Isotretinoin Laboratory Test Monitoring--A Call to Decrease Testing in an Era of High-Value, Cost-Conscious Care

Healthcarecostsare increasing faster in theUnitedStates than in any other industrialized nation, with estimates of lowvalue services—services that can be eliminated without adversely affectingpatient care—composingup to 30%ofhealth care spending.1,2(pp65-108)3 Consistent with the Physician Charter on Medical Professionalism,physicianshave a responsibility to balance individual patient needswith “care that isbasedonthewiseandcost-effectivemanagementof limitedclinical resources,” including“scrupulousavoidanceof superfluous tests.”3(p245) Specialty-specific efforts are critical to identifying the relevant areaswhere cost-effective treatment can be delivered in balance with quality and efficiency; dermatologists havebegun to join thenational efforts toward this end.4,5 In this issueof JAMADermatology, Lee andcolleagues6 present their contribution in assessing laboratory test monitoringduring isotretinoin therapy foracne,much-neededwork to create a strong foundation of information on which to establish clear recommendations for safe, high-value, costconscious care within our specialty practice. The focus of their article is isotretinoin therapy, a medicationcommonlyprescribedbydermatologists that todatehas had limited evidence-based recommendations for laboratory monitoring during treatment. A 2004 consensus statement published by the American Academy of Dermatology previously recommended periodic monitoring of serum triglyceride and cholesterol levels, as well as monitoring of transaminase levels, citing lack of consensus onmonitoring additional laboratory parameters.7 The isotretinoin package insert provides general guidelines for monitoring for lipid abnormalities before starting therapy and thenatweekly or biweekly intervals “until the lipid response to [isotretinoin] is established, which usually occurswithin 4weeks”8(p7) ormore frequently inhigh-riskpatients;weeklyor biweeklymonitoring forhepatotoxicity is also recommended.Monthly, as well as pretreatment and posttreatment, pregnancy testing for female patients is recommended in strict accordancewith the iPLEDGE registry system. Theactualpracticeofdermatologistsprescribing thismedication is poorly understood. Audience response data from attendees of a 2014AmericanAcademyofDermatology symposiumon isotretinoinmonitoring showedsignificant variability in thewaydermatologists screen.Before thepresentation, the audiencewas asked“what typeofmonitoringdoyoudo inpatients taking isotretinoin?”Of the 2343 anonymous electronic audience responses, 33% order baseline and monthly complete blood cell count, 26%order baseline andmonthly transaminase level testing, 13%order baseline andmonthly cholesterol panels, 2.4% order baseline and monthly renal function and/orelectrolyte level tests, 13.8%orderbaselineandmonthly pregnancy screening in female patients, 0.21% do not order monitoring laboratory tests inmale patients, and8.42%order laboratory test monitoring distinct from the options presented. These data suggest a clear and potentially important practice gap regarding isotretinoin monitoring. In their report, Lee and colleagues6 performed a systematic review looking at mean laboratory values and adverse events of isotretinoin therapy, capturing both analyzed data, as well as a narrative summary of the data excluded from the systematic review. The key finding was that although adverse effectswere noted for all laboratory tests (especially for cholesterol and triglyceride levels), severe or high-risk effects were rare. A critical consideration is that the authors reported changes inmeanvalueof laboratory test resultswithin a population although these are likely not themost clinically relevant to individual patient management. The second important findingwas that timingmatters; most laboratory test abnormalities were detected early (typically between 6 and 8 weeks) during the treatment period. There were no substantial late adverse effects of isotretinoin therapy, at least for the parameters measured—triglyceride and total cholesterol levels—corroboratingearlier studies, although laterdatawerenot available for all laboratory parameters examined. A thirdmajor finding of the study was that a wide range of discontinuation rateswasnoted (between0.71%and22.5%,mean5%). Importantly, thedirect relationshipbetweenabnormal laboratory test findings and therapy discontinuation was not reported. An earlier small study (515 patients, 574 isotretinoin treatment courses) published in 2013 showed that while isotretinoin adverse effects are relatively rare, the rate of laboratory test abnormality resulting in treatment discontinuation was exceedingly low—suggesting that screening may not ultimately affect the course of treatment.9 This is a central point because the latter study supports the notion of reduced laboratory testmonitoring as not only cost conscious but also providinghigh-valuecareduringisotretinointreatment.Inthepresentstudy, theauthorsused integrationof thesystematic review and thenarrative analyses to inform their clinical recommendations for individual patient care; in their conclusion, they argue that monthly laboratory screening is not indicated for patients receiving standard doses of isotretinoin. There are additional considerations in the interpretation of this study.Due to the rigorous inclusion criteria,most of the Related article page 35 Opinion