2008 Applied Pharmacoeconomics and Outcomes Research Forum

On Monday I attended the Applied Pharmacoeconomics and Outcomes Research Forum at UCSD.  I will briefly summarize the presentations of the three speakers.

Andreas Pleil, Senior Director Worldwide Medical & Outcomes Research, Pfizer.  Dr. Pliel reviewed the findings of International Society For Pharmacoeconomics and Outcomes Research (ISPOR) Task Force on Real World Data.  “Real World” data is considered to be anything outside of randomized controlled trials (RCTs).  RCTs eliminate the problem of endogeneity bias, but even RCTs have their problems.  They are expensive, and their smaller sample sizes mean that it is difficult to have meaningful sample sizes for different subpopulations.  Further, one must worry about the problem of attrition bias.

One important question is who should pay for data collection?  Should it be the pharmaceutical companies?  These firms have deep pockets but one may worry about biased reporting of evidence.  What about the government?  The government is seen as unbiased, but how much do taxpayers want to spend collecting data.  What about academics?  Academia is a great place to collect real world data but there are only a limited number of academics who may not have time to collect all this data.

Another question to ask concerns the size of the data.  An economist would say that a pharmaceutical company should collect more data if the costs from deploying a pharmaceutical too soon outweigh the potential benefits lost from delaying the drugs introduction.

T. Jeffrey White, Director of Clinical Analytic Strategies, WellPoint NExtRx.  One important aspect of having real world data is that compliance physician instructions is not as high as under a randomized controlled trial.  For instance, WellPoint found the following compliance rates:

  • Schizophrenia: 34%
  • Alzheimers: 45%
  • Anti-depressants: 46%
  • Ace-inhibitors: 57%
  • Plavix: 69%

If drug A works slightly better than drug B in an RCT, a managed care organization may still prefer drug B if compliance rates are higher.

Mark Bounthavong, Pharmacoeconomics Clinical Specialist, VA San Diego Healthcare System.  Dr. Bounthavong spoke about the San Diego VA’s decision to switch from their Alzheimers patients off of Donepezil to Galantamine.  Why did they do this?  Galantamine was found to work just as good or better than Donepezil in RCTs and in the VA’s other offices, and Galantamine’s producers were offering significant discounts.   Further, the VA would get a rebate if Galantamine gained 35% market share.

Unfortunately, after the switch was made, some patients had adverse reactions.  Over 15% of patients switched back to Donepezil (which is higher than the VA’s traditional 10% switch back).  Many patients had GI problems.  Thus, the changeover to Galantamine was made on a case-by-case basis rather than for the entire VA.

One can see that Galantamine’s producers are not stupid.  They can underprice their drug and give incentives for the VA to increase Galantamine’s market share.  After that, they can hike up the price.  Since it is costly to switch patients from one drug to another (in terms of the risk of destabilizing the patient, extra medical visits and extra administrative costs), if Galantamine would gain a large market share, it would have some pricing power.