The Prison of Precision by Stephen Vaughn MD, Ph.D

In 1999, the Institute of Medicine launched a movement towards improving the quality of patient care by publishing a document entitled, To Err Is Human:  Building A Safer Health System.”  Nearly 20 years later, the goal of the movement has not been achieved.  To be honest, it has been a spectacular failure, which has provided a major element of the explosion of the cost of healthcare in the United States, and proceeded to constantly worsen the healthcare environment of the United States.

One of the statements the IOM made was that “The decentralized and fragmented nature of the health care delivery system (some would say “nonsystem”) also contributes to unsafe conditions for patients, and serves as an impediment to efforts to improve safety.”  After twenty years, the healthcare delivery system wrestles with the problem of delivering ANY care to many in need.  Is this better?  More often than 20 years ago, it seems that we let “nature take its course” more often by the absence of care.  Is that how to reduce errors?  “the provision of care to patients by a collection of loosely affiliated organizations and providers makes it difficult to implement improved clinical information systems capable of providing timely access to complete patient information. Unsafe care is one of the prices we pay for not having organized systems of care with clear lines of accountability.”  Have we improved?

The goal of the Institute of Medicine publication was noble.  But the foresight that went into its acceptance and planning was quite limited. The ongoing principles of quality management in American healthcare were largely discarded, the rest scorned.  A decision was made to use the principles of quality in the aerospace industry as a better method of addressing quality.  It was evident that this choice was a mistake. However, bureaucratic inertia, arrogance and stubbornness led to the insistence, time and again, that the principles of the movement were itself correct and any opposition to the movement lay in character blemishes on the part of the counter-revolutionaries.

A fair comparison of the behavior of the healthcare quality movement can be made with Lenin’s collectivization of agriculture and the punishment of the rich peasants which lead to widespread starvation and death across one of the richest swaths of farmland in the world. Lenin suggested hanging the opposition. The reformers in American healthcare have of course been much more genteel.

However, the single-minded absolutism of the healthcare quality movement wasn’t the reason for the critical failure of the project.  That absolutism and rigidity arose when the reformers sensed that reform had begun to fail. Rather than honest introspection, the reformers blamed unseen opponents among the “healthcare establishment” for sabotage.

The principal assumptions accepted from the very start by the reformers were flawed.  The most fundamental error was the choice of a model for quality reform which did not fit.  The reformers were arrogant.  They took a position which was hostile and antithetical to the practice of medicine at the time, offering the unproven assumption that healthcare had never attempted to consider the preservation of quality.  In fact, they cast away proven methods of modernization and quality management which had long been in use in the practice of medicine.  Contempt and ignorance drove the intent of outsiders to start from scratch.

The authors cite “Health care is a decade or more behind other high-risk industries in its attention to ensuring basic safety. Aviation has focused extensively on building safe systems and has been doing so since World War II. Between 1990 and 1994, the U.S. airline fatality rate was less than one-third the rate experienced in mid century. In 1998, there were no deaths in the United States in commercial aviation. In health care, preventable injuries from care have been estimated to affect between three to four percent of hospital patients. Although health care may never achieve aviation’s impressive record, there is clearly room for improvement.”

The flying machine has evolved incredibly quickly over the last century or so of its existence.  The human machine has not.  There are other problems with using aerospace quality achievement as a model for healthcare quality achievement.

Not only is there a difference in the mechanics of the object of quality improvement, the process itself differs considerably.  It is absurdly obvious that the quality advancement process in aerospace made no progress up until 1900; the airplane did not exist.  In comparison, medicine is old, centuries old.  If one reads Maimonides, for example, one sees observations on the progress of medicine and the maintenance of quality that are still relevant today.  Maimonides, before 1300 AD, held strictly to empiricism, i.e. no medical treatment should be endorsed until it had been proven to be effective, no matter how compelling the conceptual basis might be.  This has always been a respectable tenet in medicine.  The reformers – like the Bolsheviks’ reform of Russian agriculture – assumed that no familiarity with a subject is necessary, if you have sufficient ideological purity.  From Patient Homes to other clever ideas, Maimonides’ principle has been abandoned by the reformers, with the expected failures.  And if the restructuring of peasant agriculture fails and causes starvation, as it did in Russia in the 1920’s – it must be the fault of the peasants, not the reformers.

Consider the foreign models of quality impressed on the American Healthcare system.  One of the most popular approaches to quality management at the time brought concepts that fit well into one particular niche, assuming that they are universally applicable.  Manned space exploration, for instance, a highly complex and critical series of processes must proceed without fault.  Expenditure of time in preparation of the spacecraft is of no consequence, nor is cost, in any sort of way beyond general principles.  What is unacceptable is failure, especially failure that loses manned personnel.  If one only has one shot to make it work, there is no room for error.

There have been at least four critical failures in the American manned space programs.  Three failures caused the loss of crew personnel, and at least one of them occurred because of a strategic mistake – deliberately cutting corners and hazarding lives in order to rush to begin a mission from political pressure.

In the passenger transport industry, the rules are simple.  Passengers are not injured or killed during proper operations; it is only when a failure of procedure occurs that passengers are injured or killed. In medicine this is not the case. 100% of our passengers on life’s journey will die at some time.  Some of them die because of avoidable error; most of them do not.  Eliminating avoidable injury of death is obviously worth great merit, but it is not as useful a measure as avoiding airplane crashes, where one is working with passenger viability numbers of 100% and 0%

The preservation of life and limb in the passenger transportation industry has no acceptable minimum.  As the saying goes, 100% is impossible, and 99% is unacceptable.  Passenger transportation operates in a zero-defects quality model.

The practice of medicine is unavoidably enmeshed in a Bayesian web.  We weigh likelihoods, and consider possibilities in the face of incomplete information.  Medicine does not operate in an environment of lower quality than the passenger transport industry; rather, it operates in a matter where what we deal with is less precise than manufacturing tolerances and enumerated likelihoods of outcomes.

Even more troubling, many common business traditions in quality are ignored in the IOM model.  It is focused on reproducing the aerospace model of quality assurance.  When one has a clearly defined sequence of tasks to execute, and the satisfactory execution of these tasks are paramount, and time and money are of relatively low importance, then medicine long ago developed a method of process which rather closely resembles the aerospace model.

In the use of especially dangerous medications, careful redundancy is practiced, checklists and critical mandatories are carefully followed.  I expect it might be rather embarrassing to the IOM that the development of exacting processes has gone on independent of the quality reformers.  Strict quality control is a common-sense control of processes that are administered by smart people who have good judgment.

It is notable that the IOM did not hold up medicine’s active models of quality management in chemotherapy and radiation therapy.  The IOM sought to use models extraneous to the practice of medicine, so to better earn a posture of legitimacy and relevance.  This has caused the IOM quality movement itself to fail catastrophically.

The practice of medicine is not the unthinking implementation of a series of standard movements and behaviors, as though practices on a factory floor.  Most of the practice of medicine is not the product of small, explicit skills.  Rather, it is the tacit comprehension, through years of exhaustive learning by book and experience, of the particulars of care of the individual patient.  All patients are the same, except in ways they are different.  This tautology, which sounds simple, is the most frustrating and rewarding element of the care of the patient.  Pretending that it is otherwise has led us to greater frustration, often worse outcomes, and a soaring cost of medical care which is mostly spent on wasteful redundancy and useless and irrelevant additional activity.

In commercial aviation, human error is absolutely unacceptable.  Extensive documentation and monitoring is used to follow the sequence of actions in air travel, and the crew of the aircraft is in a “no-win” situation.  They are tasked with the precise execution of a series of steps, as well as the outcome of successful travel.  This is their job; they have no way to excel, but only to fail.  I refer to this as the “prison of precision.”  In such areas, there are best practices only, in both the following of procedures as well as deciding when to override the standards. The only acceptable measure is success; deviation is often followed by death, and certainly with blame.  In commercial aviation, to err is human, but it is always criticized, and especially if it leads to the death of the commercial passenger.

There is so much to consider on this topic, and its damage to the practice of medicine over the last 20 years, that I cannot hope to lay it out in one essay.  If I did, it would be too long to read.  I will pick up the central points regarding the Prison of Precision in a later essay.

Get one of our books here