The Meaning of Secondary Prevention

A November letter to the editor in American Family Physician chastises that publication for misusing the term “secondary prevention,” even using it in the title of an article that was actually about tertiary prevention.

I am guilty of the same sin. I had been influenced by simplistic explanations that distinguished only two kinds of prevention: primary and secondary. I thought primary prevention was for those who didn’t yet have a disease, and secondary prevention was for those who already had the disease, to prevent recurrence or exacerbation. For example, vaccinations would be primary prevention and treatment of risk factors to prevent a second myocardial infarct would be secondary prevention.

No, there are three kinds of prevention: primary, secondary and tertiary. Primary prevention aims to prevent disease from developing in the first place. Secondary prevention aims to detect and treat disease that has not yet become symptomatic. Tertiary prevention is directed at those who already have symptomatic disease, in an attempt to prevent further deterioration, recurrent symptoms and subsequent events.

Some have suggested a 4th kind, quaternary prevention, to describe “… the set of health activities that mitigate or avoid the consequences of unnecessary or excessive interventions in the health system.” Another version is “Action taken to identify patient at risk of overmedicalisation, to protect him from new medical invasion, and to suggest to him interventions, which are ethically acceptable.” But this is not a generally accepted category.

Stedman’s medical dictionary defines secondary prevention as “interruption of any disease process before the emergence of recognized signs or diagnostic findings of the disorder.”

The Encyclopedia Britannica defines it as “early detection of disease or its precursors before symptoms appear, with the aim of preventing or curing it.”

A CME website explains the definitions used by the USPSTF:

The U.S. Preventative Services Task Forces’ Guide to Clinical Preventive Services (2d edition, 1996) describes secondary prevention measures as those that “identify and treat asymptomatic persons who have already developed risk factors or preclinical disease but in whom the condition is not clinically apparent.” (pp xli)  These activities are focused on early case finding of asymptomatic disease that occurs commonly and has significant risk for negative outcome without treatment. Screening tests are examples of secondary prevention activities, as these are done on those without clinical presentation of disease that has a significant latency period such as hyperlipidemia, hypertension, breast and prostate cancer.  With early case finding, the natural history of disease, or how the course of an illness unfolds over time without treatment, can often be altered to maximize well-being and minimize suffering.

Tertiary prevention activities involve the care of established disease, with attempts made to restore to highest function, minimize the negative effects of disease, and prevent disease-related complications.

The Library Index says “Secondary prevention, also called ‘screening,’ refers to measures that detect disease before it is symptomatic.”

The Encyclopedia of Public Health says:

Secondary prevention generally consists of the identification and interdiction of diseases that are present in the body, but that have not progressed to the point of causing signs, symptoms, and dys-function. These preclinical conditions are most often detected by disease screening (and follow-up of the findings). Examples of screening procedures that lead to the prevention of disease emergence include the Pap smear for detecting early cervical cancer, routine mammography for early breast cancer, sigmoidoscopy for detecting colon cancer, periodic determination of blood pressure and blood cholesterol levels, and screening for high blood-lead levels in persons with high occupational or other environmental exposures.

But usage is inconsistent and confusing. In addition to American Family Physician, several other organizations and publications such as the National Library of Medicine, the British Medical Journal, and the American Heart Association use the term “secondary prevention” to include patients who already have symptomatic disease.

Is this just unimportant nit-picky quibbling by self-appointed language police? I don’t think so. Science requires precise terminology and agreement about definitions. The letter to the editor in AFP gives an example of a situation where imprecision could lead to patients being harmed.  Beta blockers reduce morbidity and mortality in patients after MI, but they have been shown to increase morbidity in patients with hypertension who have not already had an MI.  If a writer recommends them for “secondary prevention” thinking he is talking about patients who have had an MI, a reader might misunderstand and give them to patients who have not yet had one.

Posted in: Science and Medicine

Leave a Comment (14) ↓

14 thoughts on “The Meaning of Secondary Prevention

  1. Draal says:

    Dr. Hall, can you offer any solution to this problem? In other words, what are the ways a new terminology in medicine gain an universally accepted definition?

    In science, I’ve seen examples of a new term being coined from one research group (e.g. metabolic engineering). It requires some traction to establish itself (i.e. multiple research groups begin to use it) before it’s established. I’ve also seen multiple terms with the same definition circulate (FHT vs F3H or ANS vs LDOX are abbreviations for the same plant enzymes; it’s a bit of a headache when writing papers or using database search functions). Usually it’s not a big deal and eventually one term wins out but it may take years for that to happen, especially in newly emerging fields.

  2. Paddy says:

    It’s a good point – we do have to be precise in our language.

  3. Sometimes words just aren’t very good, which is why they are dropped. “Inflammable,” for example.

    If “secondary prevention” is not used because the more communicative word “screening” is used instead, then it’s fairly natural for “secondary prevention” to slide over to the next prevention in the series — one which doesn’t already have a meaninful, unambiguous name attached to it.

  4. Draal says:

    “If “secondary prevention” is not used because the more communicative word “screening” is used instead, then it’s fairly natural for “secondary prevention” to slide over to the next prevention in the series — one which doesn’t already have a meaninful, unambiguous name attached to it.”

    I concur that the English language (jargon in this example) can evolve just as you described but I think the point was that having a universally accepted precise definition can be critical to avoid errors from misinterpretations. The natural progression of our language may be too slow for competing terms to battle it out before a clear winner is declared.
    Other scientific fields have established authorities for their terminology (e.g., IUPAC). Asking out of ignorance, what does the medical field have?

  5. Draal,

    Yes, persnickitiness is a virtue in science and technology. “Flammable” was chosen pretty quickly I think when it became clear that universal education about the real meaning of “inflammable” was not going to be the most effective route to communication. Was there a committee declaring “flammable” to be the preferred word?

    If people are going to predictably misunderstand or misuse a word, there are a couple of ways to get around it. If the word is misused/ misunderstood by a non-influential minority, then everyone else can simply define it every time they use it so that there is no ambiguity. This is the AFP’s letter-writer’s approach.

    Another way is to drop the ambiguous word in favour of something else less likely to be misinterpreted, which is how “flammable” came into common use. If an influential committee comes up with a useful alternative and publishes it in an influential spot, this could prod along this process.

  6. Some situations are inherently ambiguous, and there is no terminology to reliably cover it. In such situations, actual descriptions are needed. This is such a case.

    The problem is that the precise definition of secondary vs tertiary varies with the disease.

    For example, the definition that Harriet gave at the beginning for heart attack is actually correct. It is similar for stroke.

    Secondary stroke prevention is for those who have already had an event (a TIA or stroke), while tertiary prevention (a seldom-used term) is to prevent the complications of a stroke. The concept of symptomatic vs asymptomatic does not apply well to preventing events, like stroke or heart attack.

    Therefore it is simply folly to rely upon terminology and then hope that it is unambiguously defined and universally understood. To avoid miscommunication, rather than being persnickety with word use, don’t rely on terms – give specific definitions, i.e. “the prevention of strokes in those with a prior stroke or TIA” vs “the prevention of strokes in those with a-fib” or “with atherosclerotic disease” or “a hemodynamically significant (>50%) carotid stenosis” or “within 30 days of a prior stroke.”

    There is no meaningful way to split up all these situations into two categories of secondary and tertiary.

  7. Rick says:

    I think I had an NCLEX question on this.

  8. Harriet Hall says:

    Steven Novella said “The concept of symptomatic vs asymptomatic does not apply well to preventing events, like stroke or heart attack.”

    I think it does. Events are symptoms too. Primary prevention of stroke/MI (cardiovascular disease) would include things that are advisable for the entire population like having a healthy diet, exercising, not smoking, and controlling weight. Secondary prevention would include testing for risk factors like high cholesterol and high blood pressure and trying to reduce them. Tertiary prevention would apply to reducing risk after symptoms like angina or claudication or TIAs have developed or after a first stroke or MI has occurred.

    The point of the letter writer in AFP was precisely that many doctors are using the term secondary prevention in conditions like stroke and heart attack in ways that do not correspond to the precise definitions established for preventive medicine.

    Since doctors don’t all agree on exactly what secondary prevention means, Dr. Novella’s solution is the only one: explain what you are talking about instead of using a word that can be misunderstood.

  9. But Harriet, no one uses that definition with respect to stroke. I spent some time searching around, including the literature, and everything I could find uses “secondary prevention” to mean after a TIA or stroke (and possibly also with established underlying disease, like atherosclerosis or a-fib) and “tertiary prevention” to mean the prevention of complications from a stroke.

    My point is that “symptomatic” can be ambiguous with some diseases, like stroke. Does it mean symptoms from the risk factors that cause stroke, or symptoms from a stroke?

    Does having asymptomatic hypertension count as a symptom, and therefore qualify for symptomatic. What about atherosclerosis – is that “symptomatic” by definition, or only when it causes symptoms or complications, like a heart attack. Does having a heart attack make one symptomatic with respect to a stroke?

    The concept of “symptomatic” is a mess of ambiguity in this context. It may work for, say, diabetes where it is more clear that you are talking about symptoms and complications from diabetes, but not with stroke.

  10. To further clarify, what basically happens now in the stroke world is that we classify risk and prevention on two axes – primary (prior to any cerebrovascular event or symptoms) vs secondary (after a cerebrovascular event or symptoms), and separately we consider the underlying risk factors for stroke. But we don’t use risk factors to determine if prevention is primary, secondary, or tertiary. In prevention algorithms, these are considered separately, with risk factors as modifying factors determining optimal treatment.

  11. Harriet Hall says:

    @ Steven Novella

    “no one uses that definition with respect to stroke.”

    That underscores the point that the term secondary prevention is being used inconsistently. And there is still ambiguity in the usage for stroke: you said “everything I could find uses “secondary prevention” to mean after a TIA or stroke (and possibly also with established underlying disease, like atherosclerosis or a-fib).”

    Wouldn’t it be better to explain precisely what you mean than to rely only on a term that means different things in different situations and has the potential to be misunderstood? When you use your definition, you could specify “for secondary prevention of stroke in patients who have already had a stroke” And/or TIA or who have atherosclerosis or atrial fibrillation (or whatever your precise meaning).

  12. Harriet – now we are saying the same thing.

    Guidelines and studies use specific descriptions of criteria, and do not rely on terminology. But they will also refer to their specific criteria with a term, as I outlined.

    And we appear to agree that the problem is not just with the use of the terms, but the fact that the definitions are inherently ambiguous with no obvious and clear cutoff that applies across different diseases and conditions. So it can’t be fixed simply by agreeing on definition.

  13. Paddy says:

    I’m with Steven on this. A precise specification of what’s being prevented and in which patients is much more than assigning a general category which may be more than a little ambiguous however you define it.

  14. Carl Bartecchi says:

    Just a note to clarify one of Harriet’s last comments in her discussion – the one dealing with beta blockers. It is the traditional beta blockers like atenolol that have the bad rap. The
    vasodilatory beta blockers like carvedilol and nebivolol should not be lumped with the traditional beta blockers as the former have better central aortic pressure reductions and have neutral or favorable metabolic effects and are useful and effective anti-hypertensive agents. See C.V.S. Ram, Am J Cardiol 2010;106:1819-1825

Comments are closed.