November 13, 2006
From the desk of Jane Galt:Adverse selection in the VA?
Mark Kleiman asks a question:
Since the Veterans Administration, since its reform under Bill Clinton, now has the best medical-records system going and produces high-quality health care at a reasonable cost, could we move a baby step toward national health insurance by allowing non-veterans to buy into the VA system at a price equal to whatever the VA figures is its marginal cost? The initial, emotional reaction from veterans' groups might be opposition, but surely having a bigger client base would strengthen the VA system politically, against the moment — coming soon — when we're no longer at war and when, accordingly, treating veterans well starts to lose political saliency saliency.
I can see several problems with the idea. The first is that I don't think the VA system is very scalable. My impression is that the VA system is deepy tied into the university hospital system; academic doctors who, for one reason or another, value a steady paycheck and/or the opportunity to do research over the financial rewards of private practice, often do double duty as VA doctors. This keeps the quality rather higher than it ordinarily would be at VA wages. Separate them from the teaching hospitals, and the quality of care might look at lot more like your average Medicaid clinic than the system that Paul Krugman is raving about; and I don't know how many more teaching hospitals could have a VA facility slapped on top of them, or how many more doctors would be willing to practice under those conditions.
VA hospitals are also able to manage demand to meet their supply through playing with eligibility standards. In effect, my impression is that they do not have enough supply to serve their entire target population (all veterans), so they are able to keep their facilities operating at or near full capacity all the time. This helps keep average cost low. However, it also means that there isn't much slack in the system; so the marginal cost of serving extra patients will be very high, requiring new facilities be built and new staff hired. The new staff, particularly the doctors, may be higher cost or lower quality than the old staff. Nor do I think that, politically, it would be possible to turn down someone with a 30% service-related disability for care while selling care to a lifelong smoker with lung cancer; in order to expand the VA at all, you would first have to open eligibility so that all veterans were served by the system. This would be very expensive.
Also, people who are paying for their care, unlike the veterans who get deeply discounted treatment, are substantially likely to excercise their right of exit; if they get a job that offers good insurance, they'll probably drop out of the system. This means that the much vaunted preventative care savings that the VA hospitals are supposed to realise would suffer in the same way that preventative care allegedly suffers in the private system: the VA will realise substantially fewer savings from good preventative care, making it much more expensive to offer. Its electronic records system also probably wouldn't work so well, since it wouldn't get to start collecting data at 18 or 20.
But there's one criticism made by his commenters that doesn't seem very apt: adverse selection. After all, if the VA is really so much better at taking care of sick people than the rest of private industry, and if the VA really is pricing at marginal cost, then they should be able to attract a very desireable pool. It's only if the savings are fairly trivial that adverse selection becomes a problem--and if the savings are trivial, then it is more likely that the VA is a statistical anomaly than a breakthrough model.
I think the most compelling question about nationalising the VA system is probably scale: lots of things work well in smaller populations that do not gross up well to the national level, like town meetings and search parties. The VA seems to be enjoying some complementary benefits to location and government plant that cannot be replicated cheaply, or at all, nationwide. The American Legion is a powerful pressure group that checks the power of the bureaucracy; it's hard to imagine a nationwide group that could do the same thing, except possibly the AARP, and we all know what getting the AARP involved in entitlements does to cost control. And the right of exit currently functions as a substantial safety valve to the VA; if private hospitals stick around, it will undercut support for the system, and if they don't, the American middle class will mutiny at rationing, which will cause costs to further explode.
Posted by Jane Galt at November 13, 2006 2:52 PM | TrackBack | Technorati inbound linksI also wonder if any of the people raving about VA hospitals have actually ever *been* to one.
My wife's worked at VAs, and she says they are kind of nasty.
"I think the most compelling question about nationalising the VA system is probably scale"
This is true about nationalising health insurance as a whole.
Blogging is so wonderful. No need to cite any sources, just guess at all the possible reasons to dispute something you're ideologicaly opposed too ...
I think a VA like system is viable if the understanding going in is that it will be a system to provide welfare medicine. The goal, in my opinion, is not to replace the existing healthcare system, but to create a cost controlled system for those who cannot afford to pay.
The VA system is going to get lots of extra use and concomitant scrutiny with the number of wounded from Iraq, too. That should be interesting.
Here's an merely anecdotal comparison - I used to visit hospitals in Atlanta as a good work (I was *such* a nice boy). The VA was the next closest hospital to Emory's own complex and I went there a lot (once a month at least). Unless things have changed remarkably since the late 1990s I don't want to ever have to stay in a VA hospital. I'd have rather gone to Grady.
Vish-
What would you like us to say? Like Mr. Tinkler, I've been in a VA hospital. I've also been graced with 32 years of single payer medicine. Would you prefer I discuss the sergeant whose surgeon replaced the wrong knee? Or the disability rating that took three days to confirm and twelve months to receive because the paperwork was lost four times? Or the cancer that wasn't caught until it became inoperable? Or the sergeant who was cleared for deployment despite being treated for headaches, who subsequently died on deployment of an undetected aneurism?
I believe Ms McArdle is exactly right; whatever system is proposed, it cannot be a step backward. The VA will have to progress much farther than it has before it even reaches minimum acceptability.
Will the AMA remove its institutional roadblocks to the number of doctors trained each year? Do we want them to? Will the pharmaceutical companies be allowed to sell their products at a profit so that they can continue the research they need to make better products? If you have the answers to these questions, Vish, I'd love to hear them.
Phooey. A baby step in the wrong direction is still a step in the wrong direction.
"Since the Veterans Administration, since its reform under Bill Clinton, now has the best medical-records system going and produces high-quality health care at a reasonable cost..."
The premises are wrong.
1. The VA's electronic medical record system was created in the mid-1980s, not in the 1990s under Clinton. In the early 1990s (pre-Clinton), the VA added a GUI shell on top of its electronic medical record system that made it easier to use and provided better display of information.
2. The VA's EMR has had no substantial improvements in 14 years. It is far behind a number of commercial products. It's laboratory information handling is a disaster: it's 20 years behind current standards.
3. The VA does not provide high-quality healthcare at a reasonable cost. It provides medium-quality healthcare at a higher-than-necessary cost. The amount of waste and inefficiency in the VA system is greater than in most non-government hospital networks.
In the late 1990s, the Veterans Health Administration was publicly chastised because it provided variable-quality (terrible to good), high-cost care. VHA administrators took this to heart and determined to make the VHA into the best healthcare system in America. Did they kick out bad administrators, lazy employees, money-wasting managers? Did they recruit top hospital adminstrators and top-notch physicians? No to both. They looked at current indicators of quality from the Joint Commission for the Accreditation of Healthcare Organizations and made damn sure the VHA would do well on those indicators. Even if the indicators were dumb. Even if they were behind the times. Even if there were more important things to address to improve care. The result is that the VHA has an undeserved reputation as the best healthcare system in the nation.
I am a physician division director in a large VA hospital, and I see examples of waste, inefficiency, and poor care every day. VA hospitals are no better (but also no worse) than most private hospitals. Emulating the VHA will not be the panacea that Mark Kleiman or Paul Krugman claim.
Thank you for your service, Dr. T. I only have to put up with the frustrations of being a customer. You see the same frustrations without being able to change the bureaucracy that protects them. Your sacrifice is much greater than mine; I salute you.
I have worked in VA hospitals and in other hospitals. VA standard of care is poor.
Many other mental health care facilities refuse to intake patients who have come through the VA system, with the rationale that they have been too screwed up by the poor mental health care to recover under normal standards of care. My anecdotal observation only: use with caution.
I never understood what the underlying premise was that created the VA Health care system being distinct from the rest of the Health Care system.
What is the purpose of maintaining the VA outside of the rest of the system?
One thing on the medical treatment the patient often ignore is the risk involved with medical treatment: the risk mentioned here in the process of the medical service, not the risk that one would get a disease. As someone pointed out, when waiting to be operated, one always askes: when can i get cured, and how long can it take, but never ask "Doctor, what type of policy you would take to avoid the infection when operating one me?"
However, this problem can never be addressed only by the doctor nor the patients, nor both. It is like that one went to Subway for sandwich, and how many would ask the server whether s/he in the Subway to wash his/her hand? And how many
servers would be "nice" all the way?
The same applies to medical service system.Only giving the money to the patients in their medical insurance support would not solve the problem. In addition, as mentioned previously, the risk problem more prominent now in medical system is not just to control the would-diseases, but the operational beaucracy/resource allocation inefficiency.
Adverse selection, moral hazard...... the problems all exist, but not the starting points to solve the problem. coz either side of the agents within the group could perfectly control the problem, if no high criteria of ethnics are assumed.
There must a third party to be introduced.
And I feel quite interested in how VA manage their medical technology and R&D. Can anyone here provide any help for a few detailed information?
Sale: exactly. The Va system employs some 250,000 people.
The US medical system as a whole employs 13.5 million.
Does anyone seriously think that 13.5 million people can be managed without the use of markets to create incentives?
Add me to the list of people who question the very premise of this question. I have no medical knowledge of any kind, and I haven't looked into the question at all, but I just cannot believe that the VA is our nation's finest healthcare system.
It's going to take a lot more than a couple articles in the Washington Monthly to convince me of THAT. Most of my other interaction with government agencies has been less than satisfactory.
I could be persuaded that the VA has improved -- my local post office and DMV certainly have made great strides, they are now "acceptable." But the idea that it is now the BEST is almost impossible to swallow, I've never run into a government agency that can be described as the best at anything other than wasting time and money.
If we're going to add folks to the VA system, shouldn't they pay average cost, not marginal cost? After all, they are going to require more of the fixed assets.
My wife interned in a VA hospital, and thought it was horrible, the kind of place you hope you never end up in.
This was about twelve years ago. Perhaps things are much nicer now. But I wouldn't bet on it just because of a couple of articles or studies saying so.
VA facilities are only accessible to urban areas and a few rural areas.
And the campaign to paint these facilities as nirvana is way off the mark.
selection biases ...
- only healthy patients enter the system (those fit enought to be soldiers in the first place).
- there are no children in the system (children are above average consumers of health care).
- the majority of patients are men who are notoriously going-to-the-doctor adverse.
- men take up less of the health care system than women (fewer annual exams, shorter life span, no pregnancy, child-rearing issues, etc)
- non-battle injured patients are healthier. they have been given the training for a healthy, self-discplined, exercising lifestyle.
The VA was originally established to provide care to indigent veterans and a few service connected individuals (think POWs, spinal cord injury and post traumatic stress disorder, for examples). It had no (was forbidden to provide) outpatient care.
Over the years, much of this changed. Outpatient care was added because it was easier to prevent further hospitalizations. Mr Clinton permitted any veteran to get prescriptions filled at VAs with only a minimum copay ($7 per month, capped at $70 per person)
City VAs were employment bonanzas for House Representatives, but ancillary services were generally bad (janitorial, lab, nursing). Those associated with University Centers usually gave relatively good medical care. Rural VAs usually had better ancillary staff, but less knowledgible physicians. Recent changes have improved medical knowlege (care) at outlying hospitals. Staffing changes in ancillary services have generally improved in urban settings. There are still problems, however.
The VA has been a pioneer in the use of computerized medical records. This was a necessity. If you had ever seen a VA record room with its piles of unfiled lab results you would understand. The computer has essentially eliminated this problem. The VA in town has a computer system which far outstripped anything in the private sector here when it was installed. Others have come up to it since then, but not surpassed it.
Most university center VA's supply medical care about as good as is available anywhere. They do so in relatively rundown facilities which are occasionally dirty (think Kansas City several years ago)Rural VAs probably supply medical care equivalent to a medium rank private facility, but in better maintained facilities.
Overall, the VA has several problems. They haven't figured out how to do construction in a timely manner. Good administrators have good hospitals, but they have some trouble recognizing bad administrators. They ration healthcare in minor and major ways. ED drugs are written for once a week maximally. A single ACE inhibitor is generally available. Other ACE inhibitors can be written but may not be filled.
Overall, I don't think that the VA experience is scaleable to a National Health care plan. It relies too heavily on University medical expertise. Removing or diluting that would likely remove much of the "advantage" the VA has over private systems.
hmp3,
Thank you for the answers to the questions I was posing. Ultimately, I never understood why the VA system could not be accomplished through our existing Hospital networks. I would think that by so doing, it would improve both systems. As well as remind many of some of the truer costs of War. Though, that is probably the other side of the "political" calculation that created a distinct VA system, in the first place.
The VA is incredibly slow to offer outpatient care. A nephew who is VA eligible (combat veteran of Afghanistan) was having severe headaches on a regular basis. He called the local VA and was given an appointment 8 weeks later. On the day of the appointment, he was called and told it was cancelled, and the next appointment was 10 weeks later. He gave up and went to a private physician and paid out of pocket. Canada, anyone?
Use the VA model to fix Medicare (it can't get any worse) and then we can talk about national single payer.
Regarding records: The best record system is that used in regular military medecine (active duty, dependants and retirees). Not because it is so great, but because the entire rest of the country's is so bizarre. I was in shock after leaving the service and realizing that every new doctor, dentist, moved town whatever, meant that a duplicate record would be generated. It's not the waste of the duplication, but the loss of information. In military medecine, a doc (or you, yourself) can flip through your jacket and see all the care you've received for anything. That way trends and correlations can be determined. And this has little to do with electronics, just the idea of having a single jacket and taking your record from duty station to duty station.
Comments are Closed.