The above title may sound funny but with the advent of microarray technology it is entirely possible for your doctor to ask you a question like this. Namely, any array-based diagnostic test that determines a multi-component disease signature is inherently generating a statistical statement. For example:
The test results demonstrate with 90% confidence that your chances to have Disease X are 75% as detected by the marker-pattern, which is up to 85% sensitive and specific to Disease X.
The above statement may sound unintelligible and make no medical sense but if you compare the process behind this hypothetical machine-generated statement to traditional clinical diagnostics, then you realize the similarity. Namely, both are actually pattern recognition. Traditionally, diagnostics has been enabled by doctors? experience so that the more experienced the doctor, the better his of her ability to recognize symptoms and synthesize these symptoms into a disease-pattern, a.k.a. as DIAGNOSIS. Except for simple cases, doctors do pattern-recognition routinely. For example, you go to your doctor with complaints including chronic fatigue, occasional fever, head-ache, shortness of breath, and loss of appetite. Your doctor listens to you, looks at you, and analyzes your chest X-rays, blood-test etc. Depending on what your doctor finds, you might either have some pulmonary ailment or no detectable disease at all. How do doctors come up with these diagnoses? Very simply, they analyze patterns, which can greatly overlap from disease to disease but also contain specific components that determine the diagnosis and therapy.
At some point in the future, the era of diagnostic microarrays will arrive, and the clinical pattern recognition is enabled by tiny chips, which quantify molecular disease markers instead of talking to you, looking at you, listening to you, and analyzing your X-rays
Thus, the future diagnostic microarrays will potentially eliminate some of the need for traditional, doctor-enabled pattern recognition. I do not think, however, that the pattern-recognition capacity of microarrays is anywhere near to do so because (1) technical standards for diagnostic microarrays do not exist, (2) the mathematics of pattern-recognition makes sense to statisticians but not doctors and patients, and most importantly, (3) the medical community does not have surrogate biomarkers, except for infectious disease. In short, in addition to great potential, the microarray technology brings unique challenges for scientists, IVD industry, medical community, and the regulatory institutions. Some of these challenges are formulated in DIAGNOSTIC NUCLEIC ACID MICROARRAYS [MM12A, 2006 Clinical and Laboratory Standards Institute (CLSI)]. And I would like to point here to my favored one: QUALITY MANAGEMENT OF THE PRODUCTION AND PERFORMANCE OF DIAGNOSTIC MICROARRAYS.
Imagine the amount of quality control and quality assurance (QC/QA) you need for a 10-element array as opposed to a traditional ELISA test that determines only a single marker. Remember that one has to be able to QC/QA all of the array elements. The best way do so is to design a statistical QC/QA methodology, and cut back on the number of array elements as much as possible. In summary, despite great promise, diagnostic arrays will have to overcome formidable technical challenges. And before they enter routine clinical practice, the diagnostic arrays that produce multi-component diagnostic signatures will have to prove that they make clinical diagnostics more accurate and cost-effective. Unless they become capable of detecting surrogate biomarkers, which for most diseases are yet to be discovered, or provide useful information as to the best therapy for a patient, we will only be seeing experimental tests along with statistical diagnostics (see above). So, would you like to be statistically diagnosed?
The Eomix Newsletter: July 2006 (http://www.eomix.com/The_Eomix_Newsletter.htm.)
About the Author:
A molecular oncology expert. Currently, the owner and manager of Eomix, Inc. (http://www.eomix.com). As the Director of Molecular Medicine at CeMines, developed molecular pattern-recognition for cancer diagnostics. As a faculty member at the Pulmonary Division of David Geffen School of Medicine at UCLA, published a number of papers on lung cancer in highly-ranked peer-review journals including Cancer Researh, The Journal of Biological Chemistry, FASEB Journal, Science etc.
Tags: Medicine, symptoms, shortness of breath, diagnosis, lung cancer, Cancer