Failure to follow drug therapies and failure of those therapies to be effective is postulated to add billions to annual health spending, although the proof of that is sketchy. It is clearly a quality of care issue. Numerous providers and vendors have offered solutions to improve adherence and effectiveness of medication treatments. A study in the American Journal of Managed Care gives a background analysis of the potential savings from the impact of such interventions. (AJMC Article) The researchers used a commercial database of more than 10 million adult patients for the period 2011 to 2013. They focused on medications used for diabetes, hypertension and high lipid levels. They classified the study group into patients who were adherent or non-adherent to their drug therapies in the first year and/or the second year. The relative medical costs were tracked in each year, particularly for patients who switch from adherent to non-adherent during the two years or vice versa. Adherence was determined by the medication possession ratio, which basically measures whether the patient has enough of the drug on hand to be taking it as prescribed. This definition of adherence has obvious flaws, since patients could be filling prescriptions but still not taking the medicine correctly, and it may not pick up things like pill-splitting. But it probably is close enough to actual adherence to provide a fairly accurate picture of the effects.
The researchers adjusted for a number of factors to try to tease out solely the effect of non-adherence on medical costs. The sample of patients with at least one relevant drug therapy was around 860,000, of whom 61% had only one of the three study diseases and 9% had all three. Patients who went from adherent to non-adherent between the two study years were more likely to be female, live in the south, have high cholesterol, engage in fewer health-seeking behaviors, be new to the medication, be younger and have higher baseline medical costs. Patients who went from non-adherent to adherent were more likely to be male, older, have hypertension, be continuing on a medication and also had higher baseline medical costs. As might be expected, patients who became adherent showed a greater reduction in medical spend savings depending on the number of conditions they had. Patient with all three conditions who became adherent saw as much as $5300 in average annual reduced spending, depending on the diseases they became adherent to treatment for, while those with only one or two of the diseases saw much smaller savings, or in one case, high cholesterol, an increase of $365 per year, probably reflecting the cost of the drug. Patients who became non-adherent showed a similar pattern; those with three conditions had increased spending as great as an average $7946 annually, while those with only one disease had more limited spending increases, in the range of $1045 to $1706. You can also see that becoming non-adherent increases spending more than becoming adherent decreased it. The research not only suggests that adherence behavior can result in fairly significant changes in medical spending, it provides some guidance to targeting interventions to the patients most likely to show a return on the intervention and on the importance not just of getting non-adherent patients to become adherent, but to prevent adherers from slipping. It does not give any guidance on the causes of non-adherence and that is a critical piece to designing interventions that remove barriers, cost and others, to effective use of drug therapies.