Enhancing Safe Prescribing in the Neonatal Intensive Care Unit

Matthew J. Maughan

The history of medication safety and safe medication practices has been well covered over the last nearly 15 years.¹³ Of course, the basic premise of medication safety can be traced to the Hippocratic Oath’s injunction to first, “Do no harm.” In the early years of medicine, medication safety was at best a side concern. In comparison to today, medical practice in 1900 was not dramatically different from that practiced in ancient Rome. Since then, however, medication management has increased in complexity at the same time that diagnostic techniques and patient monitoring have increased.

With the dramatic increases in technology, diagnostics, and monitoring, there has been an equal explosion of medications to ease pain or treat disease. The combination of these factors has led to a modern health care system that has the potential to do as much harm as good. Few people expect to have severely grave outcomes when entering an ultramodern hospital. But that is exactly the risk patients take when hospitalization is required.

In 1999, a landmark report, To Err is Human, was published.⁶ This report documented the most likely errors that were to take place in a hospital and also gave an estimate for the number of people that would be seriously harmed in hospitals each year. Based on several studies from Colorado, New York, and Utah, the report estimated that between 44,000 and 98,000 people die each year as a result of preventable medical errors. The report also details that an estimated 7,000 people a year die from medication-related adverse events, either in or out of the hospital. In addition, nonlethal medication-related errors are estimated to cost upward of $2 billion each year.

Specific to pediatrics, there are several studies that describe the rate of medication-related problems. Raju et al. conducted a long-term, prospective study of neonatal and pediatric intensive care admissions and reported “an error rate of 1 per 6.8 admissions.”¹² This finding is significant for two reasons: First, this study represented the first time researchers had described rates as applied to pediatric patients specifically. Second, if one imagines a 49-bed neonatal intensive care unit (NICU), that rate predicts injury in seven of those patients, a much higher rate of injury than the average NICU would like to imagine. Indeed, that rate holds true for every 49 admissions: seven of those patients would be harmed. Before this study, medical practitioners would not have imagined that the harm caused to their patients by medication errors would have been that significant. Folli et al. studied more than 100,000 orders at two children’s hospitals and “a total of 479 errant medication orders were identified, of which 27 represented potentially lethal prescribing errors.”³ To put these findings in more concrete terms, for every 6 months of orders processed at a medium-sized children’s hospital, 27 lethal medication orders would be expected.

A study from 2001 reported that pediatric patients were found to be at a three times higher risk of having an adverse drug event than adult patients.⁵ This study pointed to several reasons related to dosing, dilution, patient communication, and physiology that explained the higher mortality and higher incidence of medication-related errors in pediatrics. First, weight-based dosing, required for pediatric patients, necessitates more calculations than adult dosing. Second, dispensing drugs in pediatrics is also error-prone because pharmacists often must dilute stock solutions. Third, neonatal and pediatric patients generally are not able to advocate on their own and cannot alert practitioners to adverse drug reactions. Fourth, all children, especially neonates, may have more limited internal reserves than adults with which to buffer errors.

Foundations of Safe Practices

Safe Culture

To Err is Human focuses much of its attention on creating a reporting system that allows for health care providers to anonymously report or to be able to report errors without fear of retribution from their employers, resulting in a safe culture for error reporting. As a follow-up to the report, many hospitals at this point have such a reporting system. More important than simply having a reporting system is what is done with the reports that are entered by hospital employees. The important principles of error reporting are: (1) the person who made the error needs to know that an error was made, (2) others in the work group need to know the details of the error but not the details of who made the error, (3) errors need to be reviewed in aggregate to look for trends that could point to a collective weakness in the system, and (4) any events that resulted in serious harm need to be investigated for possible root causes and implications for other parts of the hospital. In an ideal world, anonymity would not be necessary. In our current medical practice there is much to be gained from employees who have seen an error and who feel free to report unsafe conditions that they see, errors that they have almost committed, or errors they have actually committed.

Although this contention is impossible to prove, it stands to reason that every person who has worked in a hospital has made a mistake at some point. This statement should not be discouraging to current medical practitioners; on the contrary, it should be empowering. Knowing that they will personally make mistakes should cement in clinicians a relentless drive to discover the part of the process in which they participated that is broken and needs to be fixed to provide safe patient care. This knowledge should empower individuals to recognize that they are also responsible for monitoring how their errors might impact patients and co-workers. Hopefully, this knowledge will also help to engender an ability in practitioners to comfort those whose mistakes they discover. The realization that all clinicians will make mistakes should empower individual practitioners to adopt safe practices and behaviors and work to help others adopt safe practices and behaviors, with the ultimate goal of having a children’s hospital in which there are no medication-related errors.

Just Culture

As these medical error-reporting systems were being introduced, there was quite a bit of focus on having a “no blame” culture. That emphasis was warranted at the time, given that there often existed a culture that blamed, belittled, and punished those who admitted to mistakes. However, an absolute culture without blame also means that there is no culture of accountability. Instead, a safe culture must also take into account personal accountability and personal responsibility toward the larger good. It is difficult to learn from errors if errors are not shared.

David Marx has published and spoken about a different paradigm that could be useful in health care, a just culture instead of a culture of no blame.¹⁰ This paradigm can be particularly useful in pediatrics because pediatrics is an environment that is rife with complexity. The hallmark of a just culture is the determination of risk. Humans will make errors. They always have and they always will. The rate at which errors are made differs depending on the environment, the tasks performed, personal preparation for those tasks, and the choices made while performing those tasks. A just culture tries to distinguish between the errors that happen simply because clinicians are human, at-risk choices clinicians make that put them in an inherently riskier environment, and reckless choices that clinicians make. This distinction is an important feature of a safe but just culture because it differentiates between someone who is adding to the risk of a process and someone who is merely a victim of the risk of a situation. This is another reason that the just culture paradigm is important in pediatrics because an acceptance of process as a part of our daily work is essential to safe patient care. Therefore, the process itself is an item to be examined and reworked, if necessary, to produce the most optimal outcomes and thus reduce risk.

The following case study ² illustrates how a just culture could be applied to a medication-related error. In a particular children’s hospital, the policy was that all orders for electrolyte replacement needed to have two pharmacists check each order. On a particular day, an order was placed for potassium chloride. The pharmacist who received the order checked it for clinical appropriateness and then entered the order into the pharmacy system, neglecting to add diluents. The pharmacy technician who saw the label thought it looked odd that the order was for undiluted potassium chloride because he or she had never seen an order like that before, but decided to proceed with the preparation. The pharmacist who checked the order was new to hospital practice and to pediatrics and did not realize that potassium chloride is always diluted before it is given. The nurse who administered the medication did not realize from the label that there was no diluent in the syringe. The nurse did not double-check the medication with another nurse because he or she did not notice a problem with the actual dose of potassium chloride, and nurses administer a large number of electrolyte replacements on a daily basis. The patient who received the dose was fortunately sedated and paralyzed so although there was certainly venous irritation, there were no other untoward outcomes from this medication error.

Examining this case study from a just culture perspective reveals that the initial pharmacist who entered the medication order without following the accepted procedure was practicing at-risk behavior: There was a procedure to follow that was known to the pharmacist but the pharmacist believed that his or her actions were not going to increase the risk in the situation. The technician who made the dose was following the procedure for producing an electrolyte bolus and was not responsible for assuring that potassium chloride was always diluted. In this example, the technician did not make an error, even though he or she was a link in the chain of events. The pharmacist who checked the medication did not have the personal experience to know that potassium chloride is always diluted, so he or she made a human error. The nurse who administered the medication was practicing at-risk behavior because a well-developed policy for administering an electrolyte bolus existed, but he or she rationalized the behavior because the dose amount was correct and determined there was no need to further check the dose. In this example, no one was intent on harming a patient, so there was no recklessness. The appropriate coaching of all of those involved in this particular error was accomplished after all of the details were fully known.