Persistent Nondiagnosis

Brad Kittredge is an MBA/MPH student at the Haas Business School at UC, Berkeley. He is working to build an online tool to assist with complex and difficult diagnoses, which, as you’ll see, he considers among the biggest problems in medical care. He blogs more about these issues at Hyoumanity.

Every day, thousands of Americans are desperately seeking answers to complex medical conditions that doctors are unable to diagnose. Consider one example: Jenny T. is a 14 year old girl with a progressively debilitating neuro-degenerative condition that has taken her from healthy and active to nearly paralyzed in less than one year. Her parents have taken her to some of the best academic medical centers in the US, including Stanford, UCSF, and the University of Pennsylvania, but doctors have been unable to diagnose her condition, leaving Jenny and her parents desperate for answers and short on options.

Unfortunately, Jenny’s case is far from unique and persistent
nondiagnosis is a much more pervasive problem than many believe. The US
National Institutes of Health (NIH) estimate that 7,000 rare diseases
(defined as a disease that affects less than 200,000 Americans)
cumulatively affect as many as 30 million Americans. A study by the
National Organization for Rare Diseases (NORD) found that of these
patients, 36% took more than one year following their first doctor
visit to receive a diagnosis while 17% took more than six years to get
diagnosed. Mary Dunkle, VP of Communications for NORD, said that
diagnosis is the biggest problem our patients face and accounts for the
majority of the inquiries we receive. At the same time, it is not only
individuals with rare diseases who face complex and prolonged
diagnostic challenges. A number of more common conditions, such as
cancer, Type 2 Diabetes, Multiple Sclerosis, Primary Immunodeficiency
Disorders, Hepatitis, Lupus, Celiac Disease, Polycystic Ovary Syndrome,
and more frequently go undiagnosed or misdiagnosed for a prolonged
period of time. Because diagnosis is the basis for all subsequent
healthcare decisions, the absence of a timely and accurate diagnosis is
the limiting factor that prevents patients from pursuing effective

There are several factors that contribute to this pervasive problem.
First, disease and diagnosis are inherently complex as there are over
one million permutations of diseases, syndromes, and disorders that
afflict humans, each containing a distinct symptom profile and
physiology (ICD-10). Many of these conditions frequently present in a
misleading and ambiguous manner, requiring significant specialization
and experience to effectively diagnose. Second, doctors are subject to
a number of significant psychological and perceptive challenges such as
anchoring and confirmation bias, leading to frequent diagnostic
mistakes. Third, structural challenges inhibit effective management and
resolution of difficult cases. Doctors average only 12 minutes per
consultation and are not paid for additional research time, providing
no incentive to take on a difficult case and work to help a patient
find a solution. This is not a criticism of doctors, but of the
difficult structures and incentives under which they are frequently
forced to operate. These challenges also manifest through the process
of patient referral. When faced with a case they are unable to solve,
doctors refer these patients to other specialists, but these referrals
are often determined by professional or personal relationships rather
than a careful consideration of matching a patient with the best
provider. The direct result of these challenges is that, despite the
United States having the most technically advanced health care system
in the world, a large number of patients remain undiagnosed for an
extended period of time.

The implications of this are substantial in both economic and human
terms. Unable to find answers, patients continue to see doctor after
doctor in hopes that one of them will have a breakthrough insight into
their case. This leads to significant out-of-pocket expense for the
patient as well as substantial costs to the patient’s health insurer.
Delayed diagnosis also results in delayed treatment and more advanced
disease, leading to worse health outcomes and higher long-term costs.
Additional costs include lost productivity for the patient and their
family caregivers and draws on public programs such as Social Security
disability benefits. Lastly, patients facing a prolonged period of
nondiagnosis or misdiagnosis suffer severe physical, psychological, and
emotional pain and suffering.

I believe that persistent nondiagnosis, which we define as a patient
experiencing symptoms that 5 or more doctors are unable to diagnose, is
an under-recognized and under-addressed problem in our healthcare
system. Findings ways to improve the process of diagnosis and handling
of these tough cases for both patients and doctors will reduce costs,
improve health outcomes, and dramatically impact lives.

Categories: Uncategorized

Tagged as: ,

11 replies »

  1. diagnosis is essential is cancer. Because of the way cancer first develops and then spreads through the body, a delayed diagnosis can be fatal. Lung cancer, ovarian cancer and stomach cancer are three kinds that have a much lower mortality rate following a prompt diagnosi

  2. In response to rbar, the assumption that “medically unexplained symptoms” are due to “somatization” is dangerous and research suggests perhaps a key reason for the delayed and missed diagnoses. As we learn more, a greater and greater number of ‘somatization’ disorders are discovered to be a new, rare disease etc. In studies, 50-80 percent of subjects followed or reviewed for a limited time, were found to have an underlying organic disease that accounted for their symptoms.
    As for unessecary tests being needlessly run, it seems that when we don’t think there is anything ‘really’ wrong, we run tests in order to a)reassure the patient or b)cover ourselves against any later liability. This is not the same as believing there is something wrong and looking for what it is in an intelligent and aggressive fashion. The first approach leads to meaningless and repetitive tests that go in circles and so give the illusion of medical waste on patients who should be referred to a psychiatrist when really, we never honestly sought to find the answer because we didn’t believe there was a problem.
    There is no disease or disorder of somatization, its definition is so vague, its entirely useless and its a thinly veiled re-use of outdated Freudian diagnosis. Use with caution.

  3. Jenny T’s doctors should try running the details of her case through our SimulConsult Neurological Syndromes software. The software covers 1,828 neurological syndromes, which constitute a big chunk of the 7,000 rare diseases. The software and free registration are at http://www.simulconsult.com.
    The software even qualifies as a form of Health 2.0 since the database is a “computational wiki” as we discussed at http://www.aan.com/news/?event=read&article_id=5277. It is being used in training of the doctors who would be seeing such patients – see http://www.childneurologysociety.org/education/casestudies/.
    Another Health 2.0 approach to such problems is to have one’s doctor discuss the case in a Health 2.0 forum. For this case the best forum is the Child Neuro listserv (http://www-personal.umich.edu/~leber/c-n/e-mailUM.html), open only to pediatricians, neurologists and neurosurgeons.

  4. Thanks for the great comments – I agree with many of them and have a couple of reactions. First, with all of the major challenges we face in health care, it’s hard to say that diagnosis is one of THE biggest challenges we face. But, considering the difficulties we have with doing it efficiently and effectively and the role of diagnosis as the limiting factor for a patient being able to move forward with treatment, I think it receives a disproportionately small amount of attention. I have seen misdiagnosis estimates from 15-40% and believe that persistent nondiagnosis is under-recognized. Each of these cases carries with it substantial cost, both tangible and intangible, and the cumulative impact is enormous. Nonetheless, there seems to be alarmingly little attention on improving the process and tools used in diagnosis. This is certainly not to say that diagnosis will ever become entirely a science – to the contrary, I believe that diagnosis fundamentally relies, at least in part, on intuition (although I believe that clinical decision support tools like Isabel have a vital role). As mentioned, the problem is frequently helping the patient find the correct subspecialist. This is often like a needle in a haystack and nobody has the time or incentive to help the patient navigate this journey. I think we simply need more recognition of the importance of diagnosis, more emphasis on and incentive to do it thoroughly, and more tools to make it more efficient and effective.

  5. Great concise description of a difficult situation for patients and doctors alike.

  6. I largely agree with rbar’s comments. Medicine is inherently a needle-in-the-haystack challenge. That is, given a certain set of symptoms, there are maybe 25 common and unimportant explanations for that symptom set,the haystack; and maybe 5 uncommon and serious explanations – the needle. (I am just pulling these #’s out of my head for hypothetical purposes). Unlike a mechanic who is trying to diagnose your car’s problem, we cannot simply empirically replace parts till the problem goes away. Instead, we empirically do various tests until a diagnosis is found – that is, IF the tests are definitive and IF the correct tests are picked first. Add that to rbar’s cogent comment re “medically unexplained symptoms”, and you have a prescription for spending a lot of time and money finding a diagnosis, and ruling out the uncommon/serious ones, which have a low yield but high consequences.
    I applaud your efforts to develop an online tool to assist, although I do not believe this is a unique effort. The lack of a widely used one is testament to the difficulty of developing an adequate rules-based system, and to the reason why medicine is often called as much an art as science.

  7. Troy, I have a completely different take on this:
    The highest chance of diagnosing rare diseases is by channeling patients to the right subspecialty physician at an academic center. I am a nonsurgical specialist and can take, like most of my colleagues, between 40-60 minutes for a new patient. At the end, I either have a diagnosis, or at least an idea what subspecialist to send the patient to. (If that’s not the case, the patient happens to be just an absolute nightmare of a diagnostic challenge, and I tend to send those to a generalist colleague I think highly of.)
    Medical residencies (in all specialties) are doing a suboptimal job because they do not prepare residents well enough for A) what works and B) what matters. My residency and board exam taught me a lot about rare diseases, but SOME of the time would have been better spent to solidify and extend knowledge about common stuff (IMHO, it is much more important to really master the common, less common and above rare conditions, esp. the ones you can do something about, instead of wasting to much brainpower on rarities and “fascinomas”, esp. the ones you can’t treat).
    Talking about “persistent nondiagnosis”: the author’s point is well taken, although I think it is absolutely the wrong emphasis – like saying GWB was a bad bad president because he did not reform copyright law. A more important problem re. persistent nondiagnosis are what is called “medically unexplained symptoms”: common symptoms such as fatigue, dizziness, transient GI problems, chronic pain etc. that is due to somatization, fibromyalgia, depression, but accounts for billions of testing for repetitive diagnostic work ups. This is by far a larger problem than the 30 million Americans with truly rare conditions, a number that I think is likely overblown (or you have to doubt the definition: a condition affecting 200 000 in 300 Million is not that rare). NORD should be commended for doing great advocacy work, but I think the message that millions are running around with undiagnosed (and serious? and treatable?) conditions is counterproductive. What NORD could do: raise learning doctors’ awareness of “red flag” symptoms and how to channel theses patients the right way, instead of tinkering with it yourself.

  8. Some of the points are well taken. Financial incentives do play a role and directs things to interventionalists and not to brain work. However, even now the way internal medicine residencies are being morphed is to dump down the text books and programs so they have to refer to the subspecialists for their procedures. The process of using evidence based journals only as a resource, plus only computers, not classic textbooks leads to ineffectual learning. Hospitalist are a classic dumbing down of the internist and will soon be subplanted with nurse practioners or physician assistants. This will lead us all to the lowest common denominator and never push to excellence.

  9. Hi I am Aman, this article is very good. from the above I knew lot of information about new health diseases. I appreciate to you, you gave very useful information to us.