FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2938761009> ?p ?o ?g. }

• W2938761009 endingPage "896" @default.
• W2938761009 startingPage "895" @default.
• W2938761009 abstract "To the Editor: Long-term prescription opioid use is a major risk factor for addiction and overdose, and thus I read with great interest the study by Basilico et al,1 who reported that the probability of long-term opioid use did not differ by the type of opioid prescribed after orthopaedic surgery. The authors conclude: “Our analysis suggests that the notion that some opioid types are intrinsically more dangerous than others is a myth.” I write to express concern about the internal validity of the study. There are four problem areas: (1) Treatment variable: Among patients who receive more than one opioid prescription, many switch drugs to obtain a satisfactory balance between analgesia and adverse effects, a practice known as “opioid rotation.”2 In the Basilico et al study, the impact of only the first opioid was examined (“We assessed whether the initial opiate type prescribed to postoperative, opiate-naive orthopaedic trauma patients was associated with prolonged opioid use”). The authors did not estimate the effect of the successor opioids among the “switchers,” and thus ignored important differences in opioid treatments. Exclusion of those differences from the statistical analysis may have attenuated or suppressed the association between opioid type and long-term use. (2) Outcome variable: Long-term use was measured as a binary variable—presence or absence. (The authors used the terms “prolonged use,” “persistent use,” and “long-term use” interchangeably.) They defined the presence of long-term use as “the receipt of at least one opioid prescription within 90 days of injury presentation and another within 90 to 180 days postoperatively,” which means a total of two prescriptions separated by months would qualify. It is instructive to compare the definition of long-term use recommended by the Consortium to Study Opioid Risks and Trends (CONSORT), a National Institute on Drug Abuse-supported initiative. CONSORT defined long-term use as opioid treatment lasting at least 90 days, with at least 10 prescriptions and/or 120 days supply.3 Clearly, the Basilico et al definition would classify a much broader range of patients as long-term users than CONSORT, including patients whose opioid use was minimal and infrequent. By grouping patients with dissimilar prescription patterns into a single category of long-term user, the Basilico et al definition undercuts the meaning and purpose of the category. It may also incline the statistical analysis toward a finding of no difference among opioids types. (3) Secular trends: The subjects in this retrospective study had been orthopaedic trauma patients between 2002 and 2015. This was a period of large-scale change in opioid prescribing rates in the United States, with sharp growth through 2011 and then a small contraction.4 As the likelihood of receiving a prescription with renewals varied greatly over time, as did prescriber preferences for specific opioids,5 the Basilico et al analysis should have adjusted for the secular trends. The unadjusted trends may have predisposed the study to a false null. (4) Confounding by indication: Observational research of treatment alternatives is vulnerable to the strong effects of confounding by indication.6 Because the patients in the Basilico et al study were not randomly assigned to opioid treatment groups, a core concern is whether there were unmeasured ways in which the treatment groups differed before opioid treatment was received that may have affected the outcome—that is, patient-specific indications (apart from the type of injury) that would lead a physician to prescribe one opioid drug over another, such as pain sensitivity, complications from surgery, psychiatric disorders, use of antidepressants or benzodiazepines, and prognosis. Instrumental variable methods could have been used to effectively pseudo-randomize the treatment groups before estimating the association between treatment drug choice and long-term opioid use. Unless unmeasured confounding is addressed, the parameter estimates are prone to bias.7 We are in the thick of a still-growing opioid disaster caused in part by long-term prescription opioid use.8 At the same time, physicians must respond to the ongoing need to reduce suffering from pain, and therefore the question of whether one opioid is intrinsically more dangerous than another remains urgently important. Rigorous research methods are required to answer the question." @default.
• W2938761009 created "2019-04-25" @default.
• W2938761009 creator A5032414925 @default.
• W2938761009 date "2019-12-01" @default.
• W2938761009 modified "2023-09-24" @default.
• W2938761009 title "Letter to the Editor" @default.
• W2938761009 cites W1965999023 @default.
• W2938761009 cites W1969503074 @default.
• W2938761009 cites W1990355279 @default.
• W2938761009 cites W2100604639 @default.
• W2938761009 cites W2588778109 @default.
• W2938761009 cites W2594283925 @default.
• W2938761009 cites W2895064404 @default.
• W2938761009 doi "https://doi.org/10.5435/jaaos-d-19-00054" @default.
• W2938761009 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/31008875" @default.
• W2938761009 hasPublicationYear "2019" @default.
• W2938761009 type Work @default.
• W2938761009 sameAs 2938761009 @default.
• W2938761009 citedByCount "1" @default.
• W2938761009 countsByYear W29387610092019 @default.
• W2938761009 crossrefType "journal-article" @default.
• W2938761009 hasAuthorship W2938761009A5032414925 @default.
• W2938761009 hasConcept C118552586 @default.
• W2938761009 hasConcept C126322002 @default.
• W2938761009 hasConcept C170493617 @default.
• W2938761009 hasConcept C177713679 @default.
• W2938761009 hasConcept C2426938 @default.
• W2938761009 hasConcept C2778140631 @default.
• W2938761009 hasConcept C2781063702 @default.
• W2938761009 hasConcept C42219234 @default.
• W2938761009 hasConcept C48856860 @default.
• W2938761009 hasConcept C71924100 @default.
• W2938761009 hasConcept C98274493 @default.
• W2938761009 hasConceptScore W2938761009C118552586 @default.
• W2938761009 hasConceptScore W2938761009C126322002 @default.
• W2938761009 hasConceptScore W2938761009C170493617 @default.
• W2938761009 hasConceptScore W2938761009C177713679 @default.
• W2938761009 hasConceptScore W2938761009C2426938 @default.
• W2938761009 hasConceptScore W2938761009C2778140631 @default.
• W2938761009 hasConceptScore W2938761009C2781063702 @default.
• W2938761009 hasConceptScore W2938761009C42219234 @default.
• W2938761009 hasConceptScore W2938761009C48856860 @default.
• W2938761009 hasConceptScore W2938761009C71924100 @default.
• W2938761009 hasConceptScore W2938761009C98274493 @default.
• W2938761009 hasIssue "23" @default.
• W2938761009 hasLocation W29387610091 @default.
• W2938761009 hasLocation W29387610092 @default.
• W2938761009 hasOpenAccess W2938761009 @default.
• W2938761009 hasPrimaryLocation W29387610091 @default.
• W2938761009 hasRelatedWork W2067515587 @default.
• W2938761009 hasRelatedWork W2067833211 @default.
• W2938761009 hasRelatedWork W2183067126 @default.
• W2938761009 hasRelatedWork W2277036766 @default.
• W2938761009 hasRelatedWork W2408827505 @default.
• W2938761009 hasRelatedWork W2460320941 @default.
• W2938761009 hasRelatedWork W3026380818 @default.
• W2938761009 hasRelatedWork W4200501282 @default.
• W2938761009 hasRelatedWork W4322707876 @default.
• W2938761009 hasRelatedWork W46559933 @default.
• W2938761009 hasVolume "27" @default.
• W2938761009 isParatext "false" @default.
• W2938761009 isRetracted "false" @default.
• W2938761009 magId "2938761009" @default.
• W2938761009 workType "letter" @default.