The Federal Trade Commission had an interesting blog post about how they see reference pricing relate to narrow networks. Reference pricing is a wonky shiny object for health care cost control (I think it is a very good idea) where an insurer will pay X for a given condition where normal cost sharing applies to the member. If the member can find providers that will treat that condition for X or less than X, the member pays their normal deductibles and co-insurance. If the member goes to a provider that charges X+Y, the member pays deductible and coinsurance on X and then pays all of Y. The goal is to get members to move to the lower cost providers. And given recent evidence, it works.
Narrow networks work in a similar manner most of the time. A narrow network is usually a lower cost network (I can think of a couple scenarios where an insurer creates a “premium prestige” network) as the providers charge the insurers less per unit of service. If a member goes to an in-network provider, they get the lowest level of member cost sharing. If they go out of network for a PPO, they have to pay a lot, and if they go out of network for an HMO or EPO, they pay for everything. Again, the goal is get people to go to lower cost providers by a pricing mechanism. Are they effectively the same thing?
We believe there is little difference between the price reduction and quality improvement incentives associated with narrow network health plans compared with reference pricing health plans. A health insurer could create a narrow network plan to give providers an incentive to reduce their prices and improve their quality in exchange for inclusion in the network and the increased patient utilization associated with this inclusion. Alternatively, the health insurer could create a reference pricing health plan and set a relatively low reference price that would mimic the incentives and utilization of the narrow network plan. The only difference between the two plans is that, in the former, providers compete in price and quality to be included in the network, while in the latter, providers compete directly for the patients. Lost in the discussion is the possibility that some patients may prefer to delegate the responsibility for selecting low-price, high-quality providers to their insurance company instead of shouldering the burden of evaluating the relative price and quality of various providers.
I think I understand the logic of their economic argument, but I think there is a bit of a practical hole in their conclusion.
So far reference pricing works very well when the service that is referenced price is elective, widely offered, and has a long lead time. For instance Calpers reference prices hip and knee replacements. Most people are fairly indifferent as to whether or not they get a new knee on the first of the following month or the seventeenth. There is time to shop around and given the cost of the operation, there is a strong incentive to call providers and ask what they charge for what could be a $10,000 reference price bill.
Reference pricing has not worked well on low cost but time constrained services. For instance, when my son was having an asthma attack, my priority was to get him to the closest in-network urgent care clinic before it closed so that we could avoid the ER. We were working on an onset-treatment decision time line of forty five minutes. Pulling up an app (if there was such an app) to determine that the in-network urgent care 7 miles from my house was an over reference facility with a $9 surcharge that could be avoided if I went another four miles further north to an in-network below reference prive facility is not feasible. The search costs are fairly high relative to potential savings and the probability of an emergency room visit increased for each marginal mile driven.
Reference pricing fails miserably when the need is non-elective and extremely time constrained even if there is a reasonably amount of regional competition. I would assume emergency room pricing in a reference price product would be traditional fee systems, but the cases where someone enters the ER at an in-network but above the reference price hospital for significant chest pains and is then admitted for monitoring and testing over the next couple of days will lead to very large member responsibility although the best medical advice for chest pains is to get to an ER immediately. The insurer in a reference pricing model will pick up the emergency room visit charges under normal cost sharing, but the stress test, the room fee, the monitoring, the scans are all independent events that are bundled within the hospital stay. Breaking up that bundle so that a cardiac patient can leave Hospital 1 and be admitted to below reference price for the room fee at Hospital 2 and then go to three separate off-site labs for the different tests on Day 2 will be impractical.
Narrow networks simplify and would promote continuity of care in the scenario above. If a person goes to an out of network ER with significant chest pains, the insurer pays while the member pays as if they were at an in-network facility. Once the member stabilizes, most out of network hospitals are more than willing to transfer patients to in-network hospitals. That is a daily occurance, and once a member is at an in-network facility, their care is centralized and coordinated. It is a mass simplification where the hospital choice has been effectively pre-shopped by the member joining the narrow network.
Reference pricing and narrow networks are not contradictory. I could see an insurer offer a narrow network to cover non-elective, reasonably urgent and elective low cost needs while carving out a series of bundled procedures that are subject to reference pricing. If you are ambitious, that insurer could even work down to bundling and reference pricing well visits with PCPs. That would be a doozy of a plumbing and communication job but it could work as well as be very reasonably priced. But I think there are significant practical differences on the usage of reference pricing as I understand it now and narrow networks. Reference pricing is a selective cost control device that utilizes provider competition incentives while narrow networks are a more general cost control device that uses provider competition incentives.
brantl
It works, as long as the individual patient can get to the low cost provider, geographically.
Richard Mayhew
@brantl: Agreed…
The FTC post in a non-quoted part makes a very good point that all the provider price competition schemes fail miserably when there is a monopolistic or at least a very concentrated market for a particular set of providers.
Reference pricing is probably a whole lot easier to run successfully in Seattle than it is in northern Idaho; it is easier to run in Boston than in Downeast Maine etc.
lahke
Great post, Richard, but will you PLEASE learn to spell “separate”? That’s an “a” in the middle. Spells “a rat”. Is it possible your school didn’t make you read the story we all read in 6th grade, designed to fix that in the memory forever?
jl
Thanks for interesting post on reference pricing. Another issue is that Richard does not mention and I think is important is that reference pricing is disruptive. For one thing, it increases price transparency and reduces ability of drug companies and providers to supply efficient quantity of supply through price discrimination. And price discrimination can tax high income or low income depending on what method is used to identify groups that will pay different amounts. Price discrimination eases things for the little person if a third party payer picks up the tax involved in discriminatory pricing.
One of the reasons drug companies and providers hate them some reference pricing is precisely because it reduces their power to impose the price discrimination of their choice that maximizes profits, and at the same time allows maximum market penetration by getting to efficient level of output. That is a very double edged sword.
There will be more ruthless and (to patients) seemingly arbitrary design and enforcement rules about in and out of network services, since any provider will be worrying about how reference pricing rules will force them to average pricing of bundles of services in the future. It is not a magic bullet for the reasons Richard points out, and several more.
I think one of the first effects of transition to reference pricing, or even including it as part of a blended pricing system is that if reveals the problems of concentration and market power in the market for provider services.
Richard Mayhew
@jl: I’m thinking about your model… a perfectly segmented/price discrimantory market for a provider where the provider charges the maximum willingness to pay of each individual consumer that is at or above the marginal price of production seems to me to be a transfer of the consumer surplus to the provider’s private surplus. It fails misterably in a competetive market and works reasonably well in a pure monopoly.
For instance, let’s assume marginal price of 100 for a given Q and a flat price curve in the immediate area of Q. The minimum price of a good sold is $100, but the maximum price to a high willingness to pay buyer could be $200. In a perfectly competitive market with a single price, the high willingness to pay buyer gets a consumer surplus of $100. In a perfectly segmented market, that high willingness to pay buyer is paying $200 for a $100 marginal cost good, so the seller is getting a $100 monopoly transfer profit.
I can see cases where the production cost curve is very slanted, so price discrimination can lead to a higher quantity of goods supplied than a single price, but most price discrimination is a means of a provider capturing buyer and social surplus.
jl
@Richard Mayhew:
” I can see cases where the production cost curve is very slanted, so price discrimination can lead to a higher quantity of goods supplied than a single price, but most price discrimination is a means of a provider capturing buyer and social surplus. ”
I think it will usually do both, though you are right that if the elasticity of supply is very low, most of the effect will be a transfer, a capture of quasi-rents, from payer to provider, rather than changes in quantity supplied.
I think the gist of what I was trying to say is that in markets where providers have market power, the transition to reference pricing will result in some mix of higher average prices and lower quantity being supplied. And this ties into my hobbyhorse that increased competition (with appropriately strong strong market regulation) is key for real long term success of ACA.
If I can find it, I will send an article that explains my thinking.
Edit: I won’t send an article with drug company or medical group or hospital chain execs yelling that reference pricing increases price transparency, will get people killed, and is deeply immoral. You can probably find those on your own!