Tuesday, March 22, 2011

How Doctors can Avoid Malpractice Suits

     The most common cause for suing an internist, according to my malpractice insurance company, is "failure to diagnose", followed by "failure to make a timely referral". After reading most of what this company had to say, discussing events with Risk Management, and many talks with those of my patients and friends who are tort attorneys, I have developed a list of do's and dont's for practicing physicians who wish to minimize their risk of being sued. This list will probably not lead to better or more efficient or cheaper patient care, but it will reduce the malpractice exposure of any doctor who follows my recommendations to the letter. I have also thru Lexis and other search engines, verified that there is at least one malpractice case that was triggered, in part, by the physician's failure to follow one of these rules:

     1) At EVERY visit, ask the patient if he/she is suicidal, or has suicidal thoughts.

     2) Every time you prescribe any drug that acts on the central nervous system (sleeping pills, tranquilizers, anti-depressants, etc.) caution the patient not to drive or operate heavy machinery for 24 hours after taking the medicine.

     3) If any patient refuses a suggested test: mammogram, pap smear, stool for blood, fire that patient from your practice immediately. Juries have held that if the patient dies because of a refused test, you probably didn't word your advice strongly enough. The same holds true for blood tests.

     4) Never give any advice over the phone. It is basically malpractice and also a violation of most state medical laws to "prescribe", i.e. give any medical advice without examining the patient. Insist that the patient come in to see you right away, or have the patient call 911 for an ambulance to go to the ER.  I know some doctors who will call 911 if the patient refuses to.

     5) Whatever the patient's problem, after you suggest a diagnosis and treatment, have the patient see a specialist in the field of the affected organ (cardiology for hypertension, GI for abdominal pain, neurology for dizziness) and tell the patient you can't continue as the personal physician unless a specialist is seen.

    6) For any abnormal blood tests, again refer at once to a specialist.

     7) If the patient, or the patient's family suggests any test; cardiac echo, brain MRI, chest CT scan, agree at once, order the test, and again refer to the appropriate specialist.

     8) Do a stress-echo and stress-thallium and cardiology referral for any pain above the waist, an abdominal and pelvic CT scan with contrast followed by an MRI and a referral to a GI doctor and a urologist for any pain below the waist, a Chest CT scan with contrast for any cough and for any smoker with a referral to a pulmonologist and a total body bone scan for any back or hip pain, followed by referral to an orthopedist. If CT scan shows nothing, do an MRI, followed by a MRA. If headache or dizziness, head CT with contrast, MRI with gadolinium, and neuro consult. This may not be cost-efficient, but a malpractice jury never cares about that.

     9) "When in doubt, refer it out".

     Of course I never practiced medicine following the above legalistic "rules", but every doctor has to consider the risks of not following those rules every time there is patient contact.  And in America, if something "bad" happens to a patient, it must be someone's fault. The doctor no longer has to prove he/she did nothing wrong, but rather than he/she did everything that was "right".

     I suspect that there would be fewer malpractice suits if workmen's compensation laws applied: no contingency fees for lawyers, and a pre-set schedule of payments for partial or permament disabilities up to and including death. Don't forget the Texas jury that bankrupted Dow Chemical, because they believed "as a matter of fact': that silicone leaking from breast implants caused a lupus-like syndrome. Unfortunately Dow went bankrupt from the class-action suit. Two years later a careful study in the New England Journal of Medicine showed that there was absolutely no correlation, and certainly no evidence of cause-and-effect, but Dow remained in bankruptcy court, and none of the plaintiffs or lawyers had to return any of the money.

     What I am really saying is that with the best doctors, good medicine results from a combination of skills, intuition, and art. When  I walk into an ER hospital room, I often think to myself: "This patient is sick and needs a hospital admission". Then I have to invent a diagnosis, because the insurance computers don't pay for an admission diagnosis of "looks sick", which is an absurdity:just  as all you mothers out there know when your child is sick, I can see when my patient is sick. Sometimes I put down 999.90 "unknown disease", because (a) we all have at least one of those, and (b) it drives the computers wild when they try to calculate a LOS., or look for guidelines.

Saturday, March 19, 2011

The Problems With Meta-Analyses (2)

     I had written a more mathematical blog in May, 2009, denoting the logical and mathematical/statistical problems with meta-analyses, but since that time many more meta-analyses have been published, and the public has discussed these results as if they were clinical fact.  It is important to understand that the results of a meta-analysis should be presented only as a hypothetical clinical result, to be tested forwards in a properly designed clinical format, and not accepted as proven fact (such as the recent suggestion that women who ingest calcium supplements increase their risk of heart disease). In brief, a meta-analysis collects several studies of the same problem, none of which reaches clinical or statistical significance, in the hopes that the sum can be greater than its parts, and that combining non-significant studies can reach a significant result!

     Some readily understandable problems with meta-analyses:

1) You are never told which studies the author rejects as not being acceptable for his/her meta-analysis, so you cannot form your own opinion as to the validity of rejecting those particular studies.

2) The problem of the Simpson Paradox, or the Yule-Simpson Effect: sometimes all the included studies point in one direction as being clinically significant, but the meta-analysis points in exactly the opposite direction. This has been discussed, for instance, in "Chance" magazine, published by Dartmouth University, where they showed different ways of calculating Derek Jeter's batting average, with differing results, using the same data in each case.

3) There are two different statistical models or assumptions by which the analyzer combines the effects of the individual studies: the fixed effects model and the random effects model. Each model makes different assumptions about the underlying statistical distribution of observed data,, so each calculation produces different results.

4) There are two different methods for measuring the effect of the clinical  intervention: standardized mean difference or correlation. Each method produces a different end result.

5) If we look at #3 and #4, we see immediately that there are four possible combinations of analyses, leadeing to four different conclusions for the same set of studies. No one paper shows all four combinations and all four possible results.

6) Finally, the choice of what constitutes a "significant' effect in any of the included studies is purely arbitrary. When this question was studied by clinical psychologists,  no two analytical scientists reached the same conclusions of what was significant in all the included studies.

     We therefore see that the result of any meta-analysis is largely dependent on the analyzer, and the reader never has enough data to redo the analysis, so the results have to be taken on faith, which is hardly a scientific result.

     "There are three kinds of lies: Lies, Damn Lies, and Statistics"-----Mark Twain

Friday, March 11, 2011

Medical Misconceptions

     As a physicist as well as a practicing physician, I am always disturbed by medical statements that have no basis in experimental or clinical fact. Some statements are considered to be "obvious", some are practicalities stated by the government, and some have the inertia of years if not centuries of belief. I have always told my patients that if any doctor (including myself) makes a medical statement of  "fact", they should ask the doctor where they can read about it. OTOH, if the doctor says "I believe", or "I feel", then take the advice with at least 10 grains of salt. It is, of course, more difficult when your spouse tries to impress a "fact" on you, and you can't really answer "where did you get your medical degree"?. One of my patients brought in a paragraph cut out from a magazine which explained her disease and treatment in detail. Since the article was taken from the "National Enquirer", I told her that I was not a specialist in this particular field, and she needed to consult a specialist. Now that we have the internet, I often have to "unlearn" a patient before I can "learn" him or her. BTW, the only source you should trust about cancer is the web site of the National Cancer Institute. All other institutional websites have an institutional bias. For instance, the chairman of one teaching hospital wrote a book about how the only treatment for prostate cancer is surgery, so few if any urologists at that institute would suggest either radiation or watchful waiting. This raises  another  point: if you want a second opinion, always go to a different institution, since there is real institutional bias. 30 years ago, when I trained, I was told that Cornell/NY Hospital only ran digitoxin levels, and not digoxin levels, so if you were an intern at NY Hospital, you had to use digitoxin to treat your heart failure patients. Similarly, at Columbia, the ANA was run daily in the lab, and the LE Prep only weekly, so that even though the LE prep was a more definitive test for lupus, all the interns had to screen with the ANA. As a final example, the body runs on thyroid hormone. (Let's forget T3 for the purpose of this discussion.) So the real test for hypo or hyper thyroid state was free T4. However, this was a radioimmunoassay , and very expensive, so we were all taught to screen with the TSH, a pituitary hormone, and only do a free T4 if the TSH was off. This meant that the patient had to be stuck twice for blood, and the correct diagnosis was delayed.

     Brief joke: The way opticians charge for eyeglasses is as follows: When they give you your eyeglasses, they say "That will be $300". If you don't flinch, they add "That's for the frames. The lenses are another $300". If you still don't flinch, they add "Each".

     The following are some of the medical "facts" that are not based on clinical studies:

     1) Marijuana is classified as a Class I drug by the FDA, which means "no possible clinical benefit or use". Nevertheless, the U.S. government grown marijuana plants in Lexington, Ky., for those patients whose glaucoma can only be successfully treated by smoking marijuana.

     2) Ecstasy, the pleasure drug used at raves, is also classified as Class I. However, at least eight psychiatrists around the country have special license to use Ecstasy on otherwise intractable schizophrenic patients.

     3) Shades of "One Flew Over the Cuckoo's Nest": A physician in Europe received the Nobel Prize in Medicine for "proving" that pre-frontal lobotomies "cured" schizophrenia. Not so.

     4) The BMI is touted as calculating your IBW, or ideal body weight for your height. This number has no basis in fact, and is merely a derived average. No one has shown that patients with a "normal" BMI live longer, and in fact a recent study in Lancet disproved this. Also, the IBW in the Metropolitan Life Insurance tables was specifically derived for patients wearing shoes.

     5) Ideal pulse at workout = 0.8x(220-age). This is also a purely fictional number derived from healthy young men. In fact, the greater your aerobic conditioning, the lower is your resting pulse and peak pulse. Just ask any marathon runner.

     6) Eating sugar causes diabetes. Just because diabetics have high sugars, it is not true that eating sugar causes it.

     7) Too much salt causes high blood pressure. Another canard, unless you are a specially genetically formed lab rat. Normal kidneys can always clear salt. But because of this belief, many of my golfers get faint and dizzy by the 13th hole, because they sweat salt plus water, and drink only water, and dilute down the salt levels of their body. Salt in the blood defends blood pressure. The salt content of blood is approximately equal to a 50/50 mix of water and Gatorade, which is how the football coach of the University of Florida developed Gatorade.

     8) For some reason, if you go to donate blood at the Red Cross, they will refuse you if your blood pressure is "too high". This is very odd, because 100 years ago, before we had any useful anti-hypertensive medicines, the only treatment that successfully lowered blood pressure was bleeding with leeches.

     9) When I went to school the food triangle had an apex and a base, and one was protein and the other was carbohydrates. Now the food triangle taught in school is exactly the reverse. Who is/was right?

     10) Labeling calories probably does not lead to weight loss. Taco Bells in LA have had to do that for the past 13 months, but the total calories consumed by the average patron did not change.

     11)Question: If the real  reason for raising the drinking age to 21 was to keep drunk teen-age drivers off the road, why not let teenagers who do not have a driver's license drink?

     12) Why are all addicting foods vegetables (nicotine, alcohol, opium, cocaine, etc.)?

     13) The "normal" fasting blood sugar has been lowered from 140 to 120 to 110 to 100, but no one has shown any medical benefit or reduction in diabetic morbidity from these lowered definitions. OTOH,. if your FBS is 110, your life insurance premiums will now be higher.

     14) Medicare Part D now only pays for generic versions. However, many doctors are aware that for some drugs, especially those that act on the CNS (Ativan, Prozac, Wellbutrin) often only the brand name works. I have similar arguments with drug "management" systems such as Medco, who claim that only their preferred brand of stomach acid suppressant, such as Prevacid works. When I try to tell them that only Aciphex works on my patient, they don't want to hear it, although they know full well that although they have the same endpoint, since the two drugs have different chemical structures one will work when the other will not.

     15) I will close with the most common misperception of all, and one that we are all guilty of, and that is persisting in giving medical advice to our loved ones after it is evident that they will not follow it. After you tell a smoker to stop smoking or a heavy diabetic to lose weight more than once, they have received your message. It's not that they don't understand you, it's just that they don't want to follow your advice. And they know their actions may not be logical, but they still refuse. Some patients refuse to have a mammogram, or a colonoscope, or even to  see a doctor. Many patients will refuse to follow good advice, even when they agree with it, for various reasons, some of which they do not fully understand themselves. When this does happen, you should stop and ask yourselves if you would follow medical advice that a loved one gave to you if you did not believe in it or did not feel like doing it.