), anticonvulsants (n = 3477, 27.53%), and muscle relaxants (n = 2909, 23.03%). Acetaminophen and NSAIDs, including meloxicam, were more frequently used in the group with OA, while anticonvulsants and muscle relaxants were more frequently used in the group with spine disorders (Table 2). Figure 2 illustrates the various trends in medication use between 2017 and 2020.

TABLE 2 - Frequently used medication classes during hospitalization
Total (N = 12,692) OA Spine P
Polypharmacy, mean ± SD 9.68 ± 3.82 10.64 ± 3.79 8.43 ± 3.48 <0.0001
Laxative 10,347 (81.93%) 6300 (87.69%) 4047 (73.47%) <0.0001
Opioids 11,737 (92.94%) 6597 (91.83%) 5140 (93.32%) 0.002
 Opioid short acting 9977 (79.00%) 5592 (77.84%) 4385 (79.61%) 0.0160
 Opioid + AAP 4862 (38.50%) 1913 (26.63%) 2949 (53.54%) <0.0001
 Tramadol (with or without AAP) 2754 (21.81%) 2447 (34.06%) 307 (5.57%) <0.0001
 Daily MME, mean ± SD 50.31 ± 40.77 50.10 ± 37.30 50.57 ± 44.85 0.5240
Acetaminophen 6476 (51.28%) 5053 (70.34%) 1423 (25.84%) <0.0001
NSAID 3038 (24.06%) 2479 (34.51%) 559 (10.15%) <0.0001
Meloxicam 2403 (19.03%) 2357 (32.81%) 46 (0.84%) <0.0001
Anticonvulsant 3477 (27.53%) 1518 (21.13%) 1959 (35.57%) <0.0001
Muscle relaxant 2909 (23.03%) 510 (7.10%) 2399 (43.55%) <0.0001
Benzodiazepine 1,490 (11.74%) 365 (5.08%) 1,125 (20.42%) <0.0001
Electrolyte replacement 9212 (72.94%) 6079 (84.62%) 3133 (56.88%) <0.0001
Antibiotic 7922 (62.73%) 6012 (83.69%) 1910 (34.68%) <0.0001
Gastrointestinal prophylaxis 6612 (52.36%) 4061 (56.53%) 2551 (46.31%) <0.0001
Antiemetic 4720 (37.37%) 3117 (43.39%) 1603 (29.10%) <0.0001
Platelet inhibitor 4713 (37.32%) 4427 (61.62%) 286 (5.19%) <0.0001
Antihypertensive 4107 (32.52%) 2313 (32.20%) 1794 (32.57%) 0.6550
Statin 3965 (31.40%) 2560 (35.63%) 1405 (25.51%) <0.0001
Heparin 1448 (11.47%) 674 (9.38%) 774 (14.05%) <0.0001

Trends of the rate of overall and specific nonopioid analgesic utility among patients with OA (A) and spine disorders (B).

The details of medical comorbidities are described in Appendix 1 (Supplemental Digital Content 2, https://links.lww.com/PHM/B823). Patients with medical comorbidities that were recognized as having a contraindication to NSAID treatment (renal disease, peripheral vascular disease, cardiovascular disease/myocardial infarction/congestive heart failure, liver disease, and peptic ulcer disease) comprised 2187 patients or 29.11% of the total (slightly higher in spine group than OA group, 30.66% vs. 27.92%, P = 0.001).

Eleven thousand one hundred forty-four patients (87.42%) with either spine disorder or OA were discharged home after acute hospitalization. Patient characteristics associated with home discharge included younger patients (64.33 ± 11.77 vs. 72.62 ± 10.87, P < 0.0001) and male sex (44.75 vs. 33.73%, P < 0.001). The overall duration of hospitalization and percentage of intensive care unit care admission was significantly lower in the group discharged home (1.95 ± 1.97 vs. 4.45 ± 3.37 days, P < 0.0001, and 1.49% vs. 4.46%, P < 0.0001, respectively). Among the 1603 patients who were not discharged home, 1238 patients (77%) were discharged to skilled nursing facilities, 283 patients (18%) went to inpatient rehabilitation units, 37 patients left against medical advice, and 6 patients died.

A stepwise logistic regression analysis of the relationship between home discharge (dependent variable) and contributing factors (independent variables in Tables 1 and 2 including the presence of surgery during hospitalization and the effect of NSAIDs) revealed the following associations. For both the OA and spine group, nonhome discharge was associated with older age (OA: odds ratio [OR], 1.074; 95% confidence interval [CI], 1.058–1.090, and spine: OR, 1.045; CI, 1.032–1.058), patients covered by Medicaid program (OR, 2.230; CI, 1.004–4.951, and OR, 1.669; CI, 1.026–2.716), patients covered by Medicare program (OR, 2.443; CI, 1.740–3.429, and OR, 1.782; CI, 1.298–2.445) and increased length of stay (OR, 3.305; CI, 3.020–3.616, and OR, 1.309; CI, 1.266–1.354; Table 3). For patients with OA diagnoses, nonhome discharge was associated with increased body mass index (OR, 1.021; CI, 1.005–1.037), daily MME (OR, 1.004; CI, 1.001–1.007), and having received electrolyte replacement (OR, 1.661; CI, 1.272–2.168). For the spine group, nonhome discharge was associated with Black race (OR, 1.426; CI, 1.078–1.866), no insurance (OR, 2.115; CI, 1.046–4.277), increased CCI (OR, 1.163; CI, 1.095–1.235), increased number of medications (OR, 1.039; CI, 1.012–1.068), and heparin injection (OR, 2.025; CI, 1.612–2.543). Daily MME was not associated with home discharge in the spine group (P = 0.279). Overall, HACs were relatively low (n = 119, <1%), with the most common complications identified as fractures (n = 40), catheter-related urinary tract infection (n = 21), and sepsis (n = 17). For both OA and spine groups, length of stay was associated with HAC and PSI (OR, 1.672; CI, 1.190–2.347, and OR, 1.162; CI, 1.005–1.344). For the group with OA, Medicare (OR, 6.012; CI, 1.197–30.183) and benzodiazepine use (OR, 26.108; CI, 2.103–324.070) were associated with HAC and PSI. While for the spine group, increased CCI (OR, 1.310; CI, 1.031–1.665), increased length of stay (OR, 1.162; CI, 1.005–1.344), and having received electrolyte replacement (OR, 2.868; CI, 1.002–8.204) were associated with HAC and PSI (Table 4).

TABLE 3 - Stepwise logistic regression analysis of relationship between discharge destination (home vs. others) and contributing factors
Group With OA, R 2 = 0.435 Group With Spine Disorder, R 2 = 0.334
Independent Variables OR P 95% CI Independent Variables OR P 95% CI
Demographic variable
 Age 1.074 0.000 1.058–1.090 Age 1.045 0.000 1.032–1.058
 Male sex 0.655 0.000 0.535–0.800
 Other race (than Black and White) 0.395 0.001 0.231–0.675 Black 1.426 0.013 1.078–1.866
 Medicaid 2.230 0.049 1.004–4.951 Medicaid 1.669 0.039 1.026–2.716
 Medicare 2.443 0.000 1.740–3.429 Medicare 1.782 0.000 1.298–2.445
 Other insurance 1.692 0.099 0.905–3.163 No insurance 2.115 0.037 1.046–4.277
 Body mass index 1.021 0.010 1.005–1.037
Hospital course
 Any HAC 0.388 0.090 0.130–1.160 CCI 1.163 0.000 1.095–1.235
 Length of stay, d 3.305 0.000 3.020–3.616 Length of stay, d 1.309 0.000 1.266–1.354
 Daily MME per 1 MME 1.004 0.009 1.001–1.007 Polypharmacy 1.039 0.005 1.012–1.068
 Electrolyte replacement 1.661 0.000 1.272–2.168 Antibiotic 0.771 0.026 0.614–0.969
 Platelet inhibitor 0.710 0.000 0.586–0.860 Heparin 2.025 0.000 1.612–2.543
 Antiemetic 0.830 0.055 0.686–1.004 Acetaminophen 1.217 0.067 0.986–1.503
 Heparin 1.298 0.061 0.988–1.707 Statin 0.844 0.121 0.681–1.046
Bold font highlights statistical significance at P < 0.05.

TABLE 4 - Stepwise logistic regression analysis of relationship between HACs, patient safety index events, and contributing factors
Patients With OA, R 2 = 0.388 Patients With Spine Disorders, R 2 = 0.376
Independent Variables OR P 95% CI Independent Variables OR P 95% CI
Age 0.876 0.001 0.808–0.950
Medicare 6.012 0.029 1.197–30.183 CCI 1.310 0.027 1.031–1.665
Length of stay, d 1.672 0.003 1.190–2.347 Length of stay, d 1.162 0.043 1.005–1.344
Polypharmacy 0.818 0.010 0.703–0.953 Electrolyte replacement 2.868 0.049 1.002–8.204
Antiemetic 0.263 0.018 0.087–0.796 Insulin 7.481 0.070 0.848–65.961
Benzodiazepine 26.108 0.011 2.103–324.070
Muscle relaxant 8.223 0.102 0.659–102.609
Bold font highlights statistical significance at P < 0.05.


This study shows significantly high utilization (more than 90%) of opioid analgesics for the treatment of OA and spine disorders during acute hospitalization. A significantly higher rate of opioid use was observed than previously documented in other recent studies, including 13% for outpatient knee OA,8 24%–39% for hospitalized presurgical hip and knee OA,21,22 and between 20% and 55% for hospitalized presurgical spine patients.23,24 One potential argument for this finding is a selection bias with the patients representing a subset more severely affected by pain than those previously studied. Post hoc analysis showed higher daily MME in the nonsurgical group (51.03 ± 41.81 of 11,896 patients vs. 41.55 ± 26.60 of 796 surgical patients, P < 0.0001). Although this study did not assess preadmission (at home) opioid use, less than 10% of admitted patients who were not initially on an opioid during the first 24 hrs of hospital admission were subsequently prescribed an opioid during their hospitalization. This is comparatively less than the 15% of opioid-naive patients (defined as having no opioid prescription 60 days before admission) who were found to have continued opioid use after discharge in a previous study.25

Similar to other studies that followed the 2016 Center for Disease Control and Prevention Guideline for Prescribing Opioids for Chronic Pain, there seems to be an overall decreasing trend in opioid prescribing among patients with all diagnoses, excluding cancer and palliative care.26 This study, however, had the concerning finding of an upward trend of increasing MMEs prescribed for the treatment of OA and spine disorder–related pain from 2017 to 2020. In addition, the average dose of opioid medication seen within this study population (>50 MME) has been shown to be associated with a higher risk of adverse effects.26 This study showed that opioid analgesics were more frequently used in the treatment of spine disorders. Opioids with acetaminophen (AAP) were shown to be used twice as frequently in the spine group compared with the OA group, while treatment for patients in the OA group had 6 times the prescription rate of tramadol compared with the spine group. This was surprising, given the previous report of increased all-cause mortality with tramadol compared with NSAIDs in the patients with OA.27 The use of tramadol did demonstrate a downward trend of use in the OA group from 2017 to 2020 (36.39%–15.53%, P < 0.0001).

Studies looking at opioid use as a risk factor for adverse hospital outcomes have previously shown conflicting results, although these studies were limited to surgical patients only.28 When examining nonsurgical patients, hospitals with high opioid use rates have demonstrated an increased adjusted risk of a severe opioid-related adverse event per patient exposed.9 Despite the fact that most patients in this study were nonsurgical, it failed to reveal any association between opioids and any severe HACs. Previous studies have failed to demonstrate functional improvement for patients receiving opioids in the inpatient setting.29 Similar to these findings, high-dose opioid use in the patient population studied was associated with less ideal hospital outcomes, defined as nonhome discharge, with increased odds of 1.004 per 1 MME increase among patients with OA. Overly aggressive treatment of pain by opioid analgesics may produce adverse outcomes not just captured by HAC and PSI but may have more subtle impacts on patients that extend beyond the hospital stay and should be studied further.

Nonsteroidal anti-inflammatory drug use for the treatment of similar conditions in the US and other countries ranged between 35% and 56.1%, whereas our study showed NSAIDs being used for less than 11% of patients in the spine group.30,31 Overall, the low rate of NSAIDs and acetaminophen use has been shown to be a persistent trend that continues to diverge from the spine and OA-related pain treatment guidelines.8 Among the patients on opioids without a relative contraindication to use NSAIDs (3367 of 11,737 patients), either from medical comorbidities (n = 3711) or medications such as anticoagulants for deep vein thrombosis prophylaxis and antiplatelets (n = 3,367), only 14.20% (658 of 4,633) patients were prescribed NSAIDs. In addition, the patients in the spine disorder group were prescribed NSAIDs at a significantly lower rate than the OA group despite having a lower medical comorbidity index. This raises questions as to whether lower usage rates are being driven by medical decision making and patient safety or by other factors such as lack of clear guidelines for NSAIDs use on the population at risk.32 Although much has been accomplished in reducing opioid use in recent years, an increased effort to address the opioid epidemic in the inpatient setting is clearly needed. As more studies demonstrate the risks associated with opioid use, there seems to be a need to review, revise, and create better treatment algorithms and prescribing guidelines that more accurately model the patient comorbidity complexities in the acute hospital setting that can help and guide practioners to judicious use of AAP and NSAIDs for the treatment of musculoskeletal pain in the inpatient setting.29

The volume of data gathered from six different community teaching hospitals across six different states, including precise MME dosages, demographic information, and a validated medical comorbidities index system for analysis is a strength of this study and had significant power to make meaningful analysis of the impact of the opioid analgesics and other identified patient factors on patient outcomes. This specific information can help clinicians build specific guidelines based on the target population to help clinicians prepare for discharge planning.

The study population was limited to six community-based private hospitals and may not be fully applicable to nonprofit healthcare systems, such as those within academic centers or the Veterans Health Administration. Some information, including HACs, relied on a retrospective review of hospital discharge-based data including International Classification of Disease, Ninth and Tenth Revision, Clinical Modification, codes that have limited accuracy in capturing all medical problems and events. Moreover, this study could not address the potential for unrecognized coding errors. This study was limited in determining what percentage of the patients were on an opioid pain medication regimen before admission because it did not use outpatient opioid tracking tools, such as Prescription Drug Monitoring Programs. Therefore, the impact of opioid use before admission on the outcome variables was limited. The database used lacked certain baseline functional data, including the level of independence and barriers to discharge, which can be important factors in determining functional outcome and discharge disposition. This study also is unable to determine causality and indications for medication prescriptions, including contraindications for the use of or discontinuation of NSAIDs for anticipated procedures due to concerns with regard to platelet function inhibition.


In summary, opioids are widely used among patients with OA and spine disorders within the acute hospital setting despite ongoing efforts to decrease the reliance on and use of opioid medication in the treatment of noncancer-related pain disorders. Opioid analgesics and nonopioid analgesics were used at varying rates for the conditions studied with appreciable differences in their usage pattern compared with previous studies. During the course of the study from 2017 to 2020, the rate of opioid usage declined; however, during the same period, the daily MME being prescribed increased and was shown to be at a relatively high level on average (>50 MME). Several demographic and clinical characteristics were associated with HACs as well as disposition after hospitalization.


1. Mattson CL, Schieber L, Scholl L, et al.: Annual surveillance report of drug-related risks and outcomes—United States, 2017. 2017, Surveillance special report; 1. Available at: https://stacks.cdc.gov/view/cdc/47832. Accessed September 26, 2022
2. Hunter DJ, March L, Chew M: Osteoarthritis in 2020 and beyond: a Lancet Commission. Lancet 2020;396:1711–2
3. Mathieson S, Valenti L, Maher CG, et al.: Worsening trends in analgesics recommended for spinal pain in primary care. Eur Spine J 2018;27:1136–45
4. McDermott KW, Roemer M: Most Frequent Principal Diagnoses for Inpatient Stays in U.S. Hospitals, 2018. In: Healthcare Cost and Utilization Project (HCUP) Statistical Briefs [Internet]. Rockville (MD), Agency for Healthcare Research and Quality (US), 2006. Statistical Brief #277. PMID: 34428003
5. Lee SW, Patel J, Kim SY, et al.: Use of opioid analgesics in patients with chronic low back pain and knee osteoarthritis. Am J Phys Med Rehabil 2019;98:e97–8
6. Burden M, Keniston A, Wallace MA, et al.: Opioid utilization and perception of pain control in hospitalized patients: a cross-sectional study of 11 sites in 8 countries. J Hosp Med 2019;14:737–45
7. International Narcotics Control Board: Narcotic Drugs Technical Report: Estimated World Requirements for 2013–Statistics for 2011. New York, NY, United Nations, 2013
8. Gwam CU, Emara AK, Ogbonnaya IA, et al.: Addressing national opioid prescribing practices for knee osteoarthritis: an analysis of an estimated 41,389,332 patients with knee arthritis. J Am Acad Orthop Surg 2021;29:e337–44
9. Herzig SJ, Rothberg MB, Cheung M, et al.: Opioid utilization and opioid-related adverse events in nonsurgical patients in US hospitals. J Hosp Med 2014;9:73–81
10. Mapel DW, Shainline M, Paez K, et al.: Hospital, pharmacy, and outpatient costs for osteoarthritis and chronic back pain. J Rheumatol 2004;31:573–83
11. Austin SR, Wong YN, Uzzo RG, et al.: Why summary comorbidity measures such as the Charlson Comorbidity Index and Elixhauser Score Work. Med Care 2015;53:e65–72
12. Bannay A, Chaignot C, Blotière PO, et al.: The best use of the Charlson Comorbidity Index with electronic health care database to predict mortality. Med Care 2016;54:188–94
13. Tessier A, Finch L, Daskalopoulou SS, et al.: Validation of the Charlson Comorbidity Index for predicting functional outcome of stroke. Arch Phys Med Rehabil 2008;89:1276–83
14. Le TK, Montejano LB, Cao Z, et al.: Health care costs in US patients with and without a diagnosis of osteoarthritis. J Pain Res 2012;5:23
15. Mannion A, Bianchi G, Mariaux F, et al.: Can the Charlson Comorbidity Index be used to predict the ASA grade in patients undergoing spine surgery?Eur Spine J 2020;29:2941–52
16. Rahman M, Neal D, Fargen KM, et al.: Establishing standard performance measures for adult brain tumor patients: a nationwide inpatient sample database study. Neuro Oncol 2013;15:1580–8
17. Bhattacharyya L, Cecil T, Dabbah R, et al.: The value of USP public standards for therapeutic products. Pharm Res 2004;21:1725–31
18. Pommerening MJ, Landau A, Hrebinko K, et al.: An analysis of analgesia and opioid prescribing for veterans after thoracic surgery. Sci Rep 2020;10:11348
19. Dowell D, Haegerich TM, Chou R: CDC Guideline for Prescribing Opioids for Chronic Pain — United States, 2016. MMWR Recomm Rep 2016;65(No. RR-1):1–49
20. Citterio A, Franceschini M, Spizzichino L, et al.: Nontraumatic spinal cord injury: an Italian survey. Arch Phys Med Rehabil 2004;85:1483–7
21. Franklin PD, Karbassi JA, Li W, et al.: Reduction in narcotic use after primary total knee arthroplasty and association with patient pain relief and satisfaction. J Arthroplasty 2010;25:12–6
22. Hansen CA, Inacio MC, Pratt NL, et al.: Chronic use of opioids before and after total knee arthroplasty: a retrospective cohort study. J Arthroplasty 2017;32:811–7.e1
23. Armaghani SJ, Lee DS, Bible JE, et al.: Preoperative opioid use and its association with perioperative opioid demand and postoperative opioid independence in patients undergoing spine surgery. Spine 2014;39:E1524–30
24. Krebs EE, Lurie JD, Fanciullo G, et al.: Predictors of long-term opioid use among patients with painful lumbar spine conditions. J Pain 2010;11:44–52
25. Jena AB, Goldman D, Karaca-Mandic P: Hospital prescribing of opioids to Medicare beneficiaries. JAMA Intern Med 2016;176:990–7
26. Goldstick JE, Guy GP, Losby JL, et al.: Changes in initial opioid prescribing practices after the 2016 release of the CDC guideline for prescribing opioids for chronic pain. JAMA Netw Open 2021;4:e2116860
27. Zeng C, Dubreuil M, LaRochelle MR, et al.: Association of tramadol with all-cause mortality among patients with osteoarthritis. JAMA 2019;321:969–82
28. Lee D, Armaghani S, Archer KR, et al.: Preoperative opioid use as a predictor of adverse postoperative self-reported outcomes in patients undergoing spine surgery. J Bone Joint Surg Am 2014;96:e89
29. Krebs EE, Gravely A, Nugent S, et al.: Effect of opioid vs nonopioid medications on pain-related function in patients with chronic back pain or hip or knee osteoarthritis pain: the SPACE Randomized Clinical Trial. JAMA 2018;319:872–82
30. Akazawa M, Mimura W, Togo K, et al.: Patterns of drug treatment in patients with osteoarthritis and chronic low back pain in Japan: a retrospective database study. J Pain Res 2019;12:1631–48
31. Basedow M, Williams H, Shanahan EM, et al.: Australian GP management of osteoarthritis following the release of the RACGP guideline for the non-surgical management of hip and knee osteoarthritis. BMC Res Notes 2015;8:1–8
32. Atkinson TJ, Fudin J: Nonsteroidal antiinflammatory drugs for acute and chronic pain. Phys Med Rehabil Clin N Am 2020;31:219–31

Opioids; Osteoarthritis; Spine Disorders; Adverse Events; Discharge

Supplemental Digital Content

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.