Most error-reduction efforts begin in response to a serious error. Individuals and practice sites have become accustomed to reacting to an error after it has occurred. Unfortunately, only then do they consider how to prevent the same error from being repeated.Would it not be useful to be able to predict the types of errors that could occur and proactively institute preventive measures?
All too often, error potential is not included in decisions about which medications, devices, or technology to purchase. Instead, decisions are guided by cost, third-party formulary restrictions, contractual agreements with purchasing groups or vendors, and pharmaceutical marketing efforts. Input and evaluation from those who will be using the products may not be sought, and error potential may not be considered ahead of time. The ultimate result is unforeseen safety issues and errors once the product is in the hands of clinical users or health care consumers.
Failure Mode and Effects Analysis (FMEA) systematically identifies areas of potential failure and gauges what the effects would be before an error actually takes place. This proactive process can be employed to examine the use of new medications and products, as well as the design of new services and processes that may affect work flow. FMEA is best employed prior to purchase and implementation, so that preemptive action can be taken.
To cite just one example, FMEA could be used to assess the potential for error with new medications when they are first marketed or before they are prescribed or added to your inventory:
Step 1: A process-flow diagram would be designed. Then the ways the intended product would be prescribed would be explored. Who would prescribe it and for which patients? What clinical patient information would be important before the product is prescribed? How would it be procured, stored, and used, from acquisition through dispensing and administration? Who would prepare and dispense it? What information would need to be given to the patient or caregiver? How would the product be administered?
Step 2: Potential failure modes (ie, how and where systems and processes may fail and what can go wrong) would be identified while considering how the product would be used. Questions to be asked would include the following: Does the drug name look or sound like another drug name? Would a similarly spelled drug name be listed in close proximity to the intended product on computer order entry screens? Does the package label clearly express the strength or concentration? Would the product be stored near, or could it be mistaken for, another similarly packaged product? Are dosing parameters complex?
For example, an FMEA performed on quetiapine (Seroquel) might have predicted a high likelihood of mix-ups with nefazodone (Serzone). These 2 medications are likely to be prescribed by the same type of physician and for patients with similar diagnoses; are available in overlapping dosage strengths; often are administered at the same frequency; and are likely to be stored together and to appear on computerized lists in close proximity, alphabetically by brand name. Although Serzone is no longer marketed, these types of errors are still likely, because generic formulations of nefazodone are available and may be prescribed as "Serzone."
Step 3: For each failure mode, staff members would then determine the likelihood of making an error, as well as the potential consequences. What would happen to the patient if the drug were given at the wrong dose, at the wrong time, or by the wrong route? What would happen if a patient received the wrong medication or if the wrong patient received the medication?
Step 4: Staff members would consider the severity of the outcome and identify any preexisting processes in place that could help eliminate or detect the error before it reached the patient. Each process would then be evaluated for its effectiveness, based on what was learned in previous steps. For example, would obtaining additional patient information or using computer alerts, bar coding, or a double-check process catch these errors every time? Numerical values could be assigned to determine the likelihood of the occurrence, its severity, and the chance that it would be detected before causing patient harm.
Step 5: If failure modes reveal errors with significant consequences, actions would be taken to prevent the error, detect it before it reached the patient, or minimize its consequences. Such actions might include improved communication of orders by listing the indication on prescriptions or differentiating look-alike products by ordering them from different manufacturers or by storing them separately.
Although industries outside of medicine have developed elaborate FMEA scoring systems to rank items for action, a simplified FMEA process as described above can be an efficient, proactive risk management tool.
Dr. Kelly is the editor of ISMP Medication Safety Alert! Community/Ambulatory Care Edition.