Cost-effectiveness analysis can create an unlevel playing field

SAD BUT TRUE

The use of conventional cost-effectiveness analysis to estimate a value-based price can result in an “unlevel playing field” among healthcare innovations. This is explained through examples in a paper by Drs. Lou Garrison, Boshen Jiao, and Omar Dabbous.  

To illustrate this, let’s consider a simple example that assumes a treatment is priced based on a value-based price as suggested from a conventional cost-effectiveness analysis. 

Assume you have Treatment A that you conduct a conventional cost-effectiveness analysis for, and the cost-effectiveness analysis suggests that Treatment A results in 1 additional unit of a health outcome (let’s use life years for illustrative purposes) and no added costs or cost offsets (not including the cost of Treatment A) per person treated. If you were to use this analysis to suggest a value-based price for Treatment A, you would need to assume a threshold. For illustrative purposes, let’s assume a threshold of $150,000 per life year gained. So, Treatment A’s value is $150,000 per person treated ($150,000 per life year gained multiplied by 1 life year gained per person treated). 

So the “value-based prize” for the treatment per person treated is $150,000. However, to turn this prize into a price for the treatment, we must consider the duration-of treatment.

If Treatment A was a one-time treatment, the math is simple. The value-based prize of Treatment A would be paid out once and upfront—$150,000 divided by 1 is $150,000. Thus, the value-based price is $150,000. 

If Treatment A was a chronically administered treatment that a person was on for their lifetime, the math would be a little more complex. The value-based prize of Treatment A would still be $150,000 per person treated but it is now paid out over the treatment duration which spans multiple years rather than one upfront payment. Using a discount rate of 0% to keep this simple, if the person was on Treatment A for 30 years, then the value-based price would be $5,000 per year ($150,000/30 years). So over 30 years on treatment, $5,000 is paid per year to equate to the value-based prize of $150,000 per person treated.

Treatment A is assigned the same value-based prize whether it is a one-time treatment or a chronically administered treatment because it produces the same level of benefit. However, the unlevel playing field comes into play when we consider the percentage of the prize that is paid to the innovating manufacturer of Treatment A (which is the manufacturer we want to reward). 

To illustrate this, we need to consider one more piece of information—Treatment A’s market exclusivity period. We will use a market exclusivity of 13 years for this example. After that point, we assume competing copies can enter the market. Again, for simplicity purposes, we assume competing copies enter and take all the market share from the innovating manufacturer. 

Let’s go back to the one-time treatment. If the one-time treatment was administered in year 1 of the market exclusivity period, the value-based prize would be $150,000 per person treated and the manufacturer would receive 100% of that prize because it is paid out entirely over the market exclusivity period. If the one-time treatment was administered in the last year of market exclusivity, the value-based prize would be $150,000 per person treated and the manufacturer would receive 100% of that prize because it is paid out entirely over the market exclusivity period. 

Now let’s consider a chronically administered treatment. If the chronically administered treatment was started in year 1 of market exclusivity (and a patient was on it for 30 years), the value-based prize would still be $150,000 per person treated but the innovating manufacturer would receive $65,000 (annual value-based price of $5,000 multiplied by 13 years on treatment during market exclusivity). Therefore, if the treatment was started in year 1 of the market exclusivity period, the innovating manufacturer would receive 43% of the prize ($65,000/$150,000). 

If the chronically administered treatment was started in the last year of market exclusivity (and a patient was on it for 30 years), the value-based prize would still be $150,000 per person treated but the innovating manufacturer would receive $5,000 (annual value-based price of $5,000 multiplied by 1 year on treatment during market exclusivity). Therefore, if the treatment was started in the last year of market exclusivity, the innovating manufacturer would receive 3% of the prize ($5,000/$150,000).

This illustrates how the share of value to the innovating manufacturer is much greater for the one-time treatment as compared to the chronically administered treatment despite having the same benefit as it relates to health outcomes and cost. The paper by Garrison and colleagues is a must read and goes into its own set of examples and provides much greater discussion but it concludes with saying “these anomalies may be inducing distortions in the amount and types of pharmaceutical R&D”. 

Modeling drug dynamics (which includes accounting for the market exclusivity period) within cost-effectiveness analyses, as is recommended in a recent best practices report, is a way to better level the playing field.

SEEK AND DESTROY

Evergreening does little to extend the time between a branded drug’s market entry and its generic market entry and should not be used as a reason against incorporating dynamics within cost-effectiveness analyses. I am a loud advocate for incorporating dynamics within cost-effectiveness analyses (see section above and all previous newsletters). Some argue against incorporating dynamics within cost-effectiveness analyses due to uncertainty in the timing of generic entry. They cite practices like “evergreening” which they suggest prolong the exclusivity period and keep drug prices high for very long periods of time. “Evergreening” is a practice of patenting modifications (some potentially quite small) of an existing drug. This could extend the market exclusivity period and delay generic entry—which is a key component of incorporating dynamics within cost-effectiveness analyses. 

Interestingly, an analysis conducted by Emily Michiko Morris and Joshua Kresh found that patents and exclusivities after a drug’s market entry (this “evergreening” practice) did little to extend the time between a branded drug’s market entry and its generic market entry. They looked at a sample of top-selling branded small molecule drugs in 2012 and found that the average time between a branded drug’s market entry and its generic market entry was 13 years, of which only around 1.5 years was additional protection time that might have partially resulted from patents and exclusivities filed after approval. 

I am not suggesting that “evergreening” doesn’t occur. I am suggesting that we can use data, like that provided by Michiko Morris and Kresh and by many others, in economic models. This evidence on time to generic entry can be incorporated in economic models and the uncertainty and variability around it can be captured in the same manner that uncertainty and variability in other model inputs is accounted for.

TURN THE PAGE

Small molecules and biologics provide similar health benefits, but small molecules are typically cheaper and thus may be more efficient when it comes to resource allocation. Researchers from the Center for the Evaluation of Value and Risk in Health published a paper in Health Affairs titled “Small-Molecule Drugs Offer Comparable Health Benefits to Biologics and Lower Costs”. 

In their analysis, they found that small molecules and biologics provide similar health benefits, but small molecules were associated with fewer additional costs, and thus the incremental cost-effectiveness ratios were more favorable for small molecules as compared to biologics. And because I am not ready to get off my soapbox, I would gamble the vast majority of the incremental cost-effectiveness ratios in their sample did not incorporate dynamics which could potentially further increase the difference in cost-effectiveness between small molecules and biologics.

Given we are living in a world with the Inflation Reduction Act where small molecules only have nine years before a potential “negotiated” price takes effect as compared to the thirteen years that biologics have, these findings make me (and probably everyone else reading their findings) question the signals that policy makers are sending to innovators and investors. Clifford and colleagues state: “Our findings suggest that differential timelines for price negotiation could have detrimental effects, potentially leading to fewer small molecule drugs. This is concerning, as small-molecule drugs generally have more favorable cost-effectiveness profiles than biologics.”

We must do better about what signals we are sending innovators and investors with our pricing and with our policies if we are striving for an efficient and equitable system.