Reuters Events | Pharma

Thought leadership and innovation for the Pharmaceutical Industry

The journey to an alternative to the current concept of compliance proceeds through the slough of despond where the flawed status quo is lamented, but the prospect of what could be is sighted.

By Allan Showalter on Jun 9, 2009

The journey to an alternative to the current concept of compliance proceeds through the slough of despond where the flawed status quo is lamented, but the prospect of what could be is sighted.

In my last post, The Tragedy Of Patient Compliance (for the purposes of this post, "compliance," "adherence," and "concordance" are used synonymously), I made the claim that the concept of patient compliance has not produced a useful theoretical system, has not supported the creation of an organized database through which statistical correlations and trends can be discovered and tracked, and, most importantly, has not resulted in significant improvements in healthcare effectiveness, efficiency, or economies or in treatment outcome.

What Could Be - Lessons From Epidemiology

From my ongoing proselytism of this viewpoint, I am aware that such declarations, even if they are accepted as factual, fail to convey the catastrophic scale of the potential benefit lost because of this series of failures. To garner a sense of what could be, what is possible, what we're missing because compliance enhancement is typically an unrequited hope rather than a realistic expectation, compare the meager accomplishments resulting from the study of treatment adherence with those emanating from the work in epidemiology.

Hippocrates, the author of that line which has become obligatory in every patient compliance review article, "Keep watch also on the faults of the patients, which often make them lie about the taking of things prescribed, was also, according to Wikipedia, "the first person known to have examined the relationships between the occurrence of disease and environmental influences" and "coined the terms endemic (for diseases usually found in some places but not in others) and epidemic (for disease that are seen at some times but not others)." So far, then, it looks like a draw.

Since Hippocrates, however, epidemiologists have

• Developed such useful notions as "death rate," "dose-response," "herd immunity," "incubation period," and "cohort effect"
• Conceived a handy hypothesis called "germ theory" that has, on occasion, proven useful
• Demonstrated the ineffectiveness of such practices as bloodletting
• Established the causes of diseases like cholera and lung cancer and, more recently, identified the risk factors for many chronic diseases
• Set forth effective preventive practices such as quarantine and isolation (by the 1600s) and smallpox inoculations
• Worked out disease transmission vectors and implemented solutions based on them (e.g., mosquito control brought the number of yellow fever deaths in Havana from 305 to 6 in a single year)
• Provided a foundation for a public health system that has been the means of putting biomedical advances (e.g., vaccinations and antibiotics) into practice, educating the public to heath issues, and implementing hygienic programs that resulted in dramatic decreases in the mortality and morbidity rates of the infectious diseases, such as typhoid, and those diseases caused by poor nutrition, such as pellagra, that were the major health threats of the early 20th century in the US
• Developed disease eradication programs (e.g., tuberculosis and polio) that are global in scope
• Created systems of training, standardized principles (e.g., Bradford-Hill criteria) and investigational methodology, an array of government and nongovernment organizations devoted to advancing the field, and a tradition of collaboration between institutions

There is much more, but you get the idea. On the other hand, I immodestly think I was on the mark when I previously summarized the gains from compliance studies as "bupkis."

It is instructive to note that epidemiology does not have an unblemished record of getting it right the first time. In the early 1800s, a group of epidemiologists believed that cholera was caused by miasma, i.e., bad air resulting from from decayed organic matter. Moreover, there was scientific effort put forth in behalf of that theory. The UCLA School of Public Health "Competing Theories of Cholera" page points out that

One prominent supporter of the miasma theory was Dr. William Farr, then assistant commissioner for the 1851 census and a career employee of the government's General Register Office. For a while, Farr was convinced that cholera was transmitted by air. He reasoned that soil at low elevations, especially near the banks of the River Thames, contained much organic matter which produces miasmata. The concentration of such deadly miasmata would be greater at lower elevations than in communities in the surrounding hills. His calculations in 1852 seemed to support this theory.

Here's the kicker - years later, as more and more evidence mounted, Farr continued to study the transmission of cholera and, writing

a detailed report in 1866 that relied heavily on his extensive knowledge of statistics. Using death rates to justify his conclusions, he publicly acknowledged that water was the most important means of transmission, not miasmata as previously stated.

It turns out that epidemiologists are notoriously and admirably fickle, willingly abandoning theories and ideas that do not prove effective.

And, when things go wrong in epidemiology, the response is a gathering of forces to fix the problem rather than bemoan it or explain it away. The unexpected re-emergence of tuberculosis, for example, is credited, along with the emergence of HIV, with stimulating a renaissance in the field of infectious disease epidemiology, which seemed a few decades ago, doomed to extinction.

Epidemiology demonstrates by example the possibility of success in a complex field of healthcare and the tremendous impact that can result.

So, if my contention about the dearth of accomplishments in patient compliance is correct (and it is - when was the last time you read an article on patient compliance, other than a marketing piece or an ersatz news story copied and pasted from a press release, that announced the arrival of a blockbuster adherence-enhancing application or even predicted quantum advances in the field?), the question that may well now come to mind is

What The Heck Is Wrong With Patient Compliance?

If an idea like treatment adherence, with its substantial appeal to common sense (not unlike, for example, the geocentric view of the universe or the pre-Galilean belief that an objects rate of descent toward Earth is proportional to its mass), fails to produce progress or benefits after concentrated effort is expended over a prolonged period under its imprimatur, one must eventually consider the possibility that there exists some subtle but nonetheless fundamental flaw, some essential, yet nuanced error in the underlying logic of the concept itself.

Well, how about this for a problem? Today, thirty to fifty years after compliance and adherence first began to appear regularly in medical journals in the context of patients following treatment recommendations, some of the terms are pertinent to compliance and adherence, such as - oh, compliance and adherence, lack reliable definitions.

This inconsistency means, setting aside for now the matter of the validity of those definitions, no one reading the research and clinical literature on adherence, can count on "noncompliance" in one article referring to the same phenomenon when it is used in a different article. All too often, in fact, the meanings of such terms vary from one paragraph to another in the same article. (As might be imagined, the meanings of these terms are even more capriciously used in advertisements, CME course, presentations, workshops, and the lay press.)

Consequently, "compliance" currently has little meaning beyond that formed by the reader's own beliefs and biases.

How useful can the concept of compliance be if the significance of a patient being identified as compliant or noncompliant is nebulous?

In fact, my queries directed to clinicians and others with an interest in the subject about their definitions of Patient Compliance, Treatment Adherence, or the equivalent term du jour (I also offer to accept definitions of noncompliance and nonadherence) evoke one or more of the following: (a) Circular definitions that resolve to adherence is, you know, adherence. (b) A flummoxed or disdainful expression, the former not unlike that one has come to expect from ones child when the heir apparent is asked how he flunked math - again - while the latter resembles the look adorning the face of that same offspring when questioned about the possibility he may somehow bear some responsibility for the trashing of the home during his parents overnight absence. (c) I may not be able to define compliance but I know it when I see it. (d) Asking stupid questions like this could explain why you cant find a date.

Of course, some respondents opt for the statistical definition of compliance, which, as I noted in my previous post, is simple to describe and calculate for anyone who has mastered fourth grade arithmetic. I, of course, disallow such answers.

This is not, by the way, my opinion alone. For a scientific elucidation of the issue, I heartily recommend Medication Compliance and Persistence: Terminology and Definitions, by J Cramer, R. Rosenheck, G Kirk, et al., which is made available without charge by the good folks at the International Society for Pharmacoeconomics and Outcomes Research. The article is a short, straightforward, and worthwhile read, from which I have excerpted a single sentence that succinctly summarizes the core issue:

Health outcome and cost-effectiveness analyses incorporating measures of medication usage have been hampered by the lack of uniformity in standards of definitions and measurements used to describe the concepts of medication compliance or persistence.*

Ahem, I told you so.

For an experiential illustration of the inherent problems defining compliance, Ive prepared - just for you, ...

The Patient Noncompliance Recognition Pop Test

I. Which of the following patients, each of whom is prescribed the same medication to be taken at 9 AM and 9 PM every day for ten days and each of whom misses the 9 PM dose of the medication on the third day of the regimen, are noncompliant?

1. A patient who did not remember to take the 9 PM dose the third day but took the other 19 doses as prescribed.
2. A patient who did not remember to take the 9 PM dose the third day and also forgot to take 50% of all the prescribed doses.
3. An 87 year old patient with long-standing memory problems who did not remember to take the 9 PM dose the third day and whose daughter, who usually calls to remind her to take her medication, forgot to do so.
4. A 27 year old patient who did not remember to take the 9 PM dose the third day, and whose text message service, on which the patient relied as a reminder of his medication doses, was inoperative that evening because of a technical glitch.
5. A patient who did not remember to take the 9 PM dose the third day because she is 4 years old, and her parents forgot to give her the medication.
6. A patient who did not remember to take the dose because he has been in coma for six months, and his nurse gave the dose to another patient by mistake.
7. A patient who took the wrong pill, mistaking it for the correct one which it resembles.
8. A patient who took the wrong pill because his pharmacist misread the prescribing doctors handwriting and supplied the patient with the wrong medication.
9. A patient who took the right pill but took it at 5 PM instead of 9 PM because she assumed four hours wouldnt make any difference.
10. A patient who took the right pill but took it at 5 PM instead of 9 PM because the Wikipedia entry she found specified that four hours wouldnt make any difference.
11. A patient who made a conscious decision to forgo one dose because of a side-effect that was specific to a one-time set of circumstances (e.g., medication that causes drowsiness in a situation that requires the patient to drive all night to evacuate his family from an area endangered by a hurricane).
12. A patient who made a conscious decision to forgo the 9 PM dose the third day and every night thereafter because drowsiness the mornings after he took the first pills the first two nights made it difficult to arise on time for work and did not inform the prescribing doctor.
13. A patient who made a conscious decision to forgo the 9 PM dose the third day and every night thereafter because drowsiness the mornings after he took the first pills the first two nights made it difficult to arise on time for work and, after the one missed dose, did inform the prescribing doctor, who shifted the medication schedule so the 9 PM dose was no long required.
14. A patient who made a conscious decision to forgo the 9 PM dose the third day and every night thereafter because drowsiness the mornings after he took the first pills the first two nights made it difficult to arise on time for work and, after the one missed dose, did inform the prescribing doctor, who insisted that the bedtime dose was essential, from which point the patient took the bedtime dose.
15. A patient who made a conscious decision to forgo the 9 PM dose the third day and every night thereafter because drowsiness the mornings after he took the first pills the first two nights made it difficult to arise on time for work and, after the one missed dose, did inform the prescribing doctor, who insisted that the bedtime dose was essential, from which point the patient no longer told the doctor he skipped all his bedtime doses.

II. The following patients were each prescribed a medication to be taken at 9 AM and 9 PM every day, but the purposes of the medications and/or the durations for which they were to be taken varied. Each patient did not take the 9 PM dose the third day because, at 9 PM, he discovered he had left his bottle of pills in his desk at work and chose not to drive the 27 miles to his office, contact his physician or the physician on call to obtain a prescription for one pill, or to drive 4 miles to the pharmacy where he obtained the original prescription in hopes of persuading the pharmacist to supply the single pill needed. Which of these patients are noncompliant?

1. A patient who was taking a one week course of the medication for symptomatic relief of nasal congestion.
2. A patient who was taking the medication indefinitely for control of his dangerously high blood pressure which had caused one stroke.
3. A patient who was taking a two week course of the medication as chemotherapy for lung cancer that metastasized to the brain and spinal cord.

III. The essay question follows:
Some researchers and clinicians promote the notion that a patient is compliant as long as he takes enough of his medication for it to be effective. Discuss (1) Is this idea valid if and only if the patient has an accurate understanding of the minimal effective dose for the pertinent medication and disorder (if so, "adequate compliance" may require radically different degrees of persistence and precision for varying regimens, but if not, the distinction between compliance and noncompliance would seem largely independent of the patient)? (2) Under this concept, is the patient compliant if he only takes half the pills prescribed but the physician, anticipating the patient's behavior, wrote the prescription for twice the dosage needed? (3) In the context of this concept, does the term "adherent to treatment" convey any useful, generalizable information, i.e., information beyond that describing the patient's behavior in response to one specific treatment recommendation?

So, how do you think you did?

The only solace I can offer is that the questions could have been much, much more difficult. Imagine similar issues applied to treatment regimens that include multiple medications, a specific diet, an exercise schedule, clinic attendance, and required lab and imaging procedures for screening and monitoring.

Again, as I suggested in the previous post,

Compliance is a useful statistic, not a conceptual platform for improving treatment effectiveness.

Given how "noncompliance" is bandied about, I now add

Nonadherence To Treatment is not a valid medical disorder, syndrome, symptom, sign, or disease.

This is not merely a matter of rhetoric. Nor is it simple. And it is certainly not trivial.

Medical research projects are organized around definitions of the pertinent disorder, symptom, medication effect, treatments, or other parameters. By its nature, however, the definitions of medical compliance are complex, confounding, and mutually contradictory, potentially tainting research findings with embedded biases.

That a concept is vaguely defined does not, of course, mean that naming it is illegitimate, unscientific, or useless. In this case, however, the term, "compliance," has conveyed a false precision, has come to imply that noncompliance is a discrete syndrome for which a discrete solution exists.

Aha

Given these problems, it is not surprising that few gains have been made in patient compliance; it would, in fact, be surprising if significant gains had been made.

And, not only is patient compliance a failed strategy but, as the Leonard Cohen song has it, Everybody Knows. So, why are we still re-writing that same review article with the same results? Surely, there is a different approach that is potentially more effective.

But, thats the next post.

______________

*The citation given for this declaration follows: Hasford J. Biometric issues in measuring and analyzing partial compliance in clinical trials. In: Cramer, JA, Spilker, B, eds. Compliance in Medical Practice and Clinical Trials. New York: Raven Press, 1991.