psychiatric hospital—2,194 errors over 1,448 patient days, or an error rate of 1.5 errors per patient day (Grasso et al., 2003).

The committee could only find three studies in the peer-reviewed literature addressing incidence rates for medication errors arising from the use of OTC drugs. These studies (Li et al., 2000; McErlean et al., 2001; Goldman and Scolnik, 2004) showed that parents using OTC medications to treat children with fever often administer an incorrect dosage. One study of 118 caregivers treating their children with a fever reducer revealed that incorrect doses were given 47 percent of the time; another study of 248 caregivers found that 12 percent gave an overdose and 41 percent an underdose; and a third study found that of 200 patients treated for fever by a parent, 51 percent received the wrong dose. Moreover, these studies indicated that a misdose often resulted in a continued fever and an eventual trip to the emergency department.

Despite the paucity of data on OTC-related error rates, there is a growing body of literature documenting adverse OTC drug–disease and OTC drug–drug interactions. Some examples are presented in Box 3-2.

The committee could find no studies of medication error rates associated with complementary and alternative medications. There is, however, an emerging literature indicating that these medications have the potential for adverse interactions with prescription drugs (D’Arcy, 1993; Calis and Young, 2004). In particular, these types of products can interfere with the metabolism and elimination of other drugs in the body. St. John’s Wort, an herbal product commonly used to treat depression, is an example. Studies have found that St. John’s Wort impacts an enzyme that ultimately increases the oxidation of drugs (Bailey and Dresser, 2004). This action limits the bioavailability of some drugs, resulting in serious adverse effects. Specifically, studies have shown that St. John’s Wort can increase organ rejection and increase the viral load in HIV patients by limiting the effects of prescription medications (Piscitelli et al., 2000; Ruschitzka et al., 2000).

Where incidence rates have been measured systematically, medication errors have been found to be common and to occur at unacceptably high levels. Reasonably well-researched stages of the medication-use process include prescribing, dispensing, and administering in hospitals; prescribing in ambulatory clinics; dispensing in community pharmacies; and medication adherence in self-care.