Comorbid disease or illness refers to other diseases that coexist with the disease of interest, and on a population level, usually more frequently than by chance alone. The term is exacting and even enjoys a position as a Medical Subject Heading (MeSH) used in the indexing of the medical literature. Comorbidities may arise by chance, selection bias, shared environmental or genetic factors, or as part of the causal path of another condition. These associated factors are extremely important as they are often markers of disease severity (as evidenced by their judicious use in the health services research literature in chronic disease scores), can modify the risk of a given outcome, and are often undetected or ignored when attempting to understand potential adverse consequences of pharmacotherapy. A comorbid condition may imply differential risk attributable to the disorder itself (e.g. risk of suicide with comorbid depression) or suggest effects from self-treatment or pharmacotherapy that may not have otherwise come to the attention of the treating clinician or researcher. Principal to be considered among these disorders are those that may not be easily shared with a clinician including substance abuse (including alcohol) or smoking.

It is a prudent approach for the safety professional to "backtrack" by starting with the adverse event of interest and determining the possible disease and therapy mechanisms that are known to be associated with the given type of event (clinically, a differential diagnosis that enumerates the possible etiologies). Once identified, they can query what is known about the comorbidities and natural history of the disease to determine if they can explain any or all of the events under scrutiny and the degree to which they are known in the given case. This approach is more focused and can be inefficient across a number of events; however, it does provide the necessary clear direction in the search of the literature, data and follow-up often necessary with cases of interest.

Comorbidity can also be a modifier of effect: in the presence of a comorbidity, the effect of the drug of interest may be magnified or attenuated which consequently would impact the risk of a given outcome. Depression itself, whether a consequence or a contributing factor, is highly comorbid with a number of chronic medical conditions and often affects compliance, outcome and may affect the potential risk-benefit of a given therapy. Cardiovascular disease has long had a history of identifying comorbid factors as a means of determining risk of outcome (Figure 8.2). One only need look at the various risk scoring schemes that include body mass index, hypertension, cholesterol level, and family history to see how important comorbidities are in the assessment of risk.

Figure 8.2. A representation of the comorbidity of illnesses related to diabetes.

Comorbidity is of such importance in some fields that entire studies have been undertaken to examine them. Most prominent among these is the National Comorbidity Survey (Kessler, 1994), a systematic interview and screening for mental illness in a representative sample of the US population. From this study, an incredible wealth of information has emerged about the impact and natural history of mental illness and comorbid mental illness on patient outcomes.

So what does all this have to do with drug safety? Aside from its potential impact on risk, comorbidities and their treatments may modify the risk of adverse events either directly or by increasing the probability of a drug interaction or exacerbation of one illness due to the treatment for another. Comorbidities may suggest product exposures that may not be captured by pharmacy claims data because they are over-the-counter (OTC) products, nutraceuticals, vitamins or substances like alcohol, tobacco or street drugs. However, detection of comorbidities, like the index diseases themselves, is dependent on the source of information. In studies using claims data, there are the obvious issues relating to coding, ascertainment, eligibility of patients for insurance coverage for certain services; in clinical data, as in claims data, detection is dependent on the patient presenting with symptoms, the physician making the diagnosis, and the coding reflecting the encounter. This is very similar to the issues in spontaneous reporting where capture of the potential event depends on the patient presenting to the doctor with the complaint, the physician recognizing it as potentially an adverse event, and reporting it to the proper system. The precision of that coding is of interest as well, and has been shown to vary depending on the disease and the source of the data where the method of ascertainment may result in substantial differences in estimates of prevalence (Sanders, 1962). Relying on coded diagnoses in an existing database requires that the scientist have a keen understanding of the coding scheme, any restrictions on the number of codes allowed for any given site of service, in addition to how disease (or symptom), medication, procedure, hospitalization and social data are captured and recorded, and whether there are potential weaknesses given the study design. These limitations are particularly important when considering comorbidities and attempting to assess their impact.

Was this article helpful?

0 0
Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook

Post a comment