20-05-2014
Cost-Benefit Analysis of Psychological Therapy
in the Netherlands
“To what extend is providing therapy to the people that need it but do
not yet receive it a viable policy in the Netherlands?”
Vera Cuijpers 6349994/10088822 Economics
University of Amsterdam, Faculty of Economics and Business Supervisor: Lukáš Tóth
Table of contents
1. Introduction 3
2. The Improving Access to Psychological Therapy (IAPT) program 6
3. Mental health specifics of people in the Netherlands 8
4. Cost-benefit analysis of CBT in Holland 10
4.1. Method 10 4.2. Cost 12 4.3. Benefits 15 4.4. Population implications 19 5. Implications 22 6. Discussion 24 7. Appendices 26 8. Bibliography 30
1. Introduction
Bit by bit it is getting more usual and less taboo to discuss mental illness. From depression to phobias; it is on the news, friends and family discuss it together and there are more than one pamphlets about it at the doctors’ office. But what are the economic consequences of these common mental disorders? And could this be dealt with more efficiently? This is a cost-benefit analysis of mental health care provision in the Netherlands.
Since life expectancy at birth has increased immensely over the past decades the public health attention is shifting (Olshansky et al., 1991). The choices that are made in healthcare are more and more focusing on quality as well, instead of just focusing quantity, which has been the case in the past (Committee on Medical Cure and Care, 1991). This opens a more prominent role for mental healthcare, since most mental disorders are not fatal but decrease the standard of living dramatically (Melse et al., 2000).
In its research on the Global Burden of Disease, the World Health Organization (WHO) concludes that in Western Europe about 40% of all disability is due to mental illness. Compared to the rest of the world this is quite substantial (World Health Organization, 2008). Recent studies find that in the Netherlands about 42% of people has a mental disorder at some point in life (de Graaf et al., 2012a), so this is affecting a large part of the population. Even more since mental health disorders are known to have a large effect on the people surrounding the patient, on average more than is the case with physical diseases (Idstad, Røysamb, & Tambs, 2011).
Furthermore, studies by Smit et al. (2006) show that nowadays the overall costs of mental disorders are of similar heights as of physical illness, while for this type of health care the cost of illness is much more detailed and discussed. And the part of the health care budget that goes to treatment of physical illness is generally higher than the part that is spend on treatment of mental illness, as is illustrated on the next page in Figure 1 for the year 2012 (Statistics Netherlands, 2014). So a study on costs and benefits of mental healthcare could stress the importance of allocating the budget more evenly among the different parts of the health care system.
Furthermore, the Dutch government is currently dealing with very large budget cuts, which are also visible in the different parts of the health care sector. This is leading to extensive revision and major restructuring of the entire sector. Which is a good reason to look into the costs and benefits now and see if the funds could be allocated more efficiently.
Moreover, next to the health benefits of additional supply of therapy there could also be economic benefits, as is the argument of Bell et al. (2006) in their “Depression Report: A New Deal for Depression and Anxiety Disorders”. In this paper they argue that the excess economic costs of mental disorders could be substantial, among others because of extra work day losses and more visits to the doctor and hospital. Several studies examining the costs of mental illness suggest the same (Smit 2006; Layard, 2007; Cuijpers, 2006).
So it is important to do extensive research on the economic costs and benefits of therapy. Not only to improve the overall health situation, but it could simultaneously lead to more efficient allocation of government funding.
In the United Kingdom (UK) much population research on the prevalence and treatment of mental illness has been conducted, in the form of the National Psychiatric Morbidity Survey. Based on results of this survey and the studies done with the data that are retrieved from the survey, a new program was implemented in the UK, the “Improving Access to Psychological Therapies” (IAPT) program. Which basically is a large investment to increase the number of treated people with mental health problems. With the cost-benefit analysis of this paper we want to see if it could be profitable to develop a similar program in the Netherlands to increase the availability of therapy.
Therefore, the question that will be answered in this paper is: “To what extend is providing
therapy to the people that need it but do not yet receive it a viable policy in the Netherlands?
So we want to know if it would be socially profitable to make the investment of educating new therapists, develop units where people can be treated, make people aware of the availability of treatment to provide therapy to the people that have a mental disorder but do not yet have therapy. To say it differently: do the social benefits justify the social costs from an economic point of view?
Figure 1. Healthcare spending in the Netherlands (2012)
0% 20% 40% 60% 80% 100%
Hospital care Elderly care GPs
Mental health care
Pharmacies and medication Healthcare for disabled Dentists
Other
A leading article in the cost analysis of mental disorders by Smit et al. (2006) shows that the largest part of the costs of illness are the indirect costs. These are the costs that could be saved when people are successfully treated, like cost of absenteeism, additional doctors’ visits etcetera. The indirect annual per capita excess costs are €2,725 against total medical costs of €479. So the costs saved by providing therapy are multiple times as large as the costs of actually providing it. The same conclusion is drawn by Layard et al. (2007) who wrote one of the main underlying economic arguments that led to the implementation of the IAPT program in the UK.
Since the Netherlands is a rather small country the country-specific knowledge is fairly limited. However, there is more on for example the United States of America (USA), the United Kingdom and the western world as a whole (Bernert, 2004).
Although the methods, calculations and actual values differ considerably per study, it is safe to say that previous published literature concludes that the benefits will outweigh the costs. This is because in most cases prevention or treatment from the beginning is overall less expensive than dealing with the consequences after the problem had time to develop further and cause more and more problems. So the outcome of this study is also expected to show that providing therapy to the people that need it but do not yet receive it would be a viable policy in the Netherlands.
The thesis is structured as follows. First, the IAPT program in the UK will be explained. Then we will look at the specific features of mental health of people in the Netherlands. Third we get into the methods and results of the cost-benefit analysis that is applied. Subsequently we will elaborate on the implications that the outcome of the analysis will have for the program we are discussing. And finally in the discussion we mention the strengths and shortcomings of the analysis and give recommendations for further research.
2. The IAPT program
The Improving Access to Psychological Therapies (IAPT) is a government program implemented in the United Kingdom (UK). The National Health Service (NHS) invests millions of pounds in rolling out a service network all across England offering treatment to people with depression and anxiety disorders (“IAPT”, 2013). It started with the National Institute of Health and Clinical Excellence (NICE) guidelines, which say that everyone with these type of disorders should be treated with suitable therapy (NICE Guidelines, 2013). Since traditionally only medication treatment was available, a nationwide program was developed to improve the IT and workforce infrastructure of mental healthcare therapy.
There are a few ways in which IAPT will improve the mental health care system in the UK. They increase the overall availability of therapists by educating people that are already working in the mental health care. By locating the new units on spread out locations the services should become available all over the country. As well the time from referral to assessment was established which should not be more than two weeks.
In 2006 the program began in Doncaster and Newham with demonstration sites (IAPT: Demonstration Sites, 2013) that focused on improving access to psychological therapies services for adults of working age. These demonstration projects had great results in the first year (Clark et al., 2009), leading to the development of the full, nationwide IAPT program.
The program started in 2008 with the “Implementation Plan: Guidelines to National Delivery” (Department of Health, 2008). The overall program should take about six years, divided into two phases, coming to an end in 2015. Investing three hundred million pounds in the years 2008 – 2011, and another four hundred the four years after that. Ending the roll-out of the program in the year 2015.
The goal of the first phase was to train 3,600 new psychological therapists, treating 900,000 more people and provide access to services in at least half of the country by spring 2011. This goal was partly accomplished, with 3,660 newly trained cognitive behavioral therapy workers and over 600,000 people that started treatment (Gyani, Shafran, Layard, & Clark, 2013). 142 of the 151 Primary Care Trusts in England had a service from this program in at least part of their area and just over 50% of the adult population had access and 23,000 people came off sick pay or benefits (Department of Health, 2012).
The second phase is now in running, broadening the scope of the operation, according to the publication of Talking Therapies: A Four-Year Plan of Action (Department of Health, 2011) and No Health Without Mental Health (HM Government, 2011), the cross-Government
mental health strategy for people of all ages. In the four years to April 2015 the program will complete the nationwide roll-out of CBT services for adults. This means treating a minimum of 15% of the total prevalence of mental disorders, which is about six million people each year. Furthermore a stand-alone program for children and young people will be launched, as well as models of care for people with more complicated situations; long-term physical conditions, medically unexplained symptoms and severe mental illness (National Health Service, 2009).
The original justification for the IAPT program was strongly based on the idea that next to health improvement of the population, it would pay for itself, developed by a group of economists and clinical researchers (Centre for Mental Health, 2006). They argue that improving the mental health of the population will overall save the NHS a lot of money in the long run because of higher employment (The National Development Team for Inclusion, 2009) and reduction in healthcare usage (Bell et al., 2006). Evidence shows that the NHS can save up to 272 million pounds by implementing the IAPT program and the wider public sector will benefit by even more ( IAPT, 2013).
In conclusion, the IAPT program will not only improve the mental health of the British population, but will also make the spending on health care more efficient. To see if a program like this would have the same effect in the Netherlands we first take a extensive look at the mental health specifics of people in the Netherlands and then apply a cost-benefit analysis.
3. Mental health specifics of people in the Netherlands
Before starting the analysis of costs and benefits of providing therapy, we will sketch the Dutch mental health situation at this point in time. The main part of the information that is available on the Dutch mental health comes from the Netherlands Mental Health Survey and Incidence Study by the Trimbos Institute (NEMESIS). This is a study that has been done about 15 years ago, and was repeated in 2010 to be able to see the development over the years. The tests consist of three measurement times, one to two years apart. The goal of the NEMESIS studies is to gather data on all different aspects of mental and personality disorders, about the prevalence, how they originate, the course of the disorders and the consequences. Because the study has different moments of measurements it allows to study trends in for example which part of the population suffers most from mental illness and trends in health care usage among people with a disorder. Overall these kind of studies provide valuable information for developing policy and further research.
NEMESIS shows that in the Netherlands there is a relatively high prevalence of mental health disorders. Every year about 18% of the population between 18 and 64 deals with a mental disorder of some sort. On the other hand, the percentage of 12-month prevalence of mental disorders was at 23.5% in NEMESIS-1. Even though the percentage has come down in the past years the number is quite high if you compare it with other similar countries. In Europe this 12 month prevalence of mental disorders lies around 10% (Alonso et al. 2004). From this European Study of the Epidemiology of Mental Disorders (ESEMeD) of Alonso et al. (2004) it is also found that in Europe significantly more women are dealing with a disorder than men in the past year while this gender difference in the Netherlands is rather small.
The group of people that suffered from a mental disorder in the past 12 months has some characteristics that are specific to this group. First of all the unemployment rate is higher than among the normal population in the group that has had one of the disorders. Furthermore from the NEMESIS data it shows that a person in the low income class has a higher probability of having a mental disorder than people with a medium or high income. This probably correlates with the fact that people with a higher level of education have less probability of having a disorder, assuming that in general higher education leads to higher income. Finally, the study also shows that people living alone have higher chance of getting ill than people living with a partner (de Graaf et al., 2012a).
The distribution among the different disorders seems more or less the same in the Netherlands and the rest of Europe, major depression and specific phobia seem to be the most common disorders. One thing that stands out is that the Netherlands has a relatively high prevalence of alcohol abuse compared to other European countries.
In the NEMESIS study they also report on the healthcare needs of people with a disorder. The number of people with a mental disorder that reported to be in need of professional help but did not receive this is 7.2% (de Graaf et al., 2010). If you compare this with the same study done some years before (NEMESIS-1) this number has decreased.
From the people who did look for help the main part turned to the general healthcare instead of mental health care. This results in many doctors’ visits and hospital costs while this could be treated more effectively when the resources to provide the necessary therapy are directly available when needed. The large use of general healthcare to treat mental illness has not changed much since the previous NEMESIS study.
The differences between the Netherlands and other Western countries could be explained by different policies in the different countries, concerning what is covered by insurance and what not. In the Netherlands there is relatively high government influence, comparing this to for example the United States. In the Netherlands the costs of mental health care are completely, or for a large part, covered by every insurance (Rijksoverheid, 2014a). This gives more incentive to go to see a therapist than when it is not covered and thus will have to be paid by the patient itself. Especially because therapy can feel like a secondary need, if you compare it to cost that have to be made because of physical illness. Furthermore, the Netherlands has one of the highest population density in the world, which makes mental health care institutions geographically easier accessible for patients.
Another important factor for the Dutch government is the number of people that is on incapacity benefits as a consequence of mental illness. In the Netherlands about half of the people that are on incapacity benefits are unable to perform at the workplace because of psychological problems, in the year 2012 this was a little more than 400,000 people (Statistics Netherlands, 2014). This is a very high amount and since the incapacity benefits in the Netherlands are relatively high, these are a high expense account for the government.
Overall we can conclude that in the Netherlands the mental health care sector is quite developed compared to Europe and the rest of the world. Nevertheless it also seems the case that the prevalence of mental disorders is relatively high compared to other countries.
4. Cost-benefit analysis
In this chapter the methods that are used to answer the earlier stated research question, and the found results will be discussed extensively. Starting with the method that is used, and explaining the assumptions made. Second, the different social costs and benefits items will be discussed and the results will be presented. Finally a cost-benefit analysis of the mental health specifics of people in the Netherlands and the previously found cost and benefits per capita will give an economic overview of the situation in the Netherlands.
4.1. Method
To be able to answer the research question and get an idea of the economic consequences of making cognitive psychiatric therapy (CBT) more widely available, we apply cost-benefit analysis to the use of CBT. First the social costs and benefits per person will be calculated (see Table 1), these will be translated to the total Dutch population to see the overall effect of a possible program on social welfare (see Table 2).
When calculating the economic effects of providing therapy to the people that need it, we look at the social costs and social benefits. Social costs are the expenses to an entire society resulting from a change in policy. So this includes the production expenses, also called the private costs, as well as indirect expenses and other negative externalities. For social benefits this means that we look at the private benefit to society from implementing the program plus any external benefit. We look at the social costs and benefits instead of individual costs and benefits because we want to find the overall optimal outcome for society as a whole, not certain people or entities individually. The goal of this study is to see if overall society would be better off with a program for providing therapy or not, social costs and benefits are therefore the right measure to use. When talking about social costs and benefits and government decision-making, it is assumed that the government is trying to maximize social welfare, so to minimize social costs and maximize social benefits. This is because the government is looking after the overall welfare of society.
The benefits are calculated for the first year after treatment, this is because after this first year there are many uncertainties concerning these benefits. The benefits in the first year are relatively easy to predict and quite straightforward. If you try to look further into the future there could be several problems that compromise the results of the research. There is for example a chance that the disorders reoccur or get to deal with another disorder,
successful treatment does not automatically mean that a person is cured for its entire life (Churchill et al., 2001). However these long-term benefits will be much more volatile and less accurate. Additionally, because only the first year is taken into account the numbers found are not corrected for inflation and not discounted.
First of all we make the distinction between Axis-I and Axis-II disorders. Axis-I are disorders that pass, the symptoms and complaints are not visible your whole life. In comparison to Axis-II disorders, which are personality disorders and thus influence somebody’s choices and way of thinking their entire life, for example Borderline personality disorder. This exactly points out why we only look at Axis-I disorders, since the implications of these disorders for the economy change over the years, with somebody getting for example a depression, which also passes again after a certain period of time. We do not look at all the Axis-I disorders because these are simply to many, so we only include the disorders with a 1-year prevalence above 1% (this excludes for example eating and bipolar disorders). Furthermore in this study only mood disorders and anxiety disorders are taken into account. This is because other type of disorders like for example substance use disorders (e.g. alcohol abuse) have very different consequences for people’s personal lives and economic performances, furthermore these require a very different treatment. So with the methods we use this type of disorders cannot be easily compared with other types of disorders like mood and anxiety disorders.
We will calculate the costs and benefits for two types of general disorders; mood disorders and anxiety disorders. The main disorders in the category mood disorders are depression and dysthymia. For anxiety disorders we mainly talk about panic disorders and phobias. A more elaborate clarification of the disorders is attached in the Appendix I, all these disorders are defined according to the Diagnostic and Statistical Manual of Mental Disorders, version 5 (American Psychiatric Association, 2013).
To be able to calculate with these numbers we get an average number by applying weights to every disorder, based on the weights that Layard uses to compute the costs and benefits (Layard, Knapp, Clark, & Mayraz, 2007) and the overall prevalence of the disorders in the Netherlands (Graaf, Have, Gool, & Dorsselaer, 2012). These weights are shown in Appendix II.
In this cost benefit analysis we do not take into account the cost of medication. In most of the cases medication and therapy are substitutes and therapy. We don’t take into account medication, first of all because the people that use medication can be seen as people that are already treated. And we want to get an idea of what the costs and benefits are of providing treatment for the people that do not yet receive it. The second reason why we do not take into account cost of medication is because medication and therapy are substitutes in most of the cases (Chilvers et al., 2001). Since therapy yields significantly better results than medication, especially in the long run (see Figure 2.), we prefer providing therapy instead of medication. Furthermore from research done by van Schaik et al. (2004) it shows that most patients prefer to be treated with therapy than to be prescribed medication.
4.2 Costs
In this section we will elaborate on the total social costs of providing therapy to one person. There are different costs to be taken into account. When looking at the social costs of providing therapy for one person these are mainly the opportunity costs that play a fairly large
Figure 2. Six year outcome of cognitive behavior therapy for prevention of recurrent depression
role, these exist of the direct treatment costs and the costs of the time spend on going to therapy.
The treatment costs are the basic costs that are made to provide the therapy, so this includes the wage of the therapist, cost of accommodation and for example office supplies. These make up for the money and space that could not have been spend otherwise because of the provision of the therapy. We use the general costs of cognitive behavioral therapy to calculate the opportunity costs of providing therapy because this therapy is already used a lot in the Netherlands. CBT is the most and best developed therapy and the educational system for therapists is already fully in place. And it has high success rates compared to other therapies (Serfaty, Csipke, Haworth, Murad, & King,2011). So we look at the cost of one hour cognitive behavioral therapy to find the opportunity costs of treatment. The Dutch Healthcare Authority established in 2013 the maximum that one hour of therapy could cost, which is €94.40 (Dutch Healthcare Authority, 2013). We use this number to be on the safe side, although since this is a maximum it could be the case that there are some practices that offer therapy cheaper.
To get the total treatment costs per person we multiply the costs per session by the average number of sessions needed to cure a mental disorder, as shown in Table 1.a. for the lower bound, the upper bound and the average. It is difficult to put an actual number on this because it differs heavily per person and per disorder how many sessions are needed to be cured. Studying different literature yields the conclusion that normally an entire treatment takes about eight to sixteen sessions (Meuldijk, Carlier, Van Vliet, Van den Akker-Marle, &
Table 1. Per Capita Cost-Benefit Analysis Social Costs
Lower Upper Average
a. Opportunity costs of treatment € 755.20 € 1,510.40 € 1,132.80 b. Opportunity cost of time spent € 199.20 € 398.40 € 298.80 TOTAL OPPORTUNITY COSTS € 954.40 € 1,908.80 € 1,431.60
Benefits
Lower Upper Average
c. Output loss € 7,711.72
- of which absenteeism € 4,661.63
- of which presenteeism € 1,583.69
- of which extra months worked € 293.28 € 2,639.52 € 1,466.40
d. Medical costs saved € 258.19
e. Extra QALYs € 9,850.00 € 11,820.00 € 10,835.00
Zitman, 2012; Bell et al., 2006). This range is quite broad based on the fact that a very light version of a disorder requires more or less eight sessions, but a rather severe about sixteen. Overall this gives us a range for treatment costs of €755.20 to €1,510.40.
The opportunity cost of time are the costs of time that the patient uses to get to the therapy facility and the time of the actual therapy. So the costs that we take into account here are the costs of using that time to go to therapy instead of using it to do something else. How we find the value of this opportunity cost of time is shown in table 1.b.. To calculate these costs we mainly use the Dutch Manual for Costing: Methods and standard costs for economic evaluations in health care by Oostenbrink, Bouwmans, Koopmanschap, & Rutten (2004). In this manual they describe different costs that should be taken into account in economic evaluations in health care.
Table 1.b. Lower Upper Average
Opportunity cost of time € 199.20 € 398.40 € 298.80
Number of meetings 8 16 12
Hours lost per meeting 3 3 3
Cost per hour € 8.30 € 8.30 € 8.30
If a person takes the time to go to a therapy, this time cannot be spent differently, so we take this into account as a cost. Oostenbrink et al. (2004) estimates that on average one session will cost a person three hours, this includes the actual meeting with the therapist, travel time and possible waiting time. It is assumed that this time is booked off from one’s leisure time, which is valued by Oostenbrink et al. (2004) at €8.30 per hour. We multiply these with the earlier established amount of sessions per person to eventually get a range of €199.20 to €398.40.
Adding the opportunity cost of therapy and the opportunity cost of time brings the total social cost of treating one person with CBT to €1,432 on average with a range of €954 to €1,909. Of this total costs the direct treatment costs are twice as large as the indirect treatment costs.
Table 1.a. Lower Upper Average
Opportunity cost of treatment € 755.20 € 1,510.40 € 1,132.80
Number of sessions 8 16 12
4.3. Benefits
To find the social benefits of successfully treating a person dealing with a mental disorder we look at all different ways that mental illness affects the life of the individual and society as a whole. The numbers shown are values that describe the benefits for the year following the treatment, should this be successful and cure the person in question. The social benefits consist of three main accounts, there is the avoided output loss, medical costs saved and Quality adjusted life year benefits.
A lot of output for the Dutch economy as a whole is lost because of mental disorders. This output loss can be divided into three subcategories. We are talking about absenteeism when a person is physically absent at work and thus does not produce anything, and then there is presenteeism which describes the output loss for a person that is present at work but is less productive during that day because of the disorder, either in a quantitative or a qualitative way. Furthermore it is proven that a previously ill person that is now cured has a higher possibility of finding or not loosing a job, which creates more output for society.
Table 1.c. Lower Upper Average
Output loss € 7,711.72
- of which absenteeism € 4,661.63
Average days absent 15.9
Hours in a work day 8
Monetary counter value per hour € 36.66
- of which presenteeism € 1,583.69
Average days absent 5.4
Hours in a work day 8
Monetary counter value per hour € 36.66
- of which extra moths worked € 293.28 € 2,639.52 € 1,466.40
Extra months worked 0.49 0.49 0.49
Monetary counter value per hour € 36.66 € 36.66 € 36.66
Employment coefficient 0.1 0.9 0.5
The output loss because of absenteeism is calculated by first weighing the excess days absent at work because of a person having a disorder in the Netherlands, found by De Graaf, Tuithof, Dorsselaer, & Ten Have (2012b) according to the weights earlier established. This shows an average days absent of almost 16 days a year, the days absent for the different disorders can be found in Appendix III. In the Manual for Costing by Oostenbrink et al.
(2004) they show the monetary countervalue of production losses by age and gender, for the full table of counter values see Appendix IV. This we weigh according to the sample used by de Graaf et al. (2012b) which leads to an average output loss per hour of €36.66. Multiplying the number of days absent times eight hours (assumed as a normal work day) and the output loss per hour gives an average output loss because of absenteeism of €4,662, shown in table 1.c.
For the output loss because of presenteeism the calculations are more or less the same, but there is one thing that should be taken into account. The days that are found by de Graaf et al. (2012b) as days less productive either in qualitative or quantitative sense, are counted as half a workday, as is recommended by the Graaf et al. This is because these are days that a person is present at work, and just is less productive because of the disorder. Taking this into account the effective days per year that a person does not work because of his mental disorder comes to 5.4, the values for the different disorders can be found in table 4 of Appendix III. Again multiplying this by an eight hour work day and €36.66 average output loss per hour brings the total output loss because of presenteeism to €1,584 (Table 1.c.).
There is another gain in output, namely because of the higher possibility of having a job, when not having a mental disorder. Layard shows in his cost-benefit analysis that on average a person will work 0.49 extra months in the year following treatment (Layard, 2007). Layard et al. base this number on two things; that a person that is working with a mental disorder may risk losing it; timely treatment can prevent this. Moreover, if a person is out of work because of mental illness it becomes more likely to find work if they recover from their illness. So this is about the job market, losing and finding a job, which makes it different from absenteeism and presenteeism which shows output loss because of productivity.
Since the job market is volatile and it depends on employment we cannot accredit the whole extra output as a gain. It could be that there is very high unemployment and then if somebody finds a job more easily, this would be hardly any extra output. Because if this person would not have gotten the job, there would be somebody else available for it. On the other hand, if we are in a situation of full employment it could be accounted for a benefit since the output is all extra. To take this unemployment factor into account we used a coefficient, that is very high in the case of full employment and very low in the case of high unemployment. In general the Netherlands has a relatively low and stable unemployment rate, so we could say that in times of economic welfare, when we are close to full employment a large part of the produced output can be accounted for. Since there will always be some friction here the coefficient can never be exactly 1, so we assume a maximum coefficient of
0.9. In times of economic downturn on the other hand there will be higher unemployment so the coefficient will be rather small. But since the Dutch labor market is relatively stable we say the coefficient can never be exactly 0, so the lower bound is assumed 0.1. So to find the actual extra output because of extra months worked we multiply the amount of extra work done, which will be half a month, which is about 10 working days, so 80 hours if we assume a working week of 40 hours with the average hourly compensation we have used before and the unemployment coefficient. So the impact of extra months worked has the wide impact range of €293 - €2640, which yields an average addition of output of €1466.
The gain in output because of the fact that people being treated successfully are less days absent at work is much higher than the output gain because of reduced presenteeism or the higher change of finding or keeping your job. The reduced absenteeism accounts for more than half of the total excess output. This could be expected since a day absent produces nothing at all, while a day present but with reduced productivity still adds some output to the total. The total output gain adding the accounts absenteeism, presenteeism and extra work together comes to €7,711.
From research it is clear that people suffering from a mental disorder have a higher chance of being on incapacity benefits (Dutch: Arbeidsongeschiktheidsuitkering). As discussed in the chapter on the mental health characteristics of people in the Netherlands, for about half of the people on incapacity benefits the reason that they cannot work is psychological problems. When more people are likely to find a new or keep their job this reduces the number of people on incapacity benefits. Reducing this number could relieve the pressure on government current transfers, and be used to add value somewhere else. However these values will be relatively small and are partly included in the extra months worked.
The medical costs that are saved by treating a person timely are accounted for as an opportunity cost, so if these are not used to treat a person with a mental disorder they can be deployed somewhere else. To get an overview of the excess medical costs that a person with a mental disorder makes we look at the costs that are calculated by Cuijpers et al. (2006), for the full table see Appendix V. To compare the different costs of the different disorders they calculate the different categories of excess costs per capita annually. From this we find the extra direct medical costs that a person with the different disorders makes. These costs per different disorder we multiply with the earlier explained weights to get one value, which gives the medical costs saved per person treated of €258.
A way of measuring the reduced quality of life because of a disorder are the Quality Adjusted Life Years (QALY) and disability weights. QALYs are a measure that is developed to value a healthy life year (Sassi, 2006) and the disability weights show the part of the life year that you lose because of the disorder in question. A QALY for a certain illness is calculated by multiplying the prevalence of the disease by the disability weight. The disability weight is a number between 0 (perfect health) to 1 (equivalent to death) which reflects the relative severity of the disease. For example Parkinson’s disease has a disability weight of 0.68, which is relatively high while a bacterial STD has a weight of 0.07. The weights are calculated by a group of health experts based on a worldwide test where they ask hypothetical trade off questions, which outcomes they order in different categories to find the disability weights (Arnesen & Nord, 1999). In this study we take the disability weight of the disorders we are investigating, because we are looking for the life quality per person.
The disability weight for the disorders examined in this paper is 0.2 (Melse, Essink-Bot, Kramers, & Hoeymans, 2000), which means that a person with a mental disorder in the Netherlands loses 0.2 of his year because he or she cannot live optimally due to the disorder. If somebody lives 5 years with the examined disorder, this would be equivalent to dying one year prematurely. There are many different ways to value a healthy life year, and there are many different opinions about how to do this but the usually used values per healthy life year lies between €50,000 and €60,000 (Van Gils, Schoemaker, & Polder, 2014). So the value that we can put on the extra year that a successfully treated person is healthy lies between €9,850 and €11,820, as shown in table 1.e.
Taking all this together yields a benefit per successfully treated person, so a healthy life year, of on average € 18,804. The largest share of these benefits comes from the gain in quality of life. This is since the value of a happy life year is relatively high in the Netherlands compared to other countries. Furthermore the disability weights of mental disorders are quite high in comparison to other disorders and illnesses. The second largest benefit is the gain in output that consists of the three accounts earlier discussed.
Since some of the cost and benefits depend on other external factors and could change over time, there are a few scenarios that we would like to investigate further. With the
Extra QALYs € 9,850.00 € 11,820.00 € 10,835.00
QALY value € 50,000.00 € 60,000.00 € 55,000.00
different outcomes of social benefits we find here we will calculate the overall results, to see in which scenario the hypothesis hold.
The social benefit of the extra time worked is very dependent of the economy and the unemployment rate at that certain point in time. Since this depends on so many different things it is hard to say how much value can be added. That is why in Scenario 1 we take into account the lower bound, to be certain not to account for benefits that might not exist in the economic environment at the certain point in time. Adding the lower bound value of extra time worked to the other benefits, we find a value of about €17,432, we call this Scenario 1.
Furthermore, although the QALY method is by far the most used method to value the impact of a disease, there is some critique on the measurement. This is because in essence it is vary hard to express the value of a healthy life year in valuta. The full discussion falls outside the scope of this study, but as it is that QALYs are making up an extensive part of the social benefits we will calculate the outcome without the QALY benefits, to see if it in that case will hold as well. When excluding the QALY benefits the total benefits of Scenario 2 add up to €7,970.
To take a very sober scenario we put Scenario 1 and Scenario 2 together, calling in Scenario 3, so we take the lower limit of the extra time worked and leave out the QALY benefits. This brings the total benefits to €6,597.
4.4. Population implications
Table 2. Cost-benefit analysis for the Dutch population
Low Upper Average
a. Costs € 70,874,971 € 141,749,941 € 98,777,851
b. Benefits € 415,201,697
OVERALL EFFECT € 297,324,832
Now that we have found the total costs of therapy per person, we want to see what the total costs is for the Dutch population as a whole (see table 2.a.). In the Netherlands the prevalence of the investigated disorders is about 7.9%, the NEMESIS study that shows this percentage is only done for people with age between 18 and 64, so the population suitable for work. The total population suitable for work in the Netherlands is about 12 million, which shows that there are about 960,000 people with a disorder in the Netherlands. Furthermore about 7.2% of the people with a mental disorder indicate that they did not receive any
treatment while they do need it (De Graaf, Ten Have, & Dorsselaer, 2010). The table with the unmet treatment needs for the several disorders that are examined can be found in appendix VI. Translating this to the population suitable for work we find that almost 69,000 people in the Netherlands need treatment while they do not receive it yet. We multiply this with the costs per person and find a total cost of €98,777,851 to provide CBT therapy for the ones that need it.
Table 2.a. Lower Upper Average
Costs € 65,851,900 € 131,703,801 € 98,777,851
Average costs per person € 954 € 1,909 € 1,431 Population suited for work 12,205,982 12,205,982 12,205,982
People with a disorder 963,662 963,662 963,662
(7.9% of labor force)
People with unmet healthcare needs 68,998 68,998 68,998 (7.2% of people with a disorder)
In the same way we calculate the total benefits of the treatment for the whole population (see table 2.b.). Not every person treated recovers, and some people recover anyway regardless of received treatment. When correcting the numbers for these effects, the effective recovery rate of therapy is about 32% (Layard, 2007), which can be found from table 8 in Appendix VII. Using the number of people who needed treatment, found in the previous paragraph, shows that about 22,000 people will recover because of receiving the therapy. So we multiply the number of healthy people by the benefits per person and get total benefits of more than four million.
Table 2.b. Average
Benefits € 415,201,697
Average benefits per person € 18,804.91 People with unmet healthcare needs 68,998
Total recovery rate 32%
Number of people that recover 22,079
Subtracting the cost from the benefits it shows that the overall effect is positive by almost 3 million for the first year after treatment in the average situation. To look at the results on a per capita basis we look for the marginal costs and marginal benefits. The marginal cost per person treated are about €1,431 as shown earlier in the result section. To find the marginal benefits per person treated we multiply the benefits of a successfully treated
person (€18,805) with the chance that the treatment is successful (32%) which yields a marginal benefit of about €6,018 per person. Subtracting the costs from the benefits yields a positive result of €4,587 per person treated.
Earlier in the benefit section (4.3) different scenarios were discussed, changing the unemployment rate and excluding the QALYs. For these different scenarios. The outcomes of the overall effect using the numbers of the different scenarios can be found in table 3. The first scenario shows the situation when there is assumed high unemployment, which means that only a small part of the output produced as consequence of finding or keeping ones job because of treatment is extra. Overall this has a relatively small effect on the overall outcome, since this extra output is a rather small part of the total benefits. This scenario yields an overall positive effect of about 286 million. The second scenario eliminated the QALY benefits. This yields a rather high effect, since the QALYs are the highest item on the benefit account. Even though it cuts the benefits per person in half, the overall effect in this scenario is still positive with a value of almost 78 million. The third scenario shows the situation when the first and second scenario are taken together, so with high unemployment and no QALY benefits. Even in this scenario the overall benefits are still positive with a value of almost 47 million for the first year after treatment.
Table 3. Scenario outcomes
Scenario 1 Scenario 2 Scenario 3
Situation High unemployment No QALY's
High unemployment + No QALYs
Costs € 98,777,851 € 98,777,851 € 98,777,851 Benefits € 384,888,626 € 175,973,058 € 145,658,001
Average benefits per person € 17,432 € 7,970 € 6,597
5. Implications
Using the presented results, we will further explain the implications and consequences, discuss the fairness of the hypothesis as well as answer the research question.
Preventing the disorder from getting worse by providing standard therapy to the people that need it appears to be profitable in the economic sense. Looking at the results it can be concluded that treating the people that need it would generate more benefits than that it would actually cost.
This is in the long-run, when you do not take into account the initial one time investment that is needed. New therapists have to be educated, access to mental healthcare for patients should be simplified and there should be some awareness campaigns to inform people about the program. The return on investment will be increasing in the short-run, since a project like this has a rather large initial investment. With establishment of the mental health care service units, development of an implementation system and make people aware of the availability of the treatments. When all this is put in place it would only be educating an extra therapist when needed.
On the other hand, the prevalence is expected to be decreasing in the long run. Since it is assumed that with more therapists, better access and more awareness the prevalence of people that need therapy but did not receive it will decrease. Which means that it would be an option to spread out the investment, and not educate as much therapists as needed in this point in time. To avoid that a lot of money is invested in educating therapists that can only work the first few years and after that have no work because there simply are not enough patients.
According to the results in the previous section there would be on average almost 300 million in social benefits the first year after treatment. So the initial investment could be this amount and the project would break even. So startup costs, marketing costs and education costs can be up to 300 million in the first year and there would be made no losses whatsoever, when looking at overall social welfare.
Looking at this from a marginal cost and benefit perspective it shows that there is plenty of room to cover the start-up costs of the project. As shown in the cost benefit analysis section the overall marginal effect is positive with €4,587 per person. To get an idea of how large this investment would be we look at the cost of educating therapists. The costs of educating a therapist of course depends on the different kind of therapies they should be able to provide and the level of severity of the mental disorder that they should be allowed to treat, but we assume that educating one person to be allowed to apply CBT lies around €15,000 and
€17,000 (“Rino Group”, 2014). Which means that the costs of educating one therapist will be covered if this therapist treats about 4 people in the first year. Which is very likely to be the case, since in general it is reasonable to assume a therapist treats about 80 people a year (Bell et al., 2006).
In the analysis we also looked at different scenarios, when there would be times of high unemployment and when leaving out the QALY benefits. All scenarios show a positive result, even when there is assumed high unemployment and the QALYs are left out at the same time. Although the positive effect is significantly lower in the third scenario, about 47 million. Which means that in that case the initial investment should be smaller or spread out over multiple years. Which, as mentioned before, would be an option to avoid educating therapists for only the first years.
In conclusion, from looking at the cost-benefit analysis of psychological therapy it is safe to say that providing treatment for the people that need it but do not yet receive it would be a viable policy in the Netherlands, assuming that the goal is to maximize social welfare.
6. Discussion
There are some limitations to the conducted analysis, these we will discuss in this paragraph and make suggestions for further research.
With mental health problems it is not in every case that when it is successfully treated it is gone forever. In many cases it is possible that after some time it comes back, or another problem occurs (Buist-Bouwman, Ormel, Graaf, & Vollebergh,2004). In this analysis only the first year after treatment is taken into account, so recurrence of the disorder is left out. This is because the long-term is very hard to predict and there is hardly any significant research done on this topic. So it would be rather speculative to make calculations about it since there are hardly any significant numbers to calculate with.
Furthermore, there is a trend going on that more and more therapies are (partly) provided through internet interventions, and this trend is expected to continue (Riper et al., 2007). In general these internet therapies are less costly to provide, have a lower barrier to entry and increase the speed of the process. In the presented calculations, all therapies are face-to-face contact with a qualified therapist, which is more expensive. As seen in the results the direct costs of providing treatment make up the largest part of the cost, so if this could be decreased by providing therapies (partly) through the internet this could make an investment even more attractive. If the offers of internet interventions keep increasing a similar analysis could be done substituting part of the therapy by internet interventions, to see if this yields the same result.
Moreover, older people are not taken into account, because the structure of their cost-benefit analysis is completely different from the working population. However the older population segment is increasingly important when it comes to healthcare demand and corresponding costs. This is because in general it is the trend that people live longer and thus have more chance of suffering from a mental disorder. So further research focusing on the cost-benefit analysis of older people would be relevant.
Furthermore in some cases it is likely for people to have two or more mental disorders at the same time. This is left out because this would make the analysis to complex, since this group has completely different characteristics (de Graaf et al. 2012a). For further research we recommend to further extend the research by including this group in the analysis.
A limitation of the implications earlier described is that the allocation of the costs and the benefits not necessarily is evenly distributed. From the analysis it is clear that the benefits are higher than the costs of providing therapy to the ones not yet treated, so this should mean
it should be implemented. However it is not taken into account who is actually making the costs and who is benefitting. If this is not divided equally implementing it could be a problem, since there could be some parties disadvantaged which would then not support the plan.
Overall, although there are a number of limitations, the assumptions used in this analysis are applied to the costs as well as the benefits, so it is mainly the positive outcome of comparing the two is something that should be taken into account in further research.
7. Appendices
Appendix I
There are many different mental disorders that all have their different characteristics and therefore consequences. This research is limited to only two groups of mental illnesses, mood disorders and anxiety disorders. Both these categories contain different disorders, that are defined according to the Diagnostic and statistical manual of mental disorders from the American psychiatric association (2013). Below we give a short definition of the most important disorders examined in this paper.
Major depressive disorder
Major Depressive Disorder is a disorder where there occur two or more major depressive episodes. The diagnostic criteria are among other things of course a depressed mood, loss of interest or pleasure in life activities, significant weight loss or gain (unintentionally), sleeping very little or too much, recurrent thoughts of death. If a large part of these symptoms occur for more than two weeks in a row the disorder could be diagnosed as a depression.
Dysthymic disorder
Dysthymia, sometimes also called neurotic depression is a type of mood disorder that has many the same cognitive and physical problem as a major depressive disorder, however the symptoms for dysthymia are longer-lasting but less severe. The diagnostic criteria consist as well for example a depressed mood, low self-esteem, poor appetite or overeating, but for at two years.
Panic disorder
Having a panic disorder means suffering from panic attacks, these episodes peak within a few minutes. During these attacks people describe the feeling as very distressing and alarming physical symptoms occur like shaking, chest pain, dizziness etc. The consequence is that people fear the idea of another attack and often avoid situations or places where they had a panic attack before.
Generalized Anxiety disorder (GAD)
The main feature of a Generalized Anxiety Disorder (GAD) is excess irrational worry that never stops. It can concern various and multiple topics, and is often accompanied by physical
symptoms, like hot flashes, nauseasness or headaches. GAD is diagnosed when this excessive anxiety occurs for at least six months.
Specific phobia
The main feature of a specific phobia is a persistent and excessive or unreasonable fear, specifically bound to a object or situation (e.g. animal, heights, blood). The consequence is that the phobic situations are feared and avoided, interfering with normal functioning.
Appendix II
In their cost-benefit analysis Layard et al. (2007) have assigned weights to the different disorders to get an overall value for the different cost and benefit categories. These weights are; Depression: 45; Phobia: 5; Obsessive compulsive disorder: 5; Panic Disorder: 5; General anxiety disorder: 30; Post-traumatic stress disorder: 10. These weights are used in all analysis.
Appendix III
In their paper on the effect of mental and physical disorders on work performance de Graaf et al. (2012b) investigate the amount of working days lost per person specifically due to a disorder. They look at three different ways that that is possible, namely actual days absent because of a disorder, but they also look at the days of reduced quantitative functioning and reduced qualitative functioning. For none of the mental disorders that we examine the number of days of reduced quantitative functioning was statistically significant so we do not take these in to account in our calculations. The numbers of days lost in the past 12 months that they found for both categories are displayed in table 4, all these numbers are adjusted for demographics.
Table 4. Days absent Days reduced qualitative functioning
Any mental disorder 10,5 8,0
Any mood disorder 23,1 13,0
Major depression 22,8 12,1
Dysthymia 18,6 9,0
Bipolar disorder 19,5 19,1
Any anxiety disorder 10,0 9,0
Panic disorder 26,3 11,0
Agoraphobia -0,1 36,0
Social phobia 10,3 5,5
Appendix IV
Oostenbrink et al. (2004) mention in their Manual for Costing what is in general the Monetary counter value of production losses in paid labor in Euros per hour (see table 5.). We use these numbers together with the days of production loss to calculate the value of output loss because of mental illness. These values were adjusted for the demographics of the sample of de Graaf et al. (2012a) to finally get to the average monetary counter value of production of €36,66.
Table 5. Monetary counter value of production losses in paid labor (in € per hour)
Age Men Women
15-24 20,49 20,07 25-34 32,74 29,88 35-44 40,86 33,6 45-54 45,37 34,21 55-64 47,82 36,41 Appendix V
When discussing he economic costs of minor depression, Cuijpers et al. (2006) find the different costs for different disorders to be able to make a comparison, as shown in table 6. They make the distinction between direct medical, direct-non medical and indirect non-medical. Relevant for this study is the direct excess medical costs, which indicate the annual excess costs that people with a mental disorder use of direct medical costs (e.g. doctor visits, hospital costs).
Table 6. Direct excess medical costs
Minor depression 10
Major depression 478
Dysthymia 349
Any anxiety disorder 278
Appendix VI
From the NEMESIS-2 study de Graaf et al. (2012a) find the percentage of people of the people with a disorder that are in need of mental health care but did not receive this. These percentages are shown in table 7.
Table 7. Unmet healthcare needs (percentage) Mood disorders 8,7 Depression 7,9 Dysthymia 10,8 Anxiety disorders 5,9 Panic Disorder 2,8 Agoraphobia 5,7 Social phobia 10,1 Specific phobia 6,3
General Anxiety disorder 4,3
Appendix VII
In their study Layard et al. (2007) find the effectiveness of cognitive behavioral therapy for the different mental disorders, as is shown in table 8. They do this by taking several things into account; they start with the retention rate of the therapy, which is the actual percentage of people that finish the treatment. Then there is the recovery rate, which is the rate of people that actually recover after therapy. But there is a percentage of people that would recover either way, even without the therapy. So correcting for this they find a recovery rate of 32%.
Table 8. Effectiveness of CBT Retention rate Recovery rate Natural recovery rate Change in % who recover Depression 80 60 30 24 Phobia 85 70 5 55 Obsessive-compulsive disorder 80 55 5 40 Panic disorder 90 75 5 63
General anxiety disoder 80 50 20 24
PTSD 85 75 20 47
8. Bibliography
Alonso, J., Angermeyer, M. C., Bernert, S., Bruffaerts, R., Brugha, T. S., Bryson, H., . . . Vollebergh, W. (2004). Use of mental health services in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta
Psychiatrica Scandinavica, 109, 47–54.
American Psychiatric Association (2013). Diagnostic and statistical manual of mental
disorders: DSM-5. Washington, D.C: American Psychiatric Association.
Arnesen T., Nord E. (1999). The value of DALY life: problems with ethics and validity of disability adjusted life years. BMJ, 319, 1423-1425.
Bell, S., Clark, D., Knapp, M., Layard, R., Meacher, M. C., Priebe, S., . . . Wright, B. (2006). The Depression Report: A New Deal for Depression and Anxiety Disorders.London
School of Economics.
Bernert, S., (2004). Prevalence of mental disorders in Europe : results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatrica
Scandinavica, 110, 35-36.
Buist-Bouwman, M. A., Ormel, J., Graaf, R. D., & Vollebergh, W. A. (2004). Functioning after a major depressive episode: complete or incomplete recovery? Journal of Affective
Disorders, 82, 363-371.
Centre for Mental Health (2006). The Case for Psychological Therapy on the NHS. Retrieved from http://www.centreformentalhealth.org.uk/pdfs/we_need_to_talk.pdf
Chilvers, C., Dewey, M., Fielding, M., Gretton, K., Miller, V., Palmer, P., . . . Bedi, N. (2001). Antidepressant drugs and generic counselling for treatment of major depression in primary care: randomised trial with patient preference arms. British Medical
Churchill, R., Hunot, V., Corney, R., Knapp, M., McGuire, H., Tylee, A., & Wessely, S. (2001). The effectiveness of brief psychological treatment. Health Techonogy
Assessment, 5(35).
Clark, D. M., Layard, R., Smithies, R., Richards, D. A., Suckling, R., & Wright, B. (2009). Improving access to psychological therapy: Initial evaluation of two UK demonstration sites. Behaviour Research and Therapy, 45(11), 910-920.
Committee on Medical Cure and Care (1991). Report on choices in medical cure and care. The Hague: Ministry of Welfare, Health and Culture.
Cuijpers, P., Smit, F., Oostenbrink, J., R, G. D., Have, M. T., & Beekman, A. T. (2006). Economic costs of minor depression: a population-based study. Acta Psychiatr.
Scand,115, 229–236.
De Graaf, R. D., Have, M. T., Gool, C. V., & Dorsselaer, S. V. (2012a). Prevalence of mental disorders and trends from 1996 to 2009. Results from the Netherlands Mental Health Survey and Incidence Study 2. Social Psychiatry and Psychiatric Epidemiology, 47, 203-213.
De Graaf, R., Ten Have, M., & Dorsselaer, S. (2010). De Psychische Gezondheid van de
Nederlandse Bevolking. Trimbos Instituut.
De Graaf, R., Tuithof, M., Dorsselaer, S., & Ten Have, M. (2012b). Comparing the effects on work performance of mental and physical disorders. Social Psychiatry and Psychiatric
Epidemiology, 47, 1873-1883.
Department of Health (2008). IAPT Implementation Plan: National Guidelines for Regional
Delivery. Retrieved from
http://webarchive.nationalarchives.gov.uk/20130107105354/http:/www.dh.gov.uk/prod_ consum_dh/groups/dh_digitalassets/@dh/@en/documents/digitalasset/dh_083168.pdf
Department of Health (2012). IAPT Three Year Report: The first million patients. Retrieved from http://www.iapt.nhs.uk/silo/files/iapt-3-year-report.pdf
Department of Health (2011). Talking therapies: A four-year plan of action. Retrieved from http://www.iapt.nhs.uk/silo/files/talking-therapies-a-four-year-plan-of-action.pdf Dutch Healthcare Authority (2013). Tariefbeschikking: Generalistische Basis GGZ.
Retrieved from http://www.nza.nl/98174/137715/TB-CU-5069-02_Tariefbeschikking_basis_GGZ.pdf.pdf
Fava, G. A., Ruini, C., Rafanelli, M., Finos, L., Conti, S., & Grandi, S. (2004). Six-Year Outcome of Cognitive Behavior Therapy for Prevention of Recurrent Depression.
American Journal of Psychiatry, 161, 1872–1876.
Gils van, P., Schoemaker, C., & Polder, J. (2014). Hoeveel mag een gewonnen levensjaar kosten? Onderzoek naar de waardering van de QALY. Nederlands Tijdschrift voor
Geneeskunde, 158(6507).
Gyani, A., Shafran, R., Layard, R., & Clark, D. (2013). Enhancing recovery rates: Lessons from year one of IAPTq. Behavioural Research and Therapy, 51, 597-606.
HM Government (2011). No Health Without Mental Health: : A Cross-Government Mental
Health Outcomes Strategy for People of All Ages. Retrieved from
http://www.iapt.nhs.uk/silo/files/no-health-without-mental-health.pdf
IAPT | Demonstration Sites. (2013). Retrieved February 26, 2014, from
http://www.iapt.nhs.uk/about-iapt/demonstration-sites/
IAPT. (2013). Retrieved February 26, 2014, from http://www.iapt.nhs.uk/about-iapt/
Idstad, M., Røysamb, E., & Tambs, K. (2011). The effect of change in mental disorder status on change in spousal mental health: The HUNT study. Social Science & Medicine, 73, 1408-1415.
Layard, R., Knapp, M., Clark, D., & Mayraz, G. (2007). Cost-Benefit Analysis of Psychological Therapy. National Institute Economic Review, 202.
Melse, J. M., Essink-Bot, M. L., Kramers, P. G., & Hoeymans, N. (2000). A national burden of disease calculation: Dutch disability-adjusted life-years. Dutch Burden of Disease Group. American Journal of Public Health, 90(8), 1241-1247.
Meuldijk, D., Carlier, I., Van Vliet, I., Van den Akker-Marle, E., & Zitman, F. (2012). A randomized controlled trial of the efficacy and cost-effectiveness of a brief intensified cognitive behavioral therapy and/or pharmacotherapy for mood and anxiety disorders: Design and methods. Contemporary Clinical Trials, 33, 983-992.
The National Development Team for Inclusion (2009). Working Our Way to Better Mental
Health: A Framework for Action. Retrieved from
http://www.ndti.org.uk/uploads/files/Working_our_way_to_better_mental_health.pdf National Health Service (2009). NHS 2010-2015: From Good to Great. Retrieved from
http://www.official-documents.gov.uk/document/cm77/7775/7775.pdf
NICE Guidelines. (2013). Retrieved February 26, 2014, from
http://www.nice.org.uk/aboutnice/
Olshansky, S. J., Rudberg, M. A., Carnes, B. A., Cassel, C. K., & Brody, J. A. (1991). Trading Off Longer Life for Worsening HealthThe Expansion of Morbidity Hypothesis.Journal of Aging and Health, 3(2), 194-216.
Oostenbrink, J., Bouwmans, C., Koopmanschap, M., & Rutten, F. (2004). Manual for
Costing: Methods and standard costs for economic evaluations in health care (in Dutch). Diemen, Netherlands: Health Insurance Board.
Rhebergen, D., Beekman, A. T., Graaf, R. D., Nolen, W. A., Spijker, J., Hoogendijk, W. J., & Penninx, B. W. (2010). Trajectories of recovery of social and physical functioning in
major depression, dysthymic disorder and double depression: A 3-year follow-up.
Journal of Affective Disorders, 124, 148–156.
Rijksoverheid (2014a). Wie betaalt behandeling van psychische problemen? Retrieved from
http://www.rijksoverheid.nl/onderwerpen/geestelijke-gezondheidszorg/vraag-en-antwoord/wie-betaalt-de-behandeling-van-mijn-psychische-problemen.html
Rijksoverheid (2014b). Hoe hoog is de WAO-uitkering? Retrieved February 26, 2014, from
http://www.rijksoverheid.nl/onderwerpen/wia/vraag-en-antwoord/hoe-hoog-is-de-wao-uitkering.html
RINO Groep (2014). BIG- en profielopleidingen. Retrieved February 27, 2014, from http://big.rinogroep.nl/opleiding.asp?PageID=127&ID=1&pagePart=Kosten
Riper, H., Smit, F., Van der Zanden, R., Conijn, B., Kramer, J., & Mutsaers, K. (2007).
E-mental health: High tech, high touch, high trust. Utrecht: I.COM, Trimbos Instituut.
Sassi, F. (2006). Calculating QALYs, comparing QALY and DALY calculations. Oxford
University Press.
Schaik van, D., Klijnb , A., Van Hout, H., Van Marwijk, H., Beekman, A., De Haan, M., & Van Dyck, R. (2004). Patients' preferences in the treatment of depressive disorder inprimary care. General Hospital Psychiatry, 26(2), 184-189.
Serfaty, M., Csipke, E., Haworth, D., Murad, S., & King, M. (2011). A talking control for use in evaluating the effectiveness of cognitive-behavioral therapy. Behaviour Research
and Therapy, 49, 433-440.
Smit, F., Cuijpers, P., Oostenbrink, J., Batelaan, N., De Graaf, R., & Beekman, A. (2006). Costs of Nine Common Mental Disorders: Implications for Curative and Preventive Psychiatry. The Journal of Mental Health Policy and Economics, 9, 193-200.
Statistics Netherlands (Centraal Bureau voor de Statistiek) (2014). Care Expenditure and
http://http://statline.cbs.nl/StatWeb/publication/?VW=T&DM=SLNL&PA=71914NED &D1=0-23,37-45&D2=9-l&HD=101210-0925&HDR=G1&STB=T
