Featured Articles

Start the Year Off Right by Preventing These Top 10 Medication Errors and Hazards from 2020

The end of 2020 marked an important milestone for ISMP as we celebrated 25 years of publishing this newsletter, the ISMP Medication Safety Alert! Acute Care, every 2 weeks. Your willingness to voluntarily report medication errors and hazards to ISMP, and to proactively use the information we publish in the newsletter to prevent similar errors and hazards, motivates and inspires us as, together, we continue to learn about the causes of medication errors and how to prevent them.

Reflecting on our most recent year of newsletter publication, ISMP has identified the Top 10 Medication Errors and Hazards (Table 1) that appeared in the ISMP Medication Safety Alert! during 2020. The list is not based only on the most frequently reported problems or those that have caused the most serious consequences to patients, although these factors were considered. Instead, we identified errors and hazards that have been persistent and can be avoided or minimized with system and practice changes. We know it has been an extremely challenging year for healthcare providers given the current coronavirus disease 2019 (COVID-19) pandemic. Two of the Top 10 Medication Errors and Hazards are closely associated with the pandemic—errors with the COVID-19 vaccines and hazards associated with positioning infusion pumps outside of COVID-19 patients’ rooms. In addition to the two pandemic-related hazards, we believe the other eight issues warrant attention and priority in the coming year.

Table 1. Top 10 Medication Errors and Hazards from 2020

1 Prescribing, dispensing, and administering extended-release (ER) opioids to opioid-naïve patients
2 Not using smart infusion pumps with dose error-reduction systems (DERS) in perioperative settings
3 Errors with oxytocin
4 Hazards associated with positioning infusion pumps outside of COVID-19 patients’ rooms
5 Errors with the COVID-19 vaccines
6 Use of the retrospective, proxy “syringe pull-back” method of verification during pharmacy sterile compounding
7 Combining or manipulating commercially available sterile products outside the pharmacy
8 Medication loss in the tubing when administering small-volume infusions via a primary administration set
9 Wrong route (intraspinal injection) errors with tranexamic acid
10 Use of error-prone abbreviations, symbols, or dose designations


  1. Prescribing, dispensing, and administering extended-release (ER) opioids to opioid-naïve patients

Inappropriate prescribing of ER opioids to opioid-naïve patients has resulted in serious harm and death. ISMP, as well as the US Food and Drug Administration (FDA), have warned practitioners about this well known problem for decades. However, inappropriate opioid prescribing continues to occur, often due to a knowledge deficit about the dangers associated with prescribing ER opioids to opioid-naïve patients and/or not understanding the difference between opioid-naïve and opioid-tolerant. For example, in 2020, ISMP published several new reports related to prescribing fentaNYL patches to opioid-naïve, elderly patients, sometimes to treat acute pain or due to a codeine “allergy” that was a minor drug intolerance. FentaNYL patches should only be prescribed to opioid-tolerant patients for the management of pain severe enough to require daily, around-the-clock, long-term opioid treatment. This is so critical to safety that, in 2018, ISMP called for the elimination of prescribing fentaNYL patches to opioid-naïve patients and/or patients with acute pain in our Targeted Medication Safety Best Practices for Hospitals. In 2020, this Best Practice was incorporated into a new Best Practice to verify and document the patient’s opioid status (naïve vs. tolerant) and type of pain (acute vs. chronic) before prescribing and dispensing ER opioids.

To do this, ISMP first recommends establishing definitions for opioid-naïve and opioid-tolerant patients (for example, following the fentaNYL package insert definitions), and then developing and implementing a standard process for gathering and documenting each patient’s opioid status and type of pain (if pain is present). Order entry systems should default to the lowest initial starting dose and frequency when initiating orders for ER opioids, and interactive alerts should be built to confirm opioid tolerance when prescribing and dispensing ER opioids. Distinguish between true allergies and drug intolerances when collecting allergy information. Eliminate the storage of fentaNYL patches in automated dispensing cabinets (ADCs) or as unit stock in clinical locations where primarily acute pain is treated (e.g., in the emergency department [ED], operating room, post-anesthesia care unit, procedural areas). Our 2020 survey showed low compliance with many of these recommendations.

  1. Not using smart infusion pumps with dose error-reduction systems (DERS) in perioperative settings

Our updated (2020) Guidelines for Optimizing Safe Implementation and Use of Smart Infusion Pumps recommend the use of smart pumps with DERS throughout the organization, including in perioperative settings, for all infusions (including hydrating solutions) and bolus/loading doses. However, use of smart pumps with DERS by anesthesia providers in perioperative settings is limited due to barriers and unique challenges. One common barrier to optimal use of smart pumps with DERS in perioperative settings is that there may not be a clear expectation from leadership for anesthesia providers to use smart pumps with DERS. Many anesthesia providers do not understand the capabilities of smart pumps, including loading/bolus dose capabilities. Anesthesia providers may also feel that the soft and hard dose/infusion limits set in the pump are unacceptable, often because they have not been included when building the anesthesia/perioperative drug library. In many organizations, smart pumps are used in the operating room using an “anesthesia mode” setting. However, the organization may fail to understand that, in some pumps, “anesthesia mode” settings reduce all hard stops to soft stops, thereby allowing easy overrides of dosing/concentration limits that should never be bypassed.

Leadership must clearly establish that the use of smart pumps with an engaged DERS is expected in perioperative settings for all infusions and loading/bolus doses (except when the hydrating solution rate is greater than the pump allows). Involve anesthesia providers when building the smart pump library. When possible, implement upper and lower hard limits for medication doses, concentrations, infusion rates, and loading/bolus doses, and restrict the use of pumps in “anesthesia mode” if it affects individualization of infusion limits. Require anesthesia providers to use the bolus feature (if available) with hard limits for catastrophic doses, and do not allow the delivery of bolus doses by increasing the rate of the infusion. Hands-on education about how to use smart pumps with DERS, including the bolus dose feature, along with competency assessments should be implemented for all anesthesia providers. Organizations are encouraged to analyze pump data to understand any barriers to the effective use of smart pumps with DERS in the perioperative setting.

  1. Errors with oxytocin

In 2020, ISMP conducted an analysis of oxytocin errors, many of which caused hyperstimulation of the uterus, which can result in fetal distress, uterine rupture, or an emergency cesarean section. Sadly, a few maternal, fetal, and neonatal deaths have been reported. More than one-third of the reported errors were associated with look-alike vials and label confusion. For example, generic oxytocin and brand PITOCIN vials look similar to ondansetron vials from various manufacturers, which all have green caps. Several recent 10-fold dosing errors were caused by label confusion with 1, 10, and 30 mL oxytocin vials (Fresenius Kabi). The labels prominently display “10 USP units/mL,” with the total volume in the vial at the bottom of the label, causing staff to think there were only 10 units total in each vial. A few prescribing errors were caused by selection of the wrong drug on order entry screens when searching using only the first few letters of a drug name (e.g., “OXY10” for oxytocin vs. oxyCODONE; “PIT” for Pitocin vs. PITRESSIN [discontinued brand of vasopressin]). Occasionally, verbal orders for “Pitressin” were misheard as Pitocin and dispensed, or vice versa. Administration errors were often related to incomplete or omitted labels on nurse-prepared oxytocin solutions, which often led to infusion bag swaps. Numerous errors were reported in which an oxytocin bag was mixed up with either a hydrating fluid or magnesium infusion.

To prevent oxytocin errors, require prescribers to use at least five letters of a drug name when searching electronic systems. Avoid nurse-prepared oxytocin infusions and instead have pharmacy dispense oxytocin in ready-to-administer, labeled bags in standardized concentrations. Ensure oxytocin vial (and premixed infusion) labels are clear regarding the amount of drug per total volume. Employ barcode scanning technology when stocking ADCs and when preparing and administering infusions. Infuse all oxytocin solutions through a smart infusion pump with an engaged DERS. Immediately discard discontinued oxytocin infusion bags.

  1. Hazards associated with positioning infusion pumps outside of COVID-19 patients’ rooms

During the COVID-19 pandemic, some hospitals have positioned infusion pumps outside of COVID-19 patients’ rooms to conserve personal protective equipment (PPE), reduce staff exposure, and enhance the ability to hear and respond to pump alarms in a timely manner. This has been accomplished through the use of extension sets. The length and inner diameter of the long extension tubing can impact the volume of fluid needed for priming, flow rates, and the time medications and solutions take to reach a patient. Inadvertent bolus doses of medication remaining in the extension set might be administered to a patient when flushing the long tubing. Occlusion alarms may be delayed at low flow rates or become excessive at high flow rates. The long extension tubing (and electrical cords) may pose a tripping hazard and become tangled and disconnected. Barcode scanning of the patient and drug may be more challenging, and certain components of an independent double check may become difficult or impossible in some situations.

While recognizing that this is not ideal, hospitals must weigh the risk vs. benefit of positioning infusion pumps outside of COVID-19 patients’ rooms. If a decision has been made to locate pumps outside of rooms, planning and periodic reassessment of the process is a must. A special report from ECRI can help guide the selection and use of long extension sets for this purpose and includes other factors (e.g., fluid viscosity) that should be considered. Conduct periodic infusion pump rounds in the hallway to verify the accuracy of the fluids and medications infusing as well as the pump settings. Also, check that the tubing is not disconnected or a tripping hazard. Develop a temporary process that allows some components of barcode scanning and/or independent double checks to occur prior to medication administration. For example, because nurses cannot scan the barcode on the patient’s identification band, some hospitals affix the patient’s name, birthdate, and a barcode to the pump or intravenous (IV) pole located outside the room. At the end of the pandemic or when pumps are no longer located in hallways, please discontinue temporary identification measures and have staff return to the verification processes in place prior to the pandemic.

  1. Errors with the COVID-19 vaccines

Findings from the analysis of recent influenza (flu) vaccine errors can be used to prevent errors during the COVID-19 vaccine campaigns that started in December 2020. Common contributing factors associated with flu vaccine errors that could also be a risk factor for COVID-19 vaccinations include: look-alike vaccine names, labels, and packaging; unsegregated refrigerator/freezer storage; mixing/dilution errors; communication barriers with patients; not checking/documenting administration in the immunization information system (IIS); temperature excursions; and the inability to use technologies such as barcode scanning during mass immunizations.

We also reviewed early COVID-19 vaccine errors voluntarily reported to ISMP since mid-December 2020. Numerous dilution errors with the Pfizer-BioNTech vaccine have led to overdoses when too little diluent was used, often 1 mL instead of 1.8 mL. In one case, patients received the entire vial contents without dilution. In a clinic, patients received intramuscular (IM) injections of Regeneron’s casirivimab (monoclonal antibody) instead of the Moderna vaccine due to vague labeling of the monoclonal antibody, which included a product code name, not the established name. Wasted vaccines from inefficient scheduling or “no shows” were reported, as was administration of the vaccine to patients younger than indicated. A few allergic reactions were reported. (Also see the Safety Brief [to the right] on COVID-19 vaccine errors.)

When planning COVID-19 vaccine campaigns, be sure vaccination sites have enough space to assess patients before vaccination, observe them after vaccination, and treat patients who experience a reaction, all while maintaining social distancing and other pandemic measures. Provide vaccinators with a Fact Sheet for the vaccine(s) being used, and verify their competency regarding vaccine storage and preparation, patient assessment, identification of the proper vaccine injection site, administration, and emergency treatment of anaphylaxis. Ensure the vaccine scheduling process includes a reliable system to confirm appointments. Also establish a standard process for dealing with leftover doses at the end of the day. If feasible within the timeframe for vaccine stability, have the pharmacy verify the number of vaccines needed each day and dispense predrawn, labeled syringes of the vaccine. Use low dead-volume syringes/needles to withdraw as many doses as possible from the vaccine vials. At all vaccination sites, be prepared to immediately treat an allergic reaction.

  1. Use of the retrospective, proxy “syringe pull-back” method of verification during pharmacy sterile compounding

In our 2020 survey on pharmacy sterile compounding systems and practices, only half of the respondents reported it is always easy to identify with certainty which drugs, diluents, and volumes were used when verifying the preparation of compounded sterile preparations (CSPs). Respondents with the lowest confidence in the verification process cited weaknesses in the outdated, post-production proxy “syringe pull-back” method of verification. Using this method, an ingredient is injected from the syringe into the final container, and then the plunger is pulled back to the amount on the syringe that was injected. It is this “pulled-back” syringe that is checked to determine the accuracy of the amount injected. Errors may not be detected if the syringe does not reflect the actual amount added or if the syringe is not partnered with the correct container. ISMP has received multiple reports of harmful or fatal errors, mostly preparing the wrong concentration/strength or using the wrong product/diluent, that were specifically related to a failed verification system using the “syringe pull-back” method.

In 2016, our Targeted Medication Safety Best Practices for Hospitals introduced a recommendation to perform an independent verification to ensure that the proper ingredients and volumes have been prepared prior to their addition to the final container. Specifically, the Best Practice calls for the elimination of proxy methods of verifying the CSP ingredients, including the “syringe pull-back” method. Instead, the Best Practice recommends the use of technology to assist in the verification process (e.g., barcode scanning, gravimetric verification, robotics, IV workflow software) to prevent errors that would not be detected with the “syringe pull-back” method. In 2020, ISMP broadened the scope of this Best Practice to all CSPs by removing a minimum requirement to perform a prior independent verification for high-alert medications and CSPs for high-risk patients and high-risk and routes of administration. Today, we urge organizations to eliminate the “syringe pull-back” method of verification and to conduct an independent verification of all CSP ingredients prior to their addition to the final container. 

  1. Combining or manipulating commercially available sterile products outside the pharmacy

Our recent 2020 survey on admixture outside the pharmacy showed that this error-prone practice happens often during emergency situations, mostly without formal training, and that there are significant procedural deviations and challenges associated with the practice that contribute to risk. Survey respondents told us that IV push medications, IV intermittent infusions, IM injections, and IV continuous infusions were the most frequent sterile injectables prepared outside of the pharmacy, primarily by nurses, anesthesia providers, and physicians. Nearly half of the respondents told us they have not been formally trained for this complex task. The biggest concerns expressed by respondents were lack of space, rushing through the preparation process, labeling issues, mixing by memory, interruptions and distractions, and concerns about sterility and accuracy. Nearly one-third of the respondents were aware of associated errors in the past year, particularly preparation errors. 

Use the results of this survey to prompt internal discussions about the need to limit the preparation of admixtures outside the pharmacy as much as possible and how to increase the use of pharmacy- and manufacturer-prepared, ready-to-use products. If your organization did not participate in this survey, you can download it by clicking here, conduct it internally, and review the results to pinpoint your vulnerabilities and establish a plan for improvement. Your goal for 2021 should be to significantly reduce the need and frequency of admixture outside the pharmacy.

  1. Medication loss in the tubing when administering small-volume infusions via a primary administration set

When a patient has a vascular access device (e.g., saline lock) without a continuously infusing, compatible primary solution, small-volume intermittent infusions (50 to 100 mL) are often administered using a longer primary administration set (via pump or gravity) connected directly to the patient’s vascular access device. This may lead to significant underdosing because the residual volume remaining in the primary administration set may not be administered to the patient. Primary administration sets hold various amounts of residual volume in the tubing (e.g., BD Alaris pump infusion set holds about 25 mL). In one health system, about 360,000 small-volume infusions annually were likely being administered to patients at lower doses than prescribed using primary administration sets, which could have a clinical impact on patient outcomes.

To reduce the risk of lost medication in the tubing of primary administration sets, administer small-volume intermittent infusions using a shorter secondary set. It is also important to embed an appropriate carrier fluid in order sets so they will be prescribed along with the small-volume intermittent infusions, if needed. A carrier fluid is a small bag (50 to 250 mL) of compatible fluid that is used as a primary infusion to allow administration of the intermittent infusion via a secondary administration set. After the intermittent infusion is complete, the carrier fluid is infused to flush residual drug from the tubing.

  1. Wrong route (intraspinal injection) errors with tranexamic acid

ISMP continues to receive reports involving the accidental intraspinal injection of tranexamic acid instead of a local anesthetic intended for epidural or spinal anesthesia. Bupivacaine, ropivacaine, and tranexamic acid are sometimes packaged in vials with the same blue color cap. When the vials are standing upright in storage, practitioners have picked up a vial based on cap color and not noticed it was the wrong vial. Wrong route errors with tranexamic acid is the only error type repeated from our 2019 list of Top 10 Medication Errors and Hazards, and it is the only danger that rose to the level of activating the National Alert Network during 2020. Last month, FDA announced that it will be revising the tranexamic acid labeling to highlight the IV route of administration and strengthen the warnings to include the risk of wrong route errors. Accidental intraspinal injection of tranexamic acid results in severe patient harm, with a mortality rate of 50%. 

We urge practitioners to purchase these products from different manufacturers to help differentiate their appearance and/or consider alternate preparations (e.g., premixed bag, pharmacy-prepared syringes or infusions). Avoid upright storage of the vials so the labels are always visible. Store tranexamic acid vials away from other look-alike vials and add an auxiliary label to vials to highlight the IV route of administration. When possible, employ barcode scanning prior to dispensing and administration. Exela Pharma Sciences manufactures a premixed bag of 1 g/100 mL of tranexamic acid, which should be used when appropriate, or have the pharmacy prepare minibags to reduce the risk of mix-ups. Transition to NRFit syringes and connectors for local anesthetics to prevent misconnections with drugs intended for IV use.

  1. Use of error-prone abbreviations, symbols, or dose designations

Abbreviations, symbols, and certain dose designations are a convenience, a time saver, a means of fitting a word, phrase, or dose into a restricted space, and a way to avoid misspellings. However, they are sometimes misunderstood, misread, or misinterpreted, occasionally resulting in patient harm. ISMP has repeatedly published errors resulting from misinterpretation of error-prone abbreviations, symbols, and dose expressions, particularly those associated with doses/measurement units, routes of administration, drug name abbreviations, and apothecary/household abbreviations.

By the end of next week, we plan to post the 2021 updated ISMP List of Error-Prone Abbreviations, Symbols, and Dose Designations. Our updated list includes abbreviations, symbols, and dose designations that were reported to ISMP and have been frequently misinterpreted and involved in harmful or potentially harmful errors. They should NEVER be used when communicating medical information verbally, electronically, and/or in handwritten applications. In 2021, we encourage organizations to review our updated list and to use it to create or update your organization’s “Do Not Use” abbreviation list. Error-prone abbreviations, symbols, and dose designations that are included on The Joint Commission’s “Do Not Use” list (Information Management standard IM.02.02.01) are highlighted in the ISMP list, as are the error-prone abbreviations, symbols, and dose designations that are relevant mostly in handwritten communications.