Valuing inter-sectoral costs and benefits of interventions in the healthcare sector: methods for obtaining unit prices

ABSTRACT

Introduction: There is a lack of knowledge about methods for valuing health intervention-related costs and monetary benefits in the education and criminal justice sectors, also known as ‘inter-sectoral costs and benefits’ (ICBs). The objective of this study was to develop methods for obtaining unit prices for the valuation of ICBs.

Methods: By conducting an exploratory literature study and expert interviews, several generic methods were developed. The methods’ feasibility was assessed through application in the Netherlands. Results were validated in an expert meeting, which was attended by policy makers, public health experts, health economists and HTA-experts, and discussed at several international conferences and symposia.

Results: The study resulted in four methods, including the opportunity cost method (A) and valuation using available unit prices (B), self-constructed unit prices (C) or hourly labor costs (D).

Discussion: The methods developed can be used internationally and are valuable for the broad international field of HTA.

KEYWORDS:

• inter-sectoral costs and benefits
• valuation methods
• costs and cost analysis
• education
• criminal justice
• healthcare sector

Introduction

Many interventions within the health-care sector have costs and monetary benefits which spread to other sectors. These are known as ‘inter-sectoral costs and benefits’ (ICBs) [1–3]. Drost et al. identified more than 70 ICBs of health-care interventions, which were classified into five categories. These include the sectors ‘Education’ (e.g. special education), ‘Labor and Social Security’ (e.g. productivity), ‘Household & Leisure’ (e.g. informal care), ‘Criminal Justice System’ (e.g. police interventions), and a fifth category for ‘individual and family effects’ (e.g. family conflict) [1].

ICBs could contribute a considerable amount to the total monetary costs and monetary benefits (throughout this paper defined as reduction of monetary costs) of interventions in the health-care sector, with consequent serious effects on the outcomes of their economic evaluations. Earlier research has shown that including ICBs within economic evaluation studies is important and that methods are needed to value them [4–6]. However, this research on revealed preference methods for valuing ICBs was restricted mainly to productivity costs (human capital method, friction cost method) [7,8] and informal care costs (opportunity cost method, proxy good method) [4,6], while over half of the identified ICBs in our classification scheme belong to the educational sector and criminal justice sector [1]. The lack of attention to the educational and criminal justice sector is due, among other things, to the lack of methodological knowledge which is required for valuing these ICBs [1–3,9].

The importance of a monetary valuation of the aforementioned ICBs is receiving increasing recognition [1,3,9]. This is partly due to the growing interest in (social) cost–benefit analyses (CBAs) and the increased attention for ‘Health in All Policies’ (HiAP) [10–12]. This is true internationally and for the Netherlands in particular. Illustrative is the mention of the ICBs ‘special education’ and ‘legal costs’ in the updated version of the Dutch manual for costing in economic evaluations [13,14]. In addition, guidelines have recently been developed for structuring social CBAs [11,15]. For such analyses, reliable information on costs and benefits, such as educational and criminal justice ICBs, is essential. The monetary valuation of these ICBs is, therefore, more and more important.

In short, although recognition of the importance of including educational and criminal justice ICBs in economic evaluations is growing, as yet little is known within health economics literature about the methods for obtaining unit prices which can be used for valuing these ICBs. Therefore, the main aim of this study was to develop generic (internationally usable) methods to obtain unit prices which support their valuation. A second aim was to assess the feasibility of these methods by applying them in practice. Within this study, we chose the Netherlands.

Methods

Selecting relevant ICBs

Educational and criminal justice ICBs were taken from a classification scheme [1], after which a selection was made of those which could likely be expressed in monetary terms using revealed costs and reported unit prices. To distinguish between likely and not likely, those concerning changes in damages and use of services were separated from those concerning changes in behavior or functioning (which are often causally linked to these changes in damages and use of services). The second group of ICBs, which are commonly valued using stated preference methods, was excluded from further analysis. For example, as a change in ‘special educational services’ stands for a change in use of services, this ICB was considered eligible. Conversely, the ICB ‘teacher–student conflicts’ falls under change in behavior and was, therefore, excluded.

Developing and applying methods for the valuation of ICBs

The total costs and monetary benefits per ICB can be calculated by multiplying a unit price with a measured volume change. The search was, accordingly, for methods which could produce reliable unit prices.

Providing and assessing methods for the valuation of ICBs required a four-stage process. First, since the level of usability affects whether a particular method will be adopted, assessing usability was an essential part of the development process. Therefore, a study needed to be performed on the structures of the educational and criminal justice systems. Information on the Dutch educational system and criminal justice system was drawn from acknowledged, publicly available governmental reports [16,17]. Outcomes were used to break down the selected ICBs from the classification scheme into the country-specific ICBs to which methods could be applied.

Second, to develop methods for the valuation of these ICBs, a scoping review was conducted. The search was for acknowledged reports, such as the above-mentioned governmental reports, and relevant databases which provide reliable unit prices or information on the basis of which unit prices can be constructed. Therefore, the search was also for information sources containing underlying methodologies and figures containing total costs related to ICBs. Based on the information gathered, lists with unit prices were constructed.

Third, as part of this explorative process, several interviews were conducted with experts on the educational and criminal justice sectors. Among these experts were the authors of the consulted cost figures. Guided by predetermined, open-ended questions, they were asked to provide feedback on the results of the literature study. Questions were asked concerning the selection of ICBs and methods for obtaining unit prices. Apart from those interviewed, if the methodology for cost calculations was not clear, several other authors were contacted and asked to provide clarification via e-mail. Expert answers were used to make adaptations to the lists. For uniformity purposes, unit prices found and/or calculated were indexed for the year 2012 using consumer price index numbers found on Statline [18].

Fourth, to assess and improve the usability of the unit prices for the valuation of ICBs, units of prices had to reflect the standard measurement units, that is, hours, contacts, etc. Measurement units were sought using the Database of Instruments for Resource Use Measurement (DIRUM) [19] and the website of the Institute for Medical Technology Assessment (iMTA) [20], which were consulted for prominent questionnaires containing items concerning educational and criminal justice services. Measurement units were drawn from the questionnaires [21–27], after which prices were compared with these units and, if found to be necessary and possible, converted.

Expert meeting and conferences

To validate the valuation methods and resulting unit prices, an expert meeting was held in the fall of 2013. Experts were recruited based on expertise. The meeting was attended by lead authors of Dutch manuals on costing in health economic evaluations and other researchers who fulfilled the criterion of being a leading expert in the field of public health, health economics, and/or health technology assessment (HTA). Furthermore, to discuss the policy implications of this study, the meeting was attended by policymakers within the field of public health. All authors of this study were present as well.

Methods and results of the study were sent to the experts before the meeting so that they could prepare for it. Through an e-mail, all experts were asked to assess the methods. Furthermore, three subgroups were made based on the individual expertise and experience of the experts, with each being specifically asked to assess the methods’ application to the educational sector, the criminal justice sector, or policy implications. However, all experts had access to the full contents of the paper so that they could provide feedback on other parts as well.

During the semistructured meeting, the findings were presented by the lead author (RD) and discussed by all participants. The meeting was structured into four themes: (1) methods for valuing ICBs, (2) application to the educational sector, (3) application to the criminal justice sector, and (4) policy implications, which were consecutively addressed. Apart from policy implications, experts were asked to provide feedback and discuss the selection of ICBs, (inter)national usability of the valuation methods, and unit prices. In each phase, first the expert subgroup was asked to provide feedback, after which all could join the discussion. The expert meeting was recorded. For the names and affiliations of the experts, see the acknowledgements section of this article.

The analysis of the qualitative responses was conducted in several steps. Recordings were used to extract important suggestions, remarks, and comments made by participants in the expert meeting. Subsequently, these expert answers were clustered according to topic. Before adjustments were made to the methods and unit prices, strategies were discussed by the project team (RD, AP, DR, SE) until consensus was reached. Finally, a letter of reply, in which the clustered comments of the experts and strategies of how to deal with these were specified, was sent to the expert meeting participants in the beginning of 2014. Experts were given the opportunity to send final comments via e-mail, and these were used to make final adjustments.

In addition to the expert meeting, to strengthen the validity of the results, the methods for the valuation of ICBs were also presented and discussed at various workshops and conferences, including a meeting with health economists and HTA experts in Nunspeet (the Netherlands), organized by the Lowlands Health Economics Study Group (LolaHESG), at the ‘Dutch Conference on Public Health’ in The Hague (the Netherlands), and during an organized session at the ‘iHEA 10th World Congress on Health Economics’ in Dublin (Ireland).

Results

Relevant ICBs

The selection process resulted in two shortlists: one for the educational sector and one for the criminal justice sector (Table 1). As depicted in Table 1, less than half of the ICBs within the educational sector were comprised of damages or use of services and were, accordingly, considered eligible for further analysis. The relative and absolute number within the criminal justice sector is higher. This implies that, in comparison with educational ICBs, there are more criminal justice ICBs concerning damages and use of services.

Table 1. Selected ICBs.

	Educational sector	Criminal justice sector
Selected	- Problems with school entry - Grade retention - School dropout/premature leave - Special education services (schools) - Special education services (personal budget) - Special education services (transport) - Additional study guidance - Attendance officer	- (Im)material damages (victims) - Lost work of victims - Illegal untaxed income - Police services, prosecutor - Judicial expenses (lawyer) - Judicial expenses (court) - Forensic services - Forensic psychiatrist - Imprisonment - Fines and transactions - Other penalties - Lost work and belongings (offender) - Probation - Child protective services
Nonselected	- School readiness, adaption, competence, participation, engagement, attainment, productivity, meeting obligations, experience - Cognitive deficits (e.g. learning disabilities) - Change in educational outcomes, intellectual and academic performance - School teacher–student relationship - Home education, residential programs	- Violence and crime (domestic violence, unlawful threats, assaults, violence toward officials, drunk driving, sexual assaults toward adults, homicide, vandalism, theft, and child maltreatment, including physical, psychological, emotional, sexual abuse, and neglect) - Parole

Based on information drawn from yearly reports of the Ministry of Education, Culture and Science, and the Ministry of Security and Justice [16,17], the generic ICBs (Table 1) were broken down into 8 Dutch ICBs for the educational sector and 14 Dutch ICBs for the criminal justice sector (Tables 2 and 3). For example, the classification scheme provides the ICB ‘special education services’. Within the Netherlands, for some disabilities, these services can be provided in special schools and, through a personal budget, in regular schools. Therefore, a distinction was made between ‘special education (schools)’ and ‘special education (personal budget)’.

Table 2. Checklist of ICBs within the educational sector and related unit prices.

ICB	Methods^a	Unit price (euro, 2012)^b	Unit	Source^c
Problems with school entry	B, C	5092.88–22,533.71	per year	Ministry of Education, Culture and Science (Ministerie van Onderwijs, Cultuur en Wetenschap)
Grade retention		25.46–112.67	per day
School dropout/premature leave		5.42–23.97	per hour
Special education (schools)
Special education (personal budget)	B	3065.67–26,076.57	per budget	50tien Special Education (50tien Steunpunt Passend Onderwijs)
Special education (transport)	B	2123.34	per year	CPB Netherlands Bureau for Economic Policy Analysis (Centraal Planbureau)
		10.62	Per day
Additional study guidance	B	29.27	per hour	Check5 Homework Assistance (Check5 Huiswerk Begeleiding)
Attendance officer	D	40.61	per hour	Statistics Netherlands – Statline (Centraal Bureau voor de Statistiek – Statline)

^aMethods include the opportunity cost method (A), valuation based on unit prices found in acknowledged sources (B), valuation based on self-constructed unit prices (C), and valuation based on hourly labor costs (D). ICBs for which no unit prices were found were marked with an ‘E’.

^bMethods and unit prices for problems with school entry also apply to grade retention, school dropout/premature leave, and special education (schools). Ranges in unit prices result from price differences between several types of education [32].

^cAn extensive description of the sources, the methods with which unit prices were obtained and the level to which unit prices match measurement units can be found elsewhere [32].

Table 3. Checklist of ICBs within the criminal justice sector and related unit prices.

ICB	Methods^a	Unit price (euro, 2012)^b	Unit	Source^c
(Im)material damages^c	B,C	110.94–3438.76	per case	Victim Support the Netherlands (Slachtofferhulp Nederland), The Violent Offences Compensation Fund (Schadefonds Geweldsmisdrijven), Dutch Guarantee Fund (Waarborgfonds Motorverkeer)
Lost work of victims	E	–	–
Illegal untaxed income	E	–	–
Police, prosecutor	D	40.61	per hour	Statistics Netherlands – Statline (Centraal Bureau voor de Statistiek – Statline)
Judicial expenses (lawyer)	B	231.00–1455.00	per legal assistance	Ministry of Security and Justice (Ministerie van Veiligheid en Justitie), Legal Aid Board (Raad voor Rechtsbijstand)
Judicial expenses (court)	B	141.72–4003.10	per case	Ministry of Security and Justice (Ministerie van Veiligheid en Justitie), Council for the Netherlands Judiciary (Raad voor de Rechtspraak)
Forensic services	C	1420.05	per product	Netherlands Forensic Institute (Nederlands Forensisch Instituut)
Forensic psychiatrist	D	40.61	per hour	Statistics Netherlands – Statline (Centraal Bureau voor de Statistiek – Statline)
Imprisonment	B	32.79–1597.22	per day	Ministry of Security and Justice (Ministerie van Veiligheid en Justitie)
Fines and transactions	B	5.53–4545.07	per fine/transaction	Ministry of Security and Justice (Ministerie van Veiligheid en Justitie)
Other penalties	E	–	–
Lost work and belongings (offender)	E	–	–
Probation	C	1837.50–2042.74	per product	Dutch Probation Service (Reclassering Nederland), The Salvation Army (Leger des Heils)
Child protective services	E	–	–

^aMethods include the opportunity cost method (A), valuation based on unit prices found in acknowledged sources (B), valuation based on self-constructed unit prices (C), and valuation based on hourly labor costs (D). ICBs for which no unit prices were found were marked with an ‘E’.

^bRanges in unit prices result from the variety in products, such as differences in costs for several types of court cases or multiple information sources [32].

^cAn extensive description of the sources, the methods with which unit prices were obtained and the level to which unit prices match measurement units can be found elsewhere [32].

Valuing ICBs

The study resulted in four methods (referred to as methods A to D), which can be systematically and consecutively applied to produce unit prices for the valuation of ICBs (Figure 1).

Figure 1. Methods for the valuation of ICBs^a.

^aThe order in which methods are placed suggests that there is a hierarchy. However, the choice of method is also dependent on the quality of the data sources.

During the expert meeting, the idea was proposed and discussed to use the opportunity cost method for valuing ICBs (method A). Opportunity costs are the benefits omitted as a result of the time spent providing alternative services [4,28,29]. For example, to value informal care, opportunity costs equal the informal caregiver’s benefits foregone as a result of spending time on providing informal care [4]. To apply this method, information is required on both the time spent providing these services and an individual’s market wage rate or reservation wage rate for those in non-labor market participation [6]. As for informal care, this method can also be applied for valuing ICBs in the educational and criminal justice sectors. An advantage of this method is the accuracy with which costs and benefits can be valued in monetary terms, especially when micro-costing is applied to assess the economic costs. However, although micro-costing produces the most precise estimate of true costs [28], it has a disadvantage in that it is time-consuming to gather all of the required information. This is especially true when individuals’ market wages differ and when these differences are hard to retrieve, hereby rendering it a high quality but also a less feasible method to value ICBs.

When specific information on an individual’s market wages is not available or is considered too time-consuming to approximate, valuation can be based on proxy unit prices, which should be sought in acknowledged reports (method B). Key here is that these reports should be reliable, freely accessible, and transparent, and that these should be cited when used. These include, among others, governmental annual reports and annual reports of related institutions responsible for providing services related to ICBs. Unit prices can be retrieved from reports of the ministries covering policies within those ICB sectors for which unit prices are required. For example, unit prices for special education may be retrievable from the annual report of the ministry responsible for policies within the educational sector. Policies on ICBs such as these, however, can also be governed by regulatory public or private organizations outside the central government which have been granted authority and responsibility for providing services related to ICBs. An example of such an ICB is reintegration after imprisonment, which can be supervised by an independent organization. In this case, the unit price should be sought in the annual report of this organization.

If unit prices are not provided, valuation can be based on self-constructed unit prices (method C). The same reports should be searched for total costs provided in income statements and production output figures. Unit prices can then be calculated by dividing total annual costs by total annual output. When organizations provide multiple types of services, unit prices should be calculated based on the specific costs and output figures of these specific services. Given the example provided earlier, if the unit price for reintegration cannot be found using method B, the annual reports of the organization responsible for reintegration can be searched for the total costs for this specific service and the number of subjects who were provided this service.

When reports do not provide the required information, valuation can be based on hourly labor costs (method D). Acknowledged data sources should be searched for labor costs for services related to ICBs. Reliable data on labor costs is often stored and made available publicly by the government or organizations which have been given responsibility by and can be held accountable by the government. Looking at the ICB reintegration after imprisonment, the hourly labor costs of those providing reintegration services may be retrievable from an acknowledged data source. In this case, to convert hourly labor costs to the costs of reintegration, additional information is needed on the number of hours of service required to complete a reintegration trajectory.

The order in which methods are placed suggests that there is a hierarchy, with one method being dominant over or inferior to another. This was discussed during the expert meeting. Although the high level of detail to which micro-costing produces reliable cost estimates places the opportunity cost method on the top of this hierarchy – which was agreed upon during the expert meeting – the order in which the other methods were placed had to do with feasibility. Nevertheless, what should be clear is that the reliability of the data sources is important for choosing between methods B, C, and D. This means, for example, that the data source which is used for method D can sometimes be more reliable than the one used for methods B and C. In these cases, method D should be applied. We tested the usability and feasibility of methods B, C, and D, of which the results are provided below. For the results of the feasibility, when no unit price could be obtained or calculated based on the analysis of consulted sources, the ICB is marked with an ‘E’.

Usability and feasibility

To assess the usability and feasibility of methods B, C, and D, we applied them to obtain unit prices for the valuation of ICBs in the Netherlands. Application of methods B and C required acknowledged reports, such as the annual governmental reports which were consulted for background information on the Dutch educational and criminal justice system [16,17]. For method D, hourly labor costs were derived from Statline, the public website service of Statistics Netherlands [18]. A comparative analysis of regularly used questionnaires and units attached to unit prices showed that units of unit prices acquired using method D (per hour) differed from frequently found measurement units (per contact). These unit prices were not converted. Examples include the ICBs ‘police’ and ‘attendance officer’. Since the duration of contacts will differ within and between studies, these can, in combination with additional information on the duration of contacts, be used to calculate specific unit prices.

As shown in Table 2, most of the unit prices for the valuation of ICBs within the Dutch educational sector were produced using method B. Unit prices related to problems with school entry are the same as those needed to value grade retention, school dropout/premature leave, and special education (schools). Ranges in unit prices result from the variety in products. An example is the price differences between several types of education (5092.88–22,533.71 euros per year). Unit prices for these ICBs per day and per hour were calculated, respectively, by dividing the costs per year by the annual amount of schooldays and hours. The unit price needed for valuation of the activities of attendance officers could not be derived from information drawn from reports.

As shown in Table 3, for 5 out of 14 ICBs within the Dutch criminal justice sector, no unit price could be obtained or calculated. Consulted sources lacked unit prices or the information required to produce unit prices. For six out of the resulting nine other ICBs, unit prices were identified using method B. Due to the variety in types and severity of (im)material damages, both methods B and C were used. To adapt for the variety in measurement units used to express services provided by the police and prosecutors, both methods B and D were used. Ranges in unit prices result from the variety in products. An example is the price differences among several fines and transactions (5.53–4545.07 euros per fine/transaction).

Discussion

Main findings

The main objective of this study was to develop methods which support the valuation of educational and criminal justice ICBs of interventions in the health-care sector. A second objective was to apply these methods to assess their feasibility. Multiple methods were proposed, which can be applied internationally to value ICBs in monetary terms. These include the opportunity cost method (method A), valuation based on a unit price provided by an acknowledged source (method B), valuation based on a self-constructed unit price (method C), and valuation using hourly labor costs (method D). Within this particular study, method A was considered not feasible for deriving unit prices. However, overall, the combined application of methods B, C, and D turned out to be feasible. The application resulted in two lists with unit prices, one for the educational sector and one for criminal justice sector, which can be used for costing research in the Netherlands.

Limitations

The findings of the study need to be placed in the context of the study’s limitations. First, this study was conducted based on the conclusion that little is known about the methods for valuing ICBs. Since we adopted a health-care perspective and used health as a starting point for this study, this conclusion was reached from studying health-care evaluation literature only. A study of the evaluation literature in the educational and criminal justice domains could have provided us with more information regarding the impact of health outside the health-care sector.

Second, as noted earlier, the study resulted in a certain ‘hierarchy’ between valuation methods, which can be systematically and consecutively applied to produce unit prices for the valuation of ICBs. However, the variety in methods still raises the question of whether there is a ‘best’ method. Further research on the feasibility, reliability, and validity of the unit prices may provide an answer to this issue. In this regard, it might be helpful to develop measurement tools which are specifically designed to cover ICBs. This might improve the extent to which the unit prices can be applied. In response to the applicability, however, this is partly met by adapting the units of prices to measurement units.

Implications for research and policy

Although the aim was to develop generic methods, a characteristic of the products – that is, unit prices – is that these are country-specific. Part of this specificity can be explained by intercountry differences between the structures of educational and criminal justice systems. For example, differences in the way special education is organized will affect the number and actual size of related unit prices. Researchers operating outside the Netherlands could retrieve or calculate unit prices fit for use within their own country by applying the valuation methods provided in this study. Although the success of applying these methods is dependent on the availability of local or national data and might differ between countries, it is expected, within this information age, that this availability will often be sufficient. The methods provided in this study will, therefore, be valuable for the broad international field of HTA.

Including ICBs within costing research may improve the reliability of economic analyses. The types of analyses for which this might be the case are, among others, cost–effectiveness analysis (CEA), cost–utility analysis (CUA), and CBA. As for informal care [6], it is expected that the monetary costs and benefits (i.e. reduction in costs) related to the ICBs which have been selected in this study can be included in the cost side of CEA and CUA. Double counting with effect or utility outcome measures such as the QALY, which should be avoided in health economic evaluations [30], is not expected to be a serious problem. Nevertheless, in these analyses, the decision to invest time in measuring ICBs could be dependent on their expected weight relative to the total costs and benefits of an intervention. Further research could provide insight regarding the circumstances (such as types of interventions) in which ICBs should be included.

When adopting a societal perspective, which is considered to lead to the broadest representation of costs and benefits, it is important to value all affected ICBs, regardless of who pays for the costs or receives the benefits [31]. An intervention can be perceived to be not worthwhile from a narrow perspective, whereas from a societal perspective benefits can exceed costs [9]. A complete and reliable coverage of ICBs in the educational and criminal justice sector can, therefore, be of added value for making decisions on interventions in the health-care sector. This can be accomplished by adopting ICBs within social CBAs, which – as noted earlier – are becoming more important in decision-making processes.

In addition to policy making within the health-care sector, an overview of the financial impact of health interventions may draw the interest of policymakers operating outside the health-care sector for health-related issues. This, in turn, may stimulate policymakers operating outside the health-care sector into putting health-related issues on their agenda. Adopting health-related policy in policy domains outside the health-care sector is considered a step toward including HiAP [12]. Adopting a societal perspective in research, as was performed in this study, may accelerate the process of realizing HiAP.

Dissemination

We have laid out the classification scheme, valuation methods, and lists with unit prices, along with detailed descriptions of the specific processes for producing unit prices and the implications for researchers and policymakers, in a manual for conducting and assessing research on ICBs in the Netherlands [32]. The manual provides support during the classification, identification, and valuation of ICBs. Simultaneously, a plan was written for its dissemination. The contents of the manual were presented at several conferences (see section ‘Methods’) and a website has been created, in which user experiences can be shared with the authors [33]. Furthermore, the manual on ICBs will be referred to in the upcoming updated version of the official Dutch manual for costing in economic evaluations of The Dutch National Health Care Institute.

Although the unit prices depicted in this manual are limited to use only in the Netherlands, the generic methods provided in this study can be used for conducting research on ICBs in other countries as well. The application of the manual will provide further insight into their feasibility and value for specific costing research studies.

Key Issues

• Many interventions within the health-care sector have costs and monetary benefits which spread to other sectors. These are known as ‘inter-sectoral costs and benefits’ (ICBs).

• Current methods to value ICBs are restricted mainly to productivity costs and informal care costs. The lack of attention on the educational and criminal justice sectors in health economic evaluations is due, among other things, to the lack of methodological knowledge which is required for valuing these ICBs.

• The study resulted in four methods which support the valuation of ICBs. These include the opportunity cost method (A), valuation using unit prices provided by acknowledged written sources (B), valuation on the basis of self-constructed unit prices using information retrieved from these sources (C), and valuation using hourly labor costs (D). Application showed that especially methods B and C turned out to be feasible.

• Depending on the type of health-care intervention being evaluated, including ICBs within costing research may improve the reliability of economic analyses. The types of analyses for which this might be the case are, among others, CEA, CUA and CBA.

• As a complete and reliable coverage of ICBs in the educational and criminal justice sectors improves the quality of decision-supportive information, it will be of added value for making decisions on interventions in the health-care sector. Decision makers within the health-care sector should carefully consider which ICBs should be included when setting out research projects.

• Interventions in the health-care sector may serve as effective tools to reach non-health objectives (e.g. improving educational outcomes and lowering crime rates). Therefore, reporting ICBs of health interventions may draw the interest of policymakers operating outside the health-care sector for health-related issues. Adopting health-related policy in policy domains outside the health-care sector is considered a step toward HiAP.

• The valuation methods and lists with unit prices, along with detailed descriptions of the specific processes for producing unit prices and the implications for researchers and policymakers, were used to construct guidelines for conducting and assessing research on ICBs in the Netherlands.

• The generic methods provided in this study can be used to develop guidelines on ICBs in other countries as well.

Financial & competing interests disclosure

This research was supported by Grant 200400010 from the Dutch Organisation for Health Research and Development (ZonMw), The Hague. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgements

The authors would like to thank Eddy Adang (Radboud University Medical Center, Nijmegen), Lea den Broeder (National Institute for Public Health and the Environment, Bilthoven), Wim Groot (The Council for Public Health and Health Care, The Hague), Leona Hakkaart-van Roijen (Institute for Medical Technology Assessment, Rotterdam), Jack Hutten (Ministry of Education, Culture and Science, The Hague), Debora Moolenaar (Ministry of Security and Justice, The Hague), Johan Polder (National Institute for Public Health and the Environment, Bilthoven), Gerbert Romijn (Netherlands Bureau for Economic Policy Analysis, The Hague), Filip Smit (Trimbos Institute, Utrecht), Rolf van der Velden (Maastricht University, Maastricht) and Ardine de Wit (National Institute for Public Health and the Environment, Bilthoven) for their valuable feedback on the methods and results of this study. The authors would also like to thank the discussants and participants of the following conferences: ‘Dutch Conference on Public Health’ in The Hague, April 2014, ‘Lowlands Health Economists’ Studygroup’ in Oostvoorne, May 2014, ‘iHEA 10^th World Congress on Health Economics’ in Dublin, July 2014 and ‘Symposium on Social Cost-Benefit Analysis and Inter-Sectoral Costs and Benefits’ at the Ministry of Health, Welfare and Sport in The Hague, October 2014. Furthermore, the authors would like to thank Barbara Greenberg for her English editing services.