Medical Malpractice and the Courts

--by Sebastian

Useful data on medical malpractice costs in the court system is notoriously difficult to come by.  Much of the time the researchers totally ignore defense costs--which leaves an analysis of 'payouts' woefully inadequate because it assumes that a suit which doesn't lead to a payout doesn't have a cost to the doctor.  Other analyses leave you wondering if the costs are paying people who are injured or just anyone who sues.  The Harvard School of Public Health has just put out an interesting study which attempts to address some of these problems.  I am working off their direct press release (I'm not dedicated enough to spend $100 subscribing to the NEJM but if anyone finds more of the direct data please let me know) and while your mileage may vary about the conclusions offered with the data, the data itself is worth talking about. 

The factual findings are as follows:

The authors reviewed 1,452 closed claims from five malpractice insurance companies across the country. They focused on four clinical categories: surgery, obstetrics, medication and missed or delayed diagnosis, areas that collectively account for about 80% of all malpractice claims filed in the U.S. Specialist physicians in each of these clinical areas reviewed the claims and the associated medical records to determine whether the plaintiff had sustained an injury from care. If an injury had occurred, the physicians judged how likely it was to have been due to medical error.

The reviewers found that almost all of the claims involved a treatment-related injury. More than 90% involved a physical injury, which was generally severe (80% resulted in significant or major disability and 26% resulted in death). The reviewers judged that 63% of the injuries were due to error. The remaining 37% lacked evidence of error, although some were close calls.

Most claims (72%) that did not involve error did not receive compensation. When they did, the payments were lower, on average, than payments for claims that did involve error ($313,205 vs. $521,560). Among claims that involved error, 73% received compensation. “Overall, the malpractice system appears to be getting it right about three quarters of the time,” said Studdert. “That’s far from a perfect record, but it’s not bad, especially considering that questions of error and negligence can be complex.” The 27% of cases with outcomes that didn’t match their merit included claims that went unpaid even though the injury was caused by an error (16%); claims that were paid but did not involve error (10%); and claims that were paid but did not appear to involve a treatment-related injury (0.4%).

However, the study did not paint a uniformly positive picture of the current malpractice system. [--ed. That is uniformly positive?] The costs of litigating claims, including defense costs and contingency fees paid to plaintiffs’ lawyers, averaged $52,521 per claim. Overall, these administrative costs amounted to 54% of the compensation paid to plaintiffs. “Deciding negligence is a very expensive process,” said Studdert. The authors also found that it took an average of five years from injury to resolution of the claim—a long time for plaintiffs to wait for compensation and for defendants to endure the uncertainty that litigation entails.

Finally, the authors found that the claims that did not involve errors absorbed a relatively small piece of the costs of compensation. Eliminating those claims would decrease the system’s compensation and administrative costs by no more than 13% to 16%. “Many of the current tort reform initiatives, such as caps on noneconomic damages, are motivated by a perception that ‘jackpot’ awards in frivolous suits are draining the system,” explained Michelle Mello, an associate professor of health policy and law at HSPH and a co-author of the study. “But nearly 80% of the administrative costs of the malpractice system are tied to resolving claims that have merit. Finding ways to streamline the lengthy and costly processing of meritorious claims should be in the bullseye of reform efforts.”

I was struck by the following: 

37% lacked evidence of error.  I don't really understand what they mean by "some were close calls" when you use such strong language as "lacked evidence of error".  It may be clearer in the underlying paper, but I'm taking it to mean that when the case was looked at by medical professionals it definitely was not malpractice. 

If I were a doctor I wouldn't be thrilled that almost 40% of cases brought lacked evidence of error. 

Of the claims that did not involve error, 28%  were forced to pay compensation anyway.  That compensation averaged $313,205.

If I were a doctor I wouldn't be thrilled that almost 10% (0.37*0.28) of all malpractice cases were both without error and involved paying compensation averaging $313,205.  That provides an average expected cost per case (in compensation to the plaintiff) for doctors who did not commit an error of $87,697. 

The report states that 72% of claims without error did not receive compensation.  Since defense costs are not zero on such cases, it is clear that they are not taking defense costs into consideration in "compensation". 

" The costs of litigating claims, including defense costs and contingency fees paid to plaintiffs’ lawyers, averaged $52,521 per claim. Overall, these administrative costs amounted to 54% of the compensation paid to plaintiffs. "

This is a deeply odd statistic as presented.  Under normal legal accounting, the "contingency fees paid to plaintiffs' lawyers" piece would be a subset of "compensation paid to plaintiff".  Here they are treating it as if it isn't part of that.  I can't make that jibe with other statistics.  Contigency fees before trial are rarely below 20% and if the case goes to trial they tend toward 30%.  Assuming no case went to trial the average contingency fee for the cases where there was other than "no evidence of error" would be ($521,560*0.2) $112,712.  The average contingency fee for cases with no evidence of error would be ($313,205*0.2*0.28) $17,539.  [The 0.28 term represents the fact that only 28% of the no error cases received compensation].  Therefore the average contingency fee per case should be (112,712*.63+17,539*.37) $77,498.  This is greater than the reported average administration cost and we haven't even added in defense costs yet.  So as I say, the statistic as reported does not make sense to me.  [If anyone gets better access to the underlying statistics please let me know].  As such I am reluctant to rely on the statistics which are derived in their costs section. 

Defense costs are clearly not zero even in the lack of error cases.  But even without them the average expected payout for non-error cases (which is incidentally how an insurance company looks at the question) is more than $80,000 per case.  That doesn't strike me as a great outcome. 

Now just because there is a problem doesn't mean that there is a governmental remedy to the problem or that the commonly proposed solution really is a remedy to the problem.  The economic cap proposals which are most regularly touted as tort reform do not particularly address the problem of being forced to defend and pay out on cases where no medical error took place.  I suspect a large part of the problem lies not with the plaintiff's attorney, but rather with a judge who fails to quickly dispense of a frivolous case or fails to finally dismiss a case if at the end the evidence reveals that there was no error.  It may be that the costs in collecting all the right evidence cannot be easily avoided.  But when that process comes to an end, a judge should be willing to dismiss.  In practice it seems to me that judges are reluctant to dismiss cases even when the evidence is clear.  I suspect the answer is to give judges some incentive to dismiss cases or some mechanism to make it easier to dismiss cases when the scientific evidence is clear.  But I can't off-hand think of a great way to do that without endangering the plaintiff's cases where there is evidence of error.  I wonder of some sort of third-party screening system could be used instead of the battle of bought malpractice experts which currently dominate the court system.  Anyway, food for thought. 

Further interesting questions that weren't addressed:

When there is "evidence of error" was it tracked beyond a mere threshold level?  It might be interesting to divide such cases into broad categories like "minimal evidence of error mitigated by substantial evidence of non-error" "substantial evidence in both directions" and "substantial evidence of error".  My intuition is that the last two categories would have significantly higher pay-outs than the first, while the first  two would be the largest percentage of the cases.  This intuition is somewhat tempered by the highish number of non-error cases which nevertheless seem to get a substantial payout. 

In the non-error cases what does the breakdown look like per injury type?  Are medical outcomes resulting in death more likely to cause payouts in non-error cases than other types?  If so does this represent an illegitimate emotional factor among juries that might be remedied somehow? 

In error cases that do not recieve compensation, what is the injury profile like? 

How do defense costs play out over the course of the case?  Could we minimize costs by emphasizing or streamlining certain segments in the life of a lawsuit? 

UPDATE--

I have received a copy of the underlying study.  I haven't comprehensively analyzed it (more to come), but I checked out the administrative expenditures because they are A) easy to analyze and B) made absolutely no sense to me as reported in the press release. 

The press release appears to conflict with the underlying report as to:

The costs of litigating claims, including defense costs and contingency fees paid to plaintiffs’ lawyers, averaged $52,521 per claim. Overall, these administrative costs amounted to 54% of the compensation paid to plaintiffs.

This statement made no sense to me because I couldn't imagine that average defense costs were much lower on average than $40,000 (as would be strongly implied if defense plus plaintiff costs averaged $52,000) and given the other numbers $52,000 per claim can't amount to 54% of the compensation paid to the plaintiffs.

Table 3 (page 2031) Entitled "Apportionment of Total Expenditures between Claims Involving Error and Those Not Involving Error" shows 1441 claims with $72,617,594 allocated to defense costs and $131,765,574 allocated to plaintiff's costs.  This gives a rough average defense cost per case of $50,393.  This is much more in line with what I would expect. 

Comments

[liberal japonicus]

I believe that you are taking 'evidence of error' as indicating whether or not error exists, but I think in this case, it only refers to the original medical documents, and not what comes out in trial under cross examination. For some discussion of error, this chapter might be of interest. This report seems to define errors slightly differently in that close calls are grouped as a subset of errors rather than being excluded.

[hilzoy]

I have mailed Seb a copy of the paper, and am in the process of reading it. About "evidence of error": this is language from the press release, and the actual paper makes things a lot clearer. They graded the claims they looked at on a scale of 1 ("little or no evidence that an adverse outcome resulted from one or more errors") to 6 ("virtually certain evidence that an adverse outcome resulted from one or more errors".) Claims that got a score of 4 or higher were counted as due to error. "Close calls" were claims getting a score of 3 or 4; the claim that non-error cases included close calls presumably refers to the existence of claims getting a 3.

The reviewers gave 44% of claims either a 1 or a 6 (meaning: they were very, very confident about whether or not there was an error.) 30% got a 2 or a 5 (moderate confidence.) 23% were close calls (3 or 4.) The percentage of claims that got compensation increased as the score went up: 19% of 1s were compensated, 32% of 2s, 52% of 3s, and all the way up to 84% of 6s.

I will now try to work out the accounting question.

[Steve]

Hilzoy, as usual, has shed much more light on the subject. Is it a matter for concern that 19% of "1s" were compensated? Most definitely. Is it a matter for concern that only 84% of "6s" were compensated? Absolutely, but you sure wouldn't know that from the "tort reform" agenda!

Somehow, ensuring that wrongfully injured persons get compensated is not one of the goals of those who supposedly seek to improve our medical liability system. I wonder why that might be?

[Steve]

Your argument turned this

The reviewers judged that 63% of the injuries were due to error. The remaining 37% lacked evidence of error, although some were close calls.

into this

I don't really understand what they mean by "some were close calls" when you use such strong language as "lacked evidence of error". It may be clearer in the underlying paper, but I'm taking it to mean that when the case was looked at by medical professionals it definitely was not malpractice.

You can't rewrite the paper just because you don't understand it. Much of your response is attributable to the belief that 37% of the claims were out-and-out frivolous. The judgment of one reviewer in a study, particularly where their judgment is that error is a "close call," isn't nearly enough to leap to that conclusion. There's room for differences of opinion which is why not every case includes one honest expert and one lying hack.

You also rewrite the study to argue that 28% of the doctors who did not commit error were "forced" to pay compensation. As a defense lawyer, I often advise my clients to settle cases even where I feel they're right on the merits, because liability is a judgment call and you have to evaluate the degree of risk. Even though this study may make it appear as though an error/non-error decision can be made in every case, the fact is that Almighty God does not come down from Heaven to divide my cases up into two neat piles. An insurance company decides to settle because it believes there may be a finding of liability, not because it's obvious there was no error yet it feels like giving money away.

Perhaps the most remarkable thing about your analysis is that you devote a great deal of discussion to the 10% of claims that are paid despite the absence of an error, yet barely a mention of the 16% of claims that are unpaid notwithstanding the existence of an error, other than to wonder what the "injury profiles" look like. In the former set, even if we consider these costs to be completely unjustified, at a minimum the costs are spread among a large pool of insureds. Whereas the latter 16% of claims involve victims of medical error who are left with no compensation whatsoever for their injuries. Apparently that doesn't even merit mention as a social problem!

Several states already have third-party screening systems in place, where a malpractice claim is reviewed at an early stage by some combination of independent doctors, lawyers, and judges. This is certainly a reasonable tort reform idea, different from many others in that it actually aims at reducing the volume of meritless claims rather than simply reducing the volume of claims across the board. However, it would be interesting to see a similar study conducted in states where such a panel exists; I don't know what percentage of "non-error cases" would receive compensation, but I can guarantee it would still be quite far from zero!

[LizardBreath]

I agree with Steve. As further support for his point, the press release you link characterize the claims you're calling 'without error' as follows: "Although nearly one third of claims lacked clear-cut evidence of medical error..." That's not a statement that the reviewers are certain that there was no error, nor is it a statement that there was no evidence of error. It's a statement that the reviewers don't consider the evidence there was suffiecient to characterize it as 'clear-cut'.

This isn't to say that some meritless cases don't get brought, and some don't succeed -- lawsuits are like that. But it's not a different problem in the med-mal context than it is in legal actions generally, and there isn't an obviously superior alternative.

[Sebastian Holsclaw]

"Much of your response is attributable to the belief that 37% of the claims were out-and-out frivolous. The judgment of one reviewer in a study, particularly where their judgment is that error is a "close call," isn't nearly enough to leap to that conclusion. There's room for differences of opinion which is why not every case includes one honest expert and one lying hack."

True. And I would have to see the underlying paper to be absolutely sure I'm correct. But it sounds like they are using a threshold standard "lack of evidence of error". It makes sense to do research like that because weighing out conflicting evidence isn't as clear cut. So there are probably a set of non-errors which have some evidence of error which are not counted in that statistic. If included, these would probably strengthen the case for reform. There is probably another set of errors which somehow leave no evidence at all. They get counted as no evidence of error. If somehow included they would weaken the case for reform or argue for different reforms. But you can't do the flip side.

This study may have all sorts of flaws, but its approach is dramatically better than any other such study I have seen. Most other studies ignore defense costs entirely (as opposed to potentially screwing up how they are counted) and do very little to really quantify error as opposed to outcome.

[Steve]

I don't know what is going on with the timestamps, but my 9:40 comment (as well as hilzoy's 9:30 comment that I responded to) was posted well after my 10:23 comment. Weird science.

[Sebastian Holsclaw]

"But it's not a different problem in the med-mal context than it is in legal actions generally, and there isn't an obviously superior alternative."

Sure, I think lots of cases with small approaching zero risks get brought against side-defendants and that it is a huge problem generally which needs to be dealt with differently. But just because there is a huge problem generally doesn't suggest that there isn't a similar problem specifically. And I'm a conservative. I'm all about not randomly tinkering with things we think generally work. That is why I don't propose a neat solution.

[Sebastian Holsclaw]

Whoa, where did hilzoy's comment come from? I will forestall any further comment until I have read the paper.

[LizardBreath]

Same here -- when I posted my 10:29, there were only two comments in the thread -- LJ's, and Steve's 10:23.

Again, SH -- what about the fact that some meritless claims receive compensation bothers you? Do you think this is unique to the tort context? Do you have a plan for reducing payouts on meritless claims that wouldn't negatively affect people with meritorious claims?

[LizardBreath]

My last was crossposted with your 10:43, which pretty much answers my questions.

[hilzoy]

OK: here's the accounting part. (For those of you with access to the article, the crucial table is on p. 2031.) They estimate 5 figures -- total system costs, indemnity, administrative costs, and the proportion of the latter that's spent by the defense and the plaintiff. Defense plus plaintiff equals administrative costs. However, total costs does not equal admin plus indemnity, as one might expect, but defense admin plus indemnity, since the plaintiffs' lawyers are presumed to get a 35% contingency fee out of the indemnity costs. (They are working from insurance company records, and thus don't have access to actual data on plaintiffs' lawyers' fees.)

Clear? OK. Then: the total cost for administration (all cases) is 54% of the total cost for indemnity (all cases).

The plaintiff's admin costs (total for all cases) are estimated to be 131,765,574, which is 35% of total indemnity ($376,473,069.)

[st]

Seb -
When you have reviewed the study, I hope you will post again about the treatment of contingency fees, because I read this:

The costs of litigating claims, including defense costs and contingency fees paid to plaintiffs’ lawyers, averaged $52,521 per claim. Overall, these administrative costs amounted to 54% of the compensation paid to plaintiffs.

as indicating that lawyers' fees were included, while you read it as indicating that they were not ("Here they are treating it as if it isn't part of that."). I'm not saying you're wrong, as I haven't read the study either, and the language is not exactly a model of clarity, as you point out. I'm just curious, because the distinction is quite significant to the import of the findings, I think.

[hilzoy]

What's really odd about the timestamp is that I did not post my first comment at 9:30. Definitely well after 10.

[rilkefan]

Drum's comments:

"About 150 of the cases involved patients who received compensation even though there was apparently no medical error.

236 of the cases involved patients who received no compensation even though they suffered injury due to medical error."

"... nearly all cases are settled out of court."

[bleh]

I think it's also worthwhile to emphasize that, as hard as proponents of "tort reform" (read: insurance companies and their paid mouthpieces) would like to portray the tort system as badly flawed, insofar as this report's authors offer a blanket qualitative conclusion, it is that "overall, the malpractice system appears to be getting it right about three quarters of the time," which despite snarky asides, IS a pretty good record for adjudicating very complex questions of fact, feeling, and emotion, and especially so when one party to the dispute typically has the resources and incentive to delay and derail the process.

(This is no exaggeration. I saw first-hand the legal fight that was necessary to establish in the Third Circuit a precedent that insurance companies had to pay interest from the date of judgment, since they routinely simply ignored court orders to pay -- after a trial, mind -- and had to be sued yet again just to force them to comply with judgment.)

If we're looking for solutions to the "problem," perhaps we should consider regulating insurance companies -- whose products we are required BY LAW to buy, let us remember -- much more closely, e.g., by limiting their profits to a few points above inflation, requiring them to pay promptly and to avoid dilatory litigation, and so on. I think that would take care of a lot of the "malpractice insurance crisis" -- a crisis which, by the way, hasn't hurt the profits of insurance companies at all, to judge by the last few years of financial statements.

[Steve]

Drum focuses on the numerical aspect of the point I was making above, which is that in a perfect system where all claims involving error were compensated and all claims not involving error were denied, net payouts would actually be higher than they are under the present, imperfect system. Yes, insurers could save money if the meritless claims were eliminated; but they'd pay out the savings and more as a consequence of having to pay the 16% of claims that currently go unpaid even though they involve a medical error.

In other words, it's not in the financial interests of insurers to steer towards a perfect system where all claims with merit are paid. It certainly is in their interests to game the system by trying to eliminate one category of inaccuracies while doing nothing about the numerous victims of medical error who wrongfully go uncompensated; but there is no reason our public policy should favor such a one-sided "solution."

[Francis]

here's my take:

because we are human, doctors will make mistakes and juries will make mistakes even if the doctor didn't.

this is precisely doctors (and lawyers and engineers and corporate officers and directors) buy insurance -- to cover the cost of mistakes and those things perceived as mistakes.

so, if reducing insurance costs to the industry as a whole is your goal, one part of the overall puzzle is for the practice of medicine to be less of a craft and more of an applied science. the doctors would have better evidence then that their conduct was not negligent. i've read (where, no idea) that doctors have been remarkably resistant to developing the structural changes that would be necessary to bring modern applied science techniques to medical practice.

(note: so were airline pilots. but after insurance companies hired a bunch of smart people to analyze the causes of crashes, the insurers forced the airlines to develop cockpit management techniques that are, apparently, quite successful. presumably because there are many more doctors and insurers, developing structured patient management protocols is much more difficult.)

but what i find extraordinarily bizarre about SH's post is his focus on the inefficiencies in the system when there was no error, as opposed to the 16% !!! of patients who were harmed by error and received no compensation.

[Sebastian Holsclaw]

Update on average costs above.

[Sebastian Holsclaw]

"Yes, insurers could save money if the meritless claims were eliminated; but they'd pay out the savings and more as a consequence of having to pay the 16% of claims that currently go unpaid even though they involve a medical error."

On a single year basis this might be true. It isn't likely to be true on a multi-year basis because it would increase their financial incentive to minimize errors by divorcing it from the non-error cases. (If you are forced to make large payouts for both non-error cases and error cases changes you make to reduce errors are about 1/3 as effective to your bottom line as if you are forced to make large payouts only in error cases). Insurance companies can only push so many changes on to doctors. Their signal to noise ratio on the effectiveness of changes is better the more closely linked payouts are to errors.

[Steve]

Is there any evidence regarding the ability of insurance companies to reduce medical error by forcing doctors to make changes? The potential strikes me as rather limited.

I'd be rather surprised if there were a large number of potential changes that insurance companies could force upon doctors, but have not done so, because the marginal efficiency of such changes is only 66% of what it could be.

[Bernard Yomtov]

Echoing bleh somewhat, I will take "tort reform" seriously when the reformers address the behavior of insurance companies as well as plaintiffs. Complaints about misbehavior often talk about "frivolous lawsuits," and hardly ever about "frivolous defenses."

Insurance companies are well-funded, clearly, and have every incentive to delay, to try to avoid paying legitimate claims, and to drive up the plaintiff's costs. Indeed, figures onadministartive costs are a deceptive argument. If the insurer can spend $50,000 to save $100,000 in payouts then they not only save $50,000 they also get to talk about how "inefficient" the system is because they have increased administrative costs as a share of payouts.

In addition insurers have the enormous tactical advantages of being both immortal and risk-neutral. Malpractice plaintiffs will often be operating under severe time pressure because of their own health, age, and financial pressures. This puts them at a severe disadvantage in settlement negotiations.

[hilzoy]

There are actually lots of changes that would significantly reduce medical errors, though whether or not insurance companies could force them on doctors, I don't know. It has been the subject of a huge debate in the medical literature, and some of the changes are beginning to be implemented.

Basically, they involve importing lessons from other fields (e.g., aviation safety), about how to set things up to make errors less likely, and applying them to medicine. There's a certain amount of professional resistance, since doctors have always thought of themselves as autonomous professionals who do not need such things. But medicine as a whole is coming round; now, the problems seem to be coming up with funding for implementation and working out kinks in the systems. The first is a big problem -- things like electronic records (which would help a lot) cost a lot of money.

[Sebastian Holsclaw]

"Indeed, figures onadministartive costs are a deceptive argument. If the insurer can spend $50,000 to save $100,000 in payouts then they not only save $50,000 they also get to talk about how "inefficient" the system is because they have increased administrative costs as a share of payouts."

I'm not particularly a friend of insurance companies (they are often really crappy, I'm trying to figure out how to tell the story about why I decided to get out of law firm litigation without violating confidences-- an insurance decision figures prominently). But I don't see how focusing on administrative costs is deceptive. When they make up more than 50% of the total costs they count as rather important.

I think a key problem in our current tort system is the inability to efficiently eliminate bad claims without incurring enormous defense costs. This transcends the medical malpractice arena and is found in almost all multi-party lawsuit case types. Now when I say "bad claims" here I am using a specialized definition. I don't only mean claims that are frivolous in general (though they are included). I mean claims that are bad as to specific defendants.

I don't see damage caps as helping, so much of what currently passes for 'tort reform' seems silly to me. I do see an enormous cost in defending such claims. In various roles I have seen vast amounts of money spent defending claims where there is no reasonable connection between the plaintiff's harm and the defendant. These claims would probably (though not always, see the study above) fail at trial, but the huge costs are incurred long before that. Maybe there is nothing that could help. Maybe we have the system with the ideal balance of plaintiff's rights and ancillary defendant's rights. But the cost is high enough that it might be worth vistiting.

[hilzoy]

Seb: I agree completely about damage caps: they seem to me to combine not getting at the real problem (frivolous claims and their associated costs) with doing real damage to people who deserve large damages (your spouse is killed by a medical error; your damages are capped; as a result you don't come close to making up the financial costs of losing your spouse, let alone the horrific emotional costs.)

I think that we need some system for making whole people who are seriously harmed by errors (medical and otherwise); that any such system will involve human beings and will therefore have all sorts of errors and "needless" costs; and that the real question is how to minimize the needless costs without hurting the people who have legitimate claims for damages. I thought this study was pretty encouraging: for a system run by fallible human beings, the tort system seems to be doing a rather decent job. I'm sure it could be improved, but I'm equally sure that a lot of the "reforms" on offer would not improve it; and I strongly suspect that other changes (e.g., hospital policies to minimize medical errors, and probably also some sort of financial support for transition costs and for outcomes assessment) would be more effective. They would also be better, period: surely it's always better to prevent medical errors entirely than to optimize the system for paying for their costs.

[Dianne]

There's a certain amount of professional resistance, since doctors have always thought of themselves as autonomous professionals who do not need such things.

BWAHAHAHAH!! Ok, now that I've had a good or maybe evil laugh at the idea that doctors don't need error reduction techniques (hey, you try working 110 hours/week for months on end without any errors), I'll move on.

But medicine as a whole is coming round

And one of the first places that implemented a number of reforms that reduce errors, including electronic record keeping, peer error reporting and analysis committees mandated to eliminate error prone situations rather than lay blame, retractable needles (to reduce the chances of needle sticks), automatic flagging of conflicting orders, and others is...the VA system.

Yep, that evil bastion of "socialized medicine" in the US is now less error prone than virtually any private hospital. It's easier to make these changes in the context of a government run system because there's less incentive to cover up mistakes (you can't sue the government) so it's easier to figure out where the mistakes are made and how to fix them.

In general, though, I'm somewhat equivocol on the issue of tort reform. On the one hand, yes, there are frivolous lawsuits and doctors spend a lot of time (probably more than is justified, truth be told) worrying about being sued and trying to defend against the possibility. On the other, patients are in a very vunerable position with respect to doctors and maybe its good to have something to equalize the power a little. And I can't say I've never told a colleague "do you want to explain to a jury why we didn't follow standard of care?" when he or she balked at doing something for a patient (usually something time consuming after hours...the specific example I'm thinking of was plasmapheresis at midnight in a patient with TTP...necessary but not particularly pleasant for any concerned. I'm sure the patient would rather have been in Philadelphia too, but he survived. Sorry, I'm babbling.)

[Francis]

generally in this country we use a combination of the regulatory system and the tort system to provide the necessary incentives for individuals and corporations to minimize the risks they impose on others.

if transactional costs are high as a percentage of actual compensation, then a few things are going on.

1. The system is working properly. Determining whether an error has occurred is, as a matter of fact, a high percent of the total value of the error.

2. Alternatively, the barriers to entry to becoming involved in the process are so high that the parties in the system (both defense and plaintiff counsel, experts, insurance company case managers, etc.) are able to rent-seek. They are slicing off more than their fair share because there is not a true market for their respective tasks.

I'm not persuaded. John Edwards has made a ton of money as a plaintiff's attorney. But according to Sebastian he (SH) has passed the bar but never got sworn in as an attorney. There is nothing preventing SH from becoming a plaintiff's med. mal. counsel and driving the price of plaintiff representation down. (The few plaintiff's counsel I know do not make an extraordinary living. It's a tough risky job.)

Insurance carriers are among the most sophisticated purchasers of legal representation in the country. If they're paying too much (for some arguable definition of too much) for their own legal representation, they should lower the rates they're willing to pay.

3. The tort system may be inherently incapable of discovering, at a reasonable price, whether or not an error occurred. If so, then we need to create an elaborate regulatory system instead. This would likely end up looking a lot like worker's comp. and be just about as popular.

What would be interesting to know is how much the study cost, and how much agreement there was on assigning degrees of confidence to each case.

[Bernard Yomtov]

But I don't see how focusing on administrative costs is deceptive. When they make up more than 50% of the total costs they count as rather important.

It's not deceptive at all. I was merely suggesting that the insurance companies are not victims in this matter. They contribute to high administrative costs, so if we want to reduce those costs we might do well to focus on their behavior as wella stht of plaintiffs' lawyers.

I don't only mean claims that are frivolous in general (though they are included). I mean claims that are bad as to specific defendants.

Absolutely agree, especially wrt the latter. The practice of naming absolutely everyone who came anywhere the patient is awful. IANAL, but have been told that this is done because it is much easier to drop a defendant than to add one as facts come to light. Is this accurate?

[Francis]

Bernard: yep. med. mal. cases have relatively short statutes of limitation. This means that if you are not sued by a certain date, you are free and clear even if later discovered evidence establishes your wrongdoing. (this is grossly oversimplified, but is one of the drivers behind naming everybody.)

[LizardBreath]

And while overnaming of defendants is sometimes a problem, it isn't necessarily. My sister, a surgeon, was recently named in a med mal action despite the fact that her only connection to the case was having been on call the night before the patient died. She was deposed and then dismissed from the case; naming her didn't add anything significant to the costs of the case, because as a resident, she was represented by the hospital's lawyers. This is an anecdote, of course, but it does show that suing everyone in sight doesn't necessarily add noticably to the administrative costs of a med mal suit. If it's handled properly, the extraneous defendants are dropped quietly as discovery reveals them to be extraneous.

[Sebastian Holsclaw]

"She was deposed and then dismissed from the case; naming her didn't add anything significant to the costs of the case, because as a resident, she was represented by the hospital's lawyers. This is an anecdote, of course, but it does show that suing everyone in sight doesn't necessarily add noticably to the administrative costs of a med mal suit. If it's handled properly, the extraneous defendants are dropped quietly as discovery reveals them to be extraneous."

The problem is that extraneous defendants are often not dropped quickly. In many complex cases you can't possibly get through the discovery phases without plunking down huge amounts of money. Maybe I've just been remarkably unlucky, but when I've worked on on the defense side the vast majority of cases involved defending people who were just tack-on defendants. In all of these cases $30-60,000 (in fees) was spent before the defendants could get out of the case. Often an additional sum (usually in the $5-10,000 range) was paid to stop the bleeding even though everyone (including/especially the plaintiff's attorney) that the defendant would never be found liable. One plaintiff's attorney was well known to push for largish settlements from innocent parties--if you refused he would drag you until the day of trial (throwing up just enough smoke to avoid summary judgment) and then drop you without comment on the morning of the trial (after you had incurred all the expense of expert retention, expert testing and all of the trial preparation). It is the kind of thing that gives lawyers a bad name, but it happens in every city.

[Sebastian Holsclaw]

Oops, forgot to finish the thought. Your sister was quickly dropped not only because she was obviously not at fault, but also because she didn't represent a potential for more funds if dragged through the process. If she had a successful side practice or a separate source of potential funds for the plaintiff I would be unsurprised to find that she was not so quickly released from the process.

[LizardBreath]

I certainly didn't mean to say that it can't be a problem -- just that in the absence of bad faith like what you describe, it need not always be a problem.

[LizardBreath]

Your sister was quickly dropped not only because she was obviously not at fault, but also because she didn't represent a potential for more funds if dragged through the process.

And this is very true -- as a resident, she certainly wasn't a deep pocket.

[Dianne]

Thinking about this a little more: Medical malpractice cases seem to have about a 19% "false positive" occurance (successful suit where no neglegence occured) and about a 14% "false negative" occurance (unsuccessful suits in the face of obvious malpractice). Are these numbers typical of the legal system? The only other statistics on this issue I can think of immediately are the Project Innocence findings of demonstratable innocence in about 1 in 7 (about 14%) death row inmates in Illinois (if I remember the stats correctly).

So does the legal system in general have a 14-20% or so error rate? If so, is there any way these numbers could reasonably be lowered, because it seems to me pretty unacceptable that there's an approximately 15% chance that any given innocent person will be convicted and, conversely, that any given guilty person will be exonerated.

[LizardBreath]

Civil error rates aren't going to translate well to criminal cases. It's a very different system.

On the civil side, I'd say that 15-20% wrong results accords with my intuition.

[Hal]

Despite the obvious fun of a wonk orgy, the question on my mind is does this data and analysis change Sebastian's mind. It's all well and good to have a rousing good talk of percentages and percentages of percentages, but as I understand it, Sebastian is on the side of the fence that thinks the subject of this study - medical malpractice suits are one of the top 10 (maybe top 5) problems anchoring our capitalistic society like a nuclear submarine chained to an Olympic swimmer's ankle.

I know that lawyers won't ever admit anything, but I'm far more interested in how this study affects your personal views on the subject...

[Bernard Yomtov]

One question about defense costs in the study:

How, if at all, is the cost of the insurance companies' inhouse counsel calculated? I assume that they use some combination of inhouse and outside lawyers, and that the cost of the outsiders is simply their fees. But the method of assigning costs, if any, for lawyers actually employed by the companies would have a big impact on these results.

[Sebastian Holsclaw]

Typically insurance companies won't run local malpractice cases in house so nearly all of the fees (excluding in some cases coverage analysis) tend to be outside fees. But insurance companies tend to be ruthless about squeezing their lawyers for fees--I'd be very surprised if you could reduce overall costs by squeezing insurance defense lawyers.

[Steve]

My firm does a lot of insurance defense work, but we won't even touch med mal cases because the rates are so low - even compared to other types of professional liability cases. None of these cases are handled in-house by the insurance companies, though, because the policy obligates the insurer to give you a lawyer who will act as YOUR lawyer, not the insurance company's lawyer.

Even when the insurance company pays my bills, my professional duty is to YOU, the insured, not to the insurance company. I couldn't give you that kind of undivided loyalty if I were an employee of the insurance company.

[heet]

SH:"But insurance companies tend to be ruthless about squeezing their lawyers for fees--I'd be very surprised if you could reduce overall costs by squeezing insurance defense lawyers."

Steve:"Even when the insurance company pays my bills, my professional duty is to YOU, the insured, not to the insurance company. I couldn't give you that kind of undivided loyalty if I were an employee of the insurance company."

My understanding is these two ideas almost cancel each other out. The insurance companies have huge contracts w/ firms. That is the source of their tight fee schedules - they command so much business, they can shop around for firms willing to deal.

Now, another side effect of this bulk rate is the lawyer's duty that Steve mentions can be compromised. If a physician absolutely KNOWS he didn't do anything wrong and wants to fight it but the insurance firm wants to avoid the expense of going to court (why not, it isn't their reputation on the line?) the lawyer is left with a tough choice. They are torn between their duty to defend the doctor and their duty to keep the insurance company's money flowing in the door.

I'm a 2nd year med student and we have already been advised to obtain our own council in the event of a lawsuit. That way we are signalling to the insurance-paid lawyer that we are serious about our rights as a defendant.

[Steve]

It's not much of a conflict - I give the best advice I can about the risks of litigation, to both the insured and the insurer. If the two of you disagree on what to do about the risks then you need to work that out. It's not like I can force you to settle without your consent.

There are tough situations created by my need to maintain an ongoing business relationship with the insurer, but this isn't really one of them. It's not like they call me up and say "convince him to settle, or we're never sending you another case!" It just doesn't happen. They know it's outside my control.

[Sebastian Holsclaw]

"I'm a 2nd year med student and we have already been advised to obtain our own council in the event of a lawsuit. That way we are signalling to the insurance-paid lawyer that we are serious about our rights as a defendant."

That is fine in theory, but doctors rarely do it in practice because it is really expensive to hire your own lawyer at $50,000 when you already have a $50,000 hired by the insurance company.

Also the selection of counsel is subject to the insurance contract--many of them let you select your own counsel so long as they get some input and so long as they fees are not beyond what they consider reasonable. If they are, they will often let you negotiate to pay the difference.

[heet]

My info comes from lawyers friends (who don't do med mal) and a big muckety muck lawyer who teaches a medicine & law course. The former may be naive and the latter may be overzealous/paranoid (though an expert).

My lawyer friends were mostly upset at the power the insurance companies had over the firms... The prof was perhaps trying to tell us to "watch our backs".

The point that I was making is there is an incentive to do what makes the insurance company happy... I'm not suggesting every or many lawyers would ditch the defendant's interest like this but the incentive is certainly there.

Thanks for your comments, I certainly didn't know that you could negotiate some input into your representation. Another point the law prof made was we should always have our mal insurance contracts checked over for things like this...

[Bernard Yomtov]

I give the best advice I can about the risks of litigation, to both the insured and the insurer. If the two of you disagree on what to do about the risks then you need to work that out.

Wouldn't the insurer typically insist on the right to control the defense, including settling, as part of the policy? They are the ones whose money is at risk. I know I've seen that sort of provision in other types of situations involving indemnities. "If you want me to indemnify you I control the case," seems like a reasonable position to take.

[Steve]

There are court decisions on this - they have no choice but to give you an attorney that represents YOU. And it's not like I can appear in court on behalf of a client if I'm not, in fact, representing that client but actually representing their insurance company.

The most the insurer can do is require that your consent to a settlement shall not be unreasonably withheld, which is standard stuff. If you insist on rolling the dice, at some point they can cut you loose to bear the risk on your own.

[Fledermaus]

The costs of litigating claims, including defense costs and contingency fees paid to plaintiffs’ lawyers, averaged $52,521 per claim.

I haven't done any med-mal settlements but technically doesn't the plaintiff "pay" the contingency fee because it comes out of his or her damages? i.e. plaintiff has $60k of injuries but only gets $40k

[Steve]

Yes, the purpose of the figure is to measure the total administrative costs involved in adjudicating a claim.

[Sebastian Holsclaw]

Fledermaus, please read the update. The part you quote was incorrect in the press release.