Objectives: The aim of this study was to describe a comprehensive continuum that has developed in Ontario between government and key stakeholder groups, including hospitals, physicians, academic institutions, clinical epidemiologists, health economists, industry, and bioethicists to achieve evidence-based recommendations for policy development.

Methods: The various components of the comprehensive model that has evolved to develop an evidentiary platform for policy development are summarized, and the flow between these components is described.

Results: The development of the Ontario Health Technology Advisory Committee (OHTAC) and associated programs demonstrate the need to go beyond the traditional steps taken within most health technology assessment paradigms. These components include pragmatic postmarketing studies, human factors, and safety analyses, and formalized interactions with a broad spectrum of potential end-users of each technology, experts, and industry. These components, taken together with an expanded systematic review to include a range of economic analyses, and societal impacts augment the traditional systematic review processes. This approach has been found to be important in assisting decision making and has resulted in an 81 percent conversion from evidence to policy consideration for eighty-three technologies that had been assessed at the time this article was submitted.

Conclusions: The comprehensive model, centered around OHTAC, has added important new dimensions to health policy by improving its relevance to decision makers and providing an accountable and transparent basis for government to invest appropriately in health technologies. This study could also form a basis for further research into appropriate methodologies and outcome measurements as they relate to each component of this approach.

Objectives: This study sought to identify and compare various practical and current approaches of health technology assessment (HTA) priority setting.

Methods: A literature search was performed across PubMed, MEDLINE, EMBASE, BIOSIS, and Cochrane. Given an earlier review conducted by European agencies (EUR-ASSESS project), the search was limited to literature indexed from 1996 onward. We also searched Web sites of HTA agencies as well as HTAi and ISTAHC conference abstracts. Agency representatives were contacted for information about their priority-setting processes. Reports on practical approaches selected through these sources were identified independently by two reviewers.

Results: A total of twelve current priority-setting frameworks from eleven agencies were identified. Ten countries were represented: Canada, Denmark, England, Hungary, Israel, Scotland, Spain, Sweden, The Netherlands, and United States. Fifty-nine unique HTA priority-setting criteria were divided into eleven categories (alternatives; budget impact; clinical impact; controversial nature of proposed technology; disease burden; economic impact; ethical, legal, or psychosocial implications; evidence; interest; timeliness of review; variation in rates of use). Differences across HTA agencies were found regarding procedures for categorizing, scoring, and weighing of policy criteria.

Conclusions: Variability exists in the methods for priority setting of health technology assessment across HTA agencies. Quantitative rating methods and consideration of cost benefit for priority setting were seldom used. These study results will assist HTA agencies that are re-visiting or developing their prioritization methods.

Objectives: The practical significance of health technology assessment (HTA) in policy decisions or clinical practice has been challenged. Possibly, problem definitions underlying HTA do not concur sufficiently with the problem definitions held by policy makers or clinicians. We performed an in-depth case study on mebeverine, a drug prescribed to patients with irritable bowel syndrome, to explore this hypothesis.

Methods: The theoretical framework was provided by the theory of argumentative policy analysis. We analyzed documents and held semistructured interviews to collect data. We reconstructed interpretative frames to analyze actors' argumentation.

Results: The funding and usage problems relating to mebeverine were ill-structured. Actors disagreed on the information needed and the norms at stake. As a result, the problem definition shifted, and the resulting problem definitions failed to correspond with the problems perceived by the target populations.

Conclusions: To ensure that future studies on healthcare problems are useful, it is imperative that policy makers take the problem definitions of potential users into account.

Objectives: The aim of this study was to define an effective early warning system, to identify and rank the characteristics of an effective early warning system for emerging health technologies, and to evaluate current early warning systems against these characteristics.

Methods: An iterative Delphi-type process with the thirteen members of the International Information Network on New and Changing Health Technologies (EuroScan). We synthesized key characteristics that network members had graded. Members were then asked whether these characteristics were present or fulfilled in their system.

Results: The definition of an effective early warning system developed was the following: a system that identifies innovations in the field of health technology likely to have a significant impact; and disseminates information relevant to the needs of the customer which is timely, so as to enable appropriate decision making (such as resource allocation), facilitate appropriate adoption, and identify further research requirements. Five primary and eleven secondary components of effective early warning systems were identified. The five primary characteristics concerned relevance, independence, resourcing, a clear pathway for the outputs to reach decision makers, and defined customers. Although the primary characteristics were present or fulfilled to some extent in the majority of evaluated early warning systems, there was considerable variability in the presence of the secondary characteristics in the evaluated systems.

Conclusions: Our study provides a definition for an effective early warning system and a shared understanding of the important characteristics and components of such systems. This work should provide guidance to those setting up new early warning systems as well as for those managing and reviewing current systems.

Objectives: The aim of this study was to estimate thresholds for production volume, durability, and cost of care for the cost-effective adoption of liver organ replacement technologies (ORTs).

Methods: We constructed a discrete-event simulation model of the liver allocation system in the United States. The model was calibrated against UNOS data (1994–2000). Into this model, we introduced ORTs with varying durability (time to failure), cost of care, and production volume. Primary outputs of interest were time to 5 percent reduction in the waiting list and time to 5 percent increase in expected transplant volume.

Results: Model output for both calibration and validation phases closely matched published data: waiting list length (±2 percent), number of transplants (±2 percent), deaths while waiting (±5 percent), and time to transplant (±11 percent). Reducing the waiting list was dependent on both ORT durability and production volume. The longer the durability, the less production volume needed to reduce the waiting list and vice versa. However, below 250 ORT/year, durability needed to be >2 years for any significant change to be seen in the waiting list. For base-case costs, all ORT production volume and durability scenarios result in more transplants per year at less total cost of care/patient than the current system. ORTs remain cost saving until manufacturing costs are >5 times base-case costs, production is less 500 ORT/year, and durability <6 months.

Conclusions: Although there remain many technical challenges to overcome, as long as ORTs can meet these threshold criteria, they have the potential of transforming the world of end-stage liver disease.

Objectives: The aim of this study was to demonstrate how value of information analysis can measure the upper limit on returns to future research and identify the research priorities for computer-assisted total knee replacement (CAS-TKR).

Methods: Using a previous economic analysis of CAS-TKR compared with conventional TKR, the population expected value of perfect information (EVPI) was calculated using Monte Carlo simulation to provide an estimate of the upper limit on returns to future research. The population expected value of partial perfect information (EVPPI) for both individual parameters and groups of parameters was estimated to inform specific future research priorities.

Results: The UK individual EVPI would be £21.4 if the willingness to pay for one QALY (quality-adjusted life-year) were £30,000. The population EVPPI would be £8.3 million, assuming a 10-year time horizon for CAS-TKR. In this instance, the expected value of information is positively related to willingness to pay for one QALY for the range of £0 to £50,000. Although each individual parameter had an EVPPI of £0, groups of utility parameters had positive EVPPI. Population EVPPI was £5.6 million for utility parameters, £20,000 for transition probabilities relating to CAS-TKR, and £5,000 for transition probabilities related to conventional TKR.

Conclusions: The study provides evidence on which parameters further information may be of most value. Focusing research on the utility values associated with health states relating to TKR would be of greatest value.

Objectives: The National Institutes of Health (NIH) Consensus Development Conference (CDC) was instituted to provide evidence-based guidance on controversial medical issues to researchers, health practitioners, and the public; however, the degree of impact this activity has on stimulating relevant research is unclear. This study examines the impact of CDC statements on the initiation of related NIH-funded research projects.

Methods: Six CDCs from 1998 to 2001 were examined. Research initiatives related to the Conferences' topics were collected through two discrete methods: (i) the overall number of relevant pre- and postconference research activities was compiled using NIH's Information for Management, Planning, Analysis, and Coordination II (IMPAC II) and the Department of Health and Human Services' (DHHS) Computer Retrieval of Information on Scientific Projects (CRISP) grant application and award databases; (ii) for each CDC, the sponsoring institute's conference coordinator and other identified Program Directors were queried for their knowledge of new conference-specific research initiatives sponsored by their institute. The main outcome measure was the total number of requests for applications, requests for proposals, program announcements, broad agency announcements, notices, and funded investigator-initiated research program grants (RO1s) for a given Consensus topic in the 3 years before (baseline measure) and following (measure of impact) a CDC.

Results: As identified through NIH's IMPAC II and DHHS' CRISP grants and announcements databases, the total number of relevant postconference research initiatives increased for five of six CDCs when compared with baseline activity levels; research activities remained constant for the sixth. When inclusion criteria were restricted to institute-identified research initiatives, two of six CDC topics had overall increases in relevant research activity in the postconference period.

Conclusions: CDCs appear to have a positive impact on the stimulation of related NIH-funded research initiatives. Future outcomes evaluations using prospective data collection methods and more robust participation by sponsoring and cosponsoring institutes should strengthen the reliability of the association between new research initiatives on a given topic and their causal relationship to a given CDC.

Objectives: To determine the relative extent to which the National Institute for Health and Clinical Excellence (NICE) appraises new versus existing technologies, and pharmaceutical versus nonpharmaceutical health technologies.

Methods: We categorized technologies within NICE appraisals published between March 2000 and June 2006 by type and classified them as new or existing using the timeline between launch in the United Kingdom and referral to NICE. We used a 3-year postlaunch cutoff to determine whether a technology was new, with a sensitivity analysis of 1 and 5 years.

Results: We reviewed 159 technologies from 88 appraisals. Of these, 84 (53 percent) were new (sensitivity analysis 36 to 67 percent) and 75 (47 percent) were existing technologies. A total of 119 (75 percent) were pharmaceuticals, 22 (14 percent) were devices, 14 (9 percent) were procedures, and 4 (3 percent) were categorized as miscellaneous. Classification according to newness and technology type showed that 62 percent (42 to 75 percent) of the pharmaceuticals appraised were new.

Conclusions: By developing and applying a definition of new, we have found that the criticism of the bias toward new technologies is unfounded when applied to the appraisal program overall. At the same time, new pharmaceuticals are over-represented in the program compared with devices and procedures. This domination may cause inflationary pressures on the health service, but any wholesale move away from the technological frontier may be more costly.

Objectives: Implantable cardioverter defibrillator (ICD) therapy reduces the risk of sudden death in patients with ischemic cardiomyopathy, but their novelty and cost may represent barriers to utilization. The objective of this study was to examine the influence of age, gender, place of residence, and socioeconomic status on rates of ICD implantation for the primary prevention of death.

Methods: We conducted a population-based retrospective cohort study involving the entire province of Ontario, Canada. Patients were eligible if they had survived following hospitalization for heart failure from 1 January 1993, to 31 March 2004, and previously sustained an acute coronary syndrome within 5 years. Patients with an existing ICD or a documented history of cardiac arrest were excluded, as were patients who died in the hospital. Primary outcome was ICD implantation.

Results: We identified 48,426 patients hospitalized for heart failure who survived to hospital discharge. Of these, 440 received an ICD, with a gradual 30-fold increase in implantation rates over the study period (.12–3.9 percent). ICD recipients were more likely to be men (odds ratio [OR] = 4.14; 95 percent confidence interval [CI], 3.24–5.30), younger than 75 years of age (OR = 3.19; 95 percent CI, 2.57–3.96), reside in a metropolitan area (OR = 1.42; 95 percent CI, 1.04–1.9), and live in a higher socioeconomic neighborhood (OR = 1.32; 95 percent CI, 1.08–1.61).

Conclusions: Among patients with heart failure and a previous myocardial infarction, ICD use is increasing in Ontario. However, the application of this technology is characterized by major sociodemographic inequities. The causes and consequences of the pronounced age and gender discrepancies, in particular, warrant further investigation.

Objectives: The aim of this study was to evaluate the cost-effectiveness of on-site automated external defibrillators (AEDs) in the initial management of cardiac arrest in Ontario.

Methods: This was a cost-effectiveness analysis based on published literature and data from the Canadian Institute of Health Information. The participants were fictitious male and female cardiac arrest patients who were initially managed with on-site AEDs, compared with similar patients managed without on-site AEDs. This group included a subgroup of high-risk patients (i.e., heart failure and left ventricular ejection fraction <35 percent). The analysis was conducted in a variety of settings including hospitals and homes in Ontario, Canada. The main outcome evaluated was cost per quality-adjusted life-year (QALY) gained from a payer's perspective.

Results: Cost per QALY (all costs reported in Canadian dollars) was $12,768 when AEDs were deployed in hospitals, $511,766 when deployed in office buildings, $2,360,023 when deployed in apartment buildings, $87,569 when deployed in homes of high-risk patients, and $1,529,371 when deployed in homes of people older than 55 years of age.

Conclusions: Indiscriminate deployment of AEDs is not a cost-effective means of improving health outcomes of cardiac arrest. Their use should be restricted to emergency response programs, high-risk sites (such as hospitals), and high-risk patients.

Objectives: To compare the cost-effectiveness of the treatment of ankylosing spondylitis (AS) with infliximab in the United Kingdom over lifetime estimated from two different clinical trials and adjusted for clinical practice guidelines.

Methods: A cost-effectiveness model was developed to incorporate clinical, epidemiological, and economic data and allow extrapolation of trial results and incorporation of long-term treatment. Assumptions regarding treatment beyond the trials were based on open extensions from the trials and treatment guidelines by the British Society for Rheumatology. Results are presented for both the societal perspective and the perspective of the National Health Service (UK £, discounted 3.5 percent).

Results: Under the assumption that disease activity would be controlled and functional capacity would remain stable while on drug, treatment with infliximab (5 mg/kg every 6 weeks) dominates standard treatment in the societal perspective. In the National Health Service perspective, the cost per quality-adjusted life-year (QALY) gained over lifetime was £28,300 and £26,800 for the two trials. If functional capacity were to deteriorate at half the rate of untreated patients, the cost per QALY gained would be £35,300 and £34,100, respectively. The results are sensitive to the dosing regimen adopted, the discontinuation rate, and assumptions concerning disease progression while on treatment.

Conclusions: The two clinical trials yield the same cost-effectiveness results and the cost per QALY gained with treatment was found to be in the acceptable range.

Objectives: To compare the cost-effectiveness of endoscopy and empirical proton pump inhibition (PPI) therapy for management of dyspepsia in primary care.

Methods: A randomized controlled trial, including prospective collection of economic resource data, was conducted in general practice from June 2000 to August 2002, Aarhus County, Denmark. We randomly assigned 368 dyspeptic patients from thirty-two general practices to treatment with omeprazol 40 mg for 2 weeks (n=184) or endoscopy (n=184). The study adopted a societal perspective, and the year of costing was 2006. Outcome measures: days free of dyspeptic symptoms and proportion of patients with dyspepsia after 1 year based on patients' and general practitioners' (GPs) assessment. Costs were estimated from patient and GP questionnaires and from medical records.

Results: The incremental cost-effectiveness (CE) ratio for 1 day free of dyspeptic symptoms using the endoscopy strategy was €/day 154 compared with the PPI strategy. The incremental CE ratio for one person free of dyspeptic symptoms after 1 year using the endoscopy strategy was €13,905 based on the patients' evaluation, and the incremental CE ratio for one person free of predominant symptoms after 1 year was €5,990 according to the GPs' evaluation. The PPI strategy was both cheaper and more effective than the endoscopy strategy when reflux was the predominant symptom.

Conclusions: A strategy using empirical antisecretory PPI therapy should be recommended if the alternative is an endoscopy strategy for managing dyspeptic patients in general practice, especially if reflux was the predominant symptom.

Objectives: The aim of this study was to compare the generic health-related quality of life (HRQoL) instruments, the 15D and the Nottingham Health Profile (NHP), in terms of feasibility, cross-sectional construct validity, discriminatory power, and responsiveness to the change in patients with lower limb atherosclerosis disease (LLAD).

Methods: Patients who were treated conservatively (64 patients), scheduled for endovascular treatment (85 patients), or for elective surgery (31 patients) filled in the NHP and the 15D questionnaires before treatment and 12 months after. The methods of analysis included calculations of the completion rate, the multitrait–multimethod matrix, extreme group comparisons with the t-test, and calculations of “floor” and “ceiling” effects and effect sizes.

Results: Both the 15D and the NHP instruments were feasible because full data were obtained at baseline from 84.1 percent of patients and 82 percent at 12 months with the 15D and, respectively, 75.8 percent and 79 percent with the NHP. The multitrait–multimethod matrix and extreme group comparisons provided a clear convergent and discriminant evidence of the cross-sectional construct validity of the instruments. The convergent validity correlations were quite high, ranging between .40 and .682. The tendency for higher percentages at the “ceiling” and “floor” for the NHP suggests that it has less discriminatory power than the 15D on roughly comparable dimensions. The 15D and the NHP were almost equally responsive to change.

Conclusions: This study provided evidence that both the 15D and the NHP are appropriate for measuring HRQoL among patients with LLAD.

Objectives: Conducting economic evaluation in the context of clinical studies is common but has been the subject of extensive discussion due to its limitations. Various standard approaches and guidelines how such studies should be conducted have been proposed, but there is very limited information available about how common these studies are, what type of data that is collected, and how the quality of the protocols compares to the suggested standards. This study examines the prevalence and study design of health economic evaluations conducted alongside clinical trials in Sweden between 1995 and 2005.

Methods: A systematic assessment of clinical trial protocols that had arrived as applications to the Medical Products Agency between 1995 and 2005 was performed. Only protocols arriving during the first half of odd years within the time period were included.

Results: A total number of 680 protocols from 1995 to 2005 were examined, and among them, 14.4 percent included a health economic part. With the exception of year 2001, a trend toward an increased prevalence of economic evaluations next to clinical trials can be seen.

Conclusions: This study shows that economic evaluations alongside clinical trials are becoming more common, although most trials still lack a health economic part of the protocol. The information about the economic evaluation provided in the protocols is in many cases scarce, possibly due to the fact that there currently are no generally accepted and applied guidelines for economic evaluations in clinical trial protocols. Introducing requirements for detailed study plans also for the economic evaluation should improve the quality of economic evaluations alongside clinical trials.

To the Editor:

In a recent study in this Journal, Drummond and colleagues (5) attempt to address the question “Whether the standard methods of HTA are adequate for assisting decisions on patient access to and funding of orphan drugs” before proposing “a research agenda to help understand the societal value of orphan drugs and issues surrounding their development, funding, and use.”

To the Editor:

In their comment on our study, McCabe et al. make several points, with which some we agree and some we do not. Here, we respond to the most pertinent issues, using the same twelve headings.