Articles

Clinical Decision-Making: Part II

This is the second in a brief series of posts about how clinicians think. My purpose here is to elucidate how skeptical principles apply to clinical decision-making, but also as background to provide context to many of the articles we publish here.  In this installment I will review the factors that clinicians consider when deciding what tests to order for screening and when conducting a diagnostic workup.

The gunshot approach

Last week I discussed the “Dr. House” approach to medicine, using that particular TV character as an example of how medicine is often portrayed in fiction. Another aspect of the Dr. House image that is very misleading is his approach to diagnosis, which tends to be very linear. He decides what the most likely diagnosis is, then proceeds to either treat that entity or order a confirmatory diagnostic test. When that diagnosis fails, he then proceeds onto diagnosis B. A string of such failures then culminates in a flash of brilliance that allows him to make the actual obscure diagnosis and cure the patient. This approach is optimized for storytelling and drama, but is not how actual clinicians operate.

At the other end of the spectrum is what doctors often refer to as “the gunshot approach” – test for everything in hopes that you hit something. Another derogatory term that doctors throw around is “a fishing expedition,” referring to a diagnostic approach that amounts to hunting around for any possible diagnosis without having a real justification.

A more optimal approach lies somewhere between Dr. House’s serial approach and mindlessly testing for everything. Doctors tend to take a layered approach, using various criteria to decide which tests are worthwhile, which tests have to be done, and which are not justified. Often there is a tiered approach; if round one of diagnostic testing does not yield a positive result, then progressively less likely diagnoses can be pursued. When this approach is persistently negative, then there is the trick of knowing when to stop – when further testing will yield diminishing returns.

Criteria for testing

Here are the various criteria that diagnosticians use to determine which tests should be ordered.  Sometimes there are fairly strict algorithms determined by the standard of care, at other times doctors make a judgement about the cumulative impact of all of these factors combined.

Sensitivity and specificity: Sensitivity is the probability that a test will be positive if the patient actually has the diagnosis being tested for. Failure to test positive in someone with the target disease is called a false negative.  Specificity is that probability that the patient has the disease if the test comes back positive.  Testing positive in someone without the target disease is called a false positive. The more sensitive and specific an available test, the more useful it is.

How likely is the diagnosis? The probability that a diagnosis is present also dramatically affects how useful testing is. If the probability is very low then even a very specific test will be more likely to generate a false positive than a true positive. And of course, testing is more likely to give you an answer if you are looking for a diagnosis that is likely to be present.

What is the morbidity and mortality of the disease? It is more important to diagnose, or rule out, serious illness. There are, in fact, certain entities that we simply cannot afford to miss. Benign and self-limiting diseases, on the other hand, may not be worth diagnosing since they will get better on their own anyway.

How treatable is the disease you are looking for? The golden rule of diagnostic testing is this – how will the results of your test affect your management of the patient. If you don’t know the answer to that question, don’t order the test. You never want to be in a situation where you have an abnormal result and no idea what to do with it. You should have sorted that out before ordering the test.

How invasive, expensive, inconvenient, risky, or painful is the test? These factors get to the risk vs. benefit calculation of ordering a specific test.

Adding all these things together, it is clear that a doctor probably should not order an expensive, painful, highly invasive, and very nonspecific diagnostic test in order to diagnose a rare and benign entity that isn’t treatable anyway. We should order a simple and highly sensitive and specific test for a common, deadly, and curable disease.

These are two ends of the spectrum, and we encounter every possible permutation in between. Sometimes the standard of care demands a certain approach, at other times we are left to our own judgement. For example, in patients over age 50 who present with a new headache it is the standard of care to order a sedimentation rate as a screening test for temporal arteritis. Even though the diagnosis is unlikely (given everyone over 50 presenting with a headache), the test is a simple blood test, and the diagnosis is highly treatable and very severe when not treated, potentially leading to rapid and irreversible blindness. The test is highly sensitive but not very specific, so when it is positive it is usually followed up by a more invasive biopsy, which is highly specific.

On the other hand, when patients over 60 present with dementia the most likely diagnosis is Alzheimer’s disease (AD). We do not, however, perform any testing for AD, because at this time the only useful testing would be a brain biopsy, and this is not justified because it would not affect our management. Instead we order imaging, EEG, and blood tests looking for treatable causes of dementia, even those that are much less likely than AD. If the standard treatable causes are ruled out, then we make the diagnosis of “Alzheimer’s type dementia” and treat that symptomatically.  AD is a pathological diagnosis and we cannot make it without tissue, which is not worth getting at this time. There is some benefit to having a tissue diagnosis for family history purposes, but this can be obtained at autopsy with no risk to the patient.

Another way to combine all these factors is to consider the overall risks vs. benefits of several clinical approaches. Making a specific diagnosis with a laboratory test is just one approach. Sometimes it is easier and better to simply treat a probable entity rather than test for it. If the treatment is fairly benign and effective, and the test is less so, sometimes treating without testing is the better approach. Sometimes the time it would take to get the results of the test are simply too long, and treatment decisions have to be made in the meantime.

Screening tests

There are also different clinical contexts for diagnostic testing. A diagnostic test (the context above) is performed on someone who is symptomatic and in whom there is reason to suspect the specific diagnosis. A screening test is performed on a population before they are symptomatic in order to either assess the risk of developing a disease or detect a disease very early in its course when it is more treatable or to prevent morbidity.

The topic of screening has been discussed many times on SBM. The counterintuitive point that often needs to be made is that more screening is not always better. It’s possible for the negative consequences of testing to outweigh the benefits. This threshold is usually determined by how likely it is for a positive screening test to be a true positive vs. a false positive. Even for a test that’s 99% sensitive and 99% specific (which is better than most diagnostic tests), if the prevalence of the disease being screened for is one person in 1000, and you screen 100,000 people, that will result in 999 false positive tests, and 99 true positives (with one false negative). The false positives greatly outnumber the true positives. You then have to consider what the response is to a positive screening test, which may be a more invasive follow-up test, or a treatment that has its own risks. You even have to consider the anxiety and stress produced by the false positive tests, if this is for a serious or stigmatized disease.

Target populations also have to be identified. Screening tests do not necessarily have to involve the general population. They can be targeted at high risk populations, determined by age, sex, family history, or other risk factors.

Conclusion

This discussion of criteria for diagnostic testing relates to Part I of this series in which I discussed the utility of making a diagnosis at all. It is not always necessary to optimally manage a patient. I find that patients, meaning the lay public, often assume that more testing is always better, and that making a highly specific diagnosis is a prerequisite to proper management.

The reality is that diagnostic testing is just another part of the risk vs. benefit calculation at the heart of all clinical decision-making. Often the most difficult decision to make is to decide which tests not to order. It is the general experience of doctors, backed up by published data, that specialists tend to order fewer tests. With greater experience and knowledge in their specific area, they are more likely to perform a targeted workup and avoid the “gunshot approach.”

This can often be difficult to explain to an anxious patient who wants a diagnosis. This can further place a great deal of pressure on the doctor to simply order the test (combined with liability to lawsuits when unlikely outcomes turn out to be the case). These factors combine to drive up the costs of health care, a consequence that is getting increasing attention as these costs continue to rise.

One solution, which does occur but needs to have a much greater place in the practice of medicine, is published guidelines and standards. Doctors can feel more confident in not ordering a test if published guidelines tell them it is not necessary. A published standard of care also effectively shields them from lawsuits – you cannot sue for a bad outcome, only for failing to practice within the standard of care.

Diagnostic guidelines that are well-established and published need to be communicated more thoroughly to doctors in practice, and we can also do a better job at monitoring compliance to such standards.

Other entries in this series

Part I
Part III

Posted in: Science and Medicine

Leave a Comment (73) ↓

73 thoughts on “Clinical Decision-Making: Part II

  1. Janet says:

    Where does, “your insurance probably won’t pay (for surgery) unless we do this test” fit in? I had to have what I view as an extremely invasive test to confirm carpal tunnel syndrome before I could have the surgery, even though two doctors seemed certain that this was the problem (and the test confirmed). I became so distressed that they stopped the test early, but verified that I had it in both hands anyway.

    I also had to suffer through years of misery because insurance demanded that all sorts of useless remedies be tried before granting me a hysterectomy.

    I’m all for best practices and appropriate testing, but the ultimate use of these needs to be left to physicians, not to for-profit insurance companies.

  2. WilliamLawrenceUtridge says:

    I’m glad I’m healthy enough that the main outcome of this post will be to change how I watch a TV show :)

    Again, very interesting to get insights into the clinical decision making process of doctors.

  3. Janet – agreed. Insurance companies sometimes impose suboptimal management onto patients in an attempt to reduce costs. It’s not even clear if this practice reduces costs. It probably doesn’t.

    But – here is the complexity. Insurance companies also sometimes refuse to pay for tests that are not indicated. They have the algorithms in front of them and can provide a check on excessive and wasteful testing. At the very least they make doctors justify their tests. It’s a pain, but it’s not all bad.

    Insurance companies sometimes have to deal with an even higher level of interference, from state legislatures who pass ridiculous laws forcing insurance companies to cover unscientific treatments, which then forces insurance companies to write policies to mitigate stupid legislation.

    So we have politicians telling insurance companies what to tell doctors what to do.

    Perhaps what we need are expert panels if third parties who can create insurance company policies that are evidence-based and cost effective, and balance these checks with doctor-patient autonomy.

    It’s complicated, but we can do better.

  4. mousethatroared says:

    Thanks Steven Novella for another informative article! As a patient and parent of a patient, I have noticed bits and pieces of your points in observing our doctor’s decision-making and recommendation, but it is very helpful to have it pulled together with an explanation of the reasoning behind those decisions.

    Just a thought on this quote
    “One solution, which does occur but needs to have a much greater place in the practice of medicine, is published guidelines and standards. Doctors can feel more confident in not ordering a test if published guidelines tell them it is not necessary. A published standard of care also effectively shields them from lawsuits – you cannot sue for a bad outcome, only for failing to practice within the standard of care.”

    I feel published guidelines and standards is also helpful for patients to see. Doctor’s don’t always have the time to offer an in-depth explanation for why they are (or are not) ordering certain tests or the patient may not think to ask for an explanation during the office visit. Understandable guideline help the patient to check their doctor’s recommendation against a standard. If the patient see that the doctor’s recommendations match the standard, they can feel assured that the consensus of the medicine field and/or science is behind the recommendation. If they see that the recommendations are not standard and they don’t understand why, that may be a good opportunity to follow-up with the doctor to better understand the doctor’s reasoning or clarify the plan before preceding (time permitting).

    I just wanted to add, about 4 clicks before checking on this site I was looking at an table outlining the sensitivity and specify of standard auto-immune tests. I think understanding these two concepts is a very useful skill for any patient who may deal with lab tests regularly. I think I’m going to read your paragraph two more times, let it soak in for the day then check back on that table and see if I understand it any better. :)

  5. Zetetic says:

    “Gunshot Approach”?

    I’d always heard this referred to as the “Shotgun Approach”

  6. weing says:

    Could be lysdexia.

  7. cervantes says:

    “Specificity is that probability that the patient has the disease if the test comes back positive.”

    Uh, no it isn’t. This is a common misconception. Specificity is the inverse of the probability that a person who does not have the disease will come back positive. The actually probability that the person will have the disease depends on Bayes theorem.

    Suppose a test is 90% specific; and 1% of the population has the disease. If you test everybody, you will get, on average, almost 11 positive tests. (.9 * the one true positive; plus .1 * 99 true negatives = 9.9+ .9=10.8)

    However, of those approximately 11 positive tests, only one person has the disease. The true probability of a person who tests positive having the disease in this case is 1/10.8= .093, or less than 10%.

    Most physicians do not understand Bayes theorem and make this same mistake. Big problem.

  8. weing says:

    Specificity tells what is the percentage of negative test results in all those that do not have the disease in question.

  9. MTDoc says:

    @weing

    I love “lysdexia”!

    @”gunshot approach”

    Got a chuckle out of this, but we must forgive our urban brothers when it comes to understanding firearms. It is called the shotgun approach because a shotgun (colloquially called a “scatter gun”), because if you get enough shot out there you are bound to hit something. This was actually encouraged in my internship days (1961), when we were given a list of 20 lab tests to order on everyone being worked up for hypertension. On rounds you had better be able to recite any and all lab values for any and all patients on your care list. Those were the days.

  10. David Weinberg says:

    As cervantes and weing state, specificity is the likelihood that IF a patient does NOT have the disease, the test will be negative. If the specificity is .95, a patient with out the disease will have a positive result (false positive) 5% of the time.

    The probability that the patient has the disease if the test comes back positive is the Positive Predictive Value. Positive Predictive Value is a function of the Sensitivity, Specificity, and prior probability that the patient has the disease.

  11. mho says:

    Could one of you comment on the false/true test numbers if a test is repeated over time?
    How does the probability change if the people with the positive tests, are re-screened in x amount of time and get another positive test? or, negative test, with a positive on 2nd screening?

  12. cervantes says:

    PPV is indeed the single best measure of the predictive value of a test. That physicians in general do not understand this, and conflate it with the specificity, is really a catastrophe. We’re all tearing our hair out because we can’t seem to get it across. Since it depends on the prior probability, however, it will vary according to the population being tested, and will be higher if there is a good reason for prior suspicion — which may not really be quantifiable, alas. Hence if you do strongly suspect a diagnosis, and the test comes back positive, you can take that as fairly strong evidence. But if it’s a shot in the dark, most conditions have low enough prevalence that even a good test doesn’t have high PPV.

  13. mousethatroared says:

    Okay….so I have a bunch of test results and I have their sensitivity and specificity values and I have a general idea of the strength of suspicion of the diagnoses. Where can I find that prior probability calculator?

    hehe, only joking folks.

    But, how does a doctor use a prior probability? If your prior probability of having a disease is 70% how would your treat that differently than if it was 40%?

  14. MTDoc says:

    @MTR

    I know experience is a dirty word, but it does play a role here.

  15. cervantes says:

    Right MTR, you probably don’t have a prior probability calculator so, as I said, clinical judgment is still required.

    Your second question is also a good one. For example, it’s now convention to order a CT scan in cases of clinically diagnosed acute appendicitis before going to surgery. But there is no evidence that the rate of negative appendectomy has gone down, or outcomes of laparotomy (now usually replaced by laparoscopy) have improved since the introduction of CT scanning. Exactly how the test changes diagnostic thinking and treatment decision making is one key to whether it’s worth doing. (The other being, I suppose, whether that change is for the better.) In the case of acute lower right quadrant pain, even if you don’t have appendicitis, there’s a good chance you have something for which opening up your abdomen would be indicated; and that is in fact the gold standard for diagnosis of appendicitis!

    Once you have a strong suspicion of appendicitis based on clinical presentation and WBC, the CT scan probably isn’t going to change your mind no matter what it shows. So why do it?

  16. weing says:

    @MTR,
    When the doctor takes your history and examines you, he comes up with a differential diagnosis and he has a sense of how likely those diagnoses are based on his experience. So if the most likely diagnosis has about a 70% probability of being correct, then a test with a high specificity would push that to 90 percent. A highly sensitive test if negative, would push the probability of his diagnosis into the uncertain category.

  17. MTDoc says:

    @cervantes

    In my approach to RLQ abdominal pain, with appropriate physical and lab findings, the diagnosis is “acute surgical abdomen” probable appendicitis. In my day, you operated and were correct 85 to 90% of the time. Sometimes you found something else, like an intersusseption, which warranted even more prompt intervention. Now you spent a $1000, delay surgery four hours, and get the same results. There is a middle ground somewhere, but we haven’t found it yet. Also protocols can be a double edge sword. Our legal system may also hold us responsible for performing everything on the list. We don’t actually consult a recipe every time we make a diagnosis.

  18. mousethatroared says:

    MTDoc – Yes, experiences maybe teaches you not to get too excited about getting a prior probability value before you know what you would do about it…if anything. :)

    No disrespect meant to doctors on my calculator joke, I have no intentions of attempting to do their job with the help of google and online calculators…it’s just that understanding the process makes me more comfortable and enables me to use my office visits more effectively.

    For the most part reading up on the process helps me view doctors more positively. In the past I’ve mistaken some doctor’s nebulous explainations as being needlessly evasive or just downright blowing off my complaints. When I look at test results with a better knowledge of the margins of error, I can see that many of those explainations were just honest, given the uncertainty of the testing available and considering appropriate caution with the available therapies.

  19. mousethatroared says:

    weing – Thanks, that is helpful.

  20. mousethatroared says:

    Those last comments, should have been addresses to Cervantes as well. I appreciate the clarifications.

  21. Scott says:

    For example, it’s now convention to order a CT scan in cases of clinically diagnosed acute appendicitis before going to surgery. But there is no evidence that the rate of negative appendectomy has gone down, or outcomes of laparotomy (now usually replaced by laparoscopy) have improved since the introduction of CT scanning.

    In some cases the CT makes things very much worse… such as when the instructions given to the patient’s family are screwed up, so it happens four hours late, which means the results are read by a junior resident, who somehow manages to completely misread a textbook presentation and send the patient home for bed rest, so surgery is delayed till the next morning (when the patient’s PCP reads them – the immediately following call to the patient to head back to the ER right now being then followed up with tracking down said junior resident and yelling at them for a while), so the appendix in question bursts in the ER while waiting for the surgeon to come down to evaluate.

    Not that I happen to have seen that, or anything. On another note, I *do* know that when an appendix bursts, having the surgeon show up to authorize some morphine becomes a Very Good Thing. Entirely unrelated, of course.

    Just an anecdote, of course, but it illustrates quite nicely IMO that the consequences of unnecessary tests can be profound.

  22. windriven says:

    “One solution, which does occur but needs to have a much greater place in the practice of medicine, is published guidelines and standards”

    Amen.* My only concern is that guidelines and standards in some settings lead to a type of institutional stenosis. Once implemented they can become holy writ, never questioned and never reevaluated.

    * I hope that someone will come up with a neologism that captures both the simplicity and finality of amen without the religious overtone. Hear, hear is too dated. Well said is too stilted. F*** yeah is au curant – at least among the younger set – but suffers being rather coarse.

  23. MTDoc says:

    @windriven

    I agree with your basic tenets, especially since most primary care now, and more so in the future, will be delivered by PM care providers with a fraction of the training and experience that my generation of physicians was expected to have. The competent physician is aware of these “commission advisories” and considers them carefully. Every FP spends thousands of hard earned dollars every year to keep up with the latest from the Ivory Towers. Even though retired for 15 years, I spend about $500 plus each year on CME (continued medical education). However in 50+ years in this profession, I’ve seen many established practices come and go. I’m really concerned that practices, once established, especially in a government run system will become cast in stone for decades. As science oriented professionals we must not let that happen. Incidentally, I have always found your thoughts and analyses productive.

  24. windriven says:

    @MTDoc

    The interesting question to my mind is: how do we harvest the efficiency and quality control that guidelines and evidence based protocols promise while avoiding descending into a type of cookbook medicine that ends up with too many recipes for tuna casserole and jello salad.

    I’d be interested to learn your thoughts about team approaches to primary care with NPs or PAs working with and under the direct supervision of family medicine specialist physicians. It seems to me that this approach is an inevitable part of the future but MD friends and acquaintances have decidedly mixed thoughts (and all but one are in specialties other than family practice so…).

  25. MTDoc says:

    @windriven

    Your request deserves more than a quick answer. Let me think about it and see if I can give you a few observations from a primary care perspective. But I’m a bit tired this PM, so if you could check back on this thread tomorrow, I’ll try to respond more specifically.

  26. windriven says:

    @MTDoc

    Happy to wait. Many thanks. JT

  27. Great post. But there is a failure to note a diagnostic path with suspected AD:

    “On the other hand, when patients over 60 present with dementia the most likely diagnosis is Alzheimer’s disease (AD).”

    In my experience, patients don’t quite “present with dementia.” They present with memory complaints, or their family member does. This could be due to drug side effects, regular alcohol use which is not uncommon in older adults, due to delirium which can arise with life changes such as moving to a new place or being hospitalized, or due to depression which is not uncommon in the elderly, and is more common after a heart attack, and is more common when the person has other chronic medical conditions. The likelihood of substance abuse, drug side effects, depression, and AD might all be fairly equal in a 70yo presenting with memory difficulty and a few-odd recent quirks. The older adult could be seeking disability, or a family member could be trying to set them up for disability.

    So, hitting those with scattershot is a good idea.

    “AD is a pathological diagnosis and we cannot make it without tissue, which is not worth getting at this time.”

    -Limited treatment options largely make it not worth the effort to discriminate between AD and the next 2 or 3 most prevalent dementias. But if you wanted to make a fairly certain diagnosis, you can, with the help of a neurosychologist. AD can be discriminated from multi-infarct dementia, Huntington’s Chorea, Diverse Lewy Body, etc. There are even clever tests, such as the TOMM, to try to catch malingering.

    I worry that the way physicians think makes them follow the lead of the medical model, and ignoring the “biopsychosocial” model. Nothing I have said here is uncommon, unreasonable or esoteric.

  28. DugganSC says:

    @mho:

    I Am Not A Doctor, but my suspicion is that there’s a further breakdown for any test between probability of error (something went wrong when doing the test to provide a false result) and false indication (the result of the test was correct, but it doesn’t indicate what you think it does). If there’s error in the former, repeating the test will yield different results (one time, the doctor tested your urine and the second time, he accidentally tested that of a patient with a number one of from yours). If it’s the latter, you may get consistent results, but incorrect ones (the doctor finds elevated levels of a particular chemical in your blood and says you have Disease A, but your results are actually caused by Disease B, which causes the same chemical to appear).

    Incidentally, anyone looking for a medical analysis of House would probably be well served by looking at Dr. Scott’s reviews. One of the interesting factors is how often the Disease of the Week only gets missed because they either forgot to do standard diagnostics on the patient when they came in, or didn’t pay proper attention to them. House’s maxim of “everyone lies” would probably be better replaced by “people are lazy”.

    I think an interesting question is how much disclosure the doctor should give about the relative probability of correct diagnosis. When the patient asks whether a test will detect their cancer, do they just say “yes” or do they give a longer answer that Test A tests for this hormone and has a 95% specificity while Test B tests for a particular chemical and has 97% specificity, etc, and even if all of the tests come up clear, there’s only an 87% chance that that’s the correct diagnosis. And how much responsibility does the doctor have to test in multiple ways to help ensure that they eliminate cases of a test providing the correct answer to the wrong problem, or to cover for known errors in the testing methodology. It’s honest to disclose all of the risk, but humans are bad with probabilities, and having a doctor admit that a given test provides the wrong answer 10% of the time, or that it results in 1% false positives, can result in lack of patient confidence, particularly when the friendly chiropractor down the hall is offering 100% accuracy of spinal subluxation diagnosis or your money back.

  29. Meds – For the sake of this article, which was a discussion of ordering tests, I skipped over the step called phenomenology, which is what you are talking about. What I laid out is what happens once the diagnosis of dementia is made.

    But you are right, prior to that we have to determine that what the patient has is dementia, as opposed to pseudodementia from depression, malingering, anxiety, a drug side effect, sleep deprivation, etc. The history and physical is actually quite good at making this distinction, and sometimes we do use neuropsychological testing to confirm our clinical diagnosis.

    I have to disagree with you a bit about confirming AD with neuropsychological testing. These specificity of this test is not great. It is good at determining Alzheimer’s type dementia, and may give clues as to fronto-temporal dementia, multi-infarct, vs AD, etc., but I don’t think at can establish the diagnosis of AD. There are a number of biological markers that have emerged (it seems there’s a new one every month) but none have been validated as clinically useful. At present the only way to confirm the diagnosis is with tissue. This will probably change in the future, however.

  30. mousethatroared says:

    DugganSC – IME – doctor’s wont (can’t?) give you such precise percentages, but a good doctor will give a patient a general idea of the reliability of the tests being used and whether the false positive/negatives are high/low. They can tell the patient how their symptoms confirm or conflict with the testing and what conclusions that leads them to. They can say generally how certain they are of the conclusion (very certain, kinda certain, not at all certain). They can also explain how that certainty is incorporated into their assessment of risks (risks from illness, treatment or further testing) and their recommendations. Does this make sense?

    Not all doctors are going to launch into this explanation without prompting, many patients would just be overwhelmed with it. But, I think I would be hesitant to go to a doctor who couldn’t or wouldn’t have that sort of discussion about a diagnoses when asked. Just IMO.

  31. MTDoc says:

    @windriven

    Good morning. My 2 cents on team approaches to primary care. The model you illustrate “with NPs or PAs working with and under direct supervision of FPs” is exactly how the “midlevel” providers started out 40+ years ago. I’ve worked with many midlevels over the years, beginning circa 1970 with the medex program, and later with PAs and NPs, in both private practice and HMO settings. We initially had to review and “sign off” on their work and, of course, take full legal responsibility for their medical decisions. It soon became apparent that this was no more than a chart audit, sort of like making a diagnosis over the phone. Besides it took almost as long to do a decent review as it would to see the patient yourself. So what happened next was they all became de facto independent practioners, in fact, in most states are completely independent so far as supervision is concerned. The best ones did in practice try to work as part of a team, if they could catch the MD between patients. The worst ones never seemed to need advice (I’m thinking of a NP who was absolutely sold on Rolfing and magic diets.

    So the theory is good, but the devil is in the details. The MD is scheduled out two weeks with “chronic patients” leaving acute and potentially critical patients to be picked up by the midlevels. These include flu like illnesses that turn out to be bacterial meningitis or indigestion that is new onset angina. The economics are such that the physician doesn’t have time to supervise others and carry his own load. And that load is now imposed on him by his employer, keeping in mind that the era of private practice is fading fast.

    “How do we harvest the efficiency and quality control that guidelines and evidence based protocols promise while avoiding etc.,etc.”

    I don’t know, but you are asking the right question. As a primary care doc, I belong to another generation. We were trained broadly enough in those days to serve a community’s basic needs. This included OB/GYN, most general surgery, and much specialty surgery. I made a decent living with a $6.00 office call and I paid $300 for malpractice insurance. That insurance, were I to be foolish enough to offer my services to the public, would be about $40,000 a year for a limited policy.

    Didn’t mean to rant (remember I’m old), but hopefully a bit of history, biased though it may be, will prove useful to an understanding of the problem.

  32. windriven says:

    @MTDoc

    Thanks so much for taking the time to share your thoughts. I am not a physician but have spent my professional life manufacturing medical devices. I’ve had the opportunity to see some of the issues complicating health care policy close up. Others, like delivery of primary care, I know only as a consumer.

    I’m old enough to remember private practice as you knew it. The doctor who delivered my father in my grandmother’s home delivered me – though in a hospital. He also removed my tonsils and adenoids, my aunt’s gall bladder, etc. Alex Marcus, MD. RIP. I’d love for those days to return but I think they died with Andy Taylor.

    So we are left with an aging population that will necessarily require more medical attention but also with intense downward pressure on medical spending. There is no stomach in this country for the sort of reforms that will bring our per capita health care spending in line with that of, say, Australia. I expect that spending will stay in the neighborhood of 18-20% of GDP while quality, especially in primary care, decreases.

    Would it surprise you to know that there are fewer physicians per capita in the US than in most OECD countries? And Germany has twice the number of physician visits than the US. I’ll bet this will surprise you even more*:

    “Although hospital stays were relatively infrequent and
    short in the U.S., hospital spending per discharge
    far exceeded all other countries at $16,708—nearly
    triple the OECD median of $5,949. ”

    Some of the differential between US and other OECD hospitals can be explained by uncompensated care. But triple???? I think not.

    I don’t mean to rant either. But this issue just infuriates me. It is inexcusable that we spend double what many other OECD countries spend yet score only in middle ranks on quality.

    Anyway, thanks again for taking the time.

    *OECD Health Data 2010

  33. pmoran says:

    For example, it’s now convention to order a CT scan in cases of clinically diagnosed acute appendicitis before going to surgery. But there is no evidence that the rate of negative appendectomy has gone down, or outcomes of laparotomy (now usually replaced by laparoscopy) have improved since the introduction of CT scanning. Exactly how the test changes diagnostic thinking and treatment decision making is one key to whether it’s worth doing. (The other being, I suppose, whether that change is for the better.)

    It may come to depend upon the available clinical skills. Accuracy in clinical diagnosis of appendicitis is heavily dependent upon those. They also impart the confidence to subject patients who do not have signs of peritonitis to a half-day or so of observation, which will very often resolve the matter either way.

    My concern is that those clinical skills will not fully develop in the CT scan era. Why expend the necessary time and effort on that when much of the responsibility can be shifted onto a test — CT scan, or, somewhat worse for the patient in most cases, laparoscopy?

    In the case of acute lower right quadrant pain, even if you don’t have appendicitis, there’s a good chance you have something for which opening up your abdomen would be indicated; and that is in fact the gold standard for diagnosis of appendicitis!

    I can’t agree with the first part of this sentence. Especially in children and in women, self-resolving causes of RLQ pain are very common.

  34. windriven says:

    @pmoran

    So Peter, would it be feasible to prepare a diagnostic flowchart for a NP or PA that would take one through the steps for RLQ pain? Would instances of uncertainty be clear enough for a PA to recognize and to move the patient to the supervising MD for additional dx? Or would the differences be too subtle?

    There are engineering and manufacturing problems that I encounter that are easy for me to diagnose because of long experience. Other problems are easy enough for any competent technician to diagnose. Usually a competent technician knows when she is out of her depth. But I don’t know if the metaphor is valid. And the price when the technician is wrong isn’t often life or limb.

  35. MTDoc says:

    @pmoran

    “available clinical skills”. An important point. After all, there is a continuum between peritoneal “irritation” and a ruptured viscous that could all be called peritonitis. What happens, is the midlevel provider orders a $1000 test, and then calls on someone capable of making a competent decision without it. By the way, if the RLQ pain was due to mittleschmertz , I’m not sure I would order a CT, but then I never needed one.

    @windriven

    Midlevels are supposed to work on such a protocol, but such a flow chart in the present context may be a product of the hospital employer, who has a vested interest in utilizing his expensive diagnostic equipment as well as insuring that he doesn’t get sued. My flowchart for such a presenting complaint is “call a doctor”.

  36. pmoran says:

    So Peter, would it be feasible to prepare a diagnostic flowchart for a NP or PA that would take one through the steps for RLQ pain? Would instances of uncertainty be clear enough for a PA to recognize and to move the patient to the supervising MD for additional dx? Or would the differences be too subtle?

    Beyond “if they are tender in the RIF, assume appendicitis”? For the inexperienced that is probably the only safe advice.

    Mind you, even assessing tenderness takes rudimentary skills. Poke anyone briskly in the RIF (right iliac fossa) and they are likley to jump. I once saw a Australian surgical registrar (a kind of intern) test for rebound tenderness ( asign of peritonitis) in such a way that it will nearly always be positive.

  37. windriven says:

    @pmoran

    Thanks. So how is it that Australia delivers great care at half the cost per patient in the US?

  38. elburto says:

    @windriven – not pMoran and not Australian, but I suspect that the grossly overinflated cost of US healthcare has much to do with the multiple layers of authorisation and clerical support needed for even a simple dx.

    Let’s take something really simple like an ear infection.

    Here in the UK I call the surgery/go to the secure website and make an appointment with my GP.

    I arrive there and sign in via a touchscreen interface, then take a seat. After ten minutes or so the scrolling ticker on the wall says “Miss Elburto – Dr G in Room 5″

    I see him, he diagnoses the problem and prints out a prescription. I trundle around to the chemist’s shop on the next street, hand over my prescription and receive a blue and white bag of meds in return.

    I’ve seen and interacted with two humans, the GP and the pharmacy assistant. No authorisation was necessary for the appointment or the drugs. My records were updated electronically by my GP as I sat there while the script printed out.

    No cash changed hands at all, no payment was necessary at any point.

    If I hadn’t had an NHS prepayment card for my prescriptions (£104/year, unlimited use) or wasn’t exempt from charges (under 18, pregnant, in receipt of certain benefits, a pensioner, suffering certain chronic illnesses) I would have paid a standard fee of about £8 per item. Every NHS prescription item costs the same, whether it’s antibiotics or fentanyl. All contraceptive options whether prescribed (pills, patch, IUD/IUS, Depo shot, etc) or nonprescription (any type of condom, dams, replacement diaphragm etc) are exempt from charges for anyone in the UK.

    I know that the US method has more layers of authorisation and administration built in. Friends who want someone relatively simple like hormonal birth control have told me of the costs and difficulties of seeing that through. From unnecessary tests (pelvic exam and often. breast exam too, sometimes a cervical smear and STI tests as well), to difficulty getting insurance authorisation for certain brands or formulations/regimens, to multiple layers of costs.

    Here you go to see the practice nurse or to the local NHS sexual health clinic. Discuss your options and answer questions about medical history (answers can be confirmed with your electronic patient record too if you have any uncertainty) , if a hormonal contraceptive is an option then blood pressure and physical health/family history and patient needs are used to help the care provider choose a particular formulation, delivery method and schedule.

    If an implantable device is chosen it can be fitted there and then, or scheduled for whenever is convenient.

    Patient leaves with their BC of choice fitted/injected, or with a prescription (that can be filled at any high street chemist shop) and if they’ve visited a sexual health clinic, a bag of condoms so big that if inflated they could easily raise the Titanic.

    Hormonal BC for home use is given typically in 6 month supplies. When they’re about to run out you can usually go online to renew the prescription and have it delivered to your home (at no cost) or collect it within two working days. Most prescriptions can be refilled seven days before they’re due, I hear that it can be tricky to refill a prescription for HBC in the US and get it authorised in time to ensure that your protection from pregnancy is not compromised, especially if you’re away at university and not at your parents’ home address, or if your prescription is for Nokids28 generic A (Uterempty28) but your insurance now covers generic B (AtUrCervix28) only.

    Sorry that was so long, but I wanted to explain the full process of a simple medical issue from onset to treatment, and of something commonplace like acquiring contraception.

  39. elburto says:

    @Scott – Yes, morphine is just the ticket for calming down a stressed supervising physician whose underling has bollocksed up, causing a belly full of pus.

    Oh wait… the morphine’s for the patient, I see!

    I’ll never forget being that overlooked patient ( pain escalating rapidly due to a diagnostic error) and seeing a horrified urology consultant practically leap to my bedside with a syringe full of morphine. It was amazing, like someone had stripped away my spine, my head literally bounced back against the end of the trolley.

  40. windriven says:

    @elburto

    Thanks for a close look at the UK model. It is true that there are far too many layers in the US model. But when you pull apart the numbers of where money gets spent it is hospital care that really screams out. Here too, layers of bureaucracy clog the process.

    Returning to your discussion of birth control – and noting that I am not a physician – does it seem prudent to supply hormonal birth control without a pelvic or breast exam? The hormones involved can, I understand, accelerate the growth of certain tumors. It might be interesting to compare the rates of ovarian, cervical and breast cancers among reproductive age women in the two countries. Any difference certainly wouldn’t prove causation but the absence of a difference would suggest careful research into the necessity of the exams you mention.

    Prescriptions are generally quite easily obtained and filled here – depending on one’s insurance coverage. On my plan most prescriptions have just a $10 copay. But absent insurance or with a plan with limited drug coverage, cost can be crippling.

  41. DugganSC says:

    Honestly, I think part of the problem in the US is that two major steps involved, insurance and hospitals, are corporations run by CEOs whose job is to maximize money. Theoretically, that involves doing the best job they can, but too often, it involves CYA moves, obstruction, and outright feuding between the two over what tests are necessary, which are covered, and how long it takes to get them. The doctors get caught in between these two juggernauts, afraid to make a move that will result in reprimands, lawsuits, or firing. When was the last time you had a doctor glance at your CT scan results and give you an honest answer instead of frowning and telling you that you need to schedule a meeting with the neurologist, who won’t be available for another month?

    Of course, the other issue is that there’s been an arms race of charges and counter-charges. The hospitals have inflated prices on the chargemaster sheet to ensure they get some fraction from the insurance companies. The insurance companies know they’ll only pay a fraction of that value. Most patients with insurance know they’ll only pay a co-pay. And people without insurance, or with inadequate insurance, get screwed. If you thought the government was crazy with $500 hammers, just wait until you find the $18 gauze pad on your hospital bill… Short of a voice on high decreeing a unilateral reset, neither side can back down.

  42. windriven says:

    @DugganSC

    I think you’ve hit the nail pretty squarely. It is easy enough to look at the numbers and see that hospitals are a huge driver of the cost of health care (compared to physicians’ services, for instance). Since insurance as we know it isn’t a feature in many other countries the only comparison is to zero (the cost to governments being more difficult to tease out) and they certainly don’t compare favorably to zero.

  43. DugganSC says:

    {nods} Probably a more fair comparison is the Medicare prices, which are supposed to reflect actual costs, including supplies, personnel, and operating costs. The Hospitals claim that all of Medicare is shorting them by 5-10%, but then they turn around and have a 2000% markup on all of the non-Medicare bills. It’s insane.

  44. elburto says:

    @windriven – The US has a peculiar obsession with pelvic and breast exams.

    The yearly “physical” for US women is not a feature of any other system that I’m aware of. Even the ACOG now say it is unnecessary.

    The altogether bizarre phenomenon of the “baseline” cervical smear for teen girls or anyone female who has never engaged in sexual activity, as a prerequisite for a BC prescription is just mindboggling when you consider how most cervical abnormalities arise.

    A bimanual pelvic examination can’t detect much of anything really unless we’re talking about a significant issue, let alone latent malignant cells within the body. Overly frequent breast exams lead to false positives and invasive diagnostic procedures.

    In the UK women get their first cervical smear at 25, then at intervals of three to five years after that if no abnormalities are detected. Women who’ve never had sexual contact with another person do not need the test if they’re asymptomatic, so they can opt out. Anyone can refuse the test, and no prescription will be held hostage as a result.

    The UK and US have remarkably similar cervical and breast cancer detection and survival rates. Hypertension is far more of an immediate danger to a new HBC user, as is a family history of DVT, or personal history of migraine with aura, that’s why history and a BP check are so important.

    WRT solely to this yearly pelvic exam, I have talked extensively with US women about their experiences. Some are survivors of sexual abuse who live in fear of the exam, so are unable to get prescription BC. Obviously this can lead to unwanted pregnancy. The same is true for people with gender dysphoria who, again, cannot tolerate an invasive genital exam or breast exam. Again, there’s a risk of pregnancy.

    Some people simply cannot afford the expense of an appointment, exam, workup, often with STI and blood tests so again, prescription BC is out of reach. I know of women who go to Mexico and buy the pill OTC, or order from online foreign pharmacies, which is really worrying.

    The UK’s free contraception and sexual health mandate even extends to recent immigrants and foreign nationals who are students here or working.

    There’s one very simple reason, dishing out STI tests is cheaper than taking care of someone who’s acquired something like HPV or HIV. Preventing pregnancy and making access to first trimester abortions a priority is cheaper than a lifetime of NHS care for a new citizen, and prenatal and antenatal care, and labour and delivery. Giving out bags of 200 condoms with a smile and a nonjudgmental attitude increases the likelihood of people feeling unafraid to seek help for sexual health issues, so money is saved on treating conditions that have been allowed to progress out of fear.

    A Mirena intrauterine system can cost $800 in the US, which is a lot of money for something that may not work out. The women least able to afford contraception are obviously those least able to afford abortion or pregnancy care through to delivery. Insurance companies often refuse certain contraceptives to certain demographics. The constant back and forth between patient – PCP – pharmacy – insurance provider etc must add up so many dollars in manpower and paperwork, and that’s for something as simple as preventing pregnancy! I can only imagine what care of chronic conditions involves. I’ve heard horror stories about the nightmarish bureaucracy involved in pain management. I’ve been met with disbelief when I’ve said something like “The pain from $condition was agonising, so my GP prescribed morphine after nothing else worked. I sign the controlled drugs register at the pharmacy counter. if I’m not using my usual local chemist I take my passport as ID”

    Apparently this simple (for me) doctor-patient scenario is not how the same situation plays out in the US. It’s greeted with the same sort of reaction I’d expect to get if I said “Someone was giving away free fifty quid notes at the pizza shop!”

    Insurance providers insisting on so many steps, and what almost seems like a series of traps, before effective pain relief can be prescribed must add th to a small fortune.

  45. Jose A Hernandez says:

    It is true that in any screening test the problem of false positive is a significant burden, but worse is the problem of Overdiagnosis bias.

  46. Novella said: “I have to disagree with you a bit about confirming AD with neuropsychological testing. ” –I agree with you on this, regarding a confirmed diagnosis – as I as taught, and have seen via brain autopsy and histology. You comment abt phenomenolgy is great, also – when getting to particulars, this health care delivery endeavor is so complex, that it is impossible to simply make a brief comment abt an approach to a likely AD case, or almost any case, without shorting the issue i none way or another.

    I love this blog because the bloggers and commenters run over these issues, and either explore differences, or figure out agreement. With differences, ppl eventually either agree to disagree, learn something new, or if they are CAM devotees, give up and move on to plague some other blog.

  47. windriven says:

    @elburto

    You have made a number of excellent points. I must however take exception to this:

    “The UK and US have remarkably similar cervical and breast cancer detection and survival rates.”

    5 year breast cancer survival in the US is ~90% versus ~78% in the UK*. 5 year survival differential for cervical cancer isn’t as sharp but still favors the US by about 7%**. Of course it might well be that these numbers just mean that the US detects earlier so that the clock starts running earlier and American women only appear to survive longer. It is entirely possible that both populations survive for similar periods after actual onset of the disease, I just don’t have any data to either support or refute that hypothesis.

    ” I’ve heard horror stories about the nightmarish bureaucracy involved in pain management.”

    As have I. But my experience has been different. I’ve actually never sought pain management but friends and relatives who have are treated promptly and pretty aggressively. This may be a regional difference, a socio-economic difference, or something else. Lots of darkness; not much light. I wish it was easier to get reliable information.

    One area that is absolutely idiotic is that of birth control. For all our vaunted freedoms in this country we seem to be scared crapless of any body part that can’t be seen at the beach at a church social, much less the things people might do with those parts. Contraception should be available pretty much on demand. I’d much rather buy someone a pack of condoms than pay to raise a kid or two that never should have been conceived. Same with abortion. Free. If the rednecks don’t like it they should shut their whine-holes and start building extra bedrooms.

    Rant over. Thanks for listening. ;-)

    *http://www.oecd.org/els/health-systems/48098832.pdf
    **http://stats.oecd.org/index.aspx?DataSetCode=HEALTH_STAT

  48. Jan Willem Nienhuys says:

    Part of the problem with the terms specificity and sensitivity is that the words by themselves don’t suggest much. The nice thing about them is that you want both of them to be neer 100%. But the terms
    “false negative percentage” = percentage of negative (no disease tested present) results from patients that do have the disease
    and
    “false positive percentage” = percentage of positive (disease tested present) results from patients that don’t have the disease
    are almost self explanatory.

    I suspect that these terms are impopular with doctors because they emphasize that doctors can and do make mistakes.

    The prior probability that a patient has a disease is hard to obtain, because one is almost always dealing with patients that already have some complaints. Even if the prevalence of disease X in the population as a whole is low (say 0.1%), it is quite possible that given the particular symptoms the patient presents, the probability that the patient has X is 90% for one kind of symptoms and 20% for another kind of symptoms.

  49. BillyJoe says:

    “Part of the problem with the terms specificity and sensitivity is that the words by themselves don’t suggest much”

    It seems to me that the word “sensitivity” is suggestive but “specificity” is not.

    Which is why many of us seem to get the definition of “sensitvity” right but the definition of “specificity” wrong – like Steven Novella (who must be looking for a way to extricate himself from this error (: ), and Harriet Hall in a previous post (who did find a way to extricate herself from her error though not very convincingly (: )

    I suggest we can avoid this error by simply defining “specificity” AFTER defining “PPV”. Then we won’t make the mistake of defining “specificity” AS “PPV”.

  50. mousethatroared says:

    JWN
    “I suspect that these terms are impopular with doctors because they emphasize that doctors can and do make mistakes.”

    Sorry, I think I don’t understand this. How does it imply that the doctors* make mistakes to say that a test gives a certain number of false positives and negatives?

    *not that I think doctors don’t make mistakes, just that the test results aren’t their mistakes (unless they read them wrong).

  51. ebohlman says:

    5 year breast cancer survival in the US is ~90% versus ~78% in the UK*. 5 year survival differential for cervical cancer isn’t as sharp but still favors the US by about 7%**. Of course it might well be that these numbers just mean that the US detects earlier so that the clock starts running earlier and American women only appear to survive longer. It is entirely possible that both populations survive for similar periods after actual onset of the disease, I just don’t have any data to either support or refute that hypothesis.

    David Gorski wrote about this here some months ago; the data seem to point to exactly that (lead-time bias and stage migration) being the cause of the discrepancy.

  52. evilrobotxoxo says:

    Maybe Dr. Novella will cover this in later posts, but one issue I find interesting is that the concept of “diagnosis” is not always correct or optimal in medicine. In the early days of medicine, diagnosis was a big breakthrough because most illness had infectious causes. People either have syphilis or they don’t. They either have tuberculosis or they don’t. In those cases, accurate diagnosis is absolutely critical. However, today most illness is caused by chronic perturbations of normal physiology that don’t have an infectious cause. For example, one definition of hypertension is when your blood pressure is above 140/90. So that means that someone with a BP of 139/89 does not receive the diagnosis, while someone with a BP of 141/91 does? The point is that all of this testing is based on the idea that people have binary diagnoses, which is often a convenient fiction.

  53. weing says:

    “For example, one definition of hypertension is when your blood pressure is above 140/90. So that means that someone with a BP of 139/89 does not receive the diagnosis, while someone with a BP of 141/91 does?”

    No. The latter receives the diagnosis of elevated BP. Hypertension is when the average BP of three separate readings is above 140/90. At least that’s how I do it.

  54. mousethatroared says:

    windriven said “This may be a regional difference, a socio-economic difference, or something else. Lots of darkness; not much light. I wish it was easier to get reliable information.”

    Yes there may be. Also, I wonder if it’s realistic to expect to comparable quaility of service on every health care concern in a countries like the U.S. – 3,794,000 sq miles and had 313,914,040 people to the U.K. 94,058 sq miles and has 62,641,000 people.

    I appreciate that many folks outside the U.S. have criticisms of our system, which I agree, is flawed. But it does seem like they are taking the easy part of the task. It seems that people seldom offer likely political solutions for improving things….well at least solutions that wouldn’t take a time machine to implement.

    @windriven – nice rant, by the way.

  55. mousethatroared says:

    @evilrobotxoxo – Did you get a chance to read Clinical Decision Marking part I? SN does discuss diagnoses there.

  56. ConspicuousCarl says:

    If people regularly misuse the word “specificity” in the statistical sense, it might be because the statistical definition doesn’t make any sense.

    If we were good engineers and linguists but terrible farmers, and we wanted to develop an instrument to identify apples for us, we might say that redness is specific to apples and exclusive of oranges. That is, determining that the fruit is red is a way to specify that it is an apple, and exclude the possibility that it is an orange.

    But the statistical formula for specificity doesn’t include any enumeration of the thing which the test result is specific to. The ability of the test to not allow false positives in the positive results ought to be called exclusivity.

  57. Jan Willem Nienhuys says:

    mtr:

    Sorry, I think I don’t understand this. How does it imply that the doctors make mistakes

    If one orders a test, and it comes back negative one might say:
    well this test has a sensitivity of 95%. The patient probably does probably not have the disease. 95% is almost 100%.

    But this sensitivity also means that this judgement has a probability of 5% (1 in 20) of being wrong. If the other symptoms of this particular patient make it quite likely (say 80%) that the disease is present, then the total probability is 0.25 that the disease is present, or even more when the specificity is less than 100%.

    It is well known that when you present people with a choice:
    - 95% of surviving
    - 5% of dying

    they almost all vastly prefer the attractive 95% and they abhor any chance of dying. Especially if you ask two different groups and present each group with only one of the two choices, you’ll find that the question with the word ‘dying’ in it is vastly less popular. Kahneman and Tversky have done a lot of research into this.

    One would like tests to have a specificity (correct positive) and a sensitivity (correct negative) both of 100%, and then 80% and 95% etc. can lure the user or someone who orders the test into a false sense of security. If one always thinks in terms of ‘how probable is it that the result is wrong, that I am mistaken, that there is an error somewhere, then one naturally would shun terms that merely express that one (or the test one orders) is “almost” always correct.

    In any test that comes out positive, it’s not the percentage of “correct positive” that is relevant. It is the quotient
    of the correct positive percentage divided by false positive percentage. That quotient is a multiplyer. You use it to multiply the ‘prior odds’ of a positive result, giving you the ‘posterior odds’ of a positive result.

    Similarly. the quotient correct negative : false negative is the multiplyer for turning the prior odds of a negative result into the posterior odds of a negative result.

    Even though these prior odds are poorly known, one may often (I think) reasonably pick one of 100, 10, 1, 0.1, 0.01 (1 means even chance that the patient has the disease or not, 0.01 means: given all that is known before the test, there’s a chance of 1% that the patient ha

  58. Jan Willem Nienhuys says:

    OOPS – continued

    …patient hasn’t got the disease).

  59. mousethatroared says:

    JWN “If one orders a test, and it comes back negative one might say:
    well this test has a sensitivity of 95%. The patient probably does probably not have the disease. 95% is almost 100%.
    But this sensitivity also means that this judgement has a probability of 5% (1 in 20) of being wrong.”

    No, I don’t think so. The doctor is not wrong if he says that the patient is not likely to have the disease. He is just communicating the realistic probability. He is only wrong if he says that the patient does not have the disease and the patient was one of the 5% false negatives.

  60. BillyJoe says:

    JWN,

    “One would like tests to have a specificity (correct positive) and a sensitivity (correct negative) both of 100%”

    Probably just a typo coupled with a bit of imprecision but….
    sensitivity = true positive rate (percentage of those with the disease who test positive)
    specificity = true negative rate (percentage of those without the disease who test negative)

  61. mousethatroared says:

    also JWN “One would like tests to have a specificity (correct positive) and a sensitivity (correct negative) both of 100%, and then 80% and 95% etc. can lure the user or someone who orders the test into a false sense of security.”

    Yes, I do agree that a false sense of security is something that must be guarded against. I’m always skeptical of medical professionals who seem blind to the 1%, 5%, 10% chance and claim that something IS rather than probably is.

    They say don’t look for the zebra when you have a horse…but you want someone who isn’t zebra blind… (speaking from the experience of having a zebra son).

  62. Jan Willem Nienhuys says:

    Billy Joe

    Probably just a typo

    No. Just wrong. The only excuse I have that I have to look up the meaning of those two s..e..i..i..ity words each and every time I use them, and then double check to see whether I didn’t mix them up again. I omitted the double check. Wrong. Wrong. Is there an easy mnemonic? Something like ‘specificity has a p, but it’s NOT about the positives, so shun that word’ and ‘sensitivity has an n, but it is NOT about the negatives, so shun that word.’ Both silly words start with an s so shun them. A mnemonic that explains why you should forget about these words?

    mtr

    They say don’t look for the zebra when you have a horse

    Mmm. It depends where you are. If you are in a zoo (a hospital where all the difficult cases are sent to…) the prior probability of a zebra might be higher than you think. But if you work in a riding school …

  63. BillyJoe says:

    Thanks JWN for the honesty.
    I’m still waiting for Steven Novella to so likewise. ):

  64. mousethatroared says:

    JWN “Mmm. It depends where you are. If you are in a zoo (a hospital where all the difficult cases are sent to…) the prior probability of a zebra might be higher than you think. But if you work in a riding school …”

    Yes to a certain extent, but I’m not sure there are any riding schools in medicine. Meaning, I doubt there are a lot of riding schools where .1% (1% 3%) of their horses are actually zebras.

    Unless, you live in an area where zebras are prevalent and maybe they occasionally make a habit of slipping into the yard to steal feed…but then the nice things about real zebras, I’m guessing – I don’t really know anything about zebras beyond stripey and not domesticatable – is that they kick (or bite you or something) if one tries to treat them like horses. Also there is the distinctive stripey pattern. Whereas one can not count on a rare conditions conveniently attacking the doctor to call attention to itself or having a glaring color signal.

  65. Narad says:

    Is there an easy mnemonic?

    I’m pretty sure there’s a limerick in here, but I’m having trouble getting the form right.

  66. Narad says:

    OK, this is abysmal, but I got myself into it.

    Twas once a chap who was sensitive
    His stats were almost all positive
      Asked about falsity,
      With specificity
    He said, “Must you be so negative?”

  67. weing says:

    The only mnemonics I know are SNOUT, a highly SeNsitive test if negative will rule OUT, and SPIN, a highly SPecific test if positive will rule IN.

  68. agitato says:

    weing:

    Thank you for those very useful mnemonics!

  69. Chi_girl says:

    elburto- My brother lives in the UK and I am always jealous!

    To your point, depending on your age and history (previous pelvic exams/pap) there are clinics that will not do a pap and breast exam.
    Actually the guidelines have just changed here in the US, I believe pap every 3 years
    http://www.cnn.com/2012/03/14/health/brawley-cervical-cancer-screenings

    You are correct on the IUD, although I believe that differs by insurance policy.

    To anyone mentioning a 10 co-pay.. with HMO correct??

    I didn’t see this mentioned. A lot of hospitals are “buying” the local GP’s etc and are charging more for services, however when asked how they got these new numbers they were not able to give a straight answer- this is an example
    http://charlotte.cbslocal.com/2012/12/17/prices-rise-as-hospitals-buy-medical-practices/

    We are also not able to negotiate with other countries for better rates.
    I’m a patient in my 30′s and was misdiagnosed for a long time – no one bothered to really look at my MRI. I counted one time and I saw 29 physicians. I begged for a new MRI and my condition was discovered (by some tech in an open blurry MRI!) My physician “studied” this rare condition of mine and actually explained the condition incorrectly and decided to schedule me for contraindicated procedure!
    I had to find my own neurosurgeon – which I e-mailed .
    I just did my taxes and my prescriptions totaled around $1300 a month- and I have decent PPO insurance.

    That’s not including the $13k MRI and everything else..

    My mothers oncologist was also purchased by a hospital and she was given a drug she did not need, charged $5k for it. She now has terrible side effects from it (pain). Her new, respected University oncologist said the drug was not needed- ans NEVER was.

    I have worked in the medical field for 15 years and have started some clinical work this year- I am appalled at what “healthcare” is in the USA

  70. BillyJoe says:

    “That’s not including the $13k MRI”

    $300 for a brain MRI in Australia and the exchange rate is $US 1 = $AUS 1.03

  71. DugganSC says:

    @BillyJoe:
    Basically, the system as it stands now assumes that the price will be a highly elevated one. Most people with insurance will pay a small fraction of that and the insurance company itself pays only half of the chargemaster pricing. Medicaid pricing, which is what’s been agreed on as the actual price, including labor and supplies, is close to what you’re citing.

Comments are closed.