Yes, dumb down the American public, a few people keep raising their head!
American healthcare has problems, layers of problems. I agree with Mr. Brill's 26,000 word indictment of US healthcare in the March 4th issue of Time. I don't accept his implied remedies nor the suggestions of Ezra Klein and Sarah Kliff in their Wonkblog discussion, but that's just background to discussion of this case, which has important ramifications that US journalists seem unable to grasp. Paraphrasing talktothemike, I am an insider, but no fan of the American system as it is.
I have no inside information on this case. Karen Bartlett was prescribed sulindac, a nonsteroidal anti-inflammatory drug (NSAID), for shoulder pain in 2004. She developed a very severe hypersensitivity reaction, leaving her blind and badly handicapped. A jury awarded her $21 million in damages. The drug company appealed and SCOTUS has the case. The New York Times jumped in with an editorial “Hold Generic Drug Makers Accountable”. They say “Manufacturers should bear responsibility for making sure their drugs are safe and effective… If the Supreme Court shields the makers of generic drugs from consumer suits, Congress ought to amend the laws. “
Problem: All NSAIDs carry some risk of severe hypersensitivity reactions such as Ms. Bartlett suffered. Did your doctor warn you when she said, "Why don't you try some Ibuprofen?"
This nerdy wordy diary focuses on drug and health education, primarily the problem of pain and painkillers.
1. Since all NSAIDs are associated with this rare complication, severe hypersensitive skin reactions of the Stevens-Johnson/Toxic Epidermal Necrolysis group, which I'll call SJS/TEN, the questions are:
a. Is Sulindac more likely to produce this complication than other NSAIDs such as Ibuprofen? The media have publicized this dreadful but rare complication of Ibuprofen use, see Ibuprofen blinds child A California jury agreed in 2008 that the maker of Children's Motrin did not warn of danger in this case, but cleared the company of liability. However, a Philadelphia jury ordered Johnson & Johnson's McNeil Consumer Healthcare, to pay another family $10 million for their child's injuries and failure to adequately warn consumers about the toxic skin reaction she suffered, also blamed on Motrin, This Ibuprofen injury won lawsuit Most people would be horrified if they read Ibuprofen-induced hypersensitivity syndrome, by Nanau & Neuman, Translational Research 155:275, 2010. It includes 8 scary cases of severe illness ascribed to Ibuprofen but there are two important caveats: millions of people take Ibuprofen every year, so the chance of such ADRs is very low and the illnesses of some patients may not in fact have been caused by Ibuprofen. Many of them took several drugs including both Ibuprofen and acetaminophen; how do we know which was responsible? The single most reliable way to determine this is by drug challenge. After the patient recovers, a small dose of the suspect drug is given and the complication returns. This can be dangerous; I would not propose it for those with a severe reaction like Ms. Bartlett. However, some patients with this syndrome had it reproduced later by an acetaminophen challenge (example Trujillo et al, Stevens-Johnson syndrome after acetaminophen ingestion, confirmed by challenge test in an eleven-year-old patient. Allergol Immunopathol 2010; 38: 99-100.
Yes, our legal system is capricious, but that's not my interest here. I want to focus on health education, something that our schools and culture don't teach.
Adverse Drug reactions (ADRs)
can be divided into two groups
(1) patient took excessive dose The number one cause of death from NSAIDs by far is bleeding from the gastro-intestinal tract (GI bleeding). Most patients with this problem did not take excessive doses. The safe dose of any NSAID, such as Ibuprofen, decreases markedly with age. It often takes a big dose of aspirin (say 2800 milligrams or 8 tablets daily) to cause GI bleeding in a young adult, but one "baby aspirin" (81 mg) per day for a week can do this in an 82 year old. We have only rough estimates of how many deaths occur from NSAID induced GI bleeding, in many cases other factors in addition to the NSAID and age are involved. Best estimates are about 15-20,000 deaths a year. Even the CDC doesn't know if that number is increasing or decreasing.
Prescription drug overdoses now kill more Americans than auto accidents or shootings and have increased four fold in the last 15 years.These are mostly overdoses of narcotic painkillers in people under age 55. Death occurs at home and because of that and the special requirements for narcotic prescriptions, the data are much better than those for NSAID related GI bleeding, where death usually occurs in hospital and in people with known chronic disease. Over 16,000 deaths from overdoses of drugs like Vicodin and Oxycontin occurred in 2011.
(2) No excessive dose Most allergic reactions fall into this second category. However, careful studies of drug reactions find that many patients with reactions started out with high and escalating doses. It's one of several ways in which the reality of ADRs differs from the textbook picture. Textbooks divide drug hypersensitivity into 4 categories:
Type I immediate (anaphylaxis, IgE antibodies)
Type II cytotoxic rxns (cytotoxic, ex transfusion rxn, in minutes to hours, IgM or IgG antibody mediated)
Type III immune complex mediated (serum sickness, localized organ injury, time scale hours-days)
Type IV delayed hypersensitivity (TB skin test is classic example, due to sensitized cells, takes a few days to appear)
However, not all drug reactions fit into these categories; a given drug may cause rxns of several types, and many drug reactions, especially NSAID reactions, may fit best into nonspecific or non-allergic hypersensitivity (sounds like an oxymoron).
Two problems ?cause- may be infection rather than drug, can be multiple drugs acting together - and ? mechanism or basis. We need different treatments to prevent or stop different kinds of reactions. Antihistamines are minimally helpful for SJS/TEN reactions
.
b. Must doctors warn you of this risk when they advise or prescribe an NSAID? Most US NSAID use is over the counter drugs, Ibuprofen or Naproxen. Sulindac is one of 15 different NSAIDs available only by prescription. It's been in use since 1978 - I see no reason to prescribe it over less expensive NSAIDs. Acetaminophen (brand name Tylenol) is a closely related drug and is the most commonly used drug by Americans. Ibuprofen is second, and Aspirin third. Diclofenac is an NSAID used more than Ibuprofen in China and many other countries. It probably has greater risk of GI bleeding than any other NSAID in current use. GI bleeding is the most important side effect because it kills far more people than the other NSAID ADRs. Meloxicam probably has more risk of allergic skin reactions than others.
I must report child abuse and certain infections to government agencies, there is no requirement to report drug reactions. Therefore, ADR data is not precise - some papers just collect a list of patients claimed by some physician, any physician to have had an ADR. Roujeau et al published a first class study of drug induced SJS/TEN reactions in 1995, MEDICATION USE AND THE RISK OF STEVENS–JOHNSON SYNDROME OR TOXIC EPIDERMAL NECROLYSIS, N Engl J Med 1995;333:1600 in which they examined all patients to be sure of the skin diagnosis (most physicians are fuzzy about this) and personally reviewed all medication use. They compared drugs taken by the patients with those used by a larger group of patients hospitalized at the same time without ADR. The commonest drugs implicated were antibiotics, especially sulfonamide drugs, antiepileptic drugs and then NSAIDS. (Piroxicam had the worst risk ratio of all NSAIDs). Note that the risk of this dreadful ADR assocated with acetaminophen was less than 1 in French patients and 9.3 in patients from Germany, Italy and Portugal. Either acetaminophen is safe (France) or more hazardous than most NSAIDs (the other three countries). The authors discuss this and suggest that this is because of much greater acetaminophen use in France than the other countries, which might mean that those who can't tolerate it learn this at an early age before developing this severe reaction. There are other possible explanations. The point is that this data is far from precise. The large discrepancy in risk of SJS/TEN in different countries in a first class study illustrates the problem. Although the new England Journal normally requires payment for articles you can get the Roujeau article here, http://www.aardvarksfly.com/....
We now know that certain genotypes greatly increase the risk of this ADR, most notably for SJS/TEN associated with use of carbamazepine, CBZ, an anti-epileptic drug taken by about 250 of my patients. I won't start a Chinese American or person of Malay descent on CBZ because we can't test for the particular genotype, HLA-B*1502, at my hospital. The interesting sublety is that Han Chinese and their descendants have a high risk of this drug reaction to CBZ (not to NSAIDs) if they have this allele, but the same allele in Europeans and Americans doesn't seem to increase the risk, presumably because of the effects of other genes that are rare in Han Chinese. In the same way, knocking out a gene can cause severe disease in certain mouse strains but is well tolerated by others. Another HLA allele, HLA-B*5801, is a fair predictor of SJS/TEN ADRs to allopurinol, a drug used for gout. Hall found that African-Americans were much less likely to take NSAIDs in the Boston area than were Caucasians or Latinos. This probably represents social and cultural factors (in patients and doctors) more than biology (NSAID intolerance among African-Americans) but we don't know much about this. Existing databases of NSAID problems are less adequate for African-Americans than for Caucasians and Asians, Latinos are intermediate.
Radiation therapy, certain infections (AIDS for example) and some forms of cancer also increase the risk of SJS/TEN with all drugs. A few studies using Bayesian data mining (discussed in Nate Silver's book) suggest a two fold greater risk of SJS/TEN for sulindac compared to Ibuprofen, but they are shaky compared to Roujean's paper. No study suggests that sulindac is as risky as Piroxicam, meloxicam, or diclofenac, NSAIDs available in the US and prescribed by many MDs.
c. Is there evidence that this particular batch of Sulindac was spoiled or contaminated in some way? If so, we expect a disproportionate number of reactions in others who took it. All manufactured products are subject to errors and mistakes in the manufacturing process, for which manufacturers may be liable. The Sulindac taken by Ms. Bartlett was made in India. There isn't a prayer of the FDA inspecting that plant. I don't think that the FDA should allow generic drugs made outside North and Central America because they almost never inspect them. Cutting out Asian generics would leave lots of generics and drug companies would get the message that factories must be in or near to the US if they want to sell in our market. This would have some economic benefit.
d. My personal experience with SJS/TEN: I have seen about 12,000 different patients with epilepsy and almost as many with chronic migraine headaches in my long career. I had 4 patients die from this syndrome prior to 2000 when I started using IVIG for the severe cases. None of the fatal cases got the medicine from me, 3/4 got the drugs (phenobarbital and phenytoin) from emergency rooms. In the last month, two of my epileptic patients were started on phenytoin and medication that I prescribed stopped when they were taken to outside ERs. There's too much cowboy medicine in this country- I didn't see that when I was in England. Four of my patients got severe SJS/TEN from CBZ I prescribed (three were Chinese-American) requiring hospitalization for at least 5 days. This was before the first reports about HLA-B*1502. When I start a non-Chinese patient on CBZ, I give them a written sheet saying "CBZ is a high rash drug" and that about 10% of patients will develop a rash, that they should contact our hospital immediately if that happens, to avoid severe illness, with details. Once you've been on CBZ for 3 months, you won't get this ADR.
None of my migraine patients have had SJS/TEN although I use a lot of NSAIDs for migraine (Ibuprofen, Naproxen and Indomethacin, a special NSAID that has more side effects but is extremely effective for certain forms of migraine). None have had major GI bleeding but almost all are under age 30 and I make sure that they don't take the drug every day.
I'm confident that the risk of SJS/TEN is much greater for several anti-epileptic drugs (not all) than it is with NSAIDs. I never prescribe piroxicam, meloxicam or sulindac or any NSAID other than the three I mentioned. All of these drugs (Aspirin, Tylenol and NSAIDs) have some beneficial side effects. Aspirin and to a lesser extent NSAIDs like Ibuprofen reduce the risk of certain cancers, aspirin definitely reduces the risk of stroke and heart attack. Sulindac may have greater ability to prevent or stop colon cancer than other NSAIDs based on tissue culture studies. Limited human studies of Sulindac and cancer have yielded conflicting results. This kind of use would require taking it for long periods of time. I wouldn't take it to treat or prevent cancer (both of my parents had colon cancer, I have escaped it but I do have prostate cancer).
e. Who says that there is any drug anywhere without risk of serious side effects? Aspirin,Tylenol and NSAIDs kill many Americans every year. NSAIDs kill more, but statistics on this are dodgy because many ADRs have multiple causes and typically only one is selected for attention. Intentional suicidal drug overdoses are responsible for many acetaminophen deaths, but not all. Acetaminophen is the cause of many people needing liver transplants. As already noted, tens of thousands of Americans die every year from drug overdosage, the majority from narcotic painkillers.
f. Get realistic about painkillers and all drugs- this is a big problem in our culture. NSAIDs are pretty safe in people under 65 if taken for periods of weeks in moderate dosage. Opiates are often used and rarely cause hypersensitivity reactions. They can be helpful, but they cause many overdose deaths and cause constipation and falls, especially in the elderly, they often require increasing doses as use continues and they carry addiction risks. They are very helpful for a few weeks. People with painful conditions such as neuropathies, shingles, trigeminal neuralgia, fibromyalgia and other chronic pain syndromes are a special area which I can't discuss here- it's not a one paragraph issue, but it's very difficult for primary care MDs to adequately treat these patients.
My concern is with bone and joint pain, arthritis, which is so common in older people. All of us above age 60 have X ray evidence of joint disease. We can't all be treated by specialists, although a one time evaluation by a joint specialist with recommendations to the patient and the PMD is feasible. Many are lucky and aren't bothered by it- I can run five miles but there are days when I must stop and walk. I have flare ups of back pain from time to time that I've learned to control with exercise, which I start immediately as soon as it appears, I don't take any pain killer. A physical therapist friend said to me, "Isn't it better to build up your body and your muscles to control pain instead of stressing yourself with drugs?" and it worked. I have friends who use yoga in the same way. Acupuncture works for some. I recognize that many Americans don't want to use exercise, yoga or acupuncture and that everyone needs guidance to benefit from exercise or yoga. Commercial insurance typically charges ridiculous copays for physical therapy and won't cover yoga at all. Some policies cover acupuncture.
People should learn in school, when you hear about a wonder drug, be very careful. Be skeptical about the more recent Time magazine (April 1) hyping cures for cancer, and about the constant stream of cancer soundbites in the media. Yes, we are getting better results than ten years ago, but progress is slow, and recognize that drug companies and researchers have reason to hype their results. The media are saturated with inflated claims. As John Ioannidis famously put it, most big splashy results don't hold up. You can get his paper, Why most published research findings are false, at http://www.ncbi.nlm.nih.gov/.... We have useful drugs, more than 20 years ago, but very few wonder drugs and no wonder drugs for pain. Pain and pain killers remain difficult areas, only partly solved by American, Japanese or European medicine.
f. Back to the Bartlett case: Let's accept for discussion that Sulindac caused her skin reaction, although we don't know the details. She suffered a rare but well known NSAID complication. There's no evidence that the drug was defective. This NSAID complication typically occurs in older females. There's no evidence that it's more likely with generic NSAIDs or that Sulindac is truly more hazardous than Ibuprofen or Naproxen. There is evidence that many pharmacies, like CVS, overcharge for generic charges,
overcharging for generics
A humane healthcare system (we are far from that) would compensate all major problems associated with medical treatment in a no-fault way. This would not be millions of dollars and it would come from public funds. I would pay less for injury from patient selected treatment (i.e. patient bought over the counter drugs such as NSAIDs which caused serious harm) than from that associated with prescription drugs. A sensible FDA would not certify drugs from factories beyond the reach of its inspections and would inspect generic and brand name plants more frequently than it now does (it would need a bigger budget). It would require more extensive data before accepting generic delayed release drugs, since some of them have been faulty. A respectable press would not editorialize about the case without much better understanding than that shown by the New York Times, which sank to the Fox News level. A sane educational system would teach all people much more about health and the benefits of regular exercise (dancing is the most palatable form for most of us) than we now learn and would stress the fact that all drugs have side effects. We know that we may die from auto accidents, but most of us still choose to drive cars- this is analogous to drugs. If you have an older relative taking painkillers month after month for osteoarthritis, you should ask to meet with her doctor.
This is already a very long diary, so I won't discuss use of DMSO for joint pain, touted by horse trainers but looking like placebo in the few controlled human studies available, or marijuana for pain control, or the treatment of chronic pain syndromes. Each of these would require a complex diary of their own. Because chronic pain syndromes are so heterogeneous, they would require many diaries. We recognize an obligation to help pay for medical costs but have not recognized an obligation to educate people about health, to save money and improve their lives, beyond platitudes.