Wrong Way for Consumer-Driven Health Care?
Gary Claxton and colleagues, of the Kaiser Family Foundation, have written a concise analysis of the evolution in health payments from 2004 through 2015:
From 2004 to 2014, the average payments by enrollees towards deductibles rose 256% from $99 to $353, and the average payments towards coinsurance rose 107%, from $117 to $242, while average payments for copays fell by 26%, from $206 to $152. Overall, patient cost-sharing rose by 77%, from an average of $422 in 2004 to $747 in 2014. During that period, average payments by health plans rose 58%, from $2,748 to $4,354. This reflects a modest decline in the average generosity of insurance – large employer plans covered 86.7% of covered medical expenses on average in 2004, decreasing to 85.3% in 2014. Worker’s wages, meanwhile, rose by 32% from 2004 to 2014.
I would quibble with Claxton, et al’s use of the noun “generosity” to describe the share of health costs paid by insurers. Insurers pass costs through: Claims they pay are covered by premiums, which are charged to either beneficiaries or employers. If the latter, beneficiaries pay through lost wages. Plus, because claims processed and paid by insurers add administrative costs (“load”) to the costs of actual medical care, total health costs are higher. Quibbling aside, the analysis gives great insight into how the way we pay for health care has changed.
Certainly, as written, it looks like patients are bearing a much higher share of the costs directly. However, patients paid 14.6 percent of costs directly in 2014, versus 13.3 percent in 2004. There has not really been a big shift of costs directly on to patients. So-called “consumer-driven health care,” whereby patients dis-intermediate insurers and respond directly to prices presented by providers has not really taken root.
Further, while there has been significant growth in patients’ direct payments, it is increasingly malformed. In 2015, deductibles comprised 47 percent of patients’ direct payments, versus only 23 percent in 2004. However, deductibles are the crudest and least effective way for patients to increase the share of health spending they control directly. Deductibles are determined by the calendar year. They are a characteristic of what I have described as the (very flawed) heliocentric doctrine of health insurance.
Deductibles do almost nothing to contain costs incurred by the small share of very sick patients who account for most health costs. More sophisticated methods of cost sharing are required, in which all patients participate in price formation. That is the only way to get costs under control, and it looks like our system is incapable of letting such methods arise.
“Deductibles do almost nothing to contain costs”
True when you are looking at medical costs.
Not true when you are looking at administrative costs.
Thank you. What do you think deductibles do to administrative costs? (And by “administrative costs” do you mean claims processing, or other costs too?)
John, I was thinking of the cost of admin functions that are (1) “downstream” from the mostly-automated, adjudication of claims and (2) not automated to the same extent as adjudication is.
Data is captured and maintained for each claim when submitted, even if the deductible has not been satisfied and no payment is issued. Downstream effects on admin cost related to payments arise from managing, e.g., reconciliation of unmatched “issued” vs “recorded” payments; banking activities; payment audits; escheat of uncashed benefit payments; offline payment credits/debits; interaction with stop-loss insurers; payment appeals; compliance with regulatory reporting; even issuance of paper checks for some payees, and a few others. Lesser, or no expense is incurred for these activities when no payment is issued.
The impact on admin expense resulting from the deductible is Reflected in the overall admin expense which in turn is the smallest component of the total cost of insurance. But it does exist, and in absolute terms can amount to a meaningful amount of money for a large insurer.
Thank you but I am still not sure if you think the existence of a deductible increases or decreases administrative costs, independent of other factors.
“ugly and expensive scene that John F describes, where great effort goes into tracking the deductible on all illnesses”
Neither said nor described any such thing. I said this:
“The impact on admin expense resulting from the deductible is reflected in the overall admin expense which in turn is the smallest component of the total cost of insurance”
“independent of other factors.”
I’m not sure how one would separate them. Either a plan has a deductible or it doesn’t.
In either case, capture of the relevant information for every incoming claim, and calculation of the plan benefit, are highly automated; these functions are the same whether there is a deductible or not. There is no expense difference.
But the presence of a deductible means there will be claims for which no payment need be issued. Those no-payment claims do not require the downstream processing that payments receive. This creates a small difference in expense.
How much of the cost is processing the claim versus cutting the payment? The problem with deductibles is the claim still has to be processed because the provider does not know if there is a deductible or not.
Yes of course. That’s why I have been distinguishing “downstream” functions from adjudication itself. The relevant information is captured and the plan benefit is calculated for every claim; these functions are highly automated and occur whether there is a deductible or not. The presence of a deductible does not result in any expense difference among claims at the adjudication stage.
It’s the downstream functions where differences emerge (for all zero-pay claims, in fact, not just claims which had not met the deductible).
At one time, I could have told you the approximate percentage difference in cost of the downstream functions but I’ve been retired long enough I just don’t remember. But it’s small relative to the overall admin cost – which itself is a small part of the overall cost of insurance.
However, the insurer does not always know if the patient has met the deductible, because of lag in claims submission by providers. Maybe that has been solved by e-claims?
John, in the companies where I’ve worked, claims were applied against the deductible at the time they were adjudicated, i.e., when “reported” to the insurer. That way, the insurer always knows the current amount accumulated toward everyone’s deductible. I think all health insurers work this way.
The data captured from the claim in the adjudication stage includes date of service. The insurer of course knows the payment issue date for each claim. The claim-weighted time span for all claims between date of service and date of payment issue helps the insurer figure its incurred but unreported liability at any point in time e.g., quarterly and year end.
Yes, however (I understand) insurers have struggled to get physicians to submit claims in a timely way. Perhaps this has changed.
Dr Goodman made the point a couple of years ago that an ideal policy would have no deductible for the kind of care that treated or prevented a serious illness, vs a high deductible for more discretionary care.
Instead, we have the ugly and expensive scene that John F describes, where great effort goes into tracking the deductible on all illnesses.