Early in theHIVepidemic,patientswithAIDSwerenotalways diagnosed as having the disease, in some cases because their opportunistic infections (eg, disseminated histoplasmosis) were not yet recognized as AIDS-defining conditions. Sincethen,therehavebeenremarkableadvances inboththediagnosticcapabilitiesandtherapeuticoptions inthetreatmentofHIVdisease.Thisprogresshas resulted in restorationof severely ill patients’ immune systems and previously lethal HIV infection becoming a chronic disease. There has seldom been such successful and rapid progress for adiseasewithhighmorbidity and short life expectancy (eg, in the 1980s,>40%ofperinatally infectedchildrendiedbefore 4 years of age, whereas today, with optimal care, <1% die before 4 years of age1). However, the epidemic is not over, too manypeople avoid testing and too fewengage in care, andprophylaxis and treatment budgets are limited. The success of 3 ormore antiretroviral drugs (combination antiretroviral therapy [ART]) that decrease the viral capability to select multiple mutations necessary to confer resistance to 3-drugregimenshas resulted in lifelongsuppressionofHIVreplication in adherent patients. However, the outcome is less favorable in nonadherent patients, who often experience treatment failure and development of drug resistance.2 The use of combinationARTcanalso leadtomultipleadverseevents, some that are serious (eg, liver failure, cardiovascular disease), and some that cause significant morbidity (eg, diarrhea, anemia). Severalstudiessuggestedthatreductioninpillburdenandoncedailydosingincreasedadherence.3,4Althoughthesestudiesmay have been biased by comparing different drugs (eg, twicedaily lopinavir vs once-daily efavirenz) that may have different tolerability, toxicity, or both, the general consensus is that single daily dosing improves adherence.5 Inapreviousstudy,Coovadiaetal6 evaluatedwhetherchildren initially treated with ritonavir-boosted lopinavir–based therapy could be safely switched to nevirapine-based therapy soon after achieving viral load suppression and found that resistance tonevirapine selectedduringexposure to thisdrug for preventionofmother-to-child transmission led toahigher rate of virologic failure in childrenwho transitioned to nevirapine. In this issueof JAMA, Coovadia andcolleagues7 report ona randomized, open-label noninferiority trial in Johannesburg, South Africa, to evaluate whether nevirapine-exposed children who achieved initial viral suppression while receiving ritonavir-boosted lopinavir could transition to efavirenz without risk of viral failure. The authors enrolled 298 HIV-infected children aged 3 years or older (mean age, 4.3 years) exposed to nevirapine forpreventionofmother-to-child transmissionwho hadplasmaHIVRNAlevelsof less than50copies/mLwhile taking ritonavir-boosted lopinavir. Patients were randomized to switch to efavirenz (n = 150) or to continue taking ritonavirboosted lopinavir (n = 148) andwere followed up for 48weeks after randomization. The probability of viral rebound by 48 weeks was 17.6% (n = 26) in the efavirenz group and 28.4% (n = 42) in the ritonavir-boosted lopinavir group, and theprobabilitiesofviral failurewere2.7%(n = 4)withefavirenzand2.0% (n = 3) with ritonavir-boosted lopinavir. The authors conclude thatswitchingtoefavirenzvscontinuingritonavir-boosted lopinavir did not result in significantly higher rates of viral rebound or viral failure. The finding that transition to efavirenz-based therapywas better than previously observedwith transition to nevirapinebased therapy shouldnot be surprising, as several studies have previously described superior virologic and immunologic outcomeswithefavirenzvsnevirapine.8-10However, inmanycountries,nevirapineisstillcommonlyused,andsubstantialtreatment failureanddeathcontinuetooccurbecauseof the increasedrate ofvirologicfailurewithnevirapinecomparedwithefavirenz.The recentWorldHealthOrganization (WHO)guidelinescontinueto strongly recommend a nevirapine-based regimen as first-line combinationART for childrenyounger than3years if ritonavirboostedlopinavir–basedtherapyisnotfeasible.11 It ispossiblethat combinationsthat includenevirapinemaybepreferredbymany countriesbecauseoflowercost.Thisapproachmayleadtohigher failure rates,which in the long runmaybemore costly because of the increasedneed formore expensive second-line therapy. EventhoughCoovadiaetalsuggestthattheirstudy“provides evidence to support the safety and efficacy of switching to efavirenz,”7 severalquestions remain.ThecurrentWHOdosing recommendationforefavirenzmayresult inahigherproportion of children with excessive blood levels, which may lead to increased toxicity.12 Is it possible that thehighpercentageof childrenwithnightmaresandsleepingproblemsinthecurrentstudy (1 in4children)representssuchtoxicity?Howmanyofthesechildrenalsohadimpairedconcentrationandother, lessobviouscentralnervoussystemtoxiceffects?Althoughtheseadverseevents aremorecommoninthefirstfewweeksafterefavirenztreatment is started,an importantconsideration ishowmanypatientswill not adhere to the regimen during those weeks, increasing the chance for resistance to develop. Thepossible role of other factors affecting thepharmacokinetics andpharmacodynamics of efavirenz, such as nutritional status and developmental physiology(especially in those<3yearsofage, forwhomthedrugwas recently approved by the US Food and Drug Administration), Related article page 1808 Opinion