Quantophrenia in medicine and elsewhere

Share with your friends


Pitirim Sorokin had coined the term to criticize the misapplication of quantitative methods to sociology.  Murray Rothbard borrowed it from Sorokin to describe—in his usual lampooning style—the age-old fascination with mathematical modeling of economic phenomena.  In his Economic Thought Before Adam Smith he points back to Pythagoras to show the influence of number mysticism on the economic thought of Aristotle (who should have known better if we recall the quote from Ethics: “it is the mark of the educated man…”).  The same tendency grips Bacon and Petty who developed ‘political arithmetic’ in the 17th century.  Rothbard also identifies Bernoulli’s founding of ‘mathematical economics’ as leading the way to Walras’ ‘equilibrium theory,’ and so forth.

F.A. Hayek addressed the topic directly in The Counter-Revolution of Science, detailing the origin, thoughts, and influence of the “positivists” buoyed by a transcendent enthusiasm for the exact sciences.  Gilles Paquet recently published a paper titled “Quantophrenia” to review the influence of this form of numerology on public policy.

Emile Durkheim, intellectual heir to Auguste Comte and the positivists, was invoked to shed light on the theories of Geoffrey Rose, our public health maître-à-penser.  Rose’s philosophy, dominant at the IOM, WHO, CDC, etc., has spawned an entire school of epidemiology activists, some of whom have developed a mathematical model to quantify the effect of public health policy on cardiovascular death rates, the so-called IMPACT model.  Feeding this goose meta-analytic granola allows it to lay an impressive number of reports in top journals, always confirming the benefits of “population medicine.”   Those who are curious about the innards of IMPACT can peruse 40 mind-numbing pages of an e-appendix of the NEJM.  (And that is still an incomplete account of the ‘model’…)

But the frenzy for ill-fitted numerical models is not confined to public health policy and has spilled into all areas of clinical research as Alvan Feinstein was tirelessly trying to point out.  Much of his writing at the end of his career was sounding the alarm against the “distraction of quantitative models” and the ill-effects of “statistical reductionism” but to little avail.  His wish to see clinical medicine develop its particular methods of investigation rather than yield to statistical paradigms borrowed from the hard sciences has not materialized to any significant extent.  Quite the opposite, statistical alchemy now defines ‘best practice.’

Obviously not all quantification is bad and we need not give in to methodological nihilism.  Quantitative methods have shed needed light on epidemiology, pathology, and therapy.  But exactly how much light (or how much needed) is not always clear…  Taking a cue from the notion of “dispersed knowledge,” and the “wisdom of crowds” I’ve often wondered what the overall benefit would be if the use of statins was left to the discretion of “free-practicing” clinicians as opposed to adherence to guidelines and other propaganda.  Clearly physicians were using statins even before WOSCOPS, and clearly absolute benefit rates are much less than the rates mandated by current guidelines.  Who would have been most on target?  (Of course, the number of somnolent physicians who need to be prodded with guidelines is not insignificant, but they can afford to be somnolent precisely because there is no “free practice”).  This is all rhetoric, naturally.

In day-to-day medicine the cult of numbers seems to be even greater, but here it’s to satisfy the regulators.   So in the cath lab, and even when under sedation, a patient must be subjected to interrogations of the WHAT-NUMBER-IS-YOUR-CHEST-PAIN-NOW variety mandated by the safety checklists.  And the actual number matters much less than the presence of a number—any number.  To get administrators off their back, charts are filled with dutifully recorded pulses that are always 2+, capillary refills that are always < 3 secs, respiratory rates that are always 20, and other nonsense of the sort, and that’s not even mentioning the attention that is increasingly be paid to the quality “metrics.”  Rothbard also used or coined “metromania,” the obsession with mensuration…

Paquet introduced his paper with a couple of quotes.  The first one says it all…

“If it cannot be measured, it cannot be controlled.”

                                                                          Lord Kelvin

“Not everything that counts, can be counted,
and not everything that can be counted, counts.”

                                                                    Albert Einstein

It is better to solve the right problem the wrong way
than to solve the wrong problem the right way.

                                                             Richard Hamming

Leave a Reply

Your email address will not be published. Required fields are marked *