Valuable but not fully satisfactory - The benefits of health economic assessments of hospital pharmaceuticals

Vårdanalys has evaluated the benefits of health economic assessments of pharmaceutical products used in inpatient care (so-called hospital pharmaceuticals).

The Swedish Agency for Health and Care Services Analysis (Vårdanalys) has been commissioned by the Swedish Government (S2012/3857/FS) to evaluate the benefits of health economic assessments of pharmaceutical products used in inpatient care (so-called hospital pharmaceuticals). In the interim report on the commission, the suitability and effectiveness of the processes linked to the health economic assessments were considered. (Swedish Agency for Health and Care Services Analysis 2013.) This report forms the final report on the project and contains our overall assessment of the benefits of health economic assessments of hospital pharmaceuticals.


The purpose of a health economic assessment is to consider whether the cost of a certain treatment is in reasonable proportion to its benefit – in other words, whether the treatment is cost-effective. The Swedish Dental and Pharmaceutical Benefits Agency (TLV) performs health economic assessments of drugs provided on prescriptions whenever a pharmaceutical company applies for a product to be included in the pharmaceutical products benefits system. No previous equivalent assessment of cost-effectiveness has been carried out for hospital pharmaceuticals. For this reason, since the start of 2011 TLV has been commissioned by the Swedish Government (S2010/8066/HS) to run a pilot scheme within the framework of the so-called “hospital pharmaceuticals project”, whereby the authority also conducts health economic assessments of hospital pharmaceuticals.

TLV’s health economic assessment results in a knowledge base where the assessment is reported. In addition to TLV, another important player in the health economic assessment process is the “New pharmaceutical product therapies group” (NLT), a group within the Swedish Association of Local Authorities and Regions (SALAR, or “SKL” in Swedish). NLT initiates the process by deciding which pharmaceutical products should undergo a health economic assessment. TLV then takes up the case, requesting documentation from the pharmaceutical company in question and creating a health economic knowledge base. From the knowledge bases, NLT then makes a statement as to whether or not they recommend the use of the pharmaceutical product to the county councils. In certain cases, NLT will enter into a discussion on price with the pharmaceutical company concerned before giving a recommendation. The price discussion takes place in cases where the pharmaceutical product is not considered to be cost-effective at the price as stated by the pharmaceutical company to TLV in the documentation for the health economic assessment. Finally, it is a matter for each county council to decide whether and how to use the pharmaceutical product, draw up agreements with the companies, introduce the product into the medical care services it provides and offer the patients treatment.

Vårdanalys’s task of evaluating the benefits of health economic assessments of hospital pharmaceuticals includes TLV’s health economic assessment and those parts in which NLT takes part (initiation and recommendation). TLV and NLT are separate organisations and are responsible for different parts of the process. However, both play a major part in the processes surrounding health economic assessments. From now on, we will define the hospital pharmaceuticals project as involving both NLT’s and TLV’s parts, from the time at which NLT initiates the process to the time at which they submit a recommendation. We will also analyse the receipt and use of documentation and recommendations in the medical care services and among other interested parties.


The following two pages contain an account of the issues specified in the Government commission and a brief summary of our results for each issue. This applies to both the interim report and the final report. The analyses are built on a number of different sources. We have interviewed – among others – patient representatives, the pharmaceutical product managers at the county councils, pharmaceutical companies, TLV and NLT. We have also carried out a questionnaire among the country’s operations managers in the field of oncology. We have used sales statistics for the products included in the hospital pharmaceuticals project to illustrate regional differences in the use of drug use. We have also looked at the situation internationally and examined the work in other countries on health economic assessments of pharmaceutical products. It is important to bear in mind that the replies from our interviews and from the questionnaire reflect the perception of the respondent, not necessarily that of other county council representatives or patient organisations, for example.

Questions and summarised results of the final report

Evaluate the benefits of the health economic assessments of prescriptions drugs to the various interested parties, particularly the patients and county councils

Patient representatives feel that the activity delays access to new pharmaceutical products. They are critical of the project and do not feel that the hospital pharmaceuticals project has brought any reduction in the differences in access to drugs among the county councils. Pharmaceutical product managers feel that the knowledge base and recommendations match an existing need and that they can be used as the basis for decisions.

Analyse how the responsible health authorities use the knowledge bases and how the bases are implemented by the responsible health authorities

The knowledge bases and recommendations are used as the basis for decisions in the county councils’ ordered introduction processes where these exist. Elsewhere, they are forwarded to the appropriate operations managers, who can use them in their work.

Illustrate how the responsible health authorities use the assessments for the purchase of pharmaceutical products for inpatient care

For some pharmaceutical products, the knowledge bases have been used to a large extent in the national price discussions. For others, they have been used to a certain extent when the county councils have purchased the drugs.

Evaluate whether the pilot scheme has led to better use being made of existing resources for knowledge assessment as a result of the health economic assessments having been carried out by a national authority and not by various responsible health authorities

Yes. Generally speaking, the county councils did not previously perform their own health economic assessments. Carrying out the assessments at national level is a positive thing, and it is appropriate to use TLV’s expertise for this.

Evaluate the recommendations of the NLT group, which are based on TLV’s knowledge bases

The county councils feel that the recommendations are valuable, and it is our understanding that the recommendation increased the perceived benefit of the health economic assessment. We therefore consider that there should also be a recommendation in future. However, there is a lack of transparency in NLT’s work and in how the recommendations are produced. NLT believes that they provide a recommendation that is in line with TLV’s decision in cases of an application within the pharmaceutical benefits system, but we consist that there is often a lack of any references to earlier decisions, the ethical platform and relevant studies.

Evaluate the lifespan of the reliability of the knowledge bases and the accuracy of TLV’s knowledge bases and recommendations

The lifespan depends on whether or not the circumstances around the drug change, and the extent to which this happens varies from one drug to the next. In some cases, the lifespan is short due to the rapid development of new treatment methods. Accuracy depends on how well the underlying base reflects the actual clinical situation and can be increased by monitoring and updating the base if the modelled situation deviates from that of everyday clinical life.

Analyse whether the knowledge bases lead to greater equality of healthcare among and within county councils

Our overall assessment is that the hospital pharmaceuticals project has more chance of affecting the differences among county councils than the differences between population groups. The knowledge bases improve the conditions but do not guarantee greater equality of healthcare. As also indicated by the existing sales statistics, patients experience large regional differences in the use of pharmaceutical products, even those that have been through the hospital pharmaceuticals project. It is still too early to comment on the more long-term effects on regional differences.

Questions and summarised results of the initial report

Has the initiation process from NLT been coordinated and prioritised in an effective manner?

There is good coordination and communication with TLV, but it is not clear how the pharmaceutical products have been selected that will undergo a health economic assessment. NLT’s view is that initially, within the framework of the pilot scheme, they chose to test various types of drug, but their understanding is that, in principle, all new hospital pharmaceuticals with the potential to have an impact on medical care (e.g. by having a large budgetary effect or in the case of a new treatment option for a certain patient group) have been included as the scheme has developed. To the outsider, however, it is not clear on what principles the choice of drugs was made for inclusion in the hospital pharmaceuticals project, and patient organisations and pharmaceutical companies alike are questioning how the selection is made.

The duration of the period taken for case processing

Case processing times have shortened since the start of the hospital pharmaceuticals project, when cases took 10–20 months, but they remain too long. In practice, it takes at least eight months from the time at which TLV begins the health economic assessment to the time at which NLT provides a recommendation, and there are still cases where the process is not complete even though more than a year has passed.

Has TLV been given access to the material from the pharmaceutical industry that the authority requires?

All the pharmaceutical companies that were asked took part and submitted the requested material. From time to time during the investigation, additional information is required that may be more or less time-consuming to provide.

The quality of the documentation obtained from the industry, and TLV’s cost-effectiveness calculations

The material provided by the pharmaceutical companies was of a generally high quality. TLV’s knowledge bases, too, were of a mainly high quality, with TLV giving an account of the health economic assessment in a clear and structured way. TLV identifies the most important areas and then focuses on these during the analysis. The knowledge bases are relatively comprehensive, which is a step in the right direction towards greater transparency in health economic assessments. However, there is room for further development – for example, as far as the work done with clinical experts is concerned, more consistent calculation and reporting of indirect costs, and a presentation of the uncertainty factors in the analysis.

NLT’s communication with the county councils, and its recommendations for a knowledge base

There is a lack of transparency in how the recommendations are produced. There is also a lack of transparency in any price discussion between NLT and the pharmaceutical companies, and of the county councils’ assessment of the resulting agreement offer. The documentation and decision protocols are not available. On the other hand, NLT’s communication and recommendations to the county councils are open and take place through a document published on NLT’s website.

What is the companies’ view of the work being done by TLV and the benefit of the knowledge bases that the authority presents? What are the costs associated with producing the material for the pharmaceutical companies?

The pharmaceutical companies take a positive view of the fact that hospital pharmaceuticals are assessed on a health economic basis, and they regard TLV’s view as being of high quality. However, they are critical of, among other things, TLV’s handling of deviating opinions, TLV’s choice of experts and the lack of an opportunity to appeal. They are generally positive about the option of a national discussion on prices, but feel that subsequent discussions still need to be held with the individual county councils. The cost of producing the documentation is comparable with that of the application to TLV (up to several million Swedish kronor).


From a patient perspective, rapid access to a good pharmaceutical product is a central issue, irrespective of where patients live and to which county council they belong. Furthermore, from the patient perspective it is important for there to be transparency in the decision-making process surrounding the introduction and use of hospital pharmaceuticals.

There are positive aspects of the hospital pharmaceuticals project. It is of value to carry out health economic assessments of hospital pharmaceuticals at national level. Good use has been made of TLV’s expert knowledge, and the pharmaceutical companies have contributed material that has made it possible to carry out the health economic assessments. The county councils feel that it is useful to have a recommendation associated with the health economic knowledge base.

At the same time, there are major deficiencies in the hospital pharmaceuticals project, which adversely affect the benefit of the health economic assessments. The structure and processes surrounding the assessments do not work well. NLT’s work shows deficiencies in transparency and falls outside the principle of public access and the framework of the ethical platform. We therefore consider that NLT’s present structure and method of working do not meet the requirements that should be made of an organisation that deals with important pharmaceutical issues, affects the prioritisation of healthcare, and is responsible for large financial sums.

There is also room for improvement in patient involvement in the processes, and it should be possible to develop the health economic assessments in a way that would improve usability and credibility. Our assessment identifies three overall important areas for improvement.

  • The processes show a lack of transparency and predictability

The importance of transparency and predictability is emphasised in our assessment and stressed by patient representatives, county councils and pharmaceutical companies. Here, by “transparency” we mean the opportunity for insight and openness in the processes, while “predictability” refers to the possibility of being able to assess in advance what will happen. In terms of transparency, there are deficiencies in a number of parts of the processes:

• It is unclear how the pharmaceutical products included in the hospital pharmaceuticals project have been selected, and there is a lack of any developed argumentation surrounding the reasons for NLT’s recommendations.

• There is a lack of transparency about the way in which the NLT’s members are appointed, on what basis this is done and for how long the mandate of a member lasts. In addition, apart from the brief recommendation, it is not possible to examine NLT’s work.

• There is a lack of transparency about how NLT’s price discussions are conducted – in terms of who is responsible for the discussion, which parameters in TLV’s knowledge base have been discussed, and what the discussions have resulted in.

• TLV’s knowledge base is comprehensive, which helps with transparency. On the other hand, the knowledge bases require a lot of prior knowledge and are difficult to assimilate for anyone inexperienced in reading health economic analyses. This means that there is still a lack of transparency in the sense of an understanding of what happens in the health economic assessment.

• There is a lack of transparency about the way in which the county councils decide on NLT’s recommendation – whether they choose to follow it or not and what reasons lie behind the county councils’ decision. The model involving a national recommendation and decisions in the county councils in the absence of any expressed viewpoint on the recommendation on the county councils’ part make it unclear where ­– and by whom – the decision is actually being taken. In formal terms, it is the county councils that make decisions on the use of pharmaceutical products, but NLT’s recommendation can have a major impact on the decision.

The lack of predictability means, among other things, that there is a lack of any clear structures indicating which pharmaceutical products are on their way into the hospital pharmaceuticals project, when a recommendation will arrive and which pharmaceutical products are the subject of price discussions at national level. We also feel there is sparse dissemination of information and communication, which creates ambiguities and ineffectiveness. For example, there is no overall description of, or clear timeframes for, the processes.

  • The long processing times can delay access to new drugs

There is a risk that the long processing times will delay the introduction of drugs that potentially could save lives. Since the hospital pharmaceuticals project began, processing times have been cut down to some extent, but we still consider them to be too long. It is difficult to know when, and to what extent, patients would have had access to drugs that are now included in the hospital pharmaceuticals project if the hospital pharmaceuticals project had not existed. However, the long processing times most probably lead to the delayed introduction of several of the drugs in question. The challenge is to create a process that upholds the quality of the documentation but also gives patients access to good pharmaceutical products more quickly than is the case with the current process.

  • Often limited opportunities for various views to be expressed

We consider that the opportunities are currently limited for the various views to be expressed and taken into account in the assessments, and that there is a lack of patient involvement in the processes surrounding the health economic assessments. Patient organisations are invited to information and discussion meetings with TLV and NLT twice a year, but there is no patient involvement when processing the individual pharmaceutical products. TLV’s health economic assessments and NLT’s subsequent recommendations can have a major impact on those patients affected, but patients or patient organisations have no clear role in the process. The meetings of TLV and NLT are closed and there is no option to appeal against or reassess either TLV’s knowledge bases or NLT’s recommendations. It is our opinion that this results in few opportunities for patients to be involved.

The various interested parties, or various people within the same interest group, may have different views about one aspect or interpretation in the health economic assessment. For example, patient organisations may have a different view in one or more respects than the one reflected in TLV’s knowledge base. Not only that, two experts within the same area may have a different perception of how, for example, the results of a medical study should be interpreted, and the clinical situation may differ among the various county councils and clinics. As assessments and interpretations may differ, it is useful to take into account the various points of view. This increases the likelihood that the description in the health economic documentation is, and is perceived to be, a correct reflection of reality. The legitimacy of a decision or process can increase if there is the option to present different points of view.


Our starting points are that TLV should continue to carry out health economic assessments of hospital pharmaceuticals, and that the processes surrounding the hospital pharmaceuticals project need to be developed.

The figure on the next page summarises our proposals for continued development, while Chapter 4 contains a comprehensive discussion of the proposals. The proposals are divided into six areas that accompany the process for the assessment of hospital pharmaceuticals, from the initiation of the health economic assessment to the time at which the pharmaceutical product is used in medical care.

Some of the proposals can be implemented in the current operation relatively promptly, while for others a certain amount of development work is already taking place.

Other proposals aim to illustrate the various more long-term options for action, and in this way also discuss some of the challenges inherent in the operation. These proposals would entail potentially major changes to the current operation and will require further analysis and investigation before a decision can be made on them and they can be implemented. Our view is that, in future, too, there must ideally be a recommendation associated with the health economic assessment, but one issue that needs to be resolved is to decide which player is to produce the recommendation.

1. Start with the guideline principles for the creation of a legitimate process for the work done on health economic assessments of hospital pharmaceuticals

– Transparency and predictability

– Effectiveness and suitability

– Functional institutions

2. Prioritisation and initiation

Perform a health economic assessment of all new pharmaceutical products

– however, some should be prioritised for particularly fast processing on the basis of clear selection criteria

3. Health economic assessment

• Develop the health economic assessments

• Develop and clarify the knowledge bases and link the information to the introduction and monitoring

• Develop the work with clinical experts

• Create opportunities for the various partners to have a voice and for various points of view to be expressed

• Contribute to the best possible extent to public awareness

• Shorten case processing times

4. Recommendations

(preceded by a discussion on price in certain cases)

• Investigate which player should provide the recommendation: three options:

– The individual county council

– The county councils draw up joint recommendations but develop new structures and processes that are more suitable

– A state authority draws up recommendations

• Introduce clear and pronounced reasons for the recommendation and shorten the case processing time

• Develop opportunities for the county councils to negotiate together

5. Decision on use

• Communicate the reasons for decisions clearly

6. Use

• Start to use introduction protocols produced by the medical profession. These should be linked to the health economic assessment

• Introduce structured, national monitoring of the use and effect of hospital pharmaceuticals. This monitoring should be based on and link back to the health economic assessment

--> Improved conditions for more knowledge-driven and equal use of pharmaceutical products