Sometimes it is difficult to know what to make of data from even the most reputable institutions. This appears to be the case for a recent study from a Harvard research team.
A new study by Nekhlyudov et al. in the Journal of Cancer Survivorship was designed to investigate whether cost-based non-adherence to cancer medications among Medicare enrollees in the USA was similar to or different from cost-based non-adherence to medications for non-cancerous conditions.
Of course, for starters, one is tempted to wonder why one’s clinical condition would make any difference to one’s willingness to sacrifice other potential benefits at the risk of either one’s life or one’s quality of life, but perhaps that is a naïve question. (The current writer is, after all, “only” a heart attack survivor taking four medicines a day in the hope that it will prevent him having another one!)
The authors’ study was based on an analysis of data from the Medicare claims data for 2005 and from the 2005 Medicare Current Beneficiary Survey. By using these data, Nekhlyudov et al. were able to compare patient-reported non-adherence to their medications based on cost (cost-related medication non-adherence or CRN), reductions in expenditure on basic needs (so as to be able to afford medicines), and other cost reduction strategies used by Medicare beneficiaries with and without cancer.
Here are the key findings of their analysis:
- It was based on “a nationally representative sample” of 9,818 non-institutionalized Medicare enrollees.
- 1,392/9,818 enrollees (14 percent) were classified as cancer survivors.
- The cancer survivors were older, more highly educated, more likely to be male and non-Hispanic, and more likely to have multiple co-morbidities, poorer health status, and employer-paid medication coverage than the non-cancer patients.
- 10 percent of cancer survivors and 11 percent of enrollees without cancer reported CRN.
- 6 percent of cancer survivors and 9 percent of enrollees without cancer reported reductions in expenditures on basic needs so as to be able to afford their medications.
- Compared to those who did not report CRN, the 143 cancer survivors who reported CRN
- Had lower income levels (48.6 vs. 62.2 percent, p = 0.11)
- Were more likely to be African-American (13.0 versus 6.4 percent, p = 0.033)
- Were more likely to have non-employer-based medication insurance (p = 0.002)
- In adjusted analyses, CRN among members the two groups was similar, but some subgroup differences could be observed, based on gender and cancer type.
- The application of cost reduction strategies was generally similar among cancer survivors and those without cancer.
There are three important and fundamental questions about a study like this:
- How accurate are the data on which the study was based?
- How accurate is the analysis based on the data used?
- How meaningful are the results to future decisions?
We do not claim to be able to answer the first two of these questions, but we do have some comments of relevance:
- The year 2005 was the last year before the implementation of Medicare Part D (on January 1, 2006), which allowed patients to get Medicare coverage for prescription drugs administered through pharmacies. As a consequence, patients being treated with intravenous drug therapies under Medicare Part B (as part of their medical care) in 2005 were significantly preferentially covered for their drug costs compared to patients who had to pay for their own drugs at the pharmacy.
- 2005 was relatively early in the growing use of oral drugs for cancer therapy, and so it may be the case that a higher percentage of drug costs for outpatient cancer therapy in 2005 were covered under Medicare Part B than the percentage of drug costs for non-cancerous conditions in outpatient settings.
- The year 2005 was perhaps one of the most financially “sound” years in American history (at least at face value), preceding by 2 or 3 years the beginning of the crash in the housing and insurance markets that led to the current global depression.
Regardless of the accuracy of the data on which this study was based, or the quality of the analysis itself, there are serious questions about whether its results have any meaningful implications for drug compliance in the USA over the next 5 to 10 years, as we all struggle to “make ends meet” by comparison with the situation in 2005.
Let us be clear, however, that we are not criticizing the study per se. We are merely noting that any “interpretation” of these study data to suggest that non-institutionalized cancer patients on Medicare today or in the future are just as well able to afford their medications as those with other serious long-term conditions can not be justified by the data in this study. It is a historic analysis of data at a particular economic point in time, and data from such a study cannot reasonably be projected forward when there have been major changes in the underlying economy.