Bigger carrots and painful sticks to improve medication adherence

As you've probably read by now in the New England Journal of Medicine (Full Coverage for Preventive Medications after Myocardial Infarction), so-called value based insurance design, which waives co-pays for maintenance drugs, resulted in only a modest improvement in medication adherence and failed to significantly improve the primary outcome of the first major cardiovascular event or revascularization.Despite the waived co-pays and study leadership by big machers from Aetna, Harvard, CVS Caremark and the Brigham, medication adherence was still under 50 percent, an improvement of just 4 to 6 percentage points over patients who were faced with co-pays. The researchers' conclusions are as follows:

Despite the improvements in adherence that we observed, overall adherence remained low... Therefore, interventions to address other contributors to nonadherence (e.g., knowledge, attitudes, the complexity of prescribed regimens, and difficulties that patients have in accessing their medications) will be necessary to adequately address this problem.

I see things a little differently.Perhaps the trouble is that rewards for nonadherence under value based insurance design are too low and punishment is entirely absent.  Consider the following alternative study design:

• Pay those who are fully adherent $5000. If that sounds high, keep in mind that these patients incurred about $70,000 in costs on average during the follow-up period
• For those who aren't adherent, provide counseling and warnings, and a reassessment of whether their therapy is optimal. If they still aren't adherent, then cancel their insurance

Of course the second bullet point sounds terrible. But if we're serious about controlling costs shouldn't we at least contemplate punitive measures?