Effect Of Illicit Direct To Consumer Advertising On Use Of .

RESEARCH43English speakingFrench speakingDifference21First national news DTCA0200220032004200520062007YearFig 1 Number of etanercept prescriptions per 10 000population per month in Canadian provinces that arepredominantly English speaking (n 8) or French speaking (n 1).Vertical line indicates start of US advertising in January 2003.Difference between rates shown at bottom of chart; fitted trendline shows predicted differences from interrupted time seriesregression. DTCA direct to consumer advertisingStudy drugsDifferences exist between the United States andCanada in terms of availability and approval dates fordrugs.21 Consequently, we sought out drugs that wereincluded in US marketing campaigns started betweenJanuary 2003 and December 2005; not advertised onCanadian television with a brand name; and approvedfor use in Canada before US advertising, to allowestimation of the marginal effect of direct to consumeradvertising on prescribing.On the basis of these characteristics, we identifiedthree study drugs. The first eligible drug wasetanercept (Enbrel), a biological agent approved inCanada for the treatment of symptom refractoryrheumatoid arthritis. Direct to consumer advertisingfor etanercept started in January 2003, and USnetwork news advertising started in March 2003.20The second eligible drug was mometasone (Nasonex),an inhaled nasal steroid spray for symptoms ofallergy. Direct to consumer advertising for mometasone started in December 2004, and the Vanderbiltdatabase showed extensive US news advertisingstarting the same month.20 Thirdly, tegaserod (Zelnorm) is a serotonin receptor agonist approved for thetreatment of constipation predominant irritablebowel syndrome in women. When released, it wasthe only drug approved specifically for this indicationin Canada. Although direct to consumer advertisingbegan in February 2003, tegaserod’s most influentialand major campaign first aired in August 2003 andfeatured memorable written messages such as “I feelbetter” on actresses’ stomachs.6 This later campaignwas considered very successful from a marketingperspective, and even won major advertising industryawards, before the drug was withdrawn in bothCanada and the United States owing to concernabout cardiac side effects.6 22 The Vanderbilt databaseindicates that US newscast advertising for tegaserodoccurred in 1-12 September 2003 and subsequently inMarch 2004.20BMJ ONLINE FIRST bmj.comAnalysisWe used regional differences in exposure to investigatethe impact of direct to consumer advertising. As all USadvertising was in English, we hypothesised thatchanges in prescribing in Canada would be concentrated in predominantly English speaking provinces.Although French speaking Canadians watch a similaramount of television, they view much less foreignsourced television, estimated at less than 5% of allviewing.16 Consequently, we analysed the difference inprescribing rates between predominantly Englishspeaking provinces (n 8) and Quebec, where Frenchis the mother tongue for more than 80% of thepopulation.23 Quebec is also attractive as a control asit has one of the least restrictive public drug formulariesin Canada but has comparable universal healthinsurance coverage, age, sex, and income profiles tothe other provinces.24 25We used interrupted time series analysis, one of thestrongest quasi-experimental designs available, toexamine longitudinal changes in Canadian prescribingrates.26 Firstly, we calculated the difference in theprescribing rate per 10 000 population by subtractingthe rate in French speaking provinces from that inEnglish speaking provinces. We then fitted time seriesmodels to test whether a statistically significant changeoccurred in the level or trend of the difference after thestart of US advertising or US national network newsadvertising, controlling for the pre-direct to consumeradvertising level and trend. This method simultaneously controlled for any pre-advertising differencesin the absolute level of prescribing between theprovinces as well as any differences in pre-advertisingtemporal trends related to changes in the rates ofprescribing between provinces. We also did a sensitivity analysis using the ratio of English and Frenchprescribing rates instead of the difference. The resultsand interpretation of this analysis (not shown) wereconsistent with those shown below. For drugs that100Prescriptions/10 000 population/monthPrescriptions/10 000 population/monthDownloaded from bmj.com on 3 September 200880604020English speakingFrench Fig 2 Number of mometasone prescriptions per 10 000population per month in Canadian provinces that arepredominantly English speaking (n 8) or French speaking (n 1).Vertical line indicates start of US advertising in December 2004.Difference between rates shown at bottom of chart; fitted trendline shows predicted differences from interrupted time seriesregressionpage 3 of 7

RESEARCHshowed any significant impact of direct to consumeradvertising in Canada, we did a sensitivity analysisusing data from US Medicaid programmes. We alsodid this with an interrupted time series analysis butwithout an “unexposed” US control group. Althoughthis method is uncontrolled compared with theCanadian analyses, it still controlled for pre-direct toconsumer advertising trends in drug use. We used ageneralised least squares model allowing for a firstorder autoregressive correlation between consecutivemonths or quarters and excluded the advertising startmonth in Canada. We validated our use of thisautocorrelation structure by using likelihood ratiotests. Moreover, alternative models with no or longerautocorrelation structures led to results with verysimilar estimates and identical interpretations.RESULTSTable 1 describes the US advertising campaigns andCanadian approval dates for the three study drugs. Allthree drugs had large direct to consumer advertisingexpenditures, ranging from US 194 million to 314million during the study period. Pre-advertising trendsin use for each of the study drugs were generallycomparable between English speaking and Frenchspeaking provinces (figs 1, 2 and 3). We found that USdirect to consumer advertising led to increasedCanadian prescribing rates for only one of the threedrugs, tegaserod.EtanerceptFigure 1 shows the times series of monthly prescribingrates of etanercept in Canada, which were very similarin both language regions. We found that advertisinghad no statistically significant impact on the level ortrend of differences in prescribing rate between Englishspeaking and French speaking provinces (level change 0.18 prescriptions per 10 000 population, 95% confidence interval 0.39 to 0.04, P 0.10; trend change 0.03 prescriptions per 10 000 population per month, 0.06 to 0.003, P 0.07).MometasoneFigure 2 shows the monthly prescribing rates formometasone in English speaking and French speakingCanadian provinces. As with etanercept, we saw noclinically important or statistically significant change inthe level or trend of differences in prescribing ratebetween English speaking and French speakingprovinces (level change 3.61 prescriptions per10 000 population, 10.51 to 3.29, P 0.30; trendchange 0.08 prescriptions per 10 000 population permonth, 0.57 to 0.40, P 0.73).TegaserodIn contrast to the first two drugs described, US direct toconsumer advertising for tegaserod seemed to have astrong influence on Canadian prescribing. Figure 3shows the monthly prescribing rates for tegaserod. TheFebruary 2003 campaign, which contained no USpage 4 of 7Prescriptions/10 000 population/monthDownloaded from bmj.com on 3 September 20082.52.0English speakingFrench speakingDifference1.5Second national news DTCA1.0First national news DTCA0.5020032004200520062007YearFig 3 Number of tegaserod prescriptions per 10 000population per month in Canadian provinces that arepredominantly English speaking (n 8) or French speaking (n 1).Vertical lines indicate start of US advertising in February 2003and start of new TV advertising campaign in August 2003.Difference between rates shown at bottom of chart; fitted trendline shows predicted differences from interrupted time seriesregression. DTCA direct to consumer advertisingnetwork news advertising, had no significant impact onprescribing rates and was incorporated into the preadvertising period. In contrast, a level increase of 0.56prescriptions per 10 000 population (0.37 to 0.76,P 0.001) in the difference in prescribing rate betweenEnglish speaking and French speaking provincesoccurred immediately after the August 2003 campaign.We found no statistically significant change in trend( 0.003 prescriptions per 10 000 population permonth, 0.03 to 0.02, P 0.77). Overall, this representsan estimated 42% increase in the first month after directto consumer advertising. However, this difference didnot persist despite continued advertising throughoutthe study period. Within two years of direct toconsumer advertising, prescribing rates were againvirtually identical between English speaking andFrench speaking regions.Using the same start date for direct to consumeradvertising, we found a similar increase in Medicaidprescription rates of tegaserod. Figure 4 shows that thepre-advertising upward trend in tegaserod use wassubstantially higher in US Medicaid than in Canada.After national network news direct to consumeradvertising, we saw an increase in the level ofprescribing in the United States; the number ofprescriptions per 10 000 enrolees increased by 5.70(3.65 to 7.75, P 0.001). As in Canada, we found nostatistically significant change in prescribing trends( 0.62 prescriptions per 10 000 enrolees per quarter, 1.52 to 0.27, P 0.15). Overall, the estimated increasein prescribing in the first quarter of direct to consumeradvertising was 56% higher than would have beenexpected and greater than the 42% increase seen inCanada.DISCUSSIONDuring the past decade, drug manufacturers havesubstantially increased spending on direct to consumeradvertising.1 To our knowledge, this study is the firstanalysis that uses a concurrent control group toBMJ ONLINE FIRST bmj.com

RESEARCHPrescriptions/10 000 enrolees/quarterDownloaded from bmj.com on 3 September 2008302520151052003200420052006YearFig 4 Number of tegaserod prescriptions per 10 000 enroleesper quarter in US Medicaid programmes. Vertical line indicatesstart of new TV advertising campaign in third quarter of 2003.Fitted trend line represents fitted interrupted time seriesanalysis for rate of use in Medicaidevaluate the impact of such advertising on use ofspecific drugs. We found that for two of three drugs theUS direct to consumer advertising had no apparentimpact on Canadian prescribing rates, and for one drug(tegaserod) we saw a short lived effect. These mixedfindings are surprising, as we included several expensive advertising campaigns that were highly recalled byconsumers.27 28 Our empirical results raise importantquestions about whether and how prescribing trendsfor specific drugs respond to advertising directed atconsumers. Thus, they have important implications forthe ongoing debate about the benefits and harms ofdirect to consumer advertising.Possible explanationsWe believe that the differential responses to direct toconsumer advertising that we saw may be related to thecharacteristics of the drugs examined. Although all ofthe study drugs are primarily used for relievingsymptoms, they differ in important ways. For example,etanercept requires referral to a specialist and intravenous administration, making the pathway betweendirect to consumer advertising and drug use complicated. Thus, the effect of advertising probably differssubstantially from that of drugs prescribed predominantly in primary care settings. Furthermore, tegaserod, unlike the other study drugs, was the only drugapproved for its indication in Canada.29 30 In contrast,the other drugs studied all had competitors within thesame drug class. In such markets, direct to consumeradvertising might protect against drops in levels of use,rather than expanding use. Other characteristics, suchas effectiveness, may also be important. A metaanalysis of short term placebo controlled trials oftegaserod indicates that the number needed to treat forone patient to have some improvement in theirgastrointestinal symptoms is about 17, suggesting thatmost patients trying tegaserod for the first time wereunlikely to derive symptomatic benefit.30 This mayexplain, in part, why the changes in use for this drugwere short lived.Our results also suggest that when direct to consumeradvertising does increase use, a dose-response relationBMJ ONLINE FIRST bmj.comwith the level of exposure to advertising exists. Ourresults in US Medicaid programmes estimated a largerincrease than in Canada, in both absolute andpercentage terms. Although the immediate change inuse in the United States was larger than in Canada,assessing the comparative long term effect of advertising in the United States is difficult, because noconcurrent control group is available. Nevertheless,the observed US Medicaid prescribing rates returnedto the pre-direct to consumer advertising trend aroundthe same time as in Canada (mid-2005). Furthermore,use of tegaserod was both higher and growing faster inMedicaid before direct to consumer advertising,suggesting that other factors were driving thesedifferential trends. For example, we cannot rule outbetween country differences in physician directedmarketing activities.31Strengths and limitationsThe major strength of our study is the use of a strongquasi-experimental design with a comparable andconcurrent control group. Moreover, our study designcontrolled for difference in both pre-existing level andtrend and explicitly considered the timing of advertising campaigns. This method controls for differences incharacteristics between language regions of Canadathat remained constant or changed predictably overtime, such as culture or patterns of general medicalpractice. Indeed, other differences such as variation inprovincial drug reimbursement plans would bias ourresults only if they coincidentally changed when theindividual direct to consumer advertising campaignsstarted. We could find no evidence that this occurredfor any of the drugs studied. None the less, exclusionfrom provincial formularies might constrain the effectsof successful advertising campaigns. However, mostprivate insurance plans in Canada do not haveformularies and cover most of the population.32 Moreover, although Ontario and Alberta both excludedtegaserod from their public drug programmes, theeffect of direct to consumer advertising was apparent inboth provinces (data not shown).The study has other limitations. Firstly, generalisingbeyond the three drugs that met our inclusion criteria isdifficult. Secondly, we do not have information onwhether these drugs were subject to disease awarenessadvertising by companies that did not mention thebrand name. However, this would bias our results onlyif it was similarly timed, and we found no indication formometasone or etanercept of increased use coincidentwith branded direct to consumer advertising, thusmaking it unlikely. Thirdly, variation in drug coverage,the overall health system, culture, levels of exposure toadvertising, or television viewing patterns might resultin the effect of direct to consumer advertising differingbetween drugs and between countries. However, thepercentage increase in and duration of effect fortegaserod was similar in both countries. In terms ofdrug coverage, more than 60% of Canadians arecovered by generally unrestrictive employer basedprivate drug plans, and less than 20% of these plans usepage 5 of 7

RESEARCHDownloaded from bmj.com on 3 September 2008WHAT IS ALREADY KNOWN ON THIS TOPICAlthough direct to consumer advertising (DTCA) of prescription drugs remains controversial, nocontrolled studies have investigated its impact on prescribingIn the absence of such evidence, both opponents and proponents of DTCA have generallyassumed it to be highly effective at increasing the use of advertised drugsWHAT THIS STUDY ADDSDTCA campaigns seem to have mixed effectiveness; drug use did not increase for two of threedrugs studiedDespite prohibitions, DTCA can influence prescribing across national bordersThe drug (tegaserod) for which use increased with DTCA was eventually withdrawn owing tosafety concernsformularies to limit access to specific drugs.32 Althoughthe Canadian and US health systems vary substantially,studies indicate similar access to primary care andwillingness of physicians to fulfil patients’ requests forspecific prescription drugs in the two countries.15 33Overall, the striking initial effect of direct to consumeradvertising on tegaserod prescribing rates providesevidence that exposure to US advertising is sufficient toinfluence Canadian prescribing.ImplicationsThe implications of our analysis are threefold. Firstly, itindicates that illicit cross border exposure to direct toconsumer advertising has the potential to modify druguse, even where such advertising is technicallyprohibited. As advertising over global mediums suchas the internet increases, this phenomenon may grow inimportance. Secondly, to our knowledge, these resultsare the strongest evidence that direct to consumeradvertising can increase use of a drug that was removedfrom the market as a result of concerns about safety.Finally, our findings suggest that the impact of direct toconsumer advertising campaigns is mixed, as theyseem to work for some drugs and not others. If theoverall impact of direct to consumer advertising islimited or variable, then a substantial portion ofexpenditure on such advertising—borne by governments, insurers, and patients in the form of higher costsor by companies as reduced profits—may be betterspent elsewhere. Previous commentary may haveoveremphasised the impact of direct to consumeradvertising for many individual drugs for whichevidence that it increases use is either weak or nonexistent.2 Until we better understand how direct toconsumer advertising modifies prescribing for particular drugs, debates about its positive and negativeconsequences will continue to be based on conjecturerather than strong evidence.We are grateful to IMS Health Canada, ESI Canada, and Eloda for providingdata for this research. We also thank Alyce Adams, Katherine Swartz, CoryCowan, El

Sep 03, 2008 · consumer advertising campaigns came from TNS Media Intelligence. The dataset tracks advertising and estimates expenditure across several media, including tel