Chronic use of tramadol after acute pain episode: cohort study

From Wiki Journal Club
Jump to navigation Jump to search

Clinical Question

In discharged post-operative patients undergoing elective surgery in the USA (2009-18), is tramadol prescription associated with an increased risk of prolonged opiate use versus other short-acting opiates?

Bottom Line

Whilst the authors conclude tramadol is associated with an increase in prolonged opiate use post-operatively, there are plenty of confounders to dispute the validity of this conclusion.

Major Points

Using three different definitions of prolonged opiate use, the authors contend that tramadol prescription versus other short-acting opiate prescription has an increased risk ratio of 1.06-1.47 (p<0.05)

Guidelines

No guidelines have been published that reflect the results of this trial.

Design

  • Retrospective observational cohort study from American insurance administrative claims data
  • 524,318 were eligible for discharge prescription analysis, but 79,554 (15%) were excluded due to insufficient follow up time of 180 days
  • N=444,764 who had 180 days of uncensored follow up were split into mutually exclusive groups
    • 1) No opioid fill = 86,880 (20%)
    • 357,884 had opioid script on discharge (80%)
    • 2) Short acting opiate other than tramadol = 333,289 (75%) (reference group)
    • 3) Tramadol only = 13,519 (3%) (this was the group used for comparisons)
    • 4) Long-acting opiate +/- short-acting opiate = 5,619 (1%)
    • 5) Tramadol + another short-acting opiate 5,457 (1%)
  • Secondary analysis was done on N=524,318 patients who had at least 30 days of uncensored follow up to determine script numbers
    • Total by fills by drug = 212,987 Hydrocodone (53%), 150,785 Oxycodone (38%), 16,059 Tramadol (4%), 12,377 Codeine (3%), 4,831 Hydromorphone (1%), 4,825 Propoxyphene (1%)(withdrawn 2010)
  • Setting: USA commercial and Medicare Advantage insurance claims (OptumLabs Data Warehouse)
  • Timeframe: January 1, 2009 through June 30, 2018 (last surgery December 31, 2017 to allow six months follow up)
  • End-points = end of study period June 30, 2018; unenrolment from insurance, another surgical procedure (defined by claim for anaesthetic services)
  • Analysis:
    • Conversion of doses into morphine milligram equivalents (MME)
    • Logistic regression with risk ratios
  • Definition of prolonged opiate use:
    • Additional opioid use after surgery = at least one opioid fill 90-180 days after surgery
    • Persistent opioid use after surgery = any span of opioid use starting in the 180 days after surgery, lasting at least 90 days
    • CONSORT definition = opiate use starting in the 180 days after surgery that spans at least 90 days and included either 10 or more fills or 120 or more days' supply of opioids
    • The authors then stated they defined a "single episode of opioid use as a period of time during which the patient goes no more than 30 days without opioids available"

Population

Inclusion Criteria

  • Commercial and Medicare Advantage patients enrolled in a large, private, US health plan
  • Patients discharged from American hospitals after elective surgery
    • General = laparoscopy cholecystectomy +/- IOC), minimally-invasive inguinal hernia repair, open inguinal hernia repair, simple mastectomy without reconstruction, breast lumpectomy +/- axillary node biopsy, pancreaticoduodenectomy (Whipple), and parathyroidectomy
    • Orthopaedic = carpal tunnel release, knee arthroscopic meniscectomy, rotator cuff surgery, TKR, THR, lumbar laminotomy or laminectomy via posterior approach
    • Colorectal = minimally-invasive low anterior resection +/- ostomy, and partial colectomy +/- ostomy
    • Urology = minimally-invasive partial/total nephrectomy, and minimally-invasive prostatectomy
    • Thoracic = open lung lobectomy, and video-assisted thoracoscopic lung wedge resection
    • Gynaecological = minimally-invasive hysterectomy
  • Opioid fill of <1400 morphine milligram equivalents (MME)

Exclusion Criteria

  • Less than six months of continuous enrolment in medical and prescription insurance coverage before surgery
  • Less than 90 days of continuous insurance enrolment post-operatively
  • Filled prescription for opiates in the six months prior to elective surgery
  • Having treatment for opiate use disorder with buprenorphine or methadone in the 90 days after surgery
  • Having multiple unrelated procedures on the same day (to reduce complexity)
  • Inpatient stays longer than seven (7) days
  • Those who were inpatients for more than one day prior to surgery
  • Cancer patients receiving non-cancer surgeries
  • Patients receiving hospice services
  • Patients who were discharged home and stayed in a skilled nursing facility within a day of discharge
  • Opiate fills of >1400 MME (excluded top 0.5% of discharge fills)

Baseline Characteristics

Interventions

  • Tramadol vs other opiate prescription on discharge

Outcomes

  • Analysis of discharge prescriptions included patients with at least 30 days of continuous follow-up with fills <1400 MME
  • Analysis of chronic opioid use included patients with any post-surgery opioid fill and at least 180 days of continuous follow-up
  • Logistic regression was used at the individual level on the cohort with at least 180 days of uncensored follow up time
  • Reference for adjusted risk ratio comparisons for tramadol alone group was other short-acting opioid group

Primary Outcomes

Prolonged opiate use (overall | no opioid | tramadol only | other short-acting opiate | tramadol + other short-acting | any long-acting)
  • Additional use definition = 31,431 (7.1%) | 3,849 (4.4) | 1,066 (7.9) | 25,388 (7.6) | 543 (10.0) | 585 (10.4)
Adjusted RR 1.06 (1.00-1.13), p=0.049 (6% greater risk associated with tramadol group vs reference other short-acting group)
Of note, 22,779 (72%) had no opioids 31-60 days post-op, 23,630 (75%) had no opioids 61-90 days post-op, and 20,258 (64%) had no opioids 31-90 days post-op
{Despite this, patients are still in this group due to opioid fills in the 91-180 day period post-operatively (may be separate episodes rather than continuous)}
  • Persistent use definition = 4,457 (1.0%) | 314 (0.4) | 194 (1.4) | 3,559 (1.1) | 149 (2.7) | 241 (4.3)
Adjusted RR 1.47 (1.25-1.69), p<0.001 (47% greater risk associated with tramadol group vs reference other short-acting group)
488 (11%) had no opioids 31-90 days post-op
  • CONSORT definition = 2,027 (0.5%) | 175 (0.2) | 78 (0.6) | 1573 (0.5) | 71 (1.3) | 130 (2.3)
Adjusted RR 1.41 (1.08-1.75), p=0.013 (41% greater risk associated with tramadol group vs reference other short-acting group)
281 (14%) had no opioids 31-90 days post-op

Secondary Outcomes

Discharge prescription characteristics (cohort with ≥30 days' follow up time and any opioid fill)
Median amount of opioid dispensed (standardised MME) was 225 (equivalent to 30x5mg oxycodone tablets)
Tramadol comprised 16,059 scripts (4%) with a median fill of 150 (150-250) MME
Larger discharge prescriptions were associated with an increased risk of prolonged opiate use
≥500 MME was associated with nearly five times the risk of prolonged use compared with 1-199 MME using the CONSORT definition (1.5x using additional use definition, >6x using persistent use definition)

Criticisms

  • Having two cohorts is confusing (at least 30 day follow up, and 180 day follow up)
  • Tramadol prescriptions were a small amount of the overall prescriptions (3-4% depending which cohort you use)
    • Prolonged opioid use was a small percentage of this group (0.58-7.89% depending on the definition used)
    • Tramadol was also part of the the category 'tramadol + other short-acting opiates', and may also have been in the group 'any long-acting opiate'
  • The groups were not matched or compared at baseline, so may have statistically significant intrinsic differences
    • E.g. sex (tramadol 62% female), age, location (West South Central Census Division), operation type (total hip)
  • Authors are unable to determine actual opioid consumption (vs scripts filled), or capture prescriptions that weren't submitted to insurance (proposed to be small amount)
  • Authors didn't look at the different in safety profile of tramadol vs other opiates (e.g. potentially lower risk of respiratory depression, diversion)
  • There is no data as to why prescribers used tramadol instead of other opiates (and vice versa), this information may be very important (was it given to patients with a previous history of problems with opiate use?)
  • Prolonged opiate use is not necessarily synonymous with opiate abuse or addiction
  • Multiple exclusions make it difficult to extrapolate results outside of the elective post-operative setting

Funding

There was no external funding. It was supported by the Mayo Clinic Robert D and Patricia E Kern Center for the Science of Health Care Delivery.

Further Reading

Original paper: