The RAND researchers we reviewed before have come out with a follow-up study in Health Affairs of the potential impact of consumer driven plans on the American health care system. This is “Growth Of Consumer-Directed Health Plans To One-Half Of All Employer-Sponsored Insurance Could Save $57 Billion Annually,” by Amelia M. Haviland, M. Susan Marquis, Roland D. McDevitt, and Neeraj Sood.
Most of the media reports have been reasonably accurate, see The Hill and The Washington Post, but have missed the real potential in the study.
The media reports have focused on the study’s conclusion that if half the people with employer-based plans were in a consumer driven plan, the system-wide savings would be $57.1 billion. But this is a mid-range estimate that assumes an equal mix of HRA and HSA approaches. The study acknowledges that HSAs are far more cost-effective, and estimates that, if all of these people were in HSA plans the annual savings would be $73.6 billion.
Now that is a pretty big chunk of change, but even it likely underestimates the impact. The study’s authors write:
This estimate was based on cost reductions in the first year of consumer- directed plan enrollment and did not assume any reduction in cost trends for these enrollees.
First year savings are the least of it. The real value of consumer driven approaches is that trend is reduced and the savings mount up over time. The 2009 study by the American Academy of Actuaries, for one, found that the trend over time for CDHPs ranged from 12% to 17% lower than for traditional plans. (By the way, this new RAND study is one of the very few, if any, to cite the AAA study as a source.)
This difference in the experience of HRAs and HSAs is particularly important because this study relies on data from 59 large employers from 2003 to 2007. HSAs were signed into law in December 2003, and didn’t really go into effect until 2005. So the HSA experience studied here must have been very limited.
Importantly these savings do not accrue solely to employers. The study looks at out-of-pocket costs as well as premiums, and concludes that families themselves reduced their costs by over 20%. The authors write:
Savings derived from switching to consumer directed health plans could benefit both employers and employees. Inflation-adjusted health care costs have increased more rapidly than real wages in each of the past six decades. Reducing health care costs can boost output, income, and employment growth in US industries that provide employer-sponsored insurance to a large percentage of their workers.
So the savings are real and substantial, but what are the problem areas? The authors looked at several.
Are these savings due to selection? No, the authors controlled for that. They write:
Consumer-directed enrollees in these large employer plans spent less on health care than enrollees in traditional plans even in the year prior to their enrollment, and the estimates above controlled for this favorable selection into the consumer-directed plans.
Are “vulnerable populations” (high risk and/or low income) disadvantaged? They don’t seem to be. While “nonvulnerable” families decreased spending by 20.9%, those with low incomes reduced spending by 17.3% and those with chronic conditions reduced it by 14.7%.
Do people cut back on necessary care? Here it is hard to tell. Inpatient care dropped by 22.1%, outpatient by 18.2%, and prescription drugs by 16.0%. The drop in preventive care was much less, ranging from 2.7% to 4.7% for six different procedures. The authors note that prevention was fully covered in all these cases, so:
… Ongoing communication with plan enrollees may be necessary to improve their understanding and use of preventive service benefits.
One objective of consumer-directed plans is to encourage greater consumer shopping for health care. Our findings that reductions in spending occur through lower spending per episode, more use of generic versus brand-name drugs, less use of specialists, and lower inpatient hospitalization suggest that these plans do induce changes in treatment choices and not just access. Further research is required to determine whether these are appropriate changes, and our findings concerning preventive care demonstrate that more information is needed to improve consumer decision-making.
Once again, these authors have done some excellent research and we hope they continue this work. In particular, it would be very useful to find out more about the effects over time (how is trend affected?) and how effective are various forms of employee and patient education, support services, and access to information.
One of our main contentions has always been that people will navigate the system better as they become more accustomed to it. It would be useful to test that theory in real-world conditions.
Perhaps most importantly, this kind of quality research may persuade the so-called “research community” to finally put aside their partisan opposition and begin to take this trend seriously. It is the future of health care.