A 65-year-old woman with rheumatoid arthritis and chronic obstructive pulmonary disease (COPD) was admitted to a medical unit during the night with worsening shortness of breath. Orders were written at 2:00 AM for prednisone 60 mg daily and for continued administration of her daily methadone dose of 80 mg. The medications were administered by the night-shift nurse when they arrived on the unit at 6:00 AM, in accordance with a new policy that specified that all newly ordered medications be administered to patients within 4 hours.
The nurse assigned to this patient for the day shift also administered the two medications at 9:00 AM to comply with another policy regulating daily medication administration. When the day-shift nurse documented his administration of the two medications, he realized that the same medications had been administered 3 hours earlier. He immediately notified the physician, and continuous pulse oximetry and hourly vital sign checks were conducted to watch for oversedation from the methadone overdose. Fortunately, the patient was able to tolerate the extra doses of prednisone and methadone with no lasting effects.
This incident illustrates how multiple reasonable system policies can produce an unreasonable result. The policy mandating that all new medications be given within 4 hours of being ordered had just been implemented a few days before the incident, while the policy of administering all daily medications between 8:00 and 10:00 AM had been rigorously enforced for many years. Both were policies designed to enhance effectiveness and safety but, applied together, introduced a threat to patient safety.
Over the past decade, the patient safety movement has actively worked to shift the focus from practitioner blame to analyzing how system failures allow errors to reach patients. However, some health care organizations still fail to recognize that, within each system, multiple interlinked key elements can affect patient outcomes. When this linking is recognized, a robust error analysis addressing all the elements can help formulate improvements in the system as a whole. In this case, two different system policies were instituted to improve medication administration safety, and their overlap led to the error.
After analyzing serious errors, organizations try to "fix" the cause of the error by concentrating their safety strategies only on unachievable changes in the system (i.e., the "low-hanging fruit," such as a nurse failing to follow the 5 Rights). But a more robust analysis of a medication error allows reviewers to identify all the system failures that allowed an error to occur. For example, in this case, both the night and day nurses followed the 5 Rights when they administered the right medication, dose, and route to the right patient at the right time as dictated by organizational policy for drug administration. A more thorough review of the error reveals that the new policy was actually at the root cause of the error. (See Addendum for an error analysis worksheet.)
The hallmark of root cause analyses is systems thinking. Senge defined a system as "a perceived whole whose elements 'hang together' because they continually affect each other over time and operate toward a common purpose."(1) Systems thinking entails seeing wholes rather than parts. It provides a framework for seeing the changing patterns and structures that underlie complex situations. To help simplify the complexity of the medication-use system, the Institute for Safe Medication Practices (ISMP) identified 10 key elements (Table) that organizations should consider when developing system strategies to reduce or eliminate medication errors.(2) Both latent and active failures must be addressed thoroughly in analyses of errors.
Latent failures are weaknesses in the structure of an organization and often include one or more of the key elements within the system. The weaknesses are the result of both well- and ill-conceived decisions made by upper management—even the best decisions carry risks and benefits, and most involve a certain degree of compromise and trade-offs. The immediate success or failure of a strategic change can be easily identified once the change has been operationalized; however, the far-reaching effects of these decisions often go undetected until an error occurs. This is especially apparent when the strategic decision prescribes a solution that shifts the problem from one part of a system to another. Too often the individuals who "solved" the original problem are different from those who ultimately inherit the new problem. Organizations often develop error prevention strategies following a serious event that focus solely on specific risks without analyzing the unintended effects of those strategies. For example, a new policy that required nurses to immediately administer a new IV medication upon noting the order might appear to be the correct strategy in response to the late administration of a newly ordered IV vasopressin medication that resulted in patient harm. However, the root cause of the error was actually a communication failure, because the physician failed to specify that it was a "stat" order. Without thorough analysis of the error, this new strategy affected the administration of all newly ordered IV medications. The new policy then has the potential to increase the risk for other adverse events since pharmacy does not have the time to check the medication orders prior to administration. All too often the safety committee acts without the benefit of insightful input from the practitioners involved as to the cause of the error. Taken one at a time, the strategic changes are often subtle, and their effect may go unnoticed, causing no other problems. Sometimes their consequences are hidden, becoming apparent only when they occur in sequence with each other (as they did in this case) and combine with the active failure of individuals to penetrate or bypass the system's safety nets. Examples of such latent failures include the following: (i) lack of adequate drug information—no medication reconciliation process on admission and no "early dose" warning from the automated dispensing cabinet; (ii) inadequate communication about potentially conflicting medication administration policies and between health care practitioners; and (iii) a push for efficiency over safety that decreases the time for cognitive processing and double checks.
Active failures are errors committed by individuals, usually practitioners who are in direct contact with vulnerabilities and weaknesses within the structure of the organization, such as occurred in this case. The error occurs when individuals fall victim to the flaws within the system and mistakes are made. Unfortunately, we often focus our attention on these points of human fallibility. Active failures in this case included: (i) not viewing or checking the patient's complete medication history before prescribing, dispensing, and administering; (ii) failure to communicate medications administered and patient outcomes during shift handoff; and (iii) failure to educate the patient about the medications being administered. Although it may seem easier to concentrate on the human contributions to errors, the root causes of the problems often lie within latent failures and vulnerabilities. Therefore, the only effective way to decrease the likelihood of error is to strengthen the medication-use system's resistance to combinations of latent and active failures.
Safety Strategies Recommendations
To decrease the probability of future errors involving the medication-use system, this organization (and others like it) should consider the following safety strategies:
- Institute a robust medication reconciliation process that includes a column that asks when the last dose was taken/administered.
- Ask the automated dispensing cabinet vendor if the "early dose" warning can be enhanced or if they can add a "hard stop." This is a requirement that the nurse has checked and documented why an early dose is being withdrawn from the automated dispensing cabinet before the drug can be released. It can be accomplished through a drop-down list or by using free text.
- Conduct a nursing education program that stresses the importance of bringing the MAR to the bedside and monitor for adherence.
- Before adopting any medication-use policy, the organization should consider both the risks and benefits of the policy. They should also consider conducting a failure mode and effect analysis to determine the potential for failure and potential risk of harm.
- Standardize patient-centric handoff communication, and mandate that it occurs at the time of nursing shift change, breaks in care, and changes in level of care.
- If the organization functions in a hierarchical model of patient care, move toward a culture that is collaborative and supports questions pertaining to drug therapy.
- Create an environment that provides the time practitioners need to think through their patients' diagnoses and treatments and match the appropriateness of drug therapy with the patient (i.e., right drug for diagnosis, laboratory values and vital signs support the administration of the medication, etc.) rather than viewing drug distribution and administration as a task.
- Share the lessons from errors like this one, so nurses are alerted to the hazards of the specific circumstances (administering early doses and repeating them later) and, more generally, to the hazards of conflicting policies.
- Encourage patients to question their drug therapy when they are uncertain.
- Educate patients about all aspects of their drug therapy, including the drug name (generic and trade), dose, times of administration, and reason they are receiving the medication, prior to administration.
Finally, Dr. James Bagian, a safety expert and former astronaut, teaches his students that there are two ways to die on a space mission: (i) not following the rules and procedures as written; and (ii) following the rules and procedures as written (J. Gosbee, MD, MS, written communication, November 2008). Therefore, any new organizational policy must go through a rigorous process, which includes a multidisciplinary team, to ensure that the benefit to patient safety is not compromised by the risk of an unintentional outcome.
It is important to remember that the design of any new policy:
- Should not be a band-aid for a system flaw.
- Must include input from end-users.
- Must be examined for risk and pilot-tested for workability.
- Must be routinely monitored for compliance and propensity to generate workarounds.
Hedy Cohen, RN, BSN, MS Vice President Institute for Safe Medication Practices
1. Senge P. The Fifth Discipline: The Art and Practice of the Learning Organization. New York, NY: Doubleday Business; 1994. ISBN: 978-0385260954.
2. Cohen MR, ed. Medication Errors, 2nd ed. Washington, DC: APhA Publications; 2007. ISBN: 978-1582120928.
3. ASSESS - ERR™ Medication System Worksheet. Horsham, PA: Institute for Safe Medication Practices; 2006. [Available at]
Table. Key Elements That Organizations Should Consider When Developing System Strategies to Reduce or Eliminate Medication Errors.(2)
| Drug information
| Drug labeling, packaging, and nomenclature
| Drug storage, stock, and standardization
| Drug device acquisition, use, and monitoring
| Environment factors
| Competence and staff education
|Quality processes and risk management
Addendum (3). ASSESS - ERR™ Medication System Worksheet © 2006 ISMP. Reprinted with permission from the Institute for Safe Medication Practices.
Click on thumbnail for larger view.