Forecasting for generic erosion rates
*Carla Peek, head of business modeling and analytics at InPharmation, on how to create forecasts that account for the market impact of generics*
By on Feb 3, 2011Carla Peek, head of business modeling and analytics at InPharmation, on how to create forecasts that account for the market impact of generics
Generic penetration is to blame for much of Big Pharmas flagging sales growth.
As patents expire across numerous drug markets, generics have pressed in and, within years, swallowed giant swaths of market share.
This will only increase in the coming decade.
Decision Resources, a market research firm in Massachusetts, projects that generic erosion will reduce the hospital-acquired infections drug market by 40 percent in the next eight years.
Meanwhile, the generic erosion of the atypical antipsychotic drug class will cause a $2.3 billion decline in the bipolar disorder drug market over the next decade.
These dire projections continue from one drug market to the next.
Forecasters, therefore, must get better at modeling generic erosion rates when forecasting branded drug sales post-patent expiry.
Generic penetration is increasing everywhere over time, Carla Peek, head of business modeling and analytics at InPharmation, said at eyeforpharmas Forecasting Excellence Summit in Madrid.
Thus, understanding the nuances of brand erosion and the best ways to model it are key for forecasters moving forward.
An eroded landscape
While generic erosion is ubiquitous, the speed and severity with which it eats away at drug markets depends on the country, prescribing setting, therapy area, and formulation type.
On the national level, generic penetration tends to be greater in the US and the UK compared to countries on the European continent.
Its greater in freer price markets, Peek said, part of a quid pro quo with authorities.
As a result, the average basket of UK drug products experiences generic penetration around 70 percent, while the average basket of French products experiences generic penetration around 45 percent.
Looking to Asia, brands tend not to experience generic erosion in China following patent expiry since they often face generic competition from the outset, and instead continue to grow both in terms of volume and value.
On the commodity level, generic penetration varies from pure commodities to part commodities.
Pure commodities, Peek explained, are products that are in demand and the same no matter who produces them, like petrol or milk.
Anti-ulcers are a good example of pure commodities in pharma.
Part commodities, on the other hand, are products where the higher the quality, the higher the price, so the consumer is weighing quality versus price, according to Peek.
In pharma, respiratory inhalers or biosimilars fall into this latter category.
In general, pure commodities experience generic penetration of 90 percent, while part commodities contend with just 10 percent.
On the setting level, brand erosion is generally faster and more severe in hospital settings, reflecting greater brand loyalty among patients in the retail setting.
Finally, in terms of erosion by therapy areas, sales and volume erosion tends to be the greatest among infectious disease, oncology, and cardiovascular small molecule brands, according to a recent report on brand erosion from Datamonitor.
Forecasting tools, bad and good
A powerful tool for modeling how a new drug will perform against an existing class of drugs is conjoint analysis.
Using surveys, forecasters can model the decision-making process that doctors will follow when faced with competing drugs.
Different product attributeslike, say, the severity of its side effects or the convenience of its dosingwill impact doctors inclination or reluctance to prescribe the drug.
Peek warned that forecasters who use this technique to model generic erosion rates should think again.
Conjoint studies are very good when comparing oranges to oranges, classes to classes, brands to brands, but very poor when trying to compare brands with generics, she said.
When conjoint studies are done for commodity products, they typically underestimate the generic penetration.
Doctors often give part-worth scores that are unrealistic, so the various attribute ratings dont add up.
Furthermore, doctors from different countries typically give part-worth scores that are far too similar for generics considering the variation in the degree of generic penetration between countries, Peek said.
A better solution for modeling how quickly generics will capture market share is uptake curves.
One of the six principals of good forecasting is to break the problem down, forecast different components separately, and then bring them back together again, which is essentially what uptake curves do, said Peek.
In particular, InPharmation uses the Bass Model to generate its uptake curves.
The Bass Model is a mathematical form that models how a new product will spread through a user community.
When modeling the uptake of a new brand, Bass uptake curves often reveal slow adoption because people are hesitant to use a new product until others are doing likewise; in other words, theyre imitators.
An imitator is somebody you might go to restaurant with, they see your mobile phone, theyll take a look at it, ask if they can play with it, ask questions regarding its battery life, et cetera, et cetera, said Peek.
Innovators, on the other hand, are people who will walk by a mobile shop window, like the look of a phone in there, and go in and buy it without asking any questions or trying to play with it.
Generics tend to bring out the innovators in doctors because theyre already familiar with the associated brand and feel comfortable prescribing it on their own.
The Bass Model, therefore, allows you to formulate a rate of innovation for generic erosion and model it through time.
Modeling complexity
Another key to forecasting generic erosion rates accurately is accounting for their complexity.
After all, the arrival of a generic doesnt simply shrink a brands market share. Sometimes, it increases the size of the overall pie.
For instance, when a generic is launched, the attractiveness of the accompanying molecule often increases.
This varies from market to market and depends on the size of the molecule, but the impact on the molecule could be as much as 30 percent, research shows.
Likewise, when a generic launches in a therapy class, that whole therapy class can become more attractive, although Peek warned that class impact is subtle and often overestimated.
Nonetheless, its important to model a series of uptake curves that incorporate competitor launches and changes in clinical profiles, rather than a simple curve with a steady innovation rate.
In the end, Peek said, Youre looking at how to use different uptake curves to capture the essence of where generics are capturing their share from.
For more on generics and all the latest trends in forecasting, join the industrys key players at Pharma Forecasting Excellence Europe on June 14 and 15 in Berlin.