A 69-year-old man with chronic kidney disease and essential hypertension was admitted to the hospital with chest pain, headache, and accelerated hypertension. Of note, he had missed taking several days of his regular medications. Upon admission, his blood pressure was 218/126 mm Hg with a heart rate of 90 beats per minute. Physical examination revealed a patient in mild discomfort, but no distress, presence of an S4 (indicating reduced ventricular wall compliance), and otherwise normal cardiopulmonary and neurologic examination. His electrocardiograph revealed significant left ventricular hypertrophy but no acute ischemic changes. His chest radiograph demonstrated a mildly enlarged heart, but no widened mediastinum or pulmonary edema. Troponin levels were negative, and his kidney function tests were stable. After starting a nitroglycerin drip, it was decided his outpatient medications should be re-started gradually. One of his antihypertensive medications was minoxidil, and his outpatient dose of 7.5 mg per day was ordered via the electronic health record. However, midodrine 2.5 mg tablets x 3, not minoxidil 2.5 mg tablets x 3 ultimately arrived on the medical unit.
While the medication mix-up was not identified by the dispensing pharmacist, the nurse responsible for administering medications noticed that the dispensed medication differed from what had been ordered. She sought out the medical team, who quickly clarified that the intended medication was minoxidil, not midodrine. Midodrine hydrochloride is a vasopressor/antihypotensive agent used in treating orthostatic hypotension. Had this patient received midodrine in the setting of accelerated hypertension and chest pain, his condition could have worsened significantly due to this pharmacological effect.
With the ever-growing number of marketed drugs, the potential for look-alike and sound-alike (LASA) medication errors involving name pairs is increasing. LASA errors can occur between two brand names (e.g., Keppra and Kaletra), between two generic names as in the current case (e.g., midodrine and minoxidil), or between a brand and a generic name (e.g., Hespan and heparin). Consequences of LASA errors can range from no patient harm to death. In this particular case, confusion between these two products could have resulted in significant harm.
The combination of handwritten prescriptions with look-alike drug names is particular risky. The hope that computerized provider order entry (CPOE) would prevent the vast majority of cases of LASA errors has not been realized. LASA errors continue to occur in the electronic world. Similar sounding drugs typically appear near each other on computer selection screens, often with similar dosage forms and strengths and frequencies, increasing the likelihood for look-alike selection errors.
While comparing the medication order to the delivered medication, the nurse discovered that the incorrect product had been dispensed from the pharmacy. Based on the patient’s presenting condition, accelerated hypertension, and listed chronic disease state, minoxidil was indicated for this patient and correctly ordered by the provider. Midodrine is available in 2.5 mg, 5 mg, and 10 mg tablets strengths. Minoxidil tablets are 2.5 mg and 10 mg. The similarities in the first two letters of each drug name and the availability of 2.5 mg tablet strengths likely contributed to the selection error during the dispensing process in the pharmacy.
Although it is difficult to know the percentage of medication errors associated with LASA drugs, a rate of at least 1 per 1000 prescriptions has been suggested in the inpatient setting.(2) Given that the average 1000-bed hospital will have approximately 300,000 prescription orders in a month, this can add up to a significant hazard.
Notably, the near miss in the current case took place at an institution just prior to its conversion to a barcode medication administration (BCMA) system. A barcode system has the potential to greatly reduce look-alike and sound-alike medication errors when the medication has been properly selected during prescribing or order processing.
However, this benefit is limited to recognizing errors at the time of dispensing or medication administration. If the order had been entered correctly, a BCMA alert likely would have signaled a nurse that the wrong medication was in hand. However, BCMA will not catch medication errors associated with order entry or product selection errors by prescribers and pharmacists. In other words, once an order has been entered incorrectly, a barcode scan of the packaged dose will confirm the order as correct when it is not—perhaps even falsely reassuring a nurse who is wondering about what appears to be an unusual medication choice.
The Institute for Safe Medication Practices (ISMP) publishes a List of Confused Drug Names (4) as a result of error reports received through the ISMP National Medication Errors Reporting Program. These error reports are published in the ISMP Medication Safety Alert! newsletter. Hospitals should utilize this information, in combination with internal data, to identify medications that are more likely to be confused and implement strategies to prevent such errors.
Other contributing factors to LASA errors include incorrect selection from a drop-down menu; confusion during the medication reconciliation process, misheard oral orders, and incorrect product selection from the shelf or automated dispensing cabinet. The use of similar mnemonics during order entry (e.g., inadvertently using MET 500 for metronidazole when the mnemonic was built for metformin) is an additional source of error.
ISMP suggests reducing the risk of confusion by using computer alerts to make prescribers and pharmacists aware of the problematic drug name pairs, avoiding the use of oral orders or at least using read back (write down what is heard, read back to prescriber). Selective use of TALLman letters (e.g., hydrOXYzine/hydrALAzine) can differentiate the appearance of drug names on computer screen. While the use of TALLman lettering has the potential to reduce look-alike medication errors with similar generic name pairs.(5) However, studies of the efficacy of this strategy have yielded mixed results. ISMP and the Food and Drug Administration have developed lists of drug names with TALLman letters, and drug manufacturers use the FDA-sanctioned names in product labeling. The combined ISMP–FDA list is available at http://www.ismp.org/tools/tallmanletters.pdf.
Additional strategies can reduce errors associated with look-alike medication pairs, and prevent sound-alike errors as well. These strategies involve using both the brand and generic names on prescriptions and patient labels; including the indication or purpose of the medication on prescriptions and medication orders; and configuring computer selection screens to prevent look-alike names from appearing consecutively. Patients—outpatients, certainly, but even inpatients where possible—should know the name and purpose of each medication prescribed. Clinicians should inform patients of the dispensed or administered drug and allow the educated patient to question anything that appears incorrect. Printing the list of prescribed medications for the patient can be useful, allowing patients to compare the list with the dispensed medications.
Unfortunately, most of the above strategies continue to rely on human vigilance to catch errors and are, therefore, considered to have a minimal impact. Indication alerts at the time of computerized physician order entry (CPOE) have the potential to intercept drug name confusion errors.(6) In such alerts, the CPOE notifies the clinician that the patient’s electronic problem list is inconsistent with the prescribed medication (e.g., metformin prescribed for a patient not known to have diabetes). In one study, indication alerts intercepted 1.4 drug name confusion errors per 1000 alerts.(6)
The FDA’s Division of Medication Error Prevention and Analysis identifies potential problems with proposed drug names prior to approval of a New Drug Application or Abbreviated New Drug Application. In response, pharmaceutical companies prepare to prevent newly-approved medications from appearing or sounding similar to approved medications currently available.(7) However, it is becoming increasingly difficult to find names with no possibility of creating LASA medication names or products.
The FDA has an internal process for the evaluation and review of proposed drug names and their associated potential for LASA medication errors.(8) Additionally, the FDA has released new draft guidance for Best Practices in Developing Proprietary Names for Drug, which includes methods that pharmaceutical and biotechnology industries should use to determine whether a new drug looks or sounds like another currently marketed drug.(9) The new guidance employs preapproval sponsor review strategies, such as name simulation studies and use of a computer program (Phonetic and Orthographic Computer Analysis) searching. The name simulation studies also include practicing health care professionals, who evaluate the product names in a real-world environment. This new draft guidance allows the FDA to reject a name if the review determines that the name may cause confusion with other products that may result in medication errors and preventable harm. While this process still relies on human vigilance to predict potential errors, it does standardize the industry process for assessing medication names in an effort to avoid LASA-prone names.
Prior to adding a medication to formulary, ISMP encourages institutions to consider the error potential of medications by performing an internal risk assessment. This process should include an assessment of the product’s look-alike and sound-alike similarity compared to other products already on the formulary.(7) In addition, when a name mix-up is recognized to carry the potential for significant patient harm, a hard stop should be considered in the CPOE system to force the prescriber and pharmacist to confirm the drug name.
LASA medication errors can occur at any point in the medication use process, including the medication reconciliation process, order entry, and product selection. It is clear that health information technology has made the situation better, but it has not relegated these errors to the history books. While efforts from the FDA and pharmaceutical companies may aid in the prevention of future errors, many medications have look-alike and/or sound-alike similarity on the market. As a result, practitioners may benefit from performing risk reduction strategies at their practice site to help prevent LASA medication errors (see Table).
** Medication errors involving look-alike or sound-alike medications can still cause patient harm, even in environments with computerized provider order entry and barcode medication administration.
** Risk management associated with name similarity is an industry-wide obligation, with important roles for regulatory agencies and pharmaceutical companies, as well as the entire health care continuum, from practitioners to consumers.
** Before a drug is added to the formulary, the potential for look-alike or sound-alike error should be evaluated through an internal proactive risk assessment.
** Organizations should follow the recommendations in the Table to decrease their risk of look-alike and sound-alike errors.
Barcode scanning at the bedside can reduce medication errors, but only if the medication order is processed correctly.
Credit: Patent Safety Network