Featured Articles

Survey Suggests Possible Downward Trend in Identifying Key Drugs/Drug Classes as High-Alert Medications

ISMP extends our thanks to more than 200 respondents consisting of pharmacists (65%), nurses (31%), and others (4%) who completed our survey on high-alert medications between May and June 2014. We sincerely appreciate your input as we update the ISMP List of High-Alert Medications for Acute Care Settings. High-alert medications are drugs that bear a heightened risk of causing significant patient harm when they are used in error. Although mistakes may or may not be more common with these drugs, the consequences of errors are often harmful, sometimes fatal, to patients. The 2014 survey asked practitioners whether they considered the drugs on ISMP’s list to be high-alert medications, whether their practice sites considered the drugs to be high-alert medications, and whether their practice sites had implemented effective precautions for the drugs identified as high-alert medications. Highlights from the survey follow.

Practitioners’ Views

Overall results. Table 1 shows the percent of respondents who considered the drugs in our survey to be high-alert medications. These are the medications and classes of medications that are currently on the ISMP List of High-Alert Medications for Acute Care Settings. While chemotherapy topped the list in prior surveys conducted in 2007 and 2012, IV insulin and anticoagulants were most frequently considered high-alert medications by respondents of this year’s survey. Rounding out the top five high-alert medications in 2014 were chemotherapy, neuromuscular blocking agents, and epidural/intrathecal medications. Concentrated electrolyte solutions, including potassium chloride injection concentrate, hypertonic sodium chloride injection (greater than 0.9% concentration), potassium phosphate injection, and magnesium sulfate injection ranked sixth, seventh, eleventh, and twentieth, respectively. Subcutaneous insulin ranked ninth, and injectable opioids ranked twelfth. Joining oral hypoglycemics as the least likely to be considered high-alert medications were sterile water for injection, inhalation, and irrigation (100 mL or greater); IV beta-adrenergic antagonists; dialysis solutions; and radiocontrast agents.

Comparison to 2012 results. Among all 39 drugs/drug classes on ISMP’s current list, the findings were similar to our 2012 survey for 22 of the drugs. However, for 14 of the 39 drugs/drug classes, significantly fewer 2014 respondents felt these drugs should be considered high-alert medications than the 2012 respondents. The reduction was at least 5 percentage points, with the largest differences as follows:

  • Cardioplegia solution (considered a high-alert drug by 80% in 2012, 60% in 2014)
  • Beta-adrenergic antagonists, IV (54% in 2012, 35% in 2014)
  • Radiocontrast agents, IV (58% in 2012, 42% in 2014)
  • Inotropic medications, IV (66% in 2012, 53% in 2014)
  • Antiarrhythmics, IV (70% in 2012, 58% in 2014)
  • Nitroprusside sodium, IV (76% in 2012, 65% in 2014)
  • Dextrose, hypertonic, 20% or greater (74% in 2012, 64% in 2014)
  • Oxytocin, IV (71% in 2012, 61% 2014)
  • Thrombolytics (96% in 2012, 87% 2014)

While there were very small increases between 2012 and 2014 regarding individual beliefs that a drug was a high-alert medication for 9 of the 39 drugs/drug classes, increases of any significance were observed for only 2 drugs:

  • Opioids, transdermal (considered a high-alert medication by 66% in 2012, 73% in 2014)
  • Sodium chloride for injection, hypertonic, greater than 0.9% (81% in 2012, 87% in 2014)

Practice Site Adoption

Overall results. Table 1 also shows whether respondents’ practice sites considered the 39 drugs/drug classes to be high-alert medications, whether there were specific precautions in place at their practice sites, and whether respondents thought the precautions were effective. More than 90% of respondents reported that their practice sites considered anticoagulants (96%), IV insulin (95%), and chemotherapy (93%) to be high-alert medications. More than 80% of the practice sites considered potassium chloride injection concentrate (89%), subcutaneous insulin (85%), and neuromuscular blocking agents (86%) to be high-alert medications. About three-quarters of practice sites considered epidural/intrathecal medications (76%), hypertonic sodium chloride injection (75%), and potassium phosphate injection (74%) to be high-alert medications. About two-thirds (67%) of practice sites considered injectable opioids to be high-alert medications. The drugs least frequently considered high-alert drugs at practice sites included oral hypoglycemics (18%); sterile water for injection, inhalation, or irrigation (18%); dialysis solutions (22%); and liposomal drugs and conventional counterparts (28%).

Comparison to 2012 results. For about one-third of the 39 drugs/drug classes on ISMP’s list of high-alert medications, the 2014 findings regarding practice site adoption of these medications as high-alert were similar to the 2012 findings. However, for 24 of the 39 medications, significantly fewer practice sites in 2014 considered these drugs to be high-alert medications than in 2012. The reduction was at least 5 percentage points, with the largest differences as follows:

  • Cardioplegia solution (high-alert drug in 59% of practice sites in 2012, 35% in 2014)
  • Thrombolytics (93% in 2012, 70% in 2014)Dialysis solutions, peritoneal and hemodialysis (44% in 2012, 22% in 2014)
  • Radiocontrast agents, IV (46% in 2012, 29% in 2014)
  • Dextrose, hypertonic, 20% or greater (59% in 2012, 42% in 2014)
  • Beta-adrenergic antagonists, IV (44% in 2012, 29% in 2014)
  • Hypoglycemics, oral (31% in 2012, 18% in 2014)
  • Inotropic medications, IV (51% in 2012, 38% in 2014)
  • Moderate sedation agents, oral, for children (65% in 2012, 53% in 2014)
  • Nitroprusside sodium, IV (63% in 2012, 52% in 2014)
  • Antiarrhythmics, IV (55% in 2012, 44% in 2014)

The only drug that increased more than 3 percentage points between 2012 and 2014 was hypertonic sodium chloride for injection (70% in 2012, 75% in 2014).

Precautions and Effectiveness

If respondents reported that the targeted drugs were a high-alert medication at their practice sites, most also reported that there were special precautions in place to prevent errors with these drugs. For 38 of the 39 drugs, 87% or more respondents reported that special precautions were in place. Precautions were in place in only 77% of practice sites that considered oral hypoglycemics to be a high-alert medication.

If there were special precautions in place for drugs considered high-alert medications, respondents often felt the precautions were effective. For 34 of the 39 drugs, 90% or more felt the precautions were effective. The lowest levels of confidence regarding precautionary measures were associated with the remaining 5 high-alert medications: 

  • Insulin, subcutaneous (80% felt precautions were effective)
  • Opioids, transdermal (89%)
  • Opioids, injectable (IM, IV) (88%)
  • Opioids, transmucosal and oral (87%)
  • Anticoagulants (87%)

Differing Views

As with the 2012 survey, differences were again apparent between nurses’ and pharmacists’ perceptions regarding which drugs they considered to be high-alert medications (Table 1). In general, nurses more frequently identified the drugs in the survey as high-alert medications than pharmacists, particularly with these drugs/drug classes:

  • Dialysis solutions, peritoneal and hemodialysis (65% nurses, 26% pharmacists)
  • Radiocontrast agents, IV (66% nurses, 32% pharmacists)
  • Nitroprusside sodium, IV (88% nurses, 55% pharmacists)
  • Moderate sedation agents, IV (96% nurses, 64% pharmacists)
  • Moderate sedation agents, oral, for children (92% nurses, 61% pharmacists)
  • Antiarrhythmics, IV (78% nurses, 48% pharmacists)
  • Dextrose, hypertonic, 20% or greater (83% nurses, 56% pharmacists)
  • Vasopressin, IV or intraosseous (81% nurses, 56% pharmacists)

With these drugs, nurses may feel more vulnerable to harmful errors than pharmacists, particularly if the nurse is responsible for the selection and/or preparation of doses/infusions. Nurses may also have witnessed transient patient harm with these drugs in patient care areas, enhancing awareness of the potential for harm with the drugs.

On the other hand, the following examples are drugs that pharmacists identified as high-alert medications more often than nurses: 

  • Sterile water for injection, inhalation, and irrigation (44% pharmacists, 14% nurses)
  • Methotrexate, oral, non-oncologic use (71% pharmacists, 51% nurses)
  • Sodium chloride injection, hypertonic, greater than 0.9% (91% pharmacists, 76% nurses)
  • Liposomal drugs and conventional counterparts (52% pharmacists, 39% nurses)
  • Factor Xa inhibitors (80% pharmacists, 67% nurses)
  • Potassium chloride injection concentrate (95% pharmacists, 85% nurses)
  • Potassium phosphate injection (81% pharmacists, 71% nurses)

These findings suggest that perhaps pharmacists have greater awareness of the risk of harm associated with errors for these drugs than nurses. Perhaps some of these differences can also be explained by the fact that pharmacists may have sole access to several of the products for use during IV admixture.

Suggested Additions

In the survey, ISMP provided eight drugs/drug classes or other products to consider for addition to, or special emphasis on the ISMP List of High-Alert Medications for Acute Care Settings (Table 1). Insulin U-500 received an affirmative response from 98% of respondents; 92% of these respondents said their practice sites already consider this a high-alert drug, and of this group, 90% have initiated special precautions to prevent errors with insulin U-500. Half of the respondents also reported that subcutaneous EPINEPHrine is considered a high-alert drug at their practice sites. ISMP plans to add these two drugs to its high-alert medications list.

We received mostly affirmative responses to several other suggested additions in the survey. We also received several comments suggesting other medications for consideration, including IV levothyroxine, metoclopramide used in pediatrics, arginine IV, ferrous sulfate liquid and drops, sodium bicarbonate 8.4%, phytonadione, benzodiazepines, calcium IV, and regional anesthetics. We appreciate all the thought that went into making these suggestions. In an effort to keep the list manageable, we will be evaluating these possible additions carefully over the next few weeks. We will also determine whether any drugs currently on the list should be removed. We will then update our list of high-alert medications and publish it in this newsletter and on our website.

Table1
Table 1. Drugs considered to be high-alert medications

Conclusion

Again, ISMP thanks those who took the time to complete our survey on high-alert medications. Nonetheless, we are disappointed in the lower-than-usual response to this survey. In 2012, we received almost 800 survey responses. We are also concerned about what appears to be a significant downward trend in including many of the drugs on ISMP’s list of high-alert medications on practice site lists. It appears that precautions are frequently established and felt to be effective if a drug is included on the practice site list. But there seems to be numerous medications that no longer appear on practice site lists, despite long-standing risks that have led to catastrophic errors. We realize that a practice site list may not include all the medications on the ISMP list because they may not be used in a particular setting (e.g., chemotherapy). However, a reduction in practice sites that consider an injectable opioid or thrombolytics to be high-alert drugs is concerning.

Although not covered in this summary of findings, there are also significant gaps regarding the medications healthcare practitioners believe should be high-alert medications and those included on practice site lists. For example, while 78% of respondents felt opioids (IM/IV) were high-alert medications, only 67% of respondents reported that their practice sites consider this class of drugs to be high-alert medications. We urge all organizations to use these survey findings to encourage discussions about high-alert medications in your organization. Focusing on differing nursing and pharmacy perspectives regarding which drugs should be considered high-alert medications may prove especially worthwhile to enhance understanding of the perceived risks by each group. It would also be useful to evaluate any gaps in practice site adoption of safety precautions for drugs that staff perceive to be high-alert medications.