Skip to main content

Comparative safety of tramadol and other opioids following total hip and knee arthroplasty

Abstract

Background

Tramadol is increasingly used to treat acute postoperative pain among older adults following total hip and knee arthroplasty (THA/TKA). However, tramadol has a complex pharmacology and may be no safer than full opioid agonists. We compared the safety of tramadol, oxycodone, and hydrocodone among opioid-naïve older adults following elective THA/TKA.

Methods

This retrospective cohort included Medicare Fee-for-Service beneficiaries ≥ 65 years with elective THA/TKA between January 1, 2010 and September 30, 2015, 12 months of continuous Parts A and B enrollment, 6 months of continuous Part D enrollment, and no opioid use in the 6 months prior to THA/TKA. Participants initiated single-opioid therapy with tramadol, oxycodone, or hydrocodone within 7 days of discharge from THA/TKA hospitalization, regardless of concurrently administered nonopioid analgesics. Outcomes of interest included all-cause hospitalizations or emergency department visits (serious adverse events (SAEs)) and a composite of 10 surgical- and opioid-related SAEs within 90-days of THA/TKA. The intention-to-treat (ITT) and per-protocol (PP) hazard ratios (HRs) for tramadol versus other opioids were estimated using inverse-probability-of-treatment-weighted pooled logistic regression models.

Results

The study population included 2,697 tramadol, 11,407 oxycodone, and 14,665 hydrocodone initiators. Compared to oxycodone, tramadol increased the rate of all-cause SAEs in ITT analyses only (ITT HR 1.19, 95%CLs, 1.02, 1.41; PP HR 1.05, 95%CLs, 0.86, 1.29). Rates of composite SAEs were not significant across comparisons. Compared to hydrocodone, tramadol increased the rate of all-cause SAEs in the ITT and PP analyses (ITT HR 1.40, 95%CLs, 1.10, 1.76; PP HR 1.34, 95%CLs, 1.03, 1.75), but rates of composite SAEs were not significant across comparisons.

Conclusions

Postoperative tramadol was associated with increased rates of all-cause SAEs, but not composite SAEs, compared to oxycodone and hydrocodone. Tramadol does not appear to have a superior safety profile and should not be preferentially prescribed to opioid-naïve older adults following THA/TKA.

Peer Review reports

Introduction

Pain management after total hip and knee arthroplasty (THA/TKA) is a critical part of postoperative care [1]. Moderate to severe pain is typical and pain control is important to participate in rehabilitation activities and restore function. Opioid analgesics are commonly prescribed for acute postoperative pain. However, given the risk of dependence with long-term use, postoperative pain management has shifted toward opioid-sparing regimens [1]. Tramadol is a serotonin–norepinephrine reuptake inhibitor (SNRI) and prodrug that is metabolized into a weak partial opioid agonist [2]. As a result of its unique pharmacology, tramadol is perceived by many clinicians to be safer than full agonist opioids like oxycodone or hydrocodone and to even be “opioid-sparing” by reducing exposure to full opioid agonist drugs. The Drug Enforcement Administration classifies tramadol as a Schedule IV controlled substance, indicating that it is perceived to have a lower risk of addiction and misuse than Schedule II opioids like oxycodone and hydrocodone. Tramadol has therefore become increasingly popular for postoperative pain management [3].

While use of tramadol has increased substantially over time [4,5,6,7], it is a potentially risky and unpredictable drug due to a number of factors. As a prodrug, it must be metabolized from an SNRI into an opioid metabolite by the CYP2D6 enzymes in the liver. The ability to metabolize drugs through CYP2D6 varies significantly between individuals. Slower metabolizers will experience the serotonin and norepinephrine (antidepressant) effects of tramadol with very little opioid analgesia, while fast metabolizers can experience more opioid analgesia than intended by prescribers. Age-related changes to drug metabolism create more variation in metabolism of tramadol among older adults, as does use of multiple chronic medications requiring metabolism through the CYP2D6 enzymes. Tramadol’s effects are therefore unpredictable and could potentially result in more adverse events than other opioids following surgery [8]. Further, it is unclear how the real-world safety of tramadol compares to other opioids when used following high-intensity interventions, such as THA/TKA, among older adults [9]. Given the heightened risk of peri-operative morbidity in this vulnerable population, it is critical to ensure that efforts to minimize full opioid agonist exposure through the use of tramadol do not come at the expense of increased adverse events and worse health outcomes.

We assessed the safety of tramadol compared to other opioids commonly prescribed to older adults for postoperative pain following THA/TKA using national data. We hypothesized that use of tramadol would result in higher risks of all-cause and serious adverse events when compared to oxycodone or hydrocodone.

Methods

Study design and data sources

This retrospective cohort study leveraged data from the Medicare Master Beneficiary Summary File; Medicare Provider Analysis and Review (MedPAR) inpatient claims; a 20% random sample of Medicare Part B (outpatient services) and Part D (prescription drugs) claims. The timing and location of health services utilization was determined using the validated Residential History File [10]. All sources included data from January 1, 2009 to December 31, 2015 to allow for a 12-month covariate ascertainment period prior to THA/TKA procedures performed in 2010, the first calendar year in which individuals were eligible to enter the cohort. Additionally, we utilized data from the Dartmouth Atlas Hospital Tracking File and Prescription Drug Abuse Policy System. The Prescription Drug Abuse Policy System summarized legislation relevant to state prescription drug monitoring programs effective through July 1, 2016, such as circumstances where prescribers are required to check the prescription drug monitoring programs [11]. Brown University’s institutional review board approved the study (protocol# 2010002823). Due to the use of deidentified administrative data, the need for informed consent was waived.

Study participants

We designed this observational study to emulate a target trial comparing the safety of tramadol, hydrocodone, and oxycodone among U.S. Medicare beneficiaries with THA/TKA (Additional Table 1) [12]. Our study population consisted of Medicare Fee-for-Service beneficiaries aged 65 years or older on the date of elective inpatient THA/TKA procedure occurring between January 1, 2010 to September 30, 2015. THA/TKA procedures were identified based on Medicare Severity Diagnosis Related Groups codes 469 and 470 using inpatient claims. Eligible older adults were required to have 12 months of continuous Parts A and B enrollment, 6 months of continuous Part D enrollment, no opioid use in the 6 months immediately prior to the THA/TKA (Additional Table 2) [13], and no history of opioid-related inpatient hospitalizations or emergency department (ED) visits in the 12 months prior to THA/TKA. Individuals were excluded if they had Health Maintenance Organization (HMO) enrollment (Medicare Advantage) at any point prior to or during THA/TKA because these individuals' hospitalizations are not consistently reported by all hospitals [14]. We also excluded individuals with non-elective procedures, non-arthroplasty procedures, those admitted with fractures, those who were admitted for THA/TKA from a setting other than the community, and those utilizing hospice in the 12 months immediately prior to THA/TKA (Additional Table 3). Additionally, individuals initiating multiple opioids on the same day following THA/TKA (e.g., tramadol and hydrocodone); initiating opioids other than oxycodone, hydrocodone, or tramadol; or initiating opioids after seven days from THA/TKA were excluded. Lastly, we excluded beneficiaries discharged to post-acute care settings other than home with or without home health because medication dispensings through Part D cannot be observed [15]. For individuals with multiple procedures during the study period, only the first THA or TKA was included to better isolate initial opioid use.

Opioid exposure

The interventions of interest were initiation of single opioid therapy with oxycodone, hydrocodone, or tramadol within seven days following THA/TKA hospital discharge regardless of concurrently administered nonopioid analgesics or pain adjuvants (e.g., acetaminophen, nonsteroidal anti-inflammatory drugs, gabapentinoids).

Outcomes

We examined the first occurrence of two outcomes in the 90 days following THA/TKA: 1) all-cause hospitalizations and ED visits (all-cause serious adverse events [SAEs]) and 2) hospitalization and ED visits for a composite of 10 surgical- and opioid-related serious adverse events (composite SAEs) including THA/TKA revision, surgical complications [16, 17], cardiovascular events [18,19,20,21], fractures [22, 23], gastrointestinal events [24,25,26], acute liver injury [27], acute renal failure [28], opioid-related adverse events [29], delirium [30], and respiratory depression [31]. Outcomes were ascertained using International Classification of Diseases, 9th Edition-Clinical Modification (ICD-9-CM) diagnosis and procedure codes from MedPAR (inpatient hospital) and Part B Carrier (ED) claims (Additional Table 4).

Follow-up

For each eligible participant, follow-up began at THA/TKA hospital discharge. Participants were followed until an outcome of interest or censoring event occurred. In an intention-to-treat (ITT) analysis, censoring events included death; disenrollment from Medicare Parts A, B, or D; HMO enrollment; or administrative end of follow-up (December 31, 2015 or day 90 of follow-up). Two additional censoring events were included in a per-protocol (PP) analysis: treatment discontinuation (the last available days’ supply based on a gap of no more than five days between refills) and treatment switching (initiation of a different opioid).

Covariates

We measured 131 patient, hospital, and geographic characteristics prior to and concurrent with THA/TKA (Additional Table 5). We used the Medicare Master Beneficiary Summary File and Chronic Conditions Data Warehouse to ascertain beneficiary enrollment information, age, race/ethnicity [32], and medical conditions. Medication use in the 6 months prior to the THA/TKA was ascertained using Part D claims. MedPAR inpatient and Part B outpatient claims were used to calculate healthcare service utilization measures as well as the validated Combined Comorbidity Score and Claims-based Frailty Index [33, 34]. Hospital characteristics, including total beds, academic affiliation, and critical access hospital designation, were obtained from the Dartmouth Atlas Hospital Tracking file. Geographic characteristics included Department of Health and Human Services Region and “rigor” of the state prescription drug monitoring programs, which has been previously categorized using the Prescription Drug Abuse Policy System [11, 35].

Causal contrasts

We were interested in both the ITT estimand (the effect of individuals being prescribed a treatment regardless of future adherence or treatment switching) and the PP estimand (the effect of individuals adhering to and remaining on the prescribed treatment) [36].

Statistical analyses

The ITT estimand was estimated using stabilized inverse-probability-of-treatment-weighted (IPTW) pooled logistic regression models with robust error variance to adjust for potential confounding. The probabilities (i.e., propensity scores) used to construct the IPTWs were estimated via multinomial generalized boosted regression models [37, 38]. Estimated propensity scores showed substantial overlap (Additional Fig. 1). As an alternative approach, we estimated propensity scores using parametric multinomial logistic regression to assess if results differed when using non-parametric versus parametric approaches. For each subject, three probabilities were estimated, one for receiving each of the three treatments studied. The PP estimand was estimated using the same models as for the ITT estimand, but with the additional inclusion of inverse-probability-of-censoring-weights (IPCW) to reduce bias due to informative censoring. The probabilities of censoring used to construct the IPCW were estimated using logistic regression. We accounted for 131 covariates via IPTW and IPCW (Additional Table 5). Covariate balance was assessed using absolute standardized mean differences (ASMDs) with a difference of 0.1 or less considered to be indicative of sufficient covariate balance.

Pooled logistic regression models were employed for outcome estimation at the person-day level using the weighted pseudo-population after IPW [39]. Estimates were equivalent to those from discrete-time hazard models, which are interpretable as hazard ratios (HRs) with 95% confidence limits (CLs) (See Additional Appendix). The HRs provided a relative measure to compare the rates of outcomes among tramadol users versus oxycodone or hydrocodone users. We also estimated 90-day risk differences (RD) as an absolute measure using standardized survival curves [40]. Numbers needed to treat (NNT) and harm (NNH) were also estimated. Bootstrap standard errors (percentile method) for the RDs and NNTs/NNHs were estimated using 500 replications [41].

Additionally, a formal quantitative bias analysis was conducted by estimating E-values [42, 43].

Data were analyzed using SAS, version 9.4 (SAS Institute, Inc., Cary, NC), Stata, version 16 (StataCorp LLC, College Station, TX) and R, version 3.4 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Study cohort

The study population included 28,769 opioid-naïve older adults with THA/TKA: 2,697 initiating tramadol, 11,407 initiating oxycodone, and 14,665 initiating hydrocodone (Additional Fig. 2, Table 1). Prior to IPW, tramadol users were older (75–84 years old 40.7% tramadol, 29.7% oxycodone, 34.1% hydrocodone), a greater proportion were female (70.1% tramadol, 53.8% oxycodone, 57.8% hydrocodone), had THA (34.9% tramadol, 28.2% oxycodone, 27.8% hydrocodone), heart failure (16.8% tramadol, 12.9% oxycodone, 13.8% hydrocodone), and were frailer (33.1% tramadol, 27.9% oxycodone, 29.5% hydrocodone) (Table 1). Time to opioid initiation after THA/TKA and censoring events were similar across treatment groups (Additional Table 67). Multiple covariates differed at baseline based on ASMDs of > 0.1, but were well balanced across treatment groups after applying IPW (Additional Table 8). We truncated the product of the IPTW and IPCW to the 99th percentile due to the presence of some extreme IPCW (Additional Table 9). However, the product IPW distributions did not differ markedly when propensity scores were estimated using generalized boosted regression or multinomial logistic regression models.

Table 1 Characteristics of Medicare beneficiaries initiating tramadol, oxycodone, or hydrocodone after total hip and knee arthroplasty

Overall outcomes

The incidence of all-cause SAEs was 14.9% for tramadol, 13.4% for oxycodone, and 12.7% for hydrocodone in the ITT analysis and 10.3%, 9.2%, and 8.7%, respectively in the PP analysis (Additional Table 10). The incidence of composite SAEs was similar in the ITT analysis (3.1% tramadol, 2.8% oxycodone, 2.8% hydrocodone) and PP analysis (2.0% tramadol, 1.5% oxycodone, 1.7% hydrocodone).

Relative estimates of outcomes

Initiating tramadol after THA/TKA was associated with increased rates of all-cause SAEs versus oxycodone based on the ITT estimand (HR 1.19, 95%CLs 1.02, 1.41) but no significant difference was observed based on the PP estimand (HR 1.05, 95%CLs 0.86, 1.29) (Fig. 1, Table 2). Rates of composite SAEs for tramadol versus oxycodone were not significantly different across comparisons (ITT HR 1.12, 95%CLs 0.78, 1.64; PP HR 1.49, 95%CLs 0.95, 2.35).

Table 2 Hazard ratios comparing postoperative opioid use for 90-day outcomes among older Medicare beneficiaries
Fig. 1
figure 1

Crude and adjusted hazard ratios comparing postoperative tramadol versus oxycodone use for 90-day outcomes. Abbreviations: SAEs, Serious Adverse Events, SAEs; CI, Confidence Interval. Note: Outcome definitions for SAEs are listed in Additional Table 4. The unadjusted estimate is the intention-to-treat (ITT) estimand without covariate adjustment. Both the ITT and per-protocol estimates are covariate-adjusted

Compared to hydrocodone, tramadol use was associated with increased rates of all-cause SAEs based on both the ITT (HR 1.40, 95% CLs 1.10, 1.76) and PP estimands (HR 1.34, 95%CLs 1.03, 1.75) (Fig. 2, Table 2). There were no significant differences in the rates of composite SAEs for tramadol versus hydrocodone (ITT HR 1.19, 95%CLs 0.69, 2.04; PP HR 1.63, 95%CLs 0.90, 2.95).

Fig. 2
figure 2

Crude and adjusted hazard ratios comparing postoperative tramadol versus hydrocodone use for 90-day outcomes. Abbreviations: SAEs, Serious Adverse Events, SAEs; CI, Confidence Interval. Note: Outcome definitions for SAEs are listed in Additional Table 4. The unadjusted estimate is the intention-to-treat (ITT) estimand without covariate adjustment. Both the ITT and per-protocol estimates are covariate-adjusted

Absolute estimates of outcomes

Tramadol initiators did not have markedly different 90-day RDs for all-cause SAEs or composite SAEs when compared to hydrocodone or oxycodone in the ITT and PP analyses (Table 3, Additional Fig. 3–8). Similarly, NNHs and NNTs were not significantly different for all-cause SAEs or composite SAEs in the ITT and PP analyses when tramadol was compared to oxycodone or hydrocodone (Additional Table 11).

Table 3 Risk differences comparing postoperative opioid use for 90-day outcomes among older Medicare beneficiaries

Quantitative bias analysis

Across all comparisons, E-values (lower CL) for all-cause SAEs ranged from 1.28 (1.00) to 2.77 (2.10) and 1.49 (1.00) to 2.64 (1.00) for composite SAEs (Additional Table 12).

Discussion

In this nationally representative retrospective cohort study of older adults without opioid use in the 6 months prior to receiving elective THA/TKA, we found that tramadol does not have a superior safety profile compared to other opioids. Prescribing tramadol for acute postoperative pain was associated with an increased rate of all-cause SAEs compared to oxycodone or hydrocodone. Considering individuals who remained on the prescribed treatment (i.e., did not stop or switch opioids), tramadol was associated with an increased rate of all-cause SAEs versus hydrocodone only. However, there were no significant differences in the risk of all-cause SAEs based on absolute measures (i.e., RDs). No significant differences in the risk of composite SAEs were observed, possibly due to infrequent occurrence, limited statistical power, and imprecise estimates. The quantitative bias analyses suggested that a small to moderate amount of residual confounding could shift significant results to the null. Thus, interpreting results as causal effects may be imprudent.

Tramadol’s role in pain relief following THA/TKA has come under increased scrutiny, though limited guidance exists about how to optimally prescribe tramadol. Tramadol is mentioned in a 2020 guideline on the use of opioids following THA/TKA that was jointly developed by The American Association of Hip and Knee Surgeons, The American Academy of Orthopaedic Surgeons, The Hip Society, The Knee Society, and The American Society of Regional Anesthesia and Pain Medicine [9, 44]. The guideline explicitly acknowledges the limited evidence on both the postoperative adverse events and efficacy of oral tramadol, and calls for additional research. Some evidence suggests that tramadol is not more effective than other opioids [45, 46] and one study found that 11% of individuals receiving tramadol following TKA were converted to a stronger opioid due to inadequate pain relief [46]. Further, a 2020 consensus statement by the Enhanced Recovery After Surgery society recommends the use of acetaminophen and nonsteroidal anti-inflammatory drugs (NSAIDs), yet provides no guidance on tramadol use [1]. Thus, while recommendations have been made for opioid-sparing or opioid-free postoperative analgesia, guidance for treatment selection remains sparse. Such guidance is particularly important given older adults’ susceptibility to adverse events and the fact that they comprise a large, and growing, portion of the THA/TKAs performed each year. Thus, tramadol’s role in pain management among older adults requires a more nuanced evaluation.

Due to the lack of randomized controlled trials evaluating the comparative safety of tramadol relative to other opioids for acute postoperative pain, the evidence base for treatment decision-making is formed primarily from a few trials focused on chronic pain and some limited observational studies [9, 47, 48]. A small body of evidence generated among patients with osteoarthritis suggests worse outcomes, such as increased risk of all-cause mortality and hip fractures, with tramadol use compared to other opioid analgesics [49,50,51]. Short-term clinical studies in older adults with osteoarthritis prior to THA/TKA have shown a reduction in side effects when comparing tramadol to other opioids, such as respiratory depression and somnolence, but no differences in dizziness or headaches [52, 53]. Much of what is known of tramadol’s comparative safety relative to opioids or NSAIDs derives from its use to treat chronic pain resulting from osteoarthritis prior to THA/TKA [9, 54]. These studies are limited in their ability to assess tramadol’s safety compared to other opioids for postoperative pain management due to differences in pain etiology (i.e., acute vs. chronic) and absence of postoperative processes (e.g., rehabilitation). To our knowledge, our study is among the first to evaluate the safety of tramadol relative to other opioids among opioid-naïve older adults following elective THA/TKA. This study offers initial guidance as more providers shift toward opioid-sparing postoperative pain regimens for older adults. Our findings suggest that tramadol does not have a superior safety profile compared to other opioids and should not be preferentially prescribed to opioid-naïve older adults following THA/TKA. Therefore, additional research is needed to identify the safest regimen following THA/TKA that leverages multimodal analgesia and limits opioid-related adverse effects.

Limitations

We acknowledge several potential limitations. First, these results may not generalize to adults younger than 65 years, those without Medicare Fee-for-Service insurance, with a history of opioid use or opioid-related adverse events prior to THA/TKA, residing in long-term care prior to surgery, or discharged to institutional post-acute care after THA/TKA. Individuals who filled their post-surgical opioid prescription prior to the elective procedure would have been excluded from the study population. However, given the nature of our data, we were not able to ascertain the medication indication and could not definitively distinguish between opioids for post-surgical pain management versus another pain indication.

Second, the quantitative bias analysis revealed E-values less than 2.2 for adjusted ITT and PP estimands, suggesting that a small to moderate amount of residual confounding could shift significant estimates to the null. Future work could consider adjusting for additional potential confounders, such as baseline and post-baseline measures of pain, concurrent initiation of other medications (e.g., acetaminophen, non-steroidal anti-inflammatory drugs, gabapentinoids) to manage post-operative pain, patient preferences, and social support.

Third, our data are restricted to medications that were dispensed; patients may not have taken the medications as prescribed, which could have resulted in exposure misclassification. This may have led to the introduction of bias into the PP estimands if misclassification due to incomplete adherence was differential across treatment groups. Additionally, we truncated the product of the IPTW and IPCW at the 99th percentile due to the presence of some extreme IPCW. This approach reduces variance, but may not reduce bias [55, 56]. Although the means of the IPCW were close to one, the extreme weights may indicate some model misspecification or possible violations of positivity. Concerns about positivity violations are merited since we 1) included a large number of covariates in the models used to estimate the probabilities of censoring and 2) estimated the models in person-day-level discrete time data.

Fourth, we could not ascertain adverse events that did not require inpatient hospitalization or an emergency department visit, such as falls not resulting in serious injuries. Opioid-related adverse events were rare in our administrative claims data and could only be examined as a composite outcome along with surgical-related adverse events. Other potential risks of opioid therapy, such as unintentional progression to long-term use and diversion, were not evaluated in this study, but remain a particularly important area of future research [57]. It would be beneficial for future work to compare the safety of post-operative tramadol versus hydrocodone or oxycodone on opioid-related adverse events using a dataset with more detailed clinical information that is not restricted to hospital and emergency department claims.

Lastly, while immortal time bias might be a concern in this study, the time between hospital discharge and post-operative use of each of the opioids was highly similar and the amount of time was very small, so any misclassification of time and resulting bias is likely to be non-differential and minimal.

Conclusions

Our study found that initiation of postoperative tramadol following THA/TKA was associated with increased rates of all-cause SAEs compared to oxycodone and hydrocodone. However, there was no difference in the risk of composite SAEs. Since tramadol appears to be no safer than other opioids, it should not be preferentially prescribed to older adults for postoperative pain.

Availability of data and materials

The data that support the findings of this study are available from the Centers for Medicare and Medicaid Services (CMS) but restrictions apply to the availability of these data, which were used under data use agreements for the current study, and so are not publicly available. However, other researchers can establish their own data use agreement and obtain the datasets employed through the Research Data Assistance Center (ResDAC, https://resdac.org/), a CMS contractor that provides free assistance to researchers interested in CMS data.

Abbreviations

ASMDs:

Absolute standardized mean differences

CLs:

Confidence limits

ED:

Emergency department

HMO:

Health Maintenance Organization

HRs:

Hazard ratios

ITT:

Intention-to-treat

IPCW:

Inverse-probability-of-censoring-weight

IPTW:

Inverse-probability-of-treatment-weight

MedPAR:

Medicare Provider Analysis and Review

NNH:

Number needed to harm

NNT:

Number needed to treat

NSAIDs:

Nonsteroidal anti-inflammatory drugs

PP:

Per-protocol

RDs:

Risk differences

SAEs:

Serious adverse events

SNRI:

Serotonin–norepinephrine reuptake inhibitor

THA/TKA:

Total hip and knee arthroplasty

References

  1. Wainwright TW, Gill M, McDonald DA, Middleton RG, Reed M, Sahota O, et al. Consensus statement for perioperative care in total hip replacement and total knee replacement surgery: Enhanced Recovery After Surgery (ERAS(®)) Society recommendations. Acta Orthop. 2020;91(1):3–19.

    Article  PubMed  Google Scholar 

  2. Webster L, Rauck RL. Atypical opioids and their effect on respiratory drive. J Opioid Manag. 2021;17(7):109–18.

    Article  PubMed  Google Scholar 

  3. Bono JV, Robbins CE, Mehio AK, Aghazadeh M, Talmo CT. Pharmacologic pain management before and after total joint replacement of the hip and knee. Clin Geriatr Med. 2012;28(3):459–70.

    Article  PubMed  Google Scholar 

  4. Wang Y, Wu D, Chan A, Chang CH, Lee VWY, Nichol MB. Temporal trend of opioid and nonopioid pain medications: results from a national in-home survey, 2001 to 2018. Pain Rep. 2022;7(3):e1010.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Bigal LM, Bibeau K, Dunbar S. Tramadol Prescription over a 4-Year Period in the USA. Curr Pain Headache Rep. 2019;23(10):76.

    Article  PubMed  Google Scholar 

  6. DeMik DE, Carender CN, Shamrock AG, Callaghan JJ, Bedard NA. Opioid Use After Total Knee Arthroplasty: Does Tramadol Have Lower Risk Than Traditional Opioids? J Arthroplasty. 2020;35(6):1558–62.

    Article  PubMed  Google Scholar 

  7. Mullins PM, Mazer-Amirshahi M, Pourmand A, Perrone J, Nelson LS, Pines JM. Tramadol Use in United States Emergency Departments 2007–2018. J Emerg Med. 2022;62(5):668–74.

    Article  PubMed  Google Scholar 

  8. McCartney CJL, Nelligan K. Postoperative pain management after total knee arthroplasty in elderly patients: treatment options. Drugs Aging. 2014;31(2):83–91.

    Article  PubMed  Google Scholar 

  9. Hannon CP, Fillingham YA, Nam D, Courtney PM, Curtin BM, Vigdorchik J, et al. The Efficacy and Safety of Opioids in Total Joint Arthroplasty: Systematic Review and Direct Meta-Analysis. J Arthroplasty. 2020;35(10):2759-71.e13.

    Article  PubMed  Google Scholar 

  10. Intrator O, Hiris J, Berg K, Miller SC, Mor V. The residential history file: studying nursing home residents’ long-term care histories(*). Health Serv Res. 2011;46(1 Pt 1):120–37.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Prescription Drug Abuse Policy System - PDAPS 2020 [Available from: http://pdaps.org/.

  12. Hernan MA, Robins JM. Using Big Data to Emulate a Target Trial When a Randomized Trial Is Not Available. Am J Epidemiol. 2016;183(8):758–64.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Jeffery MM, Hooten WM, Hess EP, Meara ER, Ross JS, Henk HJ, et al. Opioid Prescribing for Opioid-Naive Patients in Emergency Departments and Other Settings: Characteristics of Prescriptions and Association With Long-Term Use. Ann Emerg Med. 2018;71(3):326-36.e19.

    Article  PubMed  Google Scholar 

  14. Savitz ST, Stearns SC, Groves JS, Kucharska-Newton AM, Bengtson LGS, Wruck L. Mind the Gap: Hospitalizations from Multiple Sources in a Longitudinal Study. Value in health. 2017;20(6):777–84.

    Article  PubMed  Google Scholar 

  15. Steinman MA, Zullo AR, Lee Y, Daiello LA, Boscardin WJ, Dore DD, et al. Association of β-Blockers With Functional Outcomes, Death, and Rehospitalization in Older Nursing Home Residents After Acute Myocardial Infarction. JAMA Intern Med. 2017;177(2):254–62.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Schneeweiss S, Robicsek A, Scranton R, Zuckerman D, Solomon DH. Veteran’s affairs hospital discharge databases coded serious bacterial infections accurately. J Clin Epidemiol. 2007;60(4):397–409.

    Article  PubMed  Google Scholar 

  17. Olsen MA, Ball KE, Nickel KB, Wallace AE, Fraser VJ. Validation of ICD-9-CM Diagnosis Codes for Surgical Site Infection and Noninfectious Wound Complications After Mastectomy. Infect Control Hosp Epidemiol. 2017;38(3):334–9.

    Article  PubMed  Google Scholar 

  18. Cutrona SL, Toh S, Iyer A, Foy S, Daniel GW, Nair VP, et al. Validation of acute myocardial infarction in the Food and Drug Administration’s Mini-Sentinel program. Pharmacoepidemiol Drug Saf. 2013;22(1):40–54.

    Article  PubMed  Google Scholar 

  19. Kumamaru H, Judd SE, Curtis JR, Ramachandran R, Hardy NC, Rhodes JD, et al. Validity of claims-based stroke algorithms in contemporary Medicare data: reasons for geographic and racial differences in stroke (REGARDS) study linked with medicare claims. Circ Cardiovasc Qual Outcomes. 2014;7(4):611–9.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Saczynski JS, Andrade SE, Harrold LR, Tjia J, Cutrona SL, Dodd KS, et al. A systematic review of validated methods for identifying heart failure using administrative data. Pharmacoepidemiol Drug Saf. 2012;21(Suppl 1):129–40.

    Article  PubMed  Google Scholar 

  21. Singh S, Fouayzi H, Anzuoni K, Goldman L, Min JY, Griffin M, et al. Diagnostic Algorithms for Cardiovascular Death in Administrative Claims Databases: A Systematic Review. Drug Saf. 2019;42(4):515–27.

    Article  PubMed  Google Scholar 

  22. Curtis JR, Mudano AS, Solomon DH, Xi J, Melton ME, Saag KG. Identification and validation of vertebral compression fractures using administrative claims data. Med Care. 2009;47(1):69–72.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Ray WA, Griffin MR, Fought RL, Adams ML. Identification of fractures from computerized Medicare files. J Clin Epidemiol. 1992;45(7):703–14.

    Article  CAS  PubMed  Google Scholar 

  24. Ray WA, Chung CP, Stein CM, Smalley WE, Hall K, Arbogast PG, et al. Risk of peptic ulcer hospitalizations in users of NSAIDs with gastroprotective cotherapy versus coxibs. Gastroenterology. 2007;133(3):790–8.

    Article  CAS  PubMed  Google Scholar 

  25. Abraham NS, Cohen DC, Rivers B, Richardson P. Validation of administrative data used for the diagnosis of upper gastrointestinal events following nonsteroidal anti-inflammatory drug prescription. Aliment Pharmacol Ther. 2006;24(2):299–306.

    Article  CAS  PubMed  Google Scholar 

  26. Winner M, Mooney SJ, Hershman DL, Feingold DL, Allendorf JD, Wright JD, et al. Incidence and predictors of bowel obstruction in elderly patients with stage IV colon cancer: a population-based cohort study. JAMA Surg. 2013;148(8):715–22.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Lo ReHaynesGoldbergFordeCarbonariLeidl VKDKADMKB 3rd, et al. Validity of diagnostic codes to identify cases of severe acute liver injury in the US Food and Drug Administration’s Mini-Sentinel Distributed Database. Pharmacoepidemiol Drug Saf. 2013;22(8):861–72.

    Article  Google Scholar 

  28. Waikar SS, Wald R, Chertow GM, Curhan GC, Winkelmayer WC, Liangos O, et al. Validity of International Classification of Diseases, Ninth Revision, Clinical Modification Codes for Acute Renal Failure. J Am Soc Nephrol. 2006;17(6):1688–94.

    Article  PubMed  Google Scholar 

  29. Heslin KC, Owens PL, Karaca Z, Barrett ML, Moore BJ, Elixhauser A. Trends in Opioid-related Inpatient Stays Shifted After the US Transitioned to ICD-10-CM Diagnosis Coding in 2015. Med Care. 2017;55(11):918–23.

    Article  PubMed  Google Scholar 

  30. Kim DH, Lee J, Kim CA, Huybrechts KF, Bateman BT, Patorno E, et al. Evaluation of algorithms to identify delirium in administrative claims and drug utilization database. Pharmacoepidemiol Drug Saf. 2017;26(8):945–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Jones N, Schneider G, Kachroo S, Rotella P, Avetisyan R, Reynolds MW. A systematic review of validated methods for identifying acute respiratory failure using administrative and claims data. Pharmacoepidemiol Drug Saf. 2012;21(Suppl 1):261–4.

    Article  PubMed  Google Scholar 

  32. Eicheldinger C, Bonito A. More accurate racial and ethnic codes for Medicare administrative data. Health Care Financ Rev. 2008;29(3):27–42.

    PubMed  PubMed Central  Google Scholar 

  33. Gagne JJ, Glynn RJ, Avorn J, Levin R, Schneeweiss S. A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–59.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Kim DH, Schneeweiss S, Glynn RJ, Lipsitz LA, Rockwood K, Avorn J. Measuring Frailty in Medicare Data: Development and Validation of a Claims-Based Frailty Index. J Gerontol A Biol Sci Med Sci. 2018;73(7):980–7.

    Article  PubMed  Google Scholar 

  35. Desai RJ, Jin Y, Franklin PD, Lee YC, Bateman BT, Lii J, et al. Association of Geography and Access to Health Care Providers With Long-Term Prescription Opioid Use in Medicare Patients With Severe Osteoarthritis: A Cohort Study. Arthritis & Rheumatology. 2019;71(5):712–21.

    Article  CAS  Google Scholar 

  36. Hernan MA, Robins JM. Per-Protocol Analyses of Pragmatic Trials. N Engl J Med. 2017;377(14):1391–8.

    Article  PubMed  Google Scholar 

  37. McCaffrey DF, Griffin BA, Almirall D, Slaughter ME, Ramchand R, Burgette LF. A tutorial on propensity score estimation for multiple treatments using generalized boosted models. Stat Med. 2013;32(19):3388–414.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivar Behav Res. 2011;46(3):399–424.

    Article  Google Scholar 

  39. Desai RJ, Franklin JM. Alternative approaches for confounding adjustment in observational studies using weighting based on the propensity score: a primer for practitioners. BMJ. 2019;367: l5657.

    Article  PubMed  Google Scholar 

  40. Cole SR, Hernan MA. Adjusted survival curves with inverse probability weights. Comput Methods Programs Biomed. 2004;75(1):45–9.

    Article  PubMed  Google Scholar 

  41. Altman DG. Confidence intervals for the number needed to treat. BMJ (Clinical research ed). 1998;317(7168):1309–12.

    Article  CAS  PubMed  Google Scholar 

  42. VanderWeele TJ, Ding P. Sensitivity Analysis in Observational Research: Introducing the E-Value. Ann Intern Med. 2017;167(4):268–74.

    Article  PubMed  Google Scholar 

  43. Mathur MB, Ding P, Riddell CA, VanderWeele TJ. Web Site and R Package for Computing E-values. Epidemiology (Cambridge, Mass). 2018;29(5):e45–7.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Hannon CP, Fillingham YA, Nam D, Courtney PM, Curtin BM, Vigdorchik JM, et al. Opioids in Total Joint Arthroplasty: The Clinical Practice Guidelines of the American Association of Hip and Knee Surgeons, American Society of Regional Anesthesia and Pain Medicine, American Academy of Orthopaedic Surgeons, Hip Society, and Knee Society. J Arthroplasty. 2020;35(10):2709–14.

    Article  PubMed  Google Scholar 

  45. Stubhaug A, Grimstad J, Breivik H. Lack of analgesic effect of 50 and 100 mg oral tramadol after orthopaedic surgery: a randomized, double-blind, placebo and standard active drug comparison. Pain. 1995;62(1):111–8.

    Article  PubMed  Google Scholar 

  46. Kleeman-Forsthuber L, Pollet A, Johnson RM, Boyle J, Jennings JM, Dennis DA. Evaluation of Low-Dose Versus High-Dose Opioid Pathway in Opioid-Naive Patients After Total Knee Arthroplasty. Arthroplast Today. 2022;14:81–5.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Solomon DH, Rassen JA, Glynn RJ, Garneau K, Levin R, Lee J, et al. The comparative safety of opioids for nonmalignant pain in older adults. Arch Intern Med. 2010;170(22):1979–86.

    Article  PubMed  Google Scholar 

  48. Solomon DH, Rassen JA, Glynn RJ, Lee J, Levin R, Schneeweiss S. The comparative safety of analgesics in older adults with arthritis. Arch Intern Med. 2010;170(22):1968–76.

    Article  PubMed  Google Scholar 

  49. Zeng C, Dubreuil M, LaRochelle MR, Lu N, Wei J, Choi HK, et al. Association of Tramadol With All-Cause Mortality Among Patients With Osteoarthritis. JAMA. 2019;321(10):969–82.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Wei J, Lane NE, Bolster MB, Dubreuil M, Zeng C, Misra D, et al. Association of Tramadol Use With Risk of Hip Fracture. J Bone Miner Res. 2020;35(4):631–40.

    Article  CAS  PubMed  Google Scholar 

  51. Musich S, Wang SS, Schaeffer JA, Slindee L, Kraemer S, Yeh CS. Safety Events Associated with Tramadol Use Among Older Adults with Osteoarthritis. Popul Health Manag. 2021;24(1):122–32.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Rosenthal NR, Silverfield JC, Wu SC, Jordan D, Kamin M. Tramadol/acetaminophen combination tablets for the treatment of pain associated with osteoarthritis flare in an elderly patient population. J Am Geriatr Soc. 2004;52(3):374–80.

    Article  PubMed  Google Scholar 

  53. Babul N, Noveck R, Chipman H, Roth SH, Gana T, Albert K. Efficacy and safety of extended-release, once-daily tramadol in chronic pain: a randomized 12-week clinical trial in osteoarthritis of the knee. J Pain Symptom Manage. 2004;28(1):59–71.

    Article  CAS  PubMed  Google Scholar 

  54. Aweid O, Haider Z, Saed A, Kalairajah Y. Treatment modalities for hip and knee osteoarthritis: A systematic review of safety. J Orthop Surg (Hong Kong). 2018;26(3):2309499018808669.

    Article  PubMed  Google Scholar 

  55. Sturmer T, Rothman KJ, Avorn J, Glynn RJ. Treatment effects in the presence of unmeasured confounding: dealing with observations in the tails of the propensity score distribution–a simulation study. Am J Epidemiol. 2010;172(7):843–54.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Cole SR, Hernan MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol. 2008;168(6):656–64.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Zullo AR, Danko KJ, Moyo P, Adam GP, Riester M, Kimmel MJ, et al. HRQ Comparative Effectiveness Technical Briefs. Prevention, Diagnosis, and Management of Opioids, Opioid Misuse, and Opioid Use Disorder in Older Adults. Rockville (MD): Agency for Healthcare Research and Quality (US); 2020.

    Book  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

A.R.Z is funded in part by awards R01AG077620 and R01AG065722 from the National Institute on Aging. The funder was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Author information

Authors and Affiliations

Authors

Contributions

E.B. contributed to the conception and design of the work, analysis and interpretation of data, drafting and substantively revising the work, and approval of the submitted version. M.R.R. contributed to the interpretation of data, drafting and substantively revising the work, and approval of the submitted version. F.L.B, A.J.S., S.G., V.M., and A.R.Z. contributed to the conception and/or design of the work, analysis and/or interpretation of data, substantively revising the work, and approval of the submitted version. All authors agree to be personally accountable for the authors’ own contributions and to ensure that questions related to the accuracy or integrity of any part of the work, even ones in which the author was not personally involved, are appropriately investigated, resolved, and the resolution documented in the literature.

Corresponding author

Correspondence to Melissa R. Riester.

Ethics declarations

Ethics approval and consent to participate

The Brown University Institutional Review Board approved the study (protocol# 2010002823). Due to the use of deidentified administrative data, the need for informed consent was waived.

Consent for publication

Not applicable.

Competing interests

E.B. reports prior grants from Sanofi Pasteur paid directly to Brown University for research on the epidemiology of infections and vaccinations in older nursing home residents. F.B. works as a Senior Medical Advisor for the Institute for Clinical and Economic Review with funds paid directly to Brown University. A.J.S. reports royalties from Wolters-Kluwer and Springer-Nature. S.G. reports grants from Seqirus, Sanofi; and consulting or speaker fees from Sanofi, Seqirus, Merck, and the Gerontological Society of America related to vaccines or nursing home care quality. V.M. is a paid consultant and Scientific Advisory Board chair to NaviHealth, Inc. which offers post-acute care management and services to health plans and health systems. A.R.Z. reports prior grants from Sanofi Pasteur paid directly to Brown University for research on the epidemiology of infections and vaccinations in older nursing home residents as well as respiratory syncytial virus in infants. Some authors are VA employees [S.G., V.M., A.R.Z.]. The content and views expressed in this article are those of the authors and do not necessarily reflect the position or official policies of the United States Government or the US Department of Veterans Affairs.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Bosco, E., Riester, M.R., Beaudoin, F.L. et al. Comparative safety of tramadol and other opioids following total hip and knee arthroplasty. BMC Geriatr 24, 319 (2024). https://doi.org/10.1186/s12877-024-04933-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12877-024-04933-2

Keywords