Avai lable onl ine at www.sc iencedirect .com

journal homepage: www.elsevier .com/ locate / jva l

ORIGINAL ARTICLESEconomic Evaluation

The iMTA Productivity Cost QuestionnaireA Standardized Instrument for Measuring and ValuingHealth-Related Productivity LossesClazien Bouwmans, MSc1,2,*, Marieke Krol, PhD1,2, Hans Severens, PhD1, Marc Koopmanschap, PhD1,Werner Brouwer, PhD2, Leona Hakkaart-van Roijen, PhD1,2

1Institute for Medical Technology Assessment, Erasmus University Rotterdam, Rotterdam, The Netherlands; 2Institute of HealthPolicy & Management, Erasmus University Rotterdam, Rotterdam, The Netherlands

A B S T R A C T

Background: Productivity losses often contribute significantly to thetotal costs in economic evaluations adopting a societal perspective.Currently, no consensus exists on the measurement and valuationof productivity losses. Objective: We aimed to develop a stand-ardized instrument for measuring and valuing productivity losses.Methods: A group of researchers with extensive experience inmeasuring and valuing productivity losses designed an instrumentsuitable for self-completion, building on preknowledge and evidenceon validity. The instrument was designed to cover all domains ofproductivity losses, thus allowing quantification and valuation of allproductivity losses. A feasibility study was performed to check thequestionnaire’s consistency and intelligibility. Results: The iMTAProductivity Cost Questionnaire (iPCQ) includes three modulesmeasuring productivity losses of paid work due to 1) absenteeismand 2) presenteeism and productivity losses related to 3) unpaidwork. Questions for measuring absenteeism and presenteeism werederived from existing validated questionnaires. Because validated

measures of losses of unpaid work are scarce, the questions ofthis module were newly developed. To enhance the instrument’sfeasibility, simple language was used. The feasibility studyincluded 195 respondents (response rate 80%) older than 18 years.Seven percent (n ¼ 13) identified problems while filling in theiPCQ, including problems with the questionnaire’s instructionsand routing (n ¼ 6) and wording (n ¼ 2). Five respondentsexperienced difficulties in estimating the time that would beneeded for other people to make up for lost unpaid work. Con-clusions: Most modules of the iPCQ are based on validated questionsderived from previously available instruments. The instrument isunderstandable for most of the general public.Keywords: absenteeism, presenteeism, productivity losses, unpaid work.

Copyright & 2015, International Society for Pharmacoeconomics andOutcomes Research (ISPOR). Published by Elsevier Inc.

Introduction

Economic evaluations are increasingly performed to aid decisionmakers in allocating scarce health care resources. Such evalua-tions provide information on the relative cost-effectiveness of a(new) health care intervention compared with one or more(existing) alternatives [1]. Especially when decisions regardingthe implementation or reimbursement of new interventions are(partly) based on outcomes of economic evaluations, and theseoutcomes thus influence access of patients to interventions, it isessential that these economic evaluations be conducted consis-tently and uniformly. A fundamental methodological choice,typically strongly affecting cost-effectiveness outcomes, is theperspective from which an economic evaluation is performed [2].

Evaluations from a health care perspective normally aim toinclude only those costs that fall on the health care budget (andtherefore may be seen as most relevant to health care decisionmakers). From a theoretical viewpoint, it can be argued thatoptimal (i.e., welfare maximizing) decision making is possibleonly by taking a societal perspective [3]. Evaluations conductedfrom the societal perspective aim to incorporate all relevant costsand effects, regardless of where these occur [1]. An importantcost category relevant from this perspective (but not from thehealth care perspective) is that of productivity costs. The value ofproductivity changes owing to illness and treatment oftenreflects a large part of the total costs related to health and healthcare interventions [4] and may even exceed medical costs[2]. Consequently, the decision regarding the inclusion of

1098-3015$36.00 – see front matter Copyright & 2015, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).

Published by Elsevier Inc.

http://dx.doi.org/10.1016/j.jval.2015.05.009

E-mail: bouwmans@bmg.eur.nl.

* Address correspondence to: Clazien Bouwmans, Institute for Medical Technology Assessment, Erasmus University Rotterdam, PO Box1738, 3000 DR Rotterdam, The Netherlands.

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 5 3 – 7 5 8

productivity costs in economic evaluations can have large impli-cations for final outcomes. Some have argued that performingeconomic evaluations from both the societal and the health careperspective would be most informative [5–7]. Such a two-perspective approach is already recommended in severalnational health economic guidelines [8–11]. Despite the fact thatsome of the national health economic guidelines recommend atwo-perspective approach and approximately half of the guide-lines prescribe taking a societal perspective (http://www.ispor.org/peguidelines/index.asp), in practice productivity costs areignored in the vast majority of economic evaluations [2,5,12,13].

It is unclear why productivity costs are ignored in economicevaluations so often. It has been suggested, however, that theexclusion of productivity costs in economic evaluations is relatedto a lack of guidance and standardization of measurement andvaluation of health-related productivity losses [14,15]. Indeed, mosthealth economic guidelines provide little or no guidance on how toestimate productivity costs. Moreover, golden standards regardingthe measurement and valuation of productivity losses do not exist.Specific guidance as to the types of productivity costs that shouldbe included in economic evaluations is generally limited as well[16]. In addition, validated questionnaires for measuring produc-tivity losses are scarce and not systematically applied [16].

Costs associated with health-related productivity changes arecommonly referred to as productivity costs and can be defined as“Costs associated with production loss and replacement costsdue to illness, disability and death of productive persons, bothpaid and unpaid” [17]. Currently, the importance of inclusion ofproductivity costs related to paid work in economic evaluationstaking a societal perspective is generally acknowledged. This isnot (yet) the case for unpaid work [18]. Nonetheless, unpaid workclearly is an important source of economic value contributing tooverall welfare. Given that health care interventions are oftenaimed at elderly populations who do engage in unpaid work butnot in paid work, inclusion of these costs can be highly relevantin many studies. Three types of unpaid activities can be distin-guished: household work (e.g., cooking and cleaning), care work(e.g., taking care of children, helping friends or family withcleaning and personal care), and volunteer work (e.g., helpingout in a commodity center or a sports club). Despite unpaidwork’s recognized importance, it is omitted from the vast major-ity of economic evaluations, even those that claim to take asocietal perspective [4,18,19]. In addition, complete and validatedmeasurement instruments for unpaid work are lacking. Forinstance, the measurement of unpaid work in the Health andLabor Questionnaire was limited to household tasks taken overby others [20]. The Work Productivity and Activity Impairmentquestionnaire asks about the ability to perform these tasks, withno opportunity to value in monetary terms [21].

Productivity changes in paid work may occur because ofabsence from work (absenteeism) or because of reduced produc-tivity while at work (presenteeism). Information on absence fromwork due to sick leave may be derived from existing data sources(if available), such as registrations from occupational healthservice companies. Caution is warranted in using such sources,because on important aspects such as age, health status, andoccupational background the people included in such registriesmay differ from the population included in an evaluation study.Moreover, such registrations are typically available only for sal-aried workers and not for self-employed workers. Therefore, it isadvisable to directly measure productivity losses related to paidwork, ideally by monitoring actual production. In many circum-stances, however, this is not feasible. Because of these restrictions,often patients’ self-completed questionnaires are applied in col-lecting productivity data for use in economic evaluations.

Today, various such instruments that have been developedare available. Several of these instruments have been specifically

developed to measure health-related productivity losses. Most,however, are not specifically intended or suited for use in eco-nomic evaluations [22–24]. For instance, the measurements do notalways allow directly translating productivity losses into monetaryterms or the scope may be limited to losses related to paid workonly. Additional differences between the instruments concern theformulation of separate questions, the inclusion of different typesof productivity losses, the length of the questionnaire, the recallperiod, and the (implicitly assumed) valuation method. Regardingthe latter, it is important to note the (ongoing) debates regardingpreferred valuation methods for productivity losses [14,19]. Espe-cially, the debate regarding the appropriateness of the humancapital approach and the friction cost approach received quitesome attention [25–32]. Applying these methods requires specificinformation that measurement instruments need to facilitate.Finally, not all instruments are specifically developed for self-completion in a broad population. Consequently, all these factorsmay affect the validity, completeness, generalizability, usefulness,and comparability of outcomes of these instruments.

This article reports on the development of a standardizedinstrument suitable for self-completion for measuring and valu-ing all relevant productivity losses: the iMTA Productivity CostQuestionnaire (iPCQ). For this purpose, we optimized the featuresof existing instruments. Our underlying aim was to enhance thegeneralizability and comparability of outcomes of economicevaluations.

Methods

Development

A group of well-experienced researchers in the field of measuringand valuing productivity losses for use in economic evaluations ofhealth care interventions was empanelled for the development ofa standardized instrument for measuring and valuing productiv-ity losses. This group of researchers (all working at the ErasmusUniversity Rotterdam) consisted of four health economists (L.H.R.,M. Koopmanschap, M. Krol, andW.B.) and two health scientists (C.B. and H.S.). During three interactive brainstorming sessions of 2hours, the approach for the development of the questionnairewas discussed and the main quality criteria the instrumentshould meet were drawn up. These criteria were based on up-to-date scientific knowledge and practical experience of theresearchers involved. The criteria are summarized in Table 1.

It was decided to base the standardized instrument onpreexisting questionnaires by optimizing the features of theseinstruments developed within or in cooperation with Erasmus

Table 1 – Summary of main criteria applied for thedevelopment of the iMTA Productivity CostQuestionnaire.

Objectives

Standardization: Generalizability and comparabilityBuilding on preknowledge and evidence on reliability and validity

of existing questionnairesIncluding absenteeism, presenteeism, and losses of unpaid workAllowing to quantify productivity losses in descriptive and

monetary termsSuitable for valuations based on the human capital approach and

the friction cost approachFeasible to use for adults in a broad general populationApplicable in economic evaluations of health care interventions,

independent of the disease area

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 5 3 – 7 5 8754

University: the (Short-Form) Health & Labor Questionnaire (SF-HLQ) [20] and the PROductivity and DISease Questionnaire (PRO-DISQ) [33]. A number of items included in these questionnaireshad already proved their usefulness in practice and had beenvalidated. The primary criterion for the selection of appropriatequestions for measuring different types of productivity losseswas evidence regarding the validity of the measurements.

In general, consensus exists on the relevance of productivitylosses of paid work for evaluations adopting a societal perspec-tive. Nevertheless, productivity losses at work due to healthproblems (presenteeism) are frequently omitted. Another impor-tant area that is generally ignored concerns productivity losses ofunpaid work. Still, given the aging of Western populations, agrowing proportion of patients is retired. These patients, how-ever, may still play an active role in society by performingnonpaid jobs or volunteer activities and by providing informalcare [34]. The instrument needed to be able to capture all theseaspects.

Another important prerequisite was that the questionnaireallowed for the quantification of productivity losses in descriptiveterms (e.g., the number of days of absenteeism and presenteeism,the amount of losses of/due to unpaid work) and in monetaryterms. Given the ongoing debate on whether productivity lossesare best valued with the friction cost approach or the humancapital approach [25–32,35] and the lack of practical consensus[2,5,12,13], it was considered important that the questionnairewas suitable for both approaches. To do so, questions about thefrequency and length of spells of absenteeism needed to beincluded.

The instrument should be suitable for patient self-completionto ensure efficient data collection. It was discussed whether thephrasing of the questions in the existing productivity cost instru-ments used as the foundation for the iPCQ were sufficiently easilyunderstandable for most of the general public. Difficulty oflanguage could affect the validity and feasibility of the instru-ment. Therefore, after completion of the design of the iPCQinstrument, it was sent to a specialized agency (http://www.bureautaal.nl) that rephrased the questions to language levelB1. This language level is based on the Common EuropeanFramework of Reference for Languages, which plays a centralrole in language and education policy worldwide. The CommonEuropean Framework of Reference for Languages describes lan-guage ability on a scale of levels from A1 for beginners up to C2for those who have mastered a language. Most levels areempirically validated (www.coe.int/lang-CEFR). Language levelB1 implies that the text can be comprehended by 95% of thegeneral population, including lower educated persons [36]. Toenhance the feasibility of the questionnaire, we kept the length ofthe questionnaires as short as possible.

The questionnaire has been translated into English, and in thefuture should be made available in multiple languages, allowingits use in national and international studies.

Results

General Structure of the Questionnaire

The iPCQ consists of 18 questions, of which 9 are generalquestions to collect demographic information, such as age, sex,and respondents’ work status (e.g., number of hours per week ofpaid work and number of working days a week). Productivitylosses are measured in three separate modules, allowing theoption of leaving out specific types of productivity losses whenthese are not relevant for a specific population. The question-naire adopts a recall period of 4 weeks. This period is based onresearch indicating that longer recall periods negatively affect the

accuracy of self-reported absenteeism [41]. Evidence regardingthe appropriate recall period for presenteeism, however, isindecisive [19]. In addition, the questionnaire provides an easilyunderstandable general instruction for the respondent and rout-ing instructions, allowing the respondent to easily skip questionsthat are not applicable to individual responders.

Absenteeism

The iPCQ includes three questions for measuring productivitylosses due to absence from work. The questions originate fromthe PRODISQ and the SF-HLQ and identify the occurrence ofabsenteeism from paid work and the length of absenteeism. Inaddition, the module includes a specific question to assess thestarting date of long-term absence from work. Long-term absencein this context is defined as absence starting before the recallperiod. The validity of these questions was shown in a number ofstudies [20,38,41].

Presenteeism

Three questions are included to identify health-related dimin-ished productivity at work. The questions are a combination ofquestions of the PRODISQ and the SF-HLQ. Respondents areasked whether they suffered from health problems at work andif so, for how many days (question originating from the SF-HLQ).Next, respondents are asked to rate their work performance onthese days in comparison to their functioning on normal workingdays using a 10-point rating scale. The latter question is derivedfrom the PRODISQ and based on the earlier developed Quantityand Quality (QQ) method [37]. Brouwer et al. [39] examined theconstruct and convergent validity of the alternative methodsapplied in the SF-HLQ and the PRODISQ. They concluded thatestimates of productivity losses using the Quantity scale from thePRODISQ resulted in a better cost estimation than from the so-called Osterhaus method applied in the SF-HLQ. The Quantityscale estimate was “considered to better indicate the real quan-tity of lost productivity.” Reliability of presenteeism questionswas previously examined using a test-retest design [38]. Asatisfactory agreement was found on the reported efficiencyrates (intraclass correlation coefficient of 0.729). In addition,moderate agreement was found on the reported number of daysat work while impeded (intraclass correlation coefficient 0.556).

Productivity Losses of Unpaid Work

The questions in the unpaid work module were developed duringtwo brainstorming sessions of the working group members.Measurement and valuation of losses of unpaid work in the SF-HLQ are limited to losses that are actually taken over by otherpersons, hence ignoring lost production. The members of theworking group were of the opinion that all losses should bemeasured, whether production was taken over or not, to avoidunderestimation.

To increase the measurements’ validity, unpaid tasks weredefined. Respondents are asked whether they could perform lessunpaid work, such as household work and volunteer work, as aresult of health problems. Next, they are asked to state howmany hours it would take someone else to replace this unper-formed unpaid work. The latter approach was chosen for severalreasons: 1) it ensures that all lost unpaid work is measured andnot only the work that is actually replaced by others; 2) it ensuresa reasonable estimate of required time of someone else (nothampered by health problems); and 3) it applies a “third-personcriterion” distinguishing unpaid work from leisure time. Toclarify, the third-person criterion denotes that all benefits fromactivities outside paid work that are replaceable by a third personare considered unpaid labor [40].

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 5 3 – 7 5 8 755

Valuation

To be able to produce productivity cost estimates with the humancapital approach and the friction cost approach, the absenteeismquestions were constructed in such a way that they allow fordetermining the length of individual absenteeism periods. This isnecessary because with the friction cost approach productivitylosses related to absenteeism are assumed to be restored after a“friction period,” that is, the period required to replace an illworker.

Manual

A manual was developed to provide practical support for apply-ing the instrument and for measuring and valuing productivitylosses. The manual includes examples and instructions forestimating costs of productivity losses of (long-term) absencefrom work using the friction cost approach. In addition, themanual provides a codebook for the registration of data.

In some cases, the basic questions of the iPCQ relating toabsence from work may not be feasible or appropriate, forinstance, when the health problems are characterized by episodicexacerbations (e.g., migraine), which can cause strong fluctua-tions in patients’ productivity and frequent, short episodes ofabsenteeism [20]. Such fluctuating productivity patterns may askfor a more intense monitoring of productivity changes. Therefore,the iPCQ provides an alternative version for measuring theseshort episodes of absenteeism, which was derived from theHealth and Labor Questionnaire. This version consists of dailylog reporting on presence or absence from work. The validity ofthe daily log reports was assessed previously [20]. These alter-native questions are included in the manual.

The questionnaire and the manual are available in Dutch andin English. The questionnaire is suitable for interview adminis-tration, traditional self-completion in writing, and Web-basedapplications.

Validation

The validity of several elements of the iPCQ originating from thepreexisting instruments was previously investigated. Asexplained, the questions regarding unpaid work were newlydeveloped. Although all modules need further validation, con-sequently this implies that especially for the unpaid workmodule additional research is necessary to assess its reliabilityand validity.

All questions were rephrased by a specialized agency to makethe questionnaire suitable for a broad general population. Afeasibility study was conducted as the first step in the validationprocess of the iPCQ. This study was designed to test theconsistency and intelligibility of the questionnaire. For thispurpose, a convenience sample from the general populationwas asked to complete the iPCQ and to indicate whether itconsidered certain questions unclear. In addition, respondentswere able to comment on the questionnaire’s instructions and itsrouting. Two students each randomly invited 100 adults frompublic places (e.g., shops, marketplaces, and passengers in publiclocal transport) and relatives to participate in the study. Anadditional number of respondents was recruited by colleaguesfrom our institute. The questionnaire was completed in writing.Respondents were asked to mark questions that were difficult tounderstand. The reason why the question was considered diffi-cult to answer had to be reported in a separate document. Finally,respondents were asked to comment on the questionnaire and itsstructure. Questionnaires were returned by mail. Participation inthe study was encouraged by allotting three vouchers of €50among respondents. Responsiveness and ease of administration

were also evaluated by measuring the questionnaire’s comple-tion time of 30 participants by each of the students.

The feasibility study included a total of 195 respondents 18years and older (response rate 80%).

As can be seen in Table 2, 36% of the respondents were men;25% reported a middle-level education, and 50% reported a higherlevel of education. A total of 62% were in paid profession. Tenpercent of the respondents in paid profession reported havingexperienced absenteeism in the preceding 4 weeks, and 30% hadexperienced health problems at work affecting their level ofproductivity. The number of respondents experiencing problemsin completing the iPCQ was small (n ¼ 13). Two respondentsreported that the routing from short-term absence from work tolong-term absence was unclear. This routing between thesequestions was subsequently revised. Two respondents men-tioned that explanatory statements preceding specific questionswere overlooked because of routing instructions. In response tothese comments, the placement of these instructions wasrevised. In addition, two respondents pointed out that the routingto the end of the questionnaire was unclear at some points.These routing instructions were revised. Two respondents indi-cated that the wording of the question on absence from work wasconfusing. Consequently, the wording was revised.

Five respondents experienced difficulties in estimating theamount of time that would be needed for other people to makeup for lost unpaid work. Although we acknowledge the relativedifficulty of this question, an easier yet (theoretically) correctalternative for this question is currently unavailable. For 60respondents, the time to complete the iPCQ was measured. Onaverage, the time to complete the iPCQ was 4.8 (� 2.6) minutes.

Discussion

This article reports on the development of the iPCQ, a stand-ardized instrument for measuring and valuing productivity lossesfor use in economic evaluations of health care. Where possible,the development of the instrument was based on previouslyavailable instruments. The iPCQ is a short and concise instru-ment suitable for quantifying presenteeism at and absenteeismfrom paid work as well as productivity losses related to unpaidlabor. The iPCQ has a modular structure that allows selectingrelevant components of productivity losses for individual studies.The instrument is suitable for self-reporting, online use, andface-to-face interviewing. To enhance feasibility and responsive-ness, the questions in the iPCQ are phrased in such a way thatthe questions are easily understandable for most of the generalpublic, including low-educated people. In addition, to increase itsfeasibility, also in the context of patient self-reported measure-ments, the number of questions of the instrument is limited. Onaverage, it took our convenience sample less than 5 minutes tocomplete the questionnaire.

Table 2 – Characteristics of respondents in thefeasibility study (N ¼ 195).

Characteristic Value

Age (y), mean � SD 51.4 � 17.7Sex: male, n (%) 70 (36)Education, n (%)

Low 20 (10)Secondary 71 (37)High 97 (40)

Living with partner, n (%) 130 (67)Paid work (%) 121 (62)

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 5 3 – 7 5 8756

Some aspects regarding the content of the iPCQ need to benoted. First, the iPCQ adopts a 4-week recall period, on the basisof previous research indicating less accurate reporting for longerrecall periods [41]. This period is normally shorter than theinterval of follow-up measurements in evaluation studies, inwhich measurements every 4 weeks are rare for practical rea-sons. This implies that interpolation and extrapolation of themeasured productivity losses are necessary to cover the fullevaluation period. Evidence regarding the appropriate length ofrecall periods for presenteeism is indecisive [19]. This topictherefore deserves more attention in the literature. For clarityand convenience of respondents, the recall period for presentee-ism was set equal to that of absenteeism (i.e., 4 weeks). Forreasons of reliability and comparability, we strongly advise toalign with this recall period when using the iPCQ. It is importantto realize that the validity of the measurements is influenced bythree important aspects: recall bias, measurement quantity, andmeasurement gaps. Recall bias is likely to increase with longerrecall periods. Shorter recall periods will result in an increase ineither measurement moments or measurement gaps, needinginterpolation between measurement moments. Consequently,decisions regarding recall periods should be taken with caution.Acceptable duration of measurement gaps depends on thefluctuation of productivity losses within the population ofinterest.

Second, the iPCQ does not include questions regarding com-pensation of lost productivity or regarding the effects of dimin-ished productivity on coworkers (so-called multiplier effects).Empirical research has shown that coworkers often compensateproductivity losses during regular hours [42–44] and that absen-teeism and presenteeism can negatively affect the productivity ofcoworkers in cases of team dependency [45–47]. To what extentsuch mechanisms affect final production and actual costs, how-ever, remains largely unclear. Therefore, adjusting productivitycosts for compensation mechanisms or multiplier effects ispremature. If future research provides insight into how toproperly cope with compensation mechanisms and multipliereffects, in terms of not only their measurement but also theireffect on actual costs, the iPCQ may need to be updated.

Third, regarding the measurement of unpaid work, the work-ing group did not support current ways of including this inavailable instruments. Moreover, the limited amount of availablescientific studies discussing unpaid labor offered little guidanceregarding appropriate measurement approaches. Consequently,the questions in the unpaid work module were based on expe-rience of the working group members. This implies that espe-cially this module needs additional validation. We alsoemphasize that unpaid work deserves more attention in thescientific literature in general because it is often ignored whileit may be influential especially in groups such as elderly, whooften form a large part of treated patients.

Fourth, the iPCQ was developed using a modular structurethat facilitates excluding irrelevant modules. We do advise,however, to exclude modules only after careful considerationand clear justification for doing so if one aims to conduct aneconomic evaluation from the societal perspective.

Finally, it is important to realize that the iPCQ is an instru-ment to quantify productivity losses (especially for use ineconomic evaluations). It does not measure workability or rolefunctioning.

Although these aspects are important, they are commonlyincluded in quality-of-life measures. They are therefore not seenas part of productivity per se, but as health effects and valued aspart of quality of life.

Despite critical notes, the development of the iPCQ offers avaluable comprehensive tool to enhance the comparability andgeneralizability of outcomes in economic evaluations adopting a

societal perspective. The application of standardized instrumentsis an important prerequisite for increasing the comparability ofoutcomes of economic evaluations. To test and to maintain thequality of instruments, however, it is essential to extensivelyvalidate instruments, to regularly assess the content on the basisof (new) scientific developments, and to adjust the instrumentwhen necessary. Also, in that context, the validation of the iPCQis an ongoing process. It is currently applied in several economicevaluations of interventions in different disease areas. Theresults of these studies are expected in the near future, allowingfurther assessment of the properties of the iPCQ and, wherepossible and required, improvement based on new scientificinsights.

Conclusions

The iPCQ is a concise generic instrument developed to quantifyhealth-related productivity losses of paid and unpaid work foruse in economic evaluations. It is largely based on previouslyavailable instruments and builds on the current scientific state ofplay in productivity cost measurement and valuation. The instru-ment is unique in its language use because it has been especiallydeveloped to be understandable for the vast majority of thegeneral public including low-educated people. Although manyof the items of the iPCQ have been previously validated, furthervalidation of the instrument is foreseen and encouraged. iPCQ iscurrently available in Dutch and English. Translations in otherlanguages, for example, German, French and Spanish, is expectedin the near future.

Source of financial support: The authors have no otherfinancial relationships to disclose.

R E F E R E N C E S

[1] Drummond MF, Sculpher MJ, Torrence GW, et al. Methods for theEconomic Evaluation of Health Care Programmes (3rd ed.) OxfordUniversity Press, Oxford, UK; 2005.

[2] Krol M, Papenburg J, Koopmanschap M, Brouwer W. Do productivitycosts matter? The impact of including productivity costs on theincremental costs of interventions targeted at depressive disorders.Pharmacoeconomics 2011;29:601–19.

[3] Byford S, Raftery J. Perspectives in economic evaluation. BMJ1998;316:1529–30.

[4] Krol M, Papenburg J, Tan SS, et al. A noticeable difference? Productivitycosts related to paid and unpaid work in economic evaluations onexpensive drugs [published online ahead of print April 16, 2015]. Eur JHealth Econ http://dx.doi.org/10.1007/s10198-015-0685-x.

[5] Pritchard C, Sculpher M. Productivity Costs: Principles and Practice inEconomic Evaluations. London: Office of Health Economics, 2000.

[6] Brouwer WB, van Exel NJ, Baltussen RM, Rutten FF. A dollar is a dollar-or is it? Value Health 2006;9:341–7.

[7] Jonsson B. Ten arguments for a societal perspective in the economicevaluation of medical innovations. Eur J Health Econ 2009;10:357–9.

[8] Bracco A, Krol M. Economic evaluations in European reimbursementsubmission guidelines: current status and comparisons. Expert RevPharmacoecon Outcomes Res 2013;13:579–95.

[9] Capri D, Ceci A, Terranova L, et al. Guidelines for economic evaluationsin Italy: recommendations from the Italian group ofpharmacoeconomic studies. Drug Inf J 2001;35:189–210.

[10] Norwegian Medicines Agency (Statens legemiddelverk). Guidelines onhow to conduct pharmacoeconomic analyses. In force from 1st March2012. Available from: http://www.legemiddelverket.no/English/price_and_reimbursement/application_for_reimursement/Lists/PageAttachments/default/NO/Pharmacoeconomic%20guidelines%20-%20Norway.pdf;11/08350-28. [Accessed September 15, 2014].

[11] Hjelmgren J, Berggren F, Andersson F. Health economic guidelines—similarities, differences and some implications. Value Health2001;4:225–50.

[12] Gerard K. Cost-utility in practice: a policy maker’s guide to the state ofthe art. Health Policy 1992;21:249–79.

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 5 3 – 7 5 8 757

[13] Stone PW, Chapman RH, Sandberg EA, et al. Measuring costs in cost-utility analyses: variations in the literature. Int J Technol Assess HealthCare 2000;16:111–24.

[14] Zhang W, Bansback N, Anis AH. Measuring and valuing productivityloss due to poor health: a critical review. Soc Sci Med 2011;72:185–92.

[15] Krol M, Brouwer W. How to estimate productivity costs in economicevaluations. Pharmacoeconomics 2014;32:335–44.

[16] Zhang W, Chiu JA, Bansback N, Anis AH. An update on themeasurement of productivity losses due to rheumatoid diseases. BestPract Res Clin Rheumatol 2012;26:585–97.

[17] Brouwer WB, Koopmanschap MA, Rutten FF. Productivity costs in cost-effectiveness analysis: numerator or denominator: a furtherdiscussion. Health Econ 1997;6:511–4.

[18] Krol M, Brouwer WBF. Unpaid Work in Economic Evaluations. Workingpaper In Press.

[19] Krol M, Brouwer W, Rutten F. Productivity costs in economicevaluations: past, present, future. Pharmacoeconomics 2013;31:537–49.

[20] van Roijen L, Essink-Bot M, Koopmanschap M, et al. Labor and healthstatus in economic evaluation of health care. The Health and LaborQuestionnaire. Int J Technol Assess Health Care 1996;12:405–15.

[21] Reilly MC, Zbrozek AS, Dukes EM. The validity and reproducibility of awork productivity and activity impairment instrument.Pharmacoeconomics 1993;4:353–65.

[22] Prasad M, Wahlqvist P, Shikiar R, Shih YC. A review of self-reportinstruments measuring health-related work productivity: a patient-reported outcomes perspective. Pharmacoeconomics 2004;22:225–44.

[23] Mattke S, Balakrishnan A, Bergamo G, Newberry SJ. A review ofmethods to measure health-related productivity loss. Am J Manag Care2007;13:211–7.

[24] Lofland JH, Pizzi L, Frick KD. A review of health-related workplaceproductivity loss instruments. Pharmacoeconomics 2004;22:165–84.

[25] Koopmanschap MA, Rutten FF. The impact of indirect costs onoutcomes of health care programs. Health Econ 1994;3:385–93.

[26] Koopmanschap MA, van Ineveld BM. Towards a new approach forestimating indirect costs of disease. Soc Sci Med 1992;34:1005–10.

[27] Koopmanschap M, Rutten F, van Ineveld B, van Roijen L. The frictioncost method for measuring indirect costs of disease. J Health Econ1995;14:171–89.

[28] Johannesson M, Karlsson G. The friction cost method: a comment. JHealth Econ 1997;16:249–55: discussion 257–9.

[29] Rothermich EA, Pathak DS. Productivity-cost controversies in cost-effectiveness analysis: review and research agenda. Clin Ther1999;21:255–67.

[30] Tranmer JE, Guerriere DN, Ungar WJ, Coyte PC. Valuing patient andcaregiver time: a review of the literature. Pharmacoeconomics2005;23:449–59.

[31] Brouwer WB, Koopmanschap MA. The friction-cost method:replacement for nothing and leisure for free? Pharmacoeconomics2005;23:105–11.

[32] Birnbaum H. Friction-cost method as an alternative to the human-capital approach in calculating indirect costs. Pharmacoeconomics2005;23:103–4.

[33] Koopmanschap MA. PRODISQ: a modular questionnaire on productivityand disease for economic evaluation studies. Expert Rev PharmacoeconOutcomes Res 2005;5:23–8.

[34] Sahlen KG, Lofgren C, Brodin H, et al. Measuring the value of olderpeople’s production: a diary study. BMC Health Serv Res 2012;12:4.

[35] Nyman J. Productivity costs revisited: towards a new US policy. HealthEcon 2012;21:1387–401.

[36] Language Policy Unit (Strasbourg). Common European Framework ofReference for Languages: Learning, Teaching, Assessment. Cambridge:Cambridge University Press, 2001.

[37] Meerding WJ, IJzelenberg W, Koopmanschap MA, et al. Health problemslead to considerable productivity loss at work among workers withhigh physical load jobs. J Clin Epidemiol 2005;58:517–23.

[38] Bouwmans C, De Jong K, Timman R, et al. Feasibility, reliability andvalidity of a questionnaire on healthcare consumption andproductivity loss in patients with a psychiatric disorder (TiC-P). BMCHealth Serv Res 2013;13:217.

[39] Brouwer W, Koopmanschap M, Rutten F. Productivity losses withoutabsence: measurement validation and empirical evidence. HealthPolicy 1999;48:13–27.

[40] Reid M. Economics of Household Production. New York: Wiley, 1934.[41] Severens J, Mulder J, Laheij RJ, Verbeek AL. Precision and accuracy in

measuring absence from work as a basis for calculating productivitycosts in The Netherlands. Soc Sci Med 2000;51:243–9.

[42] Severens JL, Laheij RJ, Jansen JB, et al. Estimating the cost of lostproductivity in dyspepsia. Aliment Pharmacol Ther 1998;12:919–23.

[43] Jacob-Tacken KH, Koopmanschap MA, Meerding WJ, Severens JL.Correcting for compensating mechanisms related to productivity costsin economic evaluations of health care programmes. Health Econ2005;14:435–43.

[44] Krol M, Brouwer WB, Severens JL, et al. Productivity cost calculations inhealth economic evaluations: correcting for compensationmechanisms and multiplier effects. Soc Sci Med 2012;75:1981–8.

[45] Pauly MV, Nicholson S, Xu J, et al. A general model of the impact ofabsenteeism on employers and employees. Health Econ 2002;11:221–31.

[46] Nicholson S, Pauly MV, Polsky D, et al. Measuring the effects of workloss on productivity with team production. Health Econ 2006;15:111–23.

[47] Pauly MV, Nicholson S, Polsky D, et al. Valuing reductions in on-the-jobillness: ‘presenteeism’ from managerial and economic perspectives.Health Econ 2008;17:469–85.

V A L U E I N H E A L T H 1 8 ( 2 0 1 5 ) 7 5 3 – 7 5 8758