Dealing with Multicausality

The presence of more than one "pure" etiologic factor confounds causal assessments. For example, fatigue in the patient with advanced cancer, depression, and poorly controlled diabetes may be due to one, two, or all three conditions. Although in such cases the symptom could be simply classified as uncertain in etiology, this default value wastes the diagnostic information that is available and is probably best reserved for symptoms where evidence for any probable cause is truly insufficient.

Instead, investigators can classify a symptom as multicausal, either:

• Listing the potential etiologies in a nonhierarchical fashion (Sullivan et al.,1993); or
• If one cause seems predominant, denoting primary and contributory causes (Kroenke et al.,1992).

Challenges of the Incidental Objective Finding

Another issue complicating causal assessment is the incidental objective finding, characterized by the co-occurrence of symptoms with a physical or laboratory abnormality that is either slight in its deviancy or a common finding in the general asymptomatic population.

Consider these examples from the recent past:

• Hypoglycemia was once proposed as a common etiology for a variety of physical symptoms until it was learned that similarly low levels of blood glucose occurred not infrequently in asymptomatic persons (Cahill and Soeldner,1974).
• In the 1980's, mitral valve prolapse was a popular explanation for atypical cardiopulmonary symptoms until the discovery that as much as 5% of the population has either auscultatory or echocardiographic prolapse clouded the association (Quill et al.,1988).

Currently, there remain many similar situations:

• Mild osteoarthritic X-ray changes in an older patient suffering from neck pain;
• Disc bulging on MRI in an individual experiencing back pain;
• Radiographic esophageal reflux in a person reporting nonulcer dyspepsia;
• Borderline elevations of thyroid-stimulating hormone ("subclinical hypothyroidism") in a patient complaining of persistent fatigue;
• Headaches in a person with moderately elevated blood pressure.

A variant of this phenomenon is the lab finding that does explain a discrete disorder in a certain proportion of patients but is common and nonspecific enough to be falsely labeled as the "cause" of common physical symptoms in a much larger group of patients. This has been a common problem with serologic testing for infectious agents, such as:

• "Chronic brucellosis" in the 1950s (Imboden et al.,1959);
• "Chronic mononucleosis" in the 1980s (Merlin,1986); and
• Lyme disease in the 1990s (Sigal,1996; Steere et al.,1993).

Radiologic imaging is another area where "incidentalomas" are commonly detected (Donovan and Corenblum,1995; Jensen et al.,1994; Tan and Gharib,1997).

All of this argues for caution in attributing common subjective complaints to common or minor objective abnormalities that are more likely to be coincidental than causal.

Page 17 of 26
Previous Section

Next Section