Table 1. Baseline Characteristics of Tapering Patients in Inception Cohort and Among Those With Postinduction Follow-upa.
| Characteristic | Patients, No. (%) | |
|---|---|---|
| Tapering events in inception cohort (30 255 tapers among 28 018 patients) | Tapering events with postinduction follow-up (21 515 tapers among 19 377 patients) | |
| Age category, y | ||
| 18-34.9 | 962 (3.2) | 566 (2.6) |
| 35-49.9 | 6437 (21.3) | 4449 (20.7) |
| 50-64.9 | 15 717 (52.0) | 11 479 (53.4) |
| ≥65 | 7139 (26.9) | 5021 (23.3) |
| Sex | ||
| Female | 16 309 (53.9) | 11 581 (53.8) |
| Male | 13 946 (46.1) | 9934 (46.2) |
| Educationb | ||
| High school or less | 13 560 (44.8) | 9737 (45.3) |
| More than high school | 14 886 (49.2) | 10 465 (48.6) |
| Unknown or missing | 1809 (6.0) | 1313 (6.1) |
| Rural vs urban residencec | ||
| Metropolitan or micropolitan | 28 245 (93.4) | 20 044 (93.2) |
| Small town or rural | 1933 (6.4) | 1421 (6.6) |
| Missing | 77 (0.3) | 50 (0.2) |
| Commercial insurance | 12 820 (42.4) | 8217 (38.2) |
| Elixhauser comorbiditiesd | ||
| Alcohol use disorder | 738 (2.4) | 483 (2.2) |
| Drug use disorder | 4301 (14.2) | 3030 (14.1) |
| Baseline opioid dose, MME/d | ||
| 50 to <90 | 6848 (22.6) | 4879 (22.7) |
| 90 to <150 | 7200 (23.8) | 5126 (23.8) |
| 150 to <300 | 9544 (31.6) | 6767 (31.5) |
| ≥300 | 6663 (22.0) | 4743 (22.0) |
| Coprescribed benzodiazepinee | 9078 (30.0) | 6371 (29.6) |
| Baseline year overdose eventsf | ||
| 0 | 29 502 (97.5) | 21 030 (97.8) |
| 1 | 625 (2.1) | 406 (1.9) |
| ≥2 | 128 (0.4) | 79 (0.4) |
| Baseline depression or anxietyg | 16 310 (53.9) | 11 613 (54.0) |
Abbreviation: MME, morphine milliequivalents.
Patients with tapering events in original cohort and patients with long-term follow-up both had means of 1.1 baseline periods followed by tapering events (median [range], 1.0 [1.0-4.0]). Patient characteristics at end of included baseline periods are shown, and some patients are represented by more than 1 period.
Education was estimated according to median education level of resident aged 25 years or older in patient’s US Census block.
Rurality was derived from Rural-Urban Commuting Area codes.
Elixhauser comorbidities included 27 noncancer conditions, including alcohol use disorder and drug use disorder. The depression Elixhauser indicator was not included because of its redundancy with the preexisting depression or anxiety variable.
Coprescribed benzodiazepine was defined as a concurrent benzodiazepine prescription on the date of cohort entry.
Baseline overdose events were defined by specified diagnosis codes identified on emergency department or hospital claims in baseline year.
Baseline depression or anxiety was defined by specified diagnoses identified on emergency department, hospital, or outpatient claims, or pharmacy claims for selective serotonin-reuptake inhibitor during baseline year.