Irritable bowel syndrome (IBS) has been associatedwith substantial time lost from work (absenteeism)and reduced productivity at work (presenteeism),which are the indirect costs of illness. This articlepresents a productivity model demonstrating theindirect costs associated with IBS and the reductionin those costs for a cohort of female employeeshypothetically treated with tegaserod, a new selectiveserotonin (5-hydroxytryptamine [5-HT]) type 4(5-HT4) receptor agonist, which is approved by theUS Food and Drug Administration for treatingwomen with IBS-C. The model is based on economicand epidemiologic published literature and clinicaltrial results. In this model, tegaserod treatmentresulted in $1882 in avoided lost productivity pertreated female employee. Considering only the benefitsof decreased work loss and the costs of medicaltherapy, the model predicts a benefit/cost ratio of3.75 in the base case. From an employer's perspective,medical therapy for IBS with tegaserod is costeffectiveunder a series of assumptions for thetreatment of women with IBS with constipation.
(Am J Manag Care. 2005;11:S43-
Estimates indicate that 75% of patientswith irritable bowel syndrome (IBS) arebetween 25 and 64 years of age1; thus,most patients with IBS are of working age. IBScan have a negative impact on a patient'squality of life.2-4 In addition, it imposes a substantialeconomic burden on patients and therest of society.1,5 The direct annual cost of IBSin the United States is estimated to bebetween $1.7 billion and $10 billion, excludingthe cost of prescription and over-the-counterdrugs.1,5 In addition to direct costs,IBS has been associated with substantial indirectcosts resulting from time lost from work(absenteeism) and reduced productivity atwork (presenteeism), costs that are bornelargely by the employer. In fact, indirectcosts—estimated to be as high as $20 billion1—may account for the largest proportionof the total IBS economic burden.5
Patients with IBS are typically categorizedinto subtypes based on bowel dysfunction.The categories are IBS withconstipation (IBS-C), IBS with diarrhea(IBS-D), and IBS that alternates betweenconstipation and diarrhea (IBS-A),6 which isalso known as mixed-bowel IBS. Untilrecently, IBS treatment options consisted ofsingle agents (ie, antispasmodics, fiber, laxatives,antidiarrheals, antidepressants) forspecific individual symptoms (ie, abdominalpain, constipation, diarrhea) or variouscombinations of therapies aimed at alleviatingmultiple symptoms in individualpatients. However, 2 agents are now availablein the United States that address multipleIBS symptoms.
Since its reintroduction in 2002 underrestricted conditions for use (after havingbeen withdrawn in 2000), alosetron, a serotonin(5-hydroxytryptamine [5-HT] type 3(5-HT3) receptor antagonist, has been indicatedfor use in women with severe IBS-Dthat has not responded to conventional therapy.Tegaserod, a 5-HT4receptor agonist,has been indicated for use in women withIBS-C in the United States since 2002 andhas recently been approved for men andwomen <65 years of age with chronic idiopathicconstipation. In fact, tegaserod andalosetron were the only 2 agents to receivegrade A recommendations from theAmerican College of GastroenterologyFunctional Gastrointestinal Disorder TaskForce based on the high quality of publishedevidence supporting their global efficacy inIBS-C and IBS-D, respectively.6
This article presents an economic modelof the indirect costs associated with IBS andthe treatment of IBS with tegaserod. Thepurpose of the model is 2-fold: to demonstratethe impact that indirect costs associatedwith all IBS subtypes have onemployers and to illustrate the potentialcosts and savings, in terms of indirect costs,associated with intervention for the treatmentof patients with IBS-C using tegaserodas an example of cost-saving therapy. Theworkforce receiving the hypothetical treatmentis that of Comerica Incorporated, anationwide bank that offers personal, smallbusiness, and corporate banking servicesand that has major branches in Michigan,California, Texas, and Florida.
Data were obtained from a survey of theworkforce at Comerica,7 from literature onthe epidemiology and treatment of IBS, andfrom tegaserod clinical trial results. Themodel consists of several parameters thatare specific for an employer and can bechanged to customize the model results.Values based on the Comerica workforce areused as default values where possible in themodel. The model can be used to analyzethe indirect costs associated with IBS andthe effect of treatment on these costs.
A representation of the workforce usingdemographics such as subpopulation sizes,sex, salaries, employment status (full-timevs part-time), illness (prevalence of IBS,incidence of IBS subtypes), and treatment(efficacy, costs) was constructed (Figure).Information on age and sex is based on datafrom the Comerica survey,7 and default values for salary are based on data from theBureau of Labor Statistics.8 The proportions ofmale and female patients with IBS (default valuesof 7.7% [men] and 14.5% [women]) arebased on data from the US HouseholderSurvey.9 The proportion of women with IBS-C(default value of 28%) is based on the literature.7 The proportion of patients with IBSwho seek medical care (default value of 25%)was also based on the literature.10 A samplescenario was created in which women withIBS-C seeking medical attention were prescribedtegaserod. Sixty-seven percent of thewomen were assumed to have symptom controlbased on the highest efficacy valuereported in the US tegaserod clinical trials,11which, in turn, affects the number of coursesof treatment, from 6 weeks to 54 weeks asdescribed in the "Treatment" section.
The workforce at Comerica has beenexamined for IBS prevalence and costs.7 Theworkforce is composed of 11 806 employeesaged 18 to 64 years; 90% are employed full-time,and 10% are employed part-time.7Although the age distribution of this workforceis typical, the percentage of women(77%) is greater than the norm (47%).8Wages for persons in this workforce weredistributed by age and sex, according to theaverage distribution reported by the USDepartment of Labor, Bureau of LaborStatistics, and averaged $19.82 per hour.8
Epidemiologic Data Sources
The baseline percentage of persons withIBS was based on data from the USHouseholder Survey,9 which surveyed a randomsample of 8250 households in theNational Family Opinion, Inc database. Thissurvey was chosen because it provides dataon IBS by age and sex.9
Percentages of patients with constipation(28%) and diarrhea (34%) as their primaryaltered bowel habit were based on theresults of the Comerica survey.7 Results ofthis survey are similar to those from studiesin the published literature that report thatthe 3 IBS subtypes occur with approximatelyequal frequency.12,13
The percentage of patients with IBS seekingmedical care (25%) was based on evidencefrom the published literature.11 Thisrate represents a conservative estimate ofhealthcare-seeking behavior. Other studiesin US populations have found healthcare-seekingrates as high as 45%.9,14
The impact of symptoms and adverseeffects from treatment on worker productivityhas been documented for many well-knownlong-term conditions,15 includingdiabetes,16,17 asthma,18 and migraine,19 andfor acute conditions, such as influenza.20One commonly used measure of productivityloss is the Work Productivity and ActivityImpairment (WPAI) questionnaire. TheWPAI is a productivity-based questionnairethat was developed as a general health measure,and it has been modified and validatedfor specific health conditions, includingIBS.21 It is designed to measure work impairmentfrom absenteeism (time absent fromwork) and presenteeism (reduced productivityat work) and measure daily activityimpairment, such as housework, shopping,childcare, and exercising, in the past 7days.21 WPAI outcomes are expressed asimpairment percentages, with higher numbersindicating greater impairment and lessproductivity.
Default values for absenteeism peremployee with IBS (1.7%) and presenteeismper employee with IBS (21.1%) were basedon results from the Comerica survey.7 Totallost productivity per employee with IBS(work productivity score [WPS]), whichenumerates the reduced productivity attributedto IBS as a percentage of the potentialtotal work productivity, was 21.1%, which isequivalent to working approximately 4 daysof a 5-day work week) is also based onresults of the Comerica survey.7 WPS isdefined as [(Hours absent from work + percentageof reduced productivity at work ×hours actually worked)/(hours missed becauseof ill health + hours worked)] × 100.
With a 5-day week for 50 weeks per year,a WPS of 21.1% translates into a total annualloss of 50 days because of IBS per employee.Because the model uses the averageabsenteeism and presenteeism values fromall employees included in the survey to calculatethe WPS rather than from individualemployees (Comerica), an adjustment factorwas applied to ensure that the model reflectsthe WPS for the Comerica workforce publishedin the literature.
Medications are prescribed for most (89%)patients with IBS.22 This model enables anemployer to calculate the costs and savingsassociated with a treatment option of theirchoice for a specific subpopulation of employeeswith IBS. The example used in this articleis women with IBS-C who sought medicalcare and were treated with tegaserod, whichis approved for this patient population.
The base value used in the model for thepercentage of patients with satisfactory reliefof IBS symptoms was 67%, which was basedon the highest efficacy reported in the UStegaserod clinical trials.11 A range of valuesfrom 37% (the lowest efficacy value reportedin the tegaserod clinical trials)23 to 67%are examined in the sensitivity analyses.
One assumption imposed to determinethe treatment effect was that treatment withtegaserod permits 36% of lost productivity tobe regained.24 A second assumption relatedto the extent of treatment (the annual numberof IBS symptom episodes that patientsexperience and the annual number of prescriptionsfor tegaserod written per year). USFood and Drug Administration-approvedprescribing instructions, consistent with supportingclinical trials, recommend a dosage of6 mg twice daily for 4 to 6 weeks with anadditional 4-to 6-week course as needed ifthe patient responds to therapy.23,25 Reliefgenerally occurs within the first week oftreatment for patients who respond, butsymptoms may return after treatment hasbeen stopped.11 Relief of the symptoms of IBSoccurs in 67% of patients; hence, 33% ofpatients may end therapy after the first prescription.It is not known how many patientswho respond to therapy require treatmentbeyond the refill amount. Trials have shownthat tegaserod is safe in patients receiving a12-month course of therapy.26
The outcomes of interest in this modelare the treatment costs and success rates oftherapy as measured by reductions in lostproductivity and costs to employers.Because treatment and productivity measuresoccur within the year, dollars are notdiscounted. For the purpose of this analysis,it was assumed the employer was self-insuredor was experience-rated (ie, premiumsare adjusted according to the numberand size of claims made against the employerin previous years) for healthcare costsand thus bore the burden of the drug costs.Because treatment was only provided topatients who sought care, the incrementalcost was solely that of the prescription cost.Costs associated with physician office visitsand diagnostic testing were not consideredin the model.
The cost to the employer of tegaserod6 mg twice daily was assumed to be thewholesaler acquisition cost of $4.30 perday.27 Determining the cost more preciselywould require information regarding theparticular prescription drug benefit designoffered by the employer directly or as partof a more comprehensive health plan.However, for the purposes of the model,assumptions were made based on commonfeatures of drug benefit designs, such asday supply limit per prescription (basecase, 30 days) and patient cost sharing(base case, copayment per prescription).The model is easily varied with respect today limits, mail order discounts, and copaymentor coinsurance rates.
To determine whether the base valueschosen for the model had a significantimpact on findings, a series of sensitivityanalyses was performed in which the effectsof changing the value of the relevantassumptions were tested. Key assumptionstested were wage rates, absenteeism andpresenteeism rates, medication price, efficacyof tegaserod, and productivity lossesavoided because of successful treatment.
Burden of IBS.
According to the basecase scenario, the total cost of lost productivityattributed to IBS for the employer is$12 508 421 per year, which was calculatedby adding the cost of productivity loss byfemale employees with that lost by maleemployees. For each sex, WPS (21.1%) wasmultiplied by an average working year inhours (2000 hours), average hourly compensation($18.52 for women, $24.47 for men),firm workforce (11 800), sex distribution(78% women, 22% men), and IBS prevalencerate (14.5% for women, 7.7% for men).
When the number of weeks of therapy istaken into account, the average net cost perprescription to the employer per patient is$119, and the average annual cost of treatmentto the employer per patient is $502.Each 30-day prescription entails a $10copay. Thus, a 1-month supply of tegaserodcosts $119 ($4.30/day × 30 days -$10 copay= $119). At the start of treatment, allpatients received 6 weeks of therapy, butonly those with successful results continued.It is estimated that a decreasing percentageof patients obtained refills at each 12-weekperiod. Thus, a year of treatment costs $502($4.30/day × 7 days × [100% × 6 weeks + 40%× 12 weeks + 30% × 24 weeks + 20% × 36weeks + 10% × 52 weeks] less $10 copayeach 30-day period).
Benefits from Tegaserod Intervention.
Tegaserod treatment in the base case scenarioresults in an annual savings of $1882in avoided lost productivity per treatedfemale employee. At a cost of $2.15 per dose($4.30 per day for twice-daily dosing), thetotal costs of treatment are $502 per year,resulting in a net savings cost to the employerof $1381. Therefore, the benefit/cost ratiofor the base case is 3.75 ($1882 divided by$502), indicating that the cost of treatmentis offset by productivity regained nearly 4-fold (Table).
This model is flexible.Inputs can be varied to meet the specificneeds of individual employers. A number ofmodel inputs affect the total amount of benefitsand costs but not the results per patient.For example, per patient results do notchange when the user alters the values forfirm workforce size; sex distribution; proportionof employees with IBS; proportion of personswith IBS whose primary bowel habit isconstipation, diarrhea, or alternating constipationand diarrhea; or the proportion of personsdiagnosed with IBS who seek healthcare.
However, several model inputs do resultin meaningful changes (ie, cost per patient)to benefits, costs, or both. For example, netsavings are directly correlated with wagerates; lower wage rates result in lower netsavings, whereas higher wage rates result inhigher net savings. At an hourly wage rate ofhalf the base case ($10 per full-time femaleemployee and $5 per part-time femaleemployee), the benefit/cost ratio is 1.92.Importantly, even though this ratio representsonly half the ratio of the base case scenario,it still results in net savings to theemployer. Adjusting medication price, medicationsupply limit per day, patient copaymentper prescription, and patientcoinsurance rate per prescription producessimilar results. For example, doubling themedication price still leaves the benefit/costratio at 1.78. Additionally, reducing the rateof presenteeism (impairment while atwork), which is the key to productivity lostbecause of IBS, from 20% to 10%, along witha 0% rate of absenteeism (reduced from 1.7%in the base case), leaves the benefit/costratio at 1.62.
One parameter that is obviously importantto the benefit/cost ratio is the percentage ofIBS patients who respond to tegaserod therapy.If the most conservative estimate oftegaserod efficacy (37%)23 is used, the benefit/cost ratio is 2.98. Reducing effectivenessfurther to a net effectiveness (treatment effectless placebo effect) of 18% yields a benefit/costratio of 1.49. Although the benefits are substantiallyreduced when medication efficacyis decreased, medication costs are alsoreduced—that is, with a lower treatmenteffect, fewer patients are prescribed medicationafter the first 6-or 12-week supply. Inaddition, the benefit/cost ratio is also affectedby the percentage of lost productivityregained because of symptom alleviationwith tegaserod therapy. In the base case, a36% regain rate in lost productivity wasassumed25; however, if the actual rate for agiven employer is only 25%, for example, thecost/benefit ratio will be similarly reduced(to 2.60).
Finally, in a worst case scenario thatconsiders the combination of values in percentof time absent (0%), percent of timeimpaired (10%), satisfactory relief of IBS(37%), and productivity losses avoided(25%), costs ($349) outweigh benefits($313), resulting in a benefit/cost ratio of0.90. Thus, in this case, it is possiblethat employers would not realize savingsfrom the inclusion of tegaserod in theirformularies.
This IBS productivity economic modelfeatures a number of factors that distinguishit among other models of productivity. It isthe first model on productivity to explicitlyincorporate absenteeism and presenteeismas the components of indirect costs; mostmodels just include absenteeism. Althoughit is true that presenteeism data are notoften available, they are critical to the usefulnessof this model because of the substantialrole that presenteeism plays inproductivity lost because of IBS.
This model takes advantage of the WorkerProductivity Index (WPI), a tool that measurespresenteeism (the actual decrease inproductivity of employees while they are onthe job), absenteeism, and disability. Usingthis tool in a study of 1039 employees ofFirst Card, one of the largest credit cardissuing companies in the United States,Burton and colleagues found that employeeswith digestive disorders had the lowest WPIrating compared with employees with mentalhealth disorders, respiratory disorders,injuries, musculoskeletal disorders, and cancer,indicating that productivity loss frompresenteeism is highest among employeeswith digestive disorders.16 These resultssuggest that the absenteeism/presenteeismratios reported in the previously mentionedstudies may be low. The numbers used inthe base case scenario in this model, therefore,may represent conservative estimatesof absenteeism/presenteeism attributedto IBS.
In addition, this model incorporates datafrom recent trials on the efficacy of a novelmedication for the treatment of patientswith IBS-C. Tegaserod has proven efficacyin providing global relief from the multiplesymptoms of IBS in women with IBS-C and,more recently, from the multiple symptomsof chronic (idiopathic) constipation in menand women younger than 65 years of age.Although the total number of days lost peryear per employee, calculated when theComerica workforce data (50 days per yearper employee) is applied to this model, mayseem higher than that reported by otherstudies examining productivity lost, it isimportant to remember that this modelincludes days lost to absenteeism andpresenteeism, whereas many studies evaluateabsenteeism alone. For example, theUS Householder Study found that patientswith IBS (n = 629) were absent from workan average of 13.4 days per year comparedwith 4.9 days reported by personsnot experiencing any functional gastrointestinaldisorder.9 Moreover, in a surveyof 318 patients with IBS in the UnitedStates who had been medically diagnosedwith IBS or who met diagnosticcriteria, the average number of daysaffected by ill health leading to absenteeismwas 6.4 compared with 3.0 daysfor persons without IBS.28 However, asurvey of 287 US members of theInternational Foundation for FunctionalGastrointestinal Disorders29 found that ofthe 30% of patients who reported missingsome days of work because of their IBSsymptoms, the average number of daysabsent from work (1.7 days in the 4-weekperiod before the survey) translates into22.1 days missed per patient per year,which, when added to the 39 days ofreduced productivity reported by 46% ofrespondents (for a total of 61.1 days),actually exceeds the total number usedin this model.
Other assumptions imposed on treatmentoutcomes merit review and verification infuture studies. For example, data are sparsewhen it comes to verifying the assumptionthat symptom relief leads to 36% of lost productivity regained.24 Moreover, the modelassumed full relief rather than incrementalrelief, which may or may not be appropriate.Full relief was used because of the highplacebo response reported for IBS therapies,but incremental relief may be appropriate insome patients.
Finally, in this model, an assumption wasmade regarding the percentages of patientsreceiving multiple courses of treatment andthe days they received them throughout theyear. However, these percentages may notrepresent the correct distribution of daysthroughout the year and warrant furtherinvestigation.
This model demonstrates that tegaserodis a cost-effective treatment for femaleemployees with IBS-C. The costs of treatingthese employees with tegaserod arecompensated fully by the regained productivitypreviously lost because of IBSsymptoms. Although cost-effective servicesare common, actual cost-saving therapiesare not.
In the base case scenario in this model,tegaserod saved the employer $1381 peremployee per year, for a benefit/cost ratio of3.75. Even when changes are made to a varietyof parameters (eg, reduction in hourlywage, increase in medication costs, reductionin productivity losses avoided), the sensitivityanalysis shows cost savings/offsets orno additional costs to the employer.Therefore, covering tegaserod treatment is acost-effective way for employers to reduceproductivity losses from IBS among theirfemale employees.
Burden of Gastrointestinal Diseases.
1. American Gastroenterological Association. Bethesda, Md:American Gastroenterological Association; 2001:1-86.Available at: http://www.gastro.org/clinicalRes/pdf/burden-report.pdf. Accessed November 16, 2004.
2. Frank L, Kleinman L, Rentz A, Ciesla G, Kim JJ,Zacker C. Health-related quality of life associated withirritable bowel syndrome: comparison with other chronicdiseases. 2002;24:675-689.
3. Gralnek IM, Hays RD, Kilbourne A, Naliboff B,Mayer EA. The impact of irritable bowel syndrome onhealth-related quality of life. 2000;119:654-660.
Am J Gastroenterol.
4. Koloski NA, Talley NJ, Boyce PM. The impact offunctional gastrointestinal disorders on quality of life.2000;95:67-71.
Am J Manag Care.
5. Martin R, Barron JJ, Zacker C. Irritable bowel syndrome:toward a cost-effective management approach.2001;7(suppl):S268-S275.
Am J Gastroenterol.
6. Brandt LJ, Bjorkman D, Fennerty MB, et al.Systematic review on the management of irritable bowelsyndrome in North America. 2002;97(suppl):S17-S26.
Am J Manag
7. Dean BB, Aguilar D, Barghout V, et al. Impairmentin work productivity and health-related quality of life inpatients with irritable bowel syndrome. 2005;11(suppl):S17-S26.
8. US Department of Labor, Bureau of Labor Statistics.Median weekly earnings of full-time wage and salaryworkers by union affiliation and selected characteristics.Available at: http://www.bls.gov/cps/cpsaat41.pdf.Accessed February 5, 2004.
9. Drossman DA, Li Z, Andruzzi E, et al. US householdersurvey of functional gastrointestinal disorders: prevalence,sociodemography, and health impact. 1993;38:1569-1580.
Ann Intern Med.
10. Drossman DA, Thompson WG. The irritable bowelsyndrome: review and a graduated multicomponenttreatment approach. 1992;116:1009-1016.
Aliment Pharmacol Ther.
11. Novick J, Miner P, Krause R, et al. A randomized,double-blind, placebo-controlled trial of tegaserod infemale patients suffering from irritable bowel syndromewith constipation. 2002;16:1877-1888.
12. Ringel Y, Drossman DA. Toward a positive andcomprehensive diagnosis of irritable bowel syndrome.2000;2:1-10.
Am J Med.
13. Mayer EA. Emerging disease model for functional gastrointestinaldisorders. 1999;107(suppl):2S-19S.
14. Talley NJ, Zinsmeister AR, Van Dyke C, Melton LJ3rd. Epidemiology of colonic symptoms and the irritablebowel syndrome. 1991;101:927-934.
Health Aff (Milwood).
15. Druss BG, Marcus SC, Olfson M, Tanielian T,Elinson L, Pincus HA. Comparing the national economicburden of five chronic conditions. 2001;20:233-241.
J Occup Environ Med.
16. Burton WN, Conti DJ, Chen CY, Schultz AB,Edington DW. The role of health risk factors and diseaseon worker productivity. 1999;41:863-877.
17. American Diabetes Association. Economic consequencesof diabetes mellitus in the US in 1997. 1998;21:296-309.
Allergy Clin Immunol.
18. Birnbaum HG, Berger WE, Greenberg PE, et al.Direct and indirect costs of asthma to an employer. 2002;109:264-270.
Arch Intern Med.
19. Hu XH, Markson LE, Lipton RB, Stewart WF,Berger ML. Burden of migraine in the United States: disabilityand economic costs. 1999;159:813-818.
20. Akazawa M, Sindelar JL, Paltiel AD. Economic costsof influenza-related work absenteeism. 2003;6:107-115.
Aliment Pharmacol Ther.
21. Reilly MC, Bracco A, Ricci JF, Santoro J, Stevens T.The validity and accuracy of the Work Productivity andActivity Impairment questionnaire—irritable bowel syndromeversion (WPAI:IBS). 2004;20:459-467.
Dig Dis Sci.
22. Shih YC, Barghout VE, Sandler RS, et al. Resource utilizationassociated with irritable bowel syndrome in theUnited States 1987-1997. 2002;47:1705-1715.
23. Zelnorm® (tegaserod maleate) tablets [prescribinginformation]. East Hanover, NJ: NovartisPharmaceuticals Corporation; August 2004.
Am J Gastroenterol.
24. Reilly MC, Barghout V, Ruegg P, Pecher E, Ricci JF.Tegaserod significantly reduces work productivity lossand daily activity impairment in patients with IBS withconstipation [abstract]. 2004;99:S241.
25. Rivkin A. Tegaserod maleate in the treatment of irritablebowel syndrome: a clinical review. 2003;25:1952-1974.
26. Tougas G, Snape WJ, Otten MH, et al. Long-termsafety of tegaserod in patients with constipation-predominantirritable bowel syndrome. 2002;16:1701-1708.
Red Book: 2003 Drug Topics.
27. Montvale, NJ:Medical Economics Company; 2003.
Am J Gastroenterol.
28. Hungin APS, Tack J, Mearin F, Whorwell PJ,Dennis E, Barghout V. Irritable bowel syndrome (IBS):prevalence and impact in the USA—the truth in IBS (T-IBS)survey [abstract]. 2002;97(suppl):S280-S281.
29. Hahn BA, Yan S, Strassels S. Impact of irritable bowelsyndrome on quality of life and resource use in the UnitedStates and United Kingdom. 1999;60:77-81.