HTA

What’s new in the 2023 ICER Value Assessment Framework?

On September 25, ICER released its 2023 value assessment framework. What is new? What stays the same? Here are a few key points you need to know.

  • Special cases for ultra-rare and single or short-term therapies (SST): These did not change from previous.
  • Use of real-world data.  While ICER says it is committed to using real-world data, the two use cases it provides are pretty standard: “analyses of insurance claims data to better understand health states, resource utilization, and costs, or the analysis of new data from patient surveys to provide more direct information on health utilities.”
  • Use of data in confidence.  This has been updated as of 2023. “academic-in-confidence data will be redacted from all external and public ICER documents until the earlier of: (a) publication or presentation of such data by the data owner or study investigators; (b) 12 months following the date of the public ICER meeting; or (c) for reports that are not subject to a public meeting, 12 months following report publication. Following any of these dates, ICER will unmask all redacted information from reports, presentations, and other public documents.”  Thus, in confidence information should largely be seen as evidence soon to be published at a conference or in a peer-reviewed journal but has not yet been published rather than proprietary, confidential data that stakeholders may wish never to become public.
  • Broader value elements ignored again.  ICER mentioned broader value elements such as the value of hope, preferences for health gains among those with more severe illness, and scientific spillovers.  However, these will remain unquantified in their approach. “After ongoing consideration of the potential to perform quantitative analyses of these types of additional elements of value, we believe that there are still many methodological uncertainties and concerns that suggest it is more appropriate to consider these elements through deliberation focused on a broad cost-effectiveness threshold range. There are concerns related to potential double counting between these potential additional domains and the health gains captured by the evLYG/QALY.
  • But ICER says it will consider if/how to implement GRACE in the future. ICER signals that it will “begin a special focus in coming months on considering novel ways to quantify preferences related to severity, methods that often are framed as abandoning an assumption of a linear relationship between health gain value and replacing it with a formula that can capture risk aversion, severity, and the value of insurance. We will focus on exploring the Generalized Risk-Adjusted Cost-Effectiveness framework26,27 and methods adopted by several international HTA programs that now weight health gains in relation to severity
  • Health disparities quantified using HIDI, not DCEA.  ICER uses Health Improvement Distribution Index (HIDI) index it created to quantify if a treatment might disproportionately benefit (or not) disadvantaged groups. The HIDI method is explained in their White Paper here, but basically a HIDI value > 1.0 suggests that “more health may be gained on the relative scale in the subpopulation of interest when compared to the population as a whole.”  Also, ICER uses a Participant to Disease Prevalent Ratio (PDPR) based on age, sex and race to quantify the representativeness of clinical trials.  However, distributional cost effectiveness analysis (DCEA) is not used.
  • ICER continues to use health care system (i.e., payer) as its base case.  It justifies this by saying the primary consumers of its reports (i.e., payers) prefer this output.
  • Modified societal as co-base case.  ICER does sometimes include a modified societal that includes productivity impacts, caregiver burden and potentially some other factors. ICER says that when the difference between health care system and societal perspectives is large, then modified societal will be a co-base case.
  • Estimating productivity and caregiver impacts without direct evidence. In most previous applications, if there was not direct productivity impact data at launch, ICER would assume no productivity impact.  In the latest value assessment framework, ICER will also consider indirect methods whereby productivity impacts may be available by health state and treatments ability to slow disease progression will be captured based on differential time in health state.  Also, if no caregiver data is available, ICER will assume that “assume that caregiver time spent is proportional to 75% of patient formal labor time.”
  • Discount rate remains at 3%. No change from previous.
  • Dynamic pricing not considered…yet  ICER says it will not estimates the impact of future price changes across treatments due to loss of exclusivity…except they say they will consider generic competition if a substantial price decrease is likely to occur within 12-24 months.  However, it then says, due to IRA cost growth is likely to be capped at inflation and the price of treatments are likely to fall by 75% at 9 (small-molecule) or 13 (biologic) years after launch. ICER does not commit to institute this estimate but rather will convene expert to weigh in on how to discuss.
  • Cost per QALY and cost per evLYG. Thresholds based on willingness to pay of $50k, $100k, $150k and $200k, but price benchmarks based on $100k/QALY and $150k/QALY thresholds.  They cite a preferred threshold of $104,000 per QALY (see Vanness 2021), but still the $100k/QALY and $150k/QALY are used.
  • Revised methods for single and short-term therapies (SST).  These include:
    • A 50/50 shared savings model: In this case, ICER assumes that 50% of the lifetime health system cost offsets from a new treatment are “assigned” to the health system instead of being assigned entirely to the new treatment; and
    • Cost offset maximum.  Cost offsets generated by a new treatment are capped at $150,000 per year, but are otherwise assigned entirely to the new treatment.
  • Benefits beyond health.  Ones ICER mentions include unmet need, treatment complexity; impact on caregiver quality of life, productivity, and family life; health equity; and public health impact.
  • Budget impact threshold. This will be “…calculated as double the average net budget impact for new drugs that would contribute to overall health care cost growth beyond the anticipated growth in national GDP plus an additional 1%.”

The full value assessment framework document is here.