Skip to main content
. Author manuscript; available in PMC: 2020 Nov 4.
Published in final edited form as: JAMA Netw Open. 2020 Jul 1;3(7):e2011972. doi: 10.1001/jamanetworkopen.2020.11972

Table.

Baseline Characteristics of Study Participants and Adjusted Odds Ratios of Receiving at least One Opioid Prescription Within 60 Days of Discharge After Total Hip or Knee Arthroplastya

THA (n=86,058) TKA (n=150,573)
Patient Characteristicsa N % OR 95% CI N % OR 95% CI
Year of Surgery
 2014 18940 22.01 1.00 35267 23.42 1.00
 2015 21692 25.21 1.36 1.29 1.44 38595 25.63 1.42 1.36 1.47
 2016 23228 26.99 1.66 1.57 1.75 39485 26.22 1.85 1.77 1.93
 2017 22198 25.79 1.92 1.81 2.03 37226 24.72 2.40 2.30 2.52
Level of pain at discharge
 0 5247 6.10 1.00 7316 4.86 1.00
 1–3 35372 41.10 1.38 1.28 1.49 48270 32.06 1.40 1.31 1.50
 4–6 37824 43.95 1.50 1.39 1.62 76599 50.87 1.50 1.40 1.60
 7–10 7615 8.85 1.32 1.20 1.46 18388 12.21 1.38 1.28 1.48
Age at THA/TKA admission, mean (SD) 64.48 11.17 65.99 9.59
Age Category
 <55 14906 17.32 1.00 17461 11.60 1.00
 55–64 27662 32.14 0.88 0.82 0.94 48598 32.28 0.85 0.81 0.91
 65–74 26948 31.31 0.74 0.68 0.79 54567 36.24 0.74 0.70 0.79
 75+ 16542 19.22 0.55 0.51 0.60 29947 19.89 0.57 0.54 0.62
Gender
 Female 46684 54.25 1.00 92726 61.58 1.00
 Male 39374 45.75 1.06 1.02 1.11 57847 38.42 1.04 1.01 1.07
Race/Ethnicityb
 Caucasian 75604 87.85 1.00 129232 85.83 1.00
 African American 6177 7.18 0.79 0.74 0.86 11299 7.50 0.86 0.81 0.92
 Hispanic 1660 1.93 0.60 0.53 0.68 4371 2.90 0.53 0.49 0.57
Smoking Statusb,c
 Current smoker 13479 15.66 1.00 18178 12.07 1.00
 Not currently smoking 37234 43.27 0.86 0.81 0.92 66673 44.28 0.88 0.83 0.92
 Never smoked 30762 35.75 1.07 1.01 1.15 57750 38.35 1.23 1.17 1.30
Insuranceb
 Commercial/Other (1 insurance) 38640 44.90 1.00 62580 41.56 1.00
 Medicare Only 28,622 33.26 1.28 1.21 1.37 52797 35.06 1.28 1.22 1.34
 Medicaid Only 3011 3.50 1.29 1.13 1.47 4298 2.85 1.18 1.06 1.32
 2 or More Insurances 11465 13.32 0.61 0.57 0.65 23729 15.76 0.56 0.53 0.59
Anxiety 13937 16.19 0.93 0.88 0.99 26086 17.32 1.01 0.97 1.06
Depression 16354 19.00 1.08 1.01 1.14 36101 23.98 0.98 0.94 1.02
Substance Use Disorders 4213 4.90 0.99 0.89 1.09 4979 3.31 0.92 0.85 1.01
Number of Condition Categories, mean (SD) 2.74 2.26 1.05 1.04 1.06 3.12 2.28 1.09 1.08 1.10
Condition Categoriesd
 Arrhythmia 13348 15.51 25366 16.85
 Atherosclerosis 11233 13.05 21281 14.13
 Cancer 8744 10.16 15323 10.18
 Cellulitis 2391 2.78 4743 3.15
 CHF 3506 4.07 6561 4.36
 COPD 7530 8.75 12606 8.37
 Dementia 1506 1.75 2806 1.86
 Diabetes 14116 16.40 34801 23.11
 Dialysis 95 0.11 146 0.10
 Fluid or metabolic disorder 7246 8.42 12828 8.52
 Hemolytic disorder 211 0.25 287 0.19
 Hypertension 52116 60.56 103799 68.94
 Major symptoms, abnormalities 45451 52.81 86404 57.38
 Malnutrition 406 0.47 395 0.26
 Metastatic cancer 314 0.36 371 0.25
 Morbid obesity 9086 10.56 25868 17.18
 Other injuries 11991 13.93 28550 18.96
 Paralysis 799 0.93 1300 0.86
 Pneumonia 1499 1.74 2706 1.80
 Polyneuropathy 8347 9.70 16993 11.29
 Psychiatric 4314 5.01 8528 5.66
 Renal failure 6993 8.13 13514 8.98
 Rheumatoid arthritis 6037 7.02 10428 6.93
 Severe infection 7186 8.35 13863 9.21
 Stroke 920 1.07 1693 1.12
 Ulcer 738 0.86 1219 0.81
 Vascular disorder 9261 10.76 16801 11.16
a

Patients received THA or TKA in 2014–2017 from Optum EHR data which contains structured information, such as diagnosis, procedure codes, lab results and observations, including vital signs, blood pressure, pain and body mass index (BMI). It also contains unstructured information in the form of clinical notes from office visits, consultation reports, discharge summaries and reports from nursing records, pathology, radiology and cardiology. Using a natural language processing (NLP) system, Optum staff extract these unstructured data and organize them into structured fields for clinical assessment. Clinical measurements include numeric fields from clinical notes such as blood pressure, pain and BMI. We identified THA or TKA using ICD-9, ICD-10, or CPT codes (THA: ICD-9, 81.51; ICD-10, 0SR90J9, 0SR90JA, 0SR90JZ, 0SRB0J9, 0SRB0JA, 0SRB0JZ; CPT, 27130; TKA: ICD-9, 81.54; ICD-10, 0SRC07Z, 0SRC0JZ, 0SRC0KZ, 0SRC0LZ, 0SRD07Z, 0SRD0JZ, 0SRDD0KZ, 0SRD0LZ, 0SRT07Z, 0SRT0JZ, 0SRT0KZ, 0SRU07Z, 0SRU0JZ, 0SRU0KZ, 0SRV07Z, 0SRV0JZ, 0SRV0KZ, 0SRW07Z, 0SRW0JZ, 0SRW0KZ; CPT, 27447).

b

Other or unknown categories were included in the model, but not reported.

c

Smoking status (never, former, current, other) were obtained from clinical observations found in the EHR and were taken from the observation nearest to the day of discharge, up to 6 months prior to the date of discharge. When there were multiple recorded statuses for smoking or selected, we recorded the most severe category with the “other” category being the least severe (i.e. current smoker, former smoker, never smoker, other).

d

The number of condition categories were included in the model, not as individual, separate conditions. These condition categories came from the CMS report on hospital readmission for hip or knee arthroplasty and additional conditions, including anxiety, depression, and substance use disorder. All of these condition category variables were assessed in the 12 months prior to the joint arthroplasty.