A drug-tobacco interaction is an adverse drug event.
Tobacco smoke contains thousands of chemicals, some of which have significant interactions with prescribed medications. Polycyclic aromatic hydrocarbons (PAHs) are byproducts of tobacco combustion, and are known to cause pharmacokinetic drug interactions due to the induction of certain liver enzymes (including cytochrome p450 1A2). Drugs metabolized largely by CYP 1A2 will be broken down more quickly and will achieve lower serum concentrations than expected; but because CYP 1A2 has a short half life, drug clearance will fall rapidly if a patient stops smoking. Thus patients are at risk for both reduced treatment efficacy (if the clinician is not aware of the 1A2 induction) and adverse drug events (if a therapeutic dose is not lowered when they quit smoking). This is particularly relevant to hospitalized patients, who generally stop smoking upon admission (nicotine therapy will not replace the “missing” PAHs). Relevant medications include theophylline and clozapine (both of which have narrow margins of safety) and tacrine.
Tobacco also causes pharmacodynamic drug interactions--i.e., it can amplify, counteract or otherwise modify the intended effects of a medication on bodily processes. The most concerning pharmacodynamic interaction is between tobacco and oral contraceptives. Contraceptive users who smoke have greatly elevated risks of thromboembolism and other cardiovascular events.
Drug-tobacco interactions should be included in CDS alerting systems. It is not sufficient to include tobacco in the commercially developed rules that drive alerts. If a system only checks for interactions among drugs listed in a patient’s current orders, pharmacy and allergy histories, tobacco interactions will be missed--unless the clinician enters tobacco as a medication. Instead, CDS algorithms should incorporate data on smoking habits drawn directly from the electronic record. This would only be possible in systems which use coded data entry for tobacco history. Alternatively, an alert could fire every time a tobacco-sensitive medication was prescribed to any patient, with obvious negative consequences for workflow and alert fatigue.
Source: Kroon, L. American Journal of Health-System Pharmacy. 2007;64(18):1917-1921. (http://www.medscape.com/viewarticle/562754).
Submitted by : Neff Breen