Abstract
The objective of this Health Technology Assessment (HTA) methods report was to examine and to assess decision analysis (DA) as a method to transfer and adapt international scientific evidence in HTA to the German health care context. Furthermore, we sought to develop a systematic framework to facilitate the selection, transfer, adaptation, and synthesis of these data in German HTA projects.
In this report, we review and summarise the concepts and methods of DA; present potential areas of applications, and provide a basis for the critical assessment of decision-analytic studies. The two main methods of DA, decision trees and Markov models, as well as various approaches to sensitivity analyses are described. Examples of typical situations for the use of DA in scientific evidence transfer are described, and a list of main health care domains and parameters in evidence transfer is presented. Finally, we developed a framework to transfer and apply international evidence to the national health care context. The strengths and limitations of the decision-analytic approach are critically examined.
In summary, this HTA report describes different situations, in which decision-analytic models can be useful, and demonstrates the utility of DA in transferring and applying international evidence to the national health care context. We developed a systematic instrument to transfer international evidence to the context of other countries and successfully applied this instrument in two German HTA projects. The use of this instrument is recommended in further HTA projects dealing with the application of international evidence to the German health care context. The use of decision-analytic models to transfer international evidence is endorsed. However, the limitations of DA should be clearly stated discussed transparently in all HTA reports.
Abstract
Ziele dieses HTA-Methodenberichts (HTA = Health Technology Assessment) sind die Prüfung und die Bewertung der Entscheidungsanalyse (EA) als methodisches Instrument zur Übertragung sowie zur Anpassung internationaler Evidenz aus dem Bereich HTA auf den Kontext des deutschen Gesundheitssystems. Ferner soll ein systematisches Konzept entwickelt werden, das als Orientierung bei der Auswahl, bei der Übertragung, bei der Anpassung und bei der Zusammenführung dieser Daten im Rahmen deutscher HTA-Projekte dienen kann.
Es soll ein Überblick über die Konzepte und Methoden der EA gegeben sowie neben den Anwendungsbereichen, den Möglichkeiten und den Grenzen entscheidungsanalytischer Verfahren die Grundlage für eine kritische Beurteilung entscheidungsanalytischer Studien vermittelt werden. Die beiden wichtigsten Methoden der EA sind das Entscheidungsbaumverfahren und Markov-Modelle. Zusätzlich werden die verschiedenen Formen von Sensitivitätsanalysen dargestellt. Anhand von Beispielen werden Standardsituationen für die Anwendung von EA zur Evidenzübertragung erläutert und zentrale Versorgungsbereiche und Parameter bei der Evidenzübertragung aufgezeigt. Schließlich soll ein Stufenkonzept (Framework) zur Übertragung internationaler Evidenz im Rahmen von HTA auf den nationalen Kontext entwickelt sowie die Stärken und die Limitationen des entscheidungsanalytischen Ansatzes kritisch gewürdigt und herausgearbeitet werden.
Zusammenfassend können in diesem HTA-Bericht verschiedene Situationen, in denen sich der Einsatz entscheidungsanalytischer Modelle anbietet, dargestellt und der Nutzen des Einsatzes von EA im Rahmen der Übertragung internationaler Evidenz auf den nationalen Kontext aufgezeigt werden. Es wird ein systematisches Instrumentarium zur Übertragung internationaler Evidenz auf den nationalen Versorgungskontext entwickelt, das in zwei HTA-Projekten erfolgreich angewandt und in weiteren HTA-Projekten, bei denen sich die Problematik der Übertragung von Evidenz aus ausländischen Studien auf den deutschen Kontext stellt, empfohlen wird. Der Einsatz entscheidungsanalytischer Modelle zur Evidenzübertragung wird befürwortet. Er besitzt jedoch verschiedene Limitationen, die in jedem HTA-Bericht transparent dargestellt und kritisch diskutiert werden sollen.
Executive Summary
Objectives
The objective of this Health Technology Assessment (HTA) methods report was to examine and to assess decision analysis (DA) as a method to transfer and adapt international scientific evidence in HTA to the German health care context. Furthermore, we sought to develop a systematic framework to facilitate the selection, transfer, adaptation, and synthesis of these data in German HTA projects.
Background
Due to the rapid pace of medical progress and the escalating health care costs associated with it, methods to evaluate the clinical and economic consequences of medical technologies play an important role. The gold standard for the efficacy of a health technology is the randomised controlled clinical trial. However, these studies are often time intensive, expensive, or cannot be performed for practical or ethical reasons. In addition, due to the scarce resources of the health system, the evaluation of the economic consequences of medical technologies is increasingly important. National policy makers often face the problem that the only available data is based on studies conducted in other countries which may or may not be generalisable to our health care context. While the acceptability of applying international evidence to the national context is often debatable, national data assessment is expensive and time intensive.
The use of decision-analytic models enables the combination of different evidence-based datasets, given certain assumptions. Transferring and applying evidence-based data from international studies to the German health care context, synthesizing clinical and economic data to evaluate the cost-effectiveness of a technology in the national context, and combining results from clinical studies with intermediate outcomes with results from epidemiologic observational studies with long-term outcomes requires the use of decision-analytic models.
Research Questions
The following research questions are examined in this HTA report:
What are standard situations, in which international evidence must typically be transferred and applied to the national health care context?
Which parameters must be considered in decision-analytic modelling?
Which methodological approaches are useful for German HTA projects?
What are the strengths of DA in evidence synthesis and transfer?
What are the limitations of DA as a tool to transfer evidence?
Methods
We performed a systematic literature search to identify relevant literature regarding problems and solutions of decision-analytic evidence transfer. We included studies that compared different countries or regions regarding clinical, epidemiologic, or health-economic research questions.
In this report, we review and summarise the concepts and methods of DA; present possible areas of application, and provide a basis for the critical assessment of decision-analytic studies. The two main methods of DA, decision trees and Markov models, as well as various approaches to sensitivity analyses are described. Examples of typical situations for the use of DA in scientific evidence transfer are described and a list of the main health care domains and parameters in evidence transfer is presented. Finally, we developed a framework to transfer and apply international evidence to the national health care context. The strengths and limitations of the decision-analytic approach are critically examined.
Results and conclusions
The results of the literature search show a discrepancy between the theoretical literature recommending the use of decision-analytic approaches for transferring evidence between different health care systems and the lack of actual applications of DA in the published literature. This gap demonstrates the importance of future research in this area.
In summary, this HTA report describes different situations in which decision-analytic models can be useful, and demonstrates the utility of DA in transferring and applying international evidence to the national health care context.
We developed a systematic instrument to transfer international evidence to the context of other countries. This instrument includes a list of parameters that must be considered for evidence transfer. The list comprises parameters from the following areas: sociodemography, epidemiology, health outcomes, patient preferences/quality of life, health economics, health care and technology standards, structure of the health care system, and other determinants.
Furthermore, we developed a stepwise framework to transfer and adapt evidence from other health care systems to the national context, and describe the practical steps of performing the decision-analytic evidence transfer including the different options that are available dependent on the study setting. The framework consists of the following steps:
Identification of the international evidence
Identification of the set of parameters relevant to the decision problem
Assessment of the transferability of international results using a checklist
Decision on whether a reanalysis in the German context is necessary
Decision on whether a DA is useful/necessary
Installation of an interdisciplinary expert network
Assessment of data availability for Germany and data assessment
Model development or structural model adaptation
Numeric implementation or adaptation of model parameters
Clinical and/or cost-effectiveness analysis
Sensitivity analyses
Interpretation of results and conclusions/recommendations
Identification of areas of further research
This framework has been shown to be feasible; it has been successfully applied in two German HTA reports in the fields of antiviral therapy of chronic hepatitis C and cervical cancer screening.
The use of this instrument is recommended in further HTA projects dealing with the application of international evidence to the German health care context. The use of decision-analytic models to transfer international evidence is endorsed. However, the limitations of DA should be clearly stated and discussed transparently in all HTA reports. The use of decision-analytic modelling in further HTA reports will show whether this approach can support the transfer and application of evidence in HTA and which modifications and extensions are necessary.