Objective: To examine levels of persistence andcompliance as well as the economic impact ofextended-release tolterodine (tolterodine ER) versusimmediate-and extended-release oxybutynin (oxybutyninIR or oxybutynin ER) among commerciallyinsuredpatients with overactive bladder (OAB).
Methods: Patients with OAB who initiated tolterodineER, oxybutynin IR, or oxybutynin ER betweenJanuary 2001 and December 2002 were identifiedfrom the PharMetrics Patient-Centric database; thefirst medication used in this timeframe was used fortreatment group assignment (ie, patients were only in1 group). Exploratory assessment of persistency andcompliance was conducted among all treatedpatients: subjects were matched 1:1 based on theestimated propensity score for tolterodine ER inremaining analyses. Measures included patient characteristicsas well as levels of medication, outpatientand inpatient resource utilization, and costs. Primarycomparisons were made descriptively; costs wereevaluated using generalized linear models with agamma distribution and log-link function.
Results: Compliance did not differ between tolterodineER (77.4%) and oxybutynin ER (74.3%), but waslower for oxybutynin IR (60.9%). Mean (+ standarddeviation) duration of therapy was higher for tolterodineER (139 + 132 days) versus oxybutynin ER (115 +122) and oxybutynin IR (60 + 85). Totals of 7257 and5936 matched pairs were available for tolterodine ERversus oxybutynin ER and oxybutynin IR comparisons,respectively. The mean age was 54 years in all groups;the majority was women. Utilization of outpatient andinpatient medical services was consistently loweramong tolterodine ER patients in both comparisons.Total costs were slightly lower for tolterodine ER versusoxybutynin ER ($8303 + $18 802 vs $8862 + $18 684)and oxybutynin IR ($9975 + $24 860 vs $10 521 +$22 602); differences were significant after multivariateadjustment.
Conclusions: Use of tolterodine ER results incomparable compliance to oxybutynin ER andlonger duration of use relative to either form of oxybutynin.In addition, tolterodine ER may be costeffectiverelative to oxybutynin IR or oxybutynin ERamong commercially-insured persons with OAB.
(Am J Manag Care. 2005;11:S140-S149)
Overactive bladder (OAB) is a commonand distressing medical conditionresulting in symptoms thatinclude urgency, with or without urinary frequency,and/or urgency incontinence.1 It isestimated that 16% to 17% of Americans sufferfrom OAB.2
In general, the symptoms of OAB canhave a debilitating effect on an individual bydisrupting quality of life (QOL), sleep patterns,and mental health.3 Costs associatedwith OAB also are substantial. In 2000, thetotal direct and indirect cost of OAB in theUnited States was estimated at $12.02 billion,with $9.17 billion and $2.85 billion atthe community and institution levels respectively.4 Costs are expected to escalate evenfurther as the prevalence of OAB in theUnited States increases because of the agingof the population.
The goals of OAB treatment are to reducesymptomalogy and improve QOL. There areseveral treatment options available for thiscondition, including behavioral modification,pharmacotherapy, and surgical intervention,with the most common being behavioral therapyand pharmacotherapy.5 Antimuscarinicagents have remained the mainstay of pharmacotherapyfor OAB since the launch ofoxybutynin in the 1970s. Although immediate-release oxybutynin (oxybutynin IR) hasproved effective in relieving the symptomsof OAB, it is associated with many undesirableside effects, subsequently reducingpatient adherence and compliance withtime.6 During the past several years, treatmentfor OAB has expanded to includenewer agents that reduce major side effectsand thereby improve adherence to therapy.Newer extended-release antimuscarinics,such as tolterodine, and the extended-releaseform of oxybutynin (oxybutynin ER) havedemonstrated improved efficacy, convenience,and a tolerability profile better thantheir immediate-release counterparts byoffering once-daily dosing and diminishedside effects.6-8 Several trial-based studies havebeen conducted in the past that compare efficacy,costs, and adherence in oxybutynin IRand immediate-release tolterodine (tolterodineIR) relative to their extended-releaseforms.9,10 The results of another recent analysisof data that evaluated extended-releasetolterodine (tolterodine ER) and controlled-releaseoxybutynin by using a 4-way crossoverdesign indicated that tolterodine ER was aseffective as the controlled-release oxybutyninbut with a better tolerability profile.11
Although there are numerous clinicalstudies that focus primarily on the cost andefficacy of immediate-release forms or theextended-release versions of oxybutyninand tolterodine separately,12-17 to date therehas been no large-scale retrospective studythat comprehensively assesses relevant outcomesassociated with tolterodine ER incomparison to both oxybutynin IR and oxybutyninER. Additionally, the availability ofvarious treatment options and the potentialcost implications associated with thesealternatives give rise to the need for comparisonsfrom a typical US clinical practiceperspective (ie, nonclinical trial populationand setting). The present study examinesadherence to OAB pharmacotherapy as wellas the impact of tolterodine ER versus oxybutyninIR and oxybutynin ER on the levelsof resource utilization and cost in a largecommercially-insured population withOAB.
of Diseases, Ninth Revision, Clinical
Current Procedural Terminology
Data were obtained fromthe PharMetrics Patient-Centric database,which is comprised of fully adjudicated medicaland pharmaceutical claims for morethan 46 million unique patients from 82health plans across the United States. Thedatabase includes both inpatient and outpatientdiagnoses (in  format) andprocedures (in and Health Care Common ProceduralCoding System formats) as well asboth standard and mail order prescriptionrecords. Data on prescription recordsinclude the National Drug Code as well asdays supplied and quantity dispensed. Bothhealth plan paid and charged amounts areavailable for all services rendered as well asdates of service for all claims. Additional dataelements include demographic variables (ie,age, sex, geographic region), plan type (eg,health maintenance organization [HMO],preferred provider organization [PPO]), payortype (eg, commercial, self-insured), providerspecialty, and plan enrollment dates.
All patients diagnosedwith OAB (596.5X [excluding596.53, 596.54], 788.3X, 788.41, 788.43)and at least 1 pharmacy claim for any of themedications of interest (ie, tolterodine ER,oxybutynin ER, oxybutynin IR) betweenJanuary 1, 2001, and December 31, 2002,were initially selected for inclusion in thestudy sample. The date of first medicationuse served as the patient's index date. Preindexand follow-up periods of 12 months'duration each were created in relation to theindex date. Claims for these patients werethen accumulated spanning the period ofJanuary 1, 2000, to December 31, 2003.
Patients not continuously eligible for drugand health benefits during their entire preindexand follow-up periods were excluded.Patients who had evidence of using an OABdrug of interest during the pre-index periodwere also excluded. In addition, patientsparticipating in plans without pharmacybenefits were excluded, as were patientsaged 65 years or older who were not enrolledin a Medicare "risk" plan (ie, an agreementby which a commercial health plan agrees toassume full financial risk for the coverage ofa Medicare beneficiary); the latter exclusionwas implemented to ensure that elderlypatients had complete reporting of medicaland pharmaceutical utilization.
All patients treated with tolterodine ER,oxybutynin ER, or oxybutynin IR who hadvalid information on pharmacy claims wereselected for the persistency and compliance analyses. Matched pairs of patientswere created to compare outcomes, resourceutilization, and costs in comparablegroups. To control for potential differencesbetween treatment groups, propensityscores were calculated using multivariatelogistic regression and a stepwise selectionprocedure. A propensity score represents aspecific patient's fitted probability (ie,propensity) of receiving a given treatmentoption and is calculated by summing coefficientvalues for a list of potential confoundingvariables. Use of these scoresconfers the advantage of having a singleestimate available to control for multipledifferences between treatment groups in amatched-pairs analytic design.18
Demographic and clinical characteristicsas well as pre-index variables were introducedas covariates into the model using stepwiselogistic regression techniques; those achievingsignificance at a level of <.05 wereretained. The outcome of interest was thepropensity for, or likelihood of, treatmentwith tolterodine ER. Separate propensityscores were created for each set of pair-wisecomparisons. Thus, patients treated withtolterodine ER were matched on a 1:1 basisto patients treated with oxybutynin ER andoxybutynin IR, and matching was conductedbased on a difference of + 0.01 in thisprobability.
The primary measures ofinterest for this study included persistenceand compliance as well as resource utilizationand costs of OAB-related and unrelatedservices, which were all measured over 12months of follow-up.
A number of demographic and clinicalcharacteristics, including age, sex, healthplan type (eg, HMO, PPO), geographic region,presence of selected pretreatment comorbidities(ie, Alzheimer's disease, chronicobstructive pulmonary disease, congestiveheart failure, diabetes, hypertension,ischemic heart disease, multi-infarct dementia,multiple sclerosis, Parkinson's disease,prostate hyperplasia, stroke, urinary retention,urticaria), and comorbidity score (ie,Charlson Index with Deyo modification),19were tracked for each treatment group.Among patients receiving an OAB drug ofinterest, persistence and compliance weremeasured. Persistence was calculated basedon the time (in days) from therapy initiationuntil the first discontinuation. First discontinuationwas defined as a gap in therapyexceeding 2 times the therapy days suppliedon the previous prescription. Compliancewas expressed as a ratio of the number oftherapy days supplied divided by the persistencemeasure described above. Persistencewas also calculated on a monthly basis over a12-month follow-up period from therapy initiation.Each patient's persistence durationwas recorded as a data point in the correctrange (ie, months 1-12) for when therapywas discontinued.
OAB-related and -unrelated resource utilizationand costs were assessed during thefollow-up period and included pharmaceuticalclaims indicated for OAB as well as allother pharmacy claims and medical claims(office management, emergency room, otheroutpatient, inpatient) related and unrelatedto OAB.
Where relevant, medical claims weredefined as OAB-related versus all othersbased on the presence of an OAB diagnosis.Costs were estimated based on healthplan payments for medications and servicesrendered and net of patient responsibility(ie, copayment, coinsurance, and/ordeductible). Costs were updated as necessaryusing the medical care component ofthe US Consumer Price Index. Non-OAB-related claims also were tallied andwere categorized in similar fashion (ie,pharmacy, outpatient/emergency room,inpatient).
Differences in patient characteristics,as well as costs during follow-up,were assessed using inferential statistics (ie,chi-square tests for categorical variables andWilcoxon rank-sum tests for continuous variables);<.05 was deemed to be statisticallysignificant. Generalized linear models werealso used to compare total healthcare costsbetween the matched patient subgroups. Independentvariables in the base modelsincluded age, sex, health plan type, geographicregion, treating physician specialty,pre-index comorbidities, and pre-indexCharlson Comorbidity score. All analyseswere conducted using Statistical AnalysisSoftware ([SAS], SAS Institute, Cary, NC),Version 8.2.
A total of 526 511 patients with a diagnosisof OAB or a claim for any of the OABdrugs were initially identified in thePharMetrics database during the timeframeof interest. After applying the exclusion criteria,33 067 patients remained available forstudy and had evidence of using 1 of thedrugs of interest (16 198, 8320, and 8549for tolterodine ER, oxybutynin ER, and oxybutyninIR, respectively) (Table 1). A totalof 26 770 patients had valid data for the persistencyand compliance analyses. Totals of7257 and 5936 matched pairs were availablefor the tolterodine ER versus oxybutyninER and tolterodine ER versusoxybutynin IR comparisons, respectively.The relatively low match rate for thetolterodine ER/oxybutynin IR comparisonwas likely the result of severity differencesbetween the overall cohorts, because oxybutyninIR patients were younger and hadlower levels of comorbidity relative totolterodine ER patients before matching(data not shown).
Compliance and Persistence.
In general,patients receiving medication for OABwere relatively compliant but not persistent(data not shown). The average compliancerate was 74% for the overall group;compliance was highest among tolterodineER patients (77.4%), whereas oxybutyninER and oxybutynin IR patients had lowercompliance rates of 74.3% and 61%, respectively.However, overall persistency onthese medications was low; three quartersof treated patients discontinued therapywithin 6 months. On average, patients werepersistent on therapy for 113 days (standarddeviation [SD] = 124). Again, mean(+ SD) persistence was highest among thepatients receiving tolterodine ER andlowest among patients receiving oxybutyninIR (139 [+ 132] days vs 115 [+ 122]vs 60 [+ 85] days for tolterodine ER, oxybutyninER, and oxybutynin IR, respectively).In addition, differences in persistence weresignificantly (all <.0001) in favor oftolterodine ER at months 1, 2, 3, and 12 forboth comparisons to oxybutynin ER andoxybutynin IR (data not shown).
Matched Pairs Comparisons
Demographic and Clinical Characteristics.
There were no substantive differencesin the demographic and clinicalcharacteristics in the tolterodine ER versusoxybutynin ER and the tolterodine ER versusoxybutynin IR treatment cohorts (Table 2).The average age within both groups was 54years; 19% of the patients were men in thetolterodine ER versus oxybutynin ER group,whereas 28% were men in the tolterodine ERversus oxybutynin IR matched group. Approximately21% to 23% of patients in both groupswere seen by a urologist. The most commoncomorbidities between both groups werehypertension (~31%-32%), diabetes (~11%-12%), and ischemic heart disease (~8%).Importantly, there were no observed differencesin multiple sclerosis or Parkinson's diseaseprevalence between the matched groups;the prevalence of corresponding neurogenicbladder was therefore also likely to be similar.
Resource Utilization and Cost.
Informationon healthcare utilization is presentedin Table 3 for each of the comparisonsof interest. After initiation of therapy,tolterodine ER and oxybutynin ER patientsaveraged 4 to 5 prescriptions per patient forindex therapy (4.79 vs 4.33, respectively),whereas oxybutynin IR users received significantlyfewer prescriptions (3.10 perpatient) during this period (<.0001). Inaddition, although tolterodine ER patientshad significantly higher numbers of prescriptions(<.05) for sympathomimetics,the numbers of claims for all other medicationsdid not differ; also, tolterodine ERpatients had a significantly lower utilization (<.01) of selected OAB-related servicescompared with the oxybutynin ERgroup. In comparison, tolterodine ER usersin the tolterodine ER/oxybutynin IR cohortrequired a higher number of all medicationtypes (<.05) compared with the oxybutyninIR group. Use of OAB-related medicalservices was higher in oxybutynin IR users(<.05), because most utilization was manifestedin outpatient ancillary claims. "Allother" services were also higher among theoxybutynin IR users; this was noticeableacross most inpatient and outpatient services(<.05).
Within the "all other" category, the mostcommon coded diagnoses of interest wererelated to urinary tract infections, pelvicand/or genital pain, and other nonspecificurinary symptoms. Other prevalent conditionsincluded those commonly found in thispopulation demographic (eg, hypertension,lipid disorders, arthritis and arthropathies,diabetes).
Annual total pharmacy-related costs afterinitiation of index therapy were similarbetween the tolterodine ER and the oxybutyninER groups (Table 4). However, totalOAB-related costs were significantly lower fortolterodine ER users (= .0018) versus thosein the oxybutynin ER group ($188 [$1224] vs$279 [$1783] for tolterodine ER and oxybutyninER, respectively), primarily because ofdifferences in the costs of hospitalizationand ancillary services. Additionally, total"all other" costs were significantly higher(approximately $400 on average) for the oxybutyninER group than for the tolterodine ERgroup ($5575 [$17 532] vs $5993 [$16 875]for tolterodine ER and oxybutynin ER respectively;= .0014), primarily because of higheruse of outpatient and ancillary services.This increased trend of higher outpatient andinpatient costs across all services in the oxybutyninER group was responsible for triggeringan increase of ~7% in grand totalhealthcare costs compared with the tolterodineER group ($8862 [$18 684] vs $8303[$18 802]; = .0109).
With respect to pharmacy costs, averageannual costs of tolterodine ER were muchhigher than for oxybutynin ER ($358 [$341]vs $56 [$139]; <.0001). Patients usingtolterodine ER also had significantly higher"other pharmacy" related costs comparedwith oxybutynin IR users ($2791 [$4997] vs$2204 [$3944]; <.0001). However, therewas a statistical trend toward lower OAB-relatedmedical costs among tolterodine ERpatients ($215 [$1470] vs $272 [$2433] foroxybutynin IR; = .0591).
Despite higher pharmacy costs, totalhealthcare costs were nominally lower (albeitnonsignificantly) among tolterodine ERusers ($9975 [$24 860] vs $10 521 [$22 602];= .3612), primarily because of reductionsin the costs of hospitalization and ancillaryservices.
Multivariate Analyses of Cost
Results of multivariate analyses of grandtotal costs for the tolterodine ER versusoxybutynin ER cohort indicate thatpatients in the oxybutynin ER group hadsignificantly (= .0032) higher total costscompared with tolterodine ER. In thisanalysis, men and younger patients wereassociated with reduced drug costs, whereaspatients treated by a urologist were linkedwith significantly higher total costs. Thepresence of any of the pre-index comorbiditieswas also associated with increasedcosts. Total costs were significantly lowerfor the tolterodine ER group compared withthe oxybutynin IR group (<.0001).Women in this comparison also contributedto lower total costs. Age, plan type, geography,and most pre-index comorbidities werealso important factors in determining overallcosts.
In an effort to better understand the economicoutcomes among patients receivingtolterodine ER in comparison to both oxybutyninER and oxybutynin IR in a large,commercially-insured cohort, a retrospectiveanalysis of pharmacy and medicalclaims for patients diagnosed with OAB wasundertaken. The study spanned the periodJanuary 1, 2000, to December 31, 2003, andfocused attention on resource utilization,costs, compliance, and persistency. Ofthose patients receiving OAB pharmacotherapy,tolterodine ER patients had significantlylower pharmacy utilization andused fewer outpatient and inpatient servicescompared with either oxybutynincohort. In addition, patients receivingtolterodine ER incurred lower overallhealthcare costs, suggesting that its usemay be a cost-minimizing alternative forthe management of OAB relative to oxybutynin.Although compliance and persistencewas relatively poor across treatment groups,our findings suggest that patients appear tobe somewhat more compliant and remainon therapy longer with tolterodine ER; thus,even with increased drug exposure (andcorresponding cost), tolterodine appears toretain a positive pharmacoeconomic profilerelative to oxybutynin.
We note several limitations of our analysis.First, we cannot rule out the possibilitythat differences in associated treatmentEconomic Impact of Extended-release Tolterodine versus Immediate-and Extended-release Oxybutynincosts for services may be a result of selectionbiases inherent in a retrospective databasestudy. However, to control for differencesbetween treatment cohorts, propensity-matchedpairs were created and appropriatemultivariate adjustments were employed instatistical testing procedures. Althoughunobserved differences may remain in thesepopulations, they would likely only affectthe magnitude (rather than the direction) ofour findings.
Second, for patients 65 years of age andolder, our sample was restricted to includeonly patients who were enrolled in aMedicare risk plan. These patients who areenrolled in a managed Medicare programmay differ in certain respects from the overallUS elderly population (eg, demographics,severity of disease).
In addition, this analysis was limited todirect costs only, because indirect costs arenot measurable through medical claimsdata. As is the case with any long-term disorder,indirect costs may represent a substantialburden in the management of OAB.However, because the demographic and clinicalcharacteristics of the 3 treatment groupswere similar in many respects, one wouldexpect total cost differences to persist evenif indirect costs were included.
Because the data source employed wasretrospective claims data, there was notsufficient detail to determine the contributingfactors to lower costs and better adherencein the tolterodine ER group. Lowerrates of OAB-related conditions, such as urinarytract infections and fractures, havebeen suggested for pharmacologically treatedcompared to untreated patients.20 Theconduct of new prospective economic studieswith sufficient detail on major individualcost drivers would greatly add to the evidenceon this topic.
Despite these limitations, our study hasimportant implications. Our findings suggestthat initiation of therapy with tolterodine ERhas tangible economic advantages comparedwith oxybutynin ER or oxybutynin IRamong commercially-insured persons withOAB. In addition, adherence to therapy isan important marker of the success of pharmacotherapyin controlling symptoms. Astherapies for OAB continue to evolve,prospective studies that evaluate clinicaland economic outcomes among patientsreceiving newer therapies for OAB are recommendedto confirm our findings.
Clin Drug Invest.
1. Getsios D, Caro JJ, Ishak KJ, et al. Oxybutyninextended release and tolterodine immediate release.2004;24:81-88.
World J Urol.
2. Stewart WF, Van Rooyen JB, Cundiff GW, et al.Prevalence and burden of overactive bladder in theUnited States. 2003;20:327-336.
3. Gross A, Gates C. Transdermal oxybutynin in overactivebladder. 2004;11:467-468.
4. Hu TW, Wagner TH, Bentkover JD, et al. Estimatedeconomic costs of overactive bladder in the UnitedStates. 2003;61:1123-1128.
5. Ginsberg DA. Overactive bladder: past, present andfuture. Available at: http://www.urologytimes.com/urologytimes/article/articleDetail.jspid=140399.Accessed January 19, 2005
6. Lai HH, Tim BB, Rodney AA. Selecting a medicaltherapy for overactive bladder. 2002;4(suppl4):S28-S37.
7. Chancellor MB. New frontiers in the treatment ofoveractive bladder and incontinence. 2002;4(suppl 4):S50-S56.
8. MacDiarmid S. New antimuscarinic therapy for thetreatment of mixed incontinence. 2003;5(suppl8):S18-S25.
9. Birns J, Lukkari E, Malone-Lee JG. A randomizedcontrolled trial comparing the efficacy of controlled-releaseoxybutynin tablets (10 mg once daily) with conventionaloxybutynin tablets (5 mg twice daily) inpatients whose symptoms were stabilized on 5 mg twicedaily oxybutynin. 2000;85:793-798.
10. Van Kerrebroeck P, Kreder K, Jonas U, Zinner N,Wein A. Tolterodine Study Group. Tolterodine once-daily:superior efficacy and tolerability in the treatmentof overactive bladder. 2001;57:414-421.
J Manag Care Pharm.
11. Noe L, Becker R, Williamson T, Chen D. A pharmacoeconomicmodel comparing two long-acting treatmentsfor overactive bladder. 2002;8:343-352.
Br J Urol.
12. Abrams P, Freeman R, Anderstrom C, MattiassonA. Tolterodine, a new antimuscarinic agent: as effectivebut better tolerated than oxybutynin in patients with anoveractive bladder. 1998;81:801-810.
13. Appell RA. Clinical efficacy and safety of tolterodinein the treatment of overactive bladder: a pooled analysis.1997;(50 suppl 6A):90-96;discussion 97-99.
14. Malone-Lee J, Shaffu B, Anand C, Powell C.Tolterodine: superior tolerability than and comparableefficacy to oxybutynin in individuals 50 years old orolder with overactive bladder: a randomized controlledtrial. 2001;165:1452-1456.
15. Millard R, Tuttle J, Moore K, et al. Clinical efficacyand safety of tolterodine compared to placebo in detrusoroveractivity. 1999;161:1551-1555.
Br J Urol.
16. Rentzhog L, Stanton SL, Cardozo L, Nelson E, FallM, Abrams P. Efficacy and safety of tolterodine inpatients with detrusor instability: a dose-ranging study.1998;81:42-48.
Mayo Clin Proc.
17. Appell RA, Sand P, Dmochowski R, et al.Prospective randomized controlled trial of extended-releaseoxybutynin chloride and tolterodine tartrate inthe treatment of overactive bladder: results of theOBJECT Study. 2001;76:358-363.
18. D'Agostino RB Jr. Propensity score methods for biasreduction in the comparison of a treatment to a non-randomizedcontrol group. 1998;17:2265-2281.
19. Charlson ME, Pompei P, Ales KL, MacKenzie CR. Anew method of classifying prognostic comorbidity in longitudinalstudies: development and validation. 1987;40:373-383.
Am J Manag Care.
20. Brown JS, McGhan WF, Chokroverty S.Comorbidities associated with overactive bladder.2000;6(11 suppl):S574-S579.