• Further adventures in cost-sharing: An addendum via email

    This post is jointly authored by Austin and Aaron.

    The two of us puzzled over the findings of the paper on Mayo Clinic cost-sharing Aaron posted on earlier this morning. (Go read that post first, if you haven’t already.) We exchanged emails trying to get at why primary care visits were not responsive to a decrease in cost sharing, but specialist visits and imaging utilization were responsive to an increase in cost sharing. Here are lightly edited excerpts from our exchange:

    Aaron:  I think it’s that people WANT to see specialists. They see an immediate good in it. They get a drug, or a test, or the feeling of solving a problem.

    With primary care, there’s no immediate return. You may hear something you weren’t looking for. They may tell you to lose weight or exercise or stop smoking. If you’re healthy, there’s no immediate return. So you don’t WANT it. Making it cheaper may not matter.

    Austin: That suggests something more though. Making primary care visits free at the point of service may not be enough. Maybe one has to pay people to do it. In fact, that’s essentially what the gatekeeper functions of HMOs do. They provide an additional incentive (payment, though not in dollars) to see the PC doc. You need it for the referral!

    Retail clinics can also reduce the cost (in time, if not in dollars). The time factor may be relevant to a lot of folks.

    Bottom line: There must be reasons why, according to the Mayo cost-sharing study, people don’t increase primary care visits when copays go down but they do decrease specialty use when copays go up. Understanding those reasons is crucial for building on the study and applying the results for policy. If primary care visits are important and if people are not receiving enough primary care, then the study suggests we need to do more than just lower copays. We may actually have to pay for people to use more primary care. That payment need not be in dollars. Other incentives may work too.

    • http://www.rand.org/pubs/external_publications/EP20110048.html
      –in above RAND, introduction of CDHP with high deductibles dropped overall service use, BUT, with preventative services fully covered, there was a corresponding drop there as well.  

    • Some thoughts

      Paying most people won’t work for the same reason eliminating the cost did not work. It is internalized as a relatively small gain for a service people aren’t exactly dying to consume.

      The study examined Mayo clinic employees, the majority of which would probably not change behavior for an extra $10 in their pocket when it comes to a PC visit – they are not exactly buying groceries.

      Speciality visits on the other hand start to cause some pain. They are taking $25 plus co-insurance out of their pocket – perhaps $15 to $30 more depending on what is done in the visit (it would help to get some actual costs) – which is felt as roughly 2x the final dollar amount.

      So you are looking at $10 gain vs. $80 to $110 loss – accounting for loss aversion. I don’t understand why this is surprising in the population studied.

      Further, I suspect people’s marginal rate of substitution of perceived qualtity (PCP) for time is huge.

      I’d suggest adding a surcharge to any patient who goes directly to a specialist visit similar to the surcharge on out-of-network visits in a PPO. The surcharge will be internalized as a loss to those who forgoe PC visit first, i.e hurt more.

      Notice I said goes directly to a specialist, a referral is not required from the PC visit. That has caused lots of problems in the past.

      It does nothing to eliminate choice should the patient want to go directly to specialist (at extra cost) or go without a referral, and if desired, provides the patient with an informed agent (the PCP) for specialist selection and to help counterbalance the specialists information advantage.

    • Speaking as an N of one I tend to want to see a specialist when I think the medical issue I’m concerned with is serious/urgent/obviously-wrong and I head for the GP when I’ve got something relatively routine to deal with that I’m not particularly worried about.

      Trivial Explanation:
      -Price sensitivity declines as anxiety/pain/symptom-specifity increase.

      This conjecture seems to fit the data pretty well. Is there something I’m missing?

    • Another N of one–except that I’m also a provider (mental health, not physical) and hear about doctor visits from my clients. And what I hear a lot is that primary-care physicians have incredibly busy schedules, spend maybe fifteen minutes with you, ask about the problem you came in for, check briefly, and are gone (after leaving a prescription if you’re lucky). Specialists, on the other hand, have an hour or so allotted to each patient and–contrary to many images–go over everything that might be relevant.

      So people are happier with seeing specialists because they perceive that they are getting more “bang for the buck” and/or more personal attention.

      This is not, of course, universal. My primary-care doc tends to spend an hour or so per patient even for routine visits; if she’s got a lot of people booked, she rushes between rooms checking in with one, then another, but somehow it works out. She’s always behind schedule–at the end of the day, my appointment may happen two hours after it was scheduled–but I wouldn’t consider changing.

      For what it’s worth…

    • Like most behavioral issues in health the key issue is simple. Since you can’t trade health directly for money, discount rates are not driven down to market interest rates and the operational discount rates for most people are around 20% or more, this makes them extremely impatient. And in turn it affects how people value specialist vs. PCP visits.

      Behavioral issue #2 is loss aversion, so expected utility is not linear.

      I think those two issues together explain your bottom line.