Session 2.2 – Evidence Assessment Bodies and Priority Setting
Wednesday October 29 15:30-17:00
An International Comparison of Revealed Preference for Health Gains: A Retrospective Analysis of Drug Funding Decisions in 3 Countries
(Abstract 0099)Anthony Harris1, Jing Jing Li1, Braden Manns2, Fiona Shrive2
1Monash University, Australia, 2University of Calgary, Canada
Australia has been a pioneer in the systematic evaluation of drugs before public funding on the Pharmaceutical Benefits Scheme (PBS). In 1992 it became the first country with mandatory evaluation of both the effectiveness and the value for money of new drugs on the national public insurance scheme. Bodies in Canada (CDR) and the UK (NICE) have subsequently adopted systems of evaluation with similar characteristics.
The study examines factors that have influenced decisions to recommend that a drug be listed on the PBS compared to decisions in Canada and England. There are clear differences in the rate of rejection for funding across jurisdictions. Variation in decisions and the differences in factors that are considered important might reveal differences in attitudes indicative of community attitudes. For example one jurisdiction may have a strong preference for life saving drugs over life enhancing drugs, or may be more accepting of high costs of health gains in some indications (eg cancer) compared to others (eg arthritis).
Hypotheses on factors determining recommendations and differences across countries are tested in a binary choice regression modelling framework controlling for those factors identified above that vary within and across countries. Separate models for each country are estimated, and where appropriate fixed or random effects models are used to test and adjust for variation across a drug and countries. The paper presents recently completed work in Australia a proposed methodology for the trination comparison, and preliminary results.
Assessing the Quality of Evidence for Decision-making (QED): A new instrument
(Abstract 0104)Jing Jing Li, Anthony Harris, Geoffrey Chin
Monash University, Australia
Policy makers need to know how much confidence they can place in evidence used in prioritising health care spending. Systematic and explicit methods of making judgements can reduce inconsistencies and improve communication. In the context of evaluating the value for money of public insurance coverage decisions of drugs in Australia, we have devised a system for appraising the level and quality of clinical and economic evidence presented to decision makers. Compared to existing methods, our instrument, the Quality of Evidence in Decision-making (QED) is based on a broader range of attributes. The criteria included in the QED are: strength of evidence (level, statistical precision, and quality), size of the effect and relevance of the evidence (comparator and population). The quality of evidence is further sub-divided into 12 elements including: the adequacy of study selection, randomisation, blinding, follow up, patient attrition, baseline patient characteristics, sample size, dosage regimens, acceptable outcomes, and methods of analyses including subgroup and post hoc analyses. A score is calculated for each QED attribute. We tested the QED on 254 drug submissions to the Australian Pharmaceutical Benefits Advisory Committee (PBAC) between 1994 and 2004. We show that the QED can be used to provide a relevant assessment tool for decision makers.
A Tri-Nation Comparison of Pharmacoeconomic Submissions
(Abstract 0130)Braden Manns1, Tony Harris2, Fiona Clement3, Karen Lee4
1University of Calgary, Canada, 2Monash University, Australia, 3University of Newcastle, United Kingdom, 4Canadian Agency for Drugs and Technology in Health, Canada
Background: The Common Drug Review (CDR) was established in Canada to provide formulary listing recommendations to the participating public drug plans. Although set up with differing mandates, the National Institute for Clinical and Health Excellence (NICE) in the United Kingdom and the Pharmaceutical Benefits Advisory Committee (PBAC) in Australia appear to have similar roles.
Objective: To compare the submissions received by each committee and describe the similarities and differences among review processes and recommendations made.
Methods: Through the public access website maintained by each committee a list of each submission considered was compiled. Data were collected from the implementation of publicly accessible information to Dec 31, 2007 (NICE - Feb. 2001, CDR - Jan. 2003 and PBAC - July 2005). Each drug submission for a unique indication was considered a separate submission.
Results: NICE has reviewed 157 unique drug/indication combinations, while CDR and PBAC have each reviewed 84 and 153 drug/indication combinations, respectively. A positive listing recommendation was given for 89% of NICE submissions, 49% of CDR submissions and 53% of PBAC submissions. The list rate did not vary substantially when subgroups were considered based on type of disease, goal of treatment, or if the drug was considered first in its class.
Policy Implications: There is variation in the list rate of the 3 countries. This may, in part, be due to the mandate of each committee within the drug reimbursement system.
Can Preferences from Discrete Choice Studies Translate into Meaningful Public Involvement in Priority Setting? An Application to Health Technology Appraisal Decisions in the UK.
(Abstract 0078)Colin Green
University of Exeter, Peninsula Medical School, United Kingdom
Background: The UK NHS seeks to take into account the views of the general public when setting healthcare priorities. But there is little evidence available, and difficulties translating evidence. The use of public preferences collected in ‘social value’ surveys using discrete choice methods, offers an opportunity to help with the challenge of including public views in priority setting.
Objective: To test the use of information on preferences from discrete choice surveys to estimate the social value of healthcare interventions in a way that may help in a range of priority setting processes.
Methods: The translation of data from a social discrete choice study (reporting data from 250 face-to-face interviews) to a technology appraisal setting. Preference data from the study are mapped onto the characteristics of health technologies appraised by UK NICE, and predicted social preferences are compared with decisions made on the funding of health technologies for the NHS.
Results: A broad set of social value data were derived. The social value data could be mapped onto existing deliberative processes. The retrospective comparison found 10 of 13 judgments consistent with predicted preferences and appraisal judgments (4=yes, 6=no judgments).
Policy implications: The application demonstrates how such data can be helpful across a broad range of appraisal environments, and that preference data can be used in a simple and policy-relevant manner to make the academic practical, and to offer an opportunity for meaningful public involvement in priority setting across all levels of the UK NHS.