When evaluating results of any “medical” intervention, including proactive health management (PHM) efforts involving medical care providers, the question must always be asked: which “results” can logically be attributed to the effort, as contrasted to any number of confounding factors at work?  In the past, both vendors and customers of PHM interventions have tended to ascribe everything good that happens after the intervention to their efforts, where at least some such results may be totally unrelated.

Once the PHM evaluation has avoided or controlled for the most common causes of overestimating results, namely “regression to the mean” with before/after comparisons, and self-selection bias with side-by-side comparisons, a third control is often recommended: “therapeutic specificity”.  This involves making sure that the changes in valuable metrics discovered can logically be attributed to, i.e. have some known connection with the intervention.

The origin of this demand for therapeutic specificity was curative medicine, where rigorous clinical control studies are needed to evaluate new drugs and medical procedures.  Unless the physiological effects of such treatments are known to produce specific consequences, any other consequences discovered in patients so treated should not be considered to result from such treatment.  For example, if asthma patients are found to improve their IQ scores after treatment with a particular drug, this improvement should not be attributed to the drug unless there is an established connection between it and mental performance effects.

In evaluating diabetes management, for example, one proponent of therapeutic specificity made a strong case for not counting among the effects of such a disease management effort any reductions in hospital admissions or other sickness care use and expense not directly linked to diabetes.  Specifically: “It should not be assumed that an intervention targeting one disease will impact another”. [A. Linden “What Will It Take for DM to Demonstrate an ROI?”, Disease Management Colloquium May 2006 (www.ehcca.com)]

When considering the cost savings from diabetes DM, evaluations may focus solely on sickness care use/expense reductions where diabetes was the sole cause, include it plus known co-morbidities, or at least conditions known to be directly related thereto.  But “therapeutic specificity applies to the “therapy” used, not to the disease targeted alone.  For example, diabetes is not merely a disease, but a known risk factor for a host of other disease, of the eyes, hands and feet, kidneys and heart.  For this reason, diabetes DM often includes explicit attention to blood pressure and cholesterol, in addition to blood glucose control.

Moreover, since diabetes is known to be “reversible” through lifestyle changes alone, nor merely medications, diabetes DM often includes lifestyle modification efforts.  If these are successful, patients are often able to control this chronic condition without relying on medications anymore, saving themselves and third-party payors significant prescription drug costs.  The same has been shown to be true for heart disease, and may be true even for prostate cancer. [D. Ornish, et al. “Can Lifestyle Changes Reverse Prostate Cancer?” Preventive Medicine Research Institute April 2002 (www.pmri.org)]

The “rule” of therapeutic specificity is clearly intended to apply to the therapy, not necessarily the disease involved, particularly with diabetes, since it is also a risk factor for so many other diseases.  The focus in evaluation should be ensuring that the DM or other PHM intervention is logically related to any results claimed to be due to it, even when the results are not directly related to the disease specifically targeted.

Moreover, medications often have unexpected, even unintended consequences, that can either be negative “side effects” to watch out for, or “serendipitous” added benefits that add to their value.  For example, many of the statin drugs used in reducing cholesterol levels have been linked to reductions in cancer risk.  Recent research in the UK has indicated that controlling people’s heart disease reduces older people’s cognitive function declines. [A. Harding “Heart Disease Tied to Mental Decline” ReutersHealth.com July 27, 2007]

The rule of therapeutic specificity only argues that no connection between an intervention and a result should be assumed, not that none should be credited to a disease-specific or risk-specific intervention.  Fortunately, there are many good “side effects” of both medical and lifestyle interventions, and both should be counted in both benefits and costs of DM and PHM interventions.  While care is needed to avoid “over-attribution” of positive effects to such interventions, equal care is needed to be sure of avoiding “under-attribution” as well.

If there are effects from treating diseases or risk conditions that go beyond the specific disease being addressed, these can justify counting at least the risk-reduction effects of DM and PHM interventions.  This can easily be expressed as a probable percentage reduction in sickness care costs, absences, productivity and performance impairment related to the calculated risk reduction effects.  While a deliberately myopic insistence on limited effects may prevent over-attribution and over-investment in DM/PHM, a more realistic and open-minded approach can prevent both under-attribution and under-measurement, and thereby under-investment.