Home Support ISMP Newsletters Webinars Report Errors Educational Store Consulting FAQ Tools About Us Contact Us
ISMP Facebook

Evidence builds: lack of focus on human factors allows error-prone devices

From the July 28, 1999 issue

PROBLEM: Flawed design of medication delivery devices promotes user error, often resulting in adverse patient outcomes. Many are aware of a design flaw, present in over 200,000 infusion pumps currently used in US hospitals, that allows solutions to "free-flow" into patients if the tubing is removed from the pump without closing the set's gravity flow control clamp. Likewise, there are numerous design flaws with other devices, from syringes and administration sets to complex infusion pumps. Typically, practitioners using these devices are blamed for errors. Yet, we continue to receive error reports which show that some manufacturers and regulatory authorities fail to pay attention to user needs and human factors (the interrelationships between humans, the tools they use, and the environments in which they work1). Consider the following examples.

Abbott Lifecare 4100 PCA Plus II Infuser: A recent fatality occurred after a practitioner misprogrammed the drug concentration for a loading dose of morphine. When programming the device, the pump defaults to its minimum setting of 0.1 mg/mL. If the practitioner fails to adjust the concentration and accepts the 0.1 mg/mL default value, the pump calculates a delivery volume that results in a ten-fold overdose if using 1 mg/mL concentration morphine syringes. A fifty-fold overdose would occur if using the 5 mg/mL concentration syringes. Further, setting dose limits will not prevent overinfusions, as the pump believes it is delivering the proper volume/dose based on the concentration selected during programming. Erroneous low concentration settings pose the highest risk of patient harm when administering loading doses (typically larger doses than those self-administered) and when using higher concentrations of the drug (5 mg/mL). Additional fatalities have been previously described.2

Baxter AP-II infusion pump: When the pump alarmed to signify that the total volume of a 250 mL epidural infusion of fentanyl and bupivicaine had infused, a nurse found that the bag currently hanging was still full. While no solution had infused over the past 20 hours, the pump had been running and recording the volume as infused without detecting an "upstream" occlusion. While the manufacturer intends to release a new model that will detect "upstream" occlusions, there are no warnings to alert practitioners to this device failure in current models.

Medfusion syringe pump model 2010: A physician asked a nurse to press the "prime" button on a pump already delivering a continuous fentanyl infusion to a neonate. He wanted to give a bolus dose to manage pain before placing an additional venous line in the infant. The physician, rather than ordering a specific dose, erroneously used this imprecise and unsafe method which delivers an unknown quantity of drug and solution. However, the pump's design should not allow the "prime" button to be activated during infusion delivery. After 16 seconds, an alarm will sound, but if the button remains depressed, the fluid will continue to flow. In this case, the nurse accidentally pressed the "prime" button on the wrong pump, which was delivering a potassium chloride infusion of 0.1 mEq/Kg. The infant received an unknown amount of the concentrated electrolyte via IV bolus. Luckily, the infant did not sustain any harm.

Add-Vantage system, Viaflex Plus system, dual chambered IV containers and high alert drug concentrates in IV additive syringes: Errors of omission are frequently reported when nurses forget to activate and mix drugs administered intermittently via the AddVantage or Viaflex Plus systems. Until 1996, heparin was available in a dual chamber bag that required users to break a seal between the drug and 5% dextrose injection diluent to provide the final admixture. However, staff working on busy medical-surgical units either did not know that mixing was required or they forgot to activate the container. In one case, a patient developed impaired circulation from a DVT and had to have his leg amputated after staff twice neglected to activate the container. The patient received no heparin; only dextrose from the bottom chamber. As reported in our October 7, 1998, issue, similarly designed bags of amino acid and dextrose have likewise been misused. Also, before 1993, lidocaine concentrate syringes were available from three manufacturers to facilitate admixture preparation. However, the syringe needle also fit the "Y" site of IV administration sets. Before the products were discontinued, there were at least 70 deaths or brain injuries stemming from direct IV injection of the drug concentrate. Recently, ISMP became aware of a manufacturer that was considering a similar syringe design to package another concentrated cardiac medication. We expressed our concern to FDA and, to our knowledge, this is no longer being considered.

Epi-PEN syringes designed to deliver epinephrine for emergency treatment of allergic reactions: A recent study showed that 10 % of patients and family members did not know to remove the cap before injection; 62% did not know to hold the syringe in place for at least 10 seconds; and 46% did not know that it could be used without removing clothing.3 In our October 22, 1997, issue, we discussed a case where a nurse held the pen upside down and pushed against the black tip which houses the needle, believing this would uncap the device. However, the black tip is supposed to be placed against the thigh for injection. The spring-loaded needle automatically injected the nurse's thumb instead of the patient's thigh.

SAFE PRACTICE RECOMMENDATION: The above examples are painful evidence that careful application of human factors and user testing during device design is crucial to identify and correct device attributes that increase the potential for human error. Until FDA demands better designed devices and forces redesign or removal of unsafe devices currently on the market, implementing the following recommendations can enhance safety. Use PCA pumps that default to a high concentration or zero setting and strive to use a single concentration of analgesics. Avoid stocking and using multiple concentrations of the same drug that differ by a factor of ten (e.g. 0.5 mg/mL and 5 mg/mL of morphine). Carefully position syringe labels so that important drug information can be readily seen during pump set-up and infusion. Monitor patients frequently and have antidotes readily available. Consider using oximetry for PCA patients and those on continuous narcotic infusions so that errors can be detected more rapidly and consequences can be minimized. Avoid using pumps with a "prime" function that can "bolus" during infusions. When new devices are purchased, use Failure Mode and Effects Analysis (ask what could possibly go wrong) to identify design flaws and all points where user error might occur so that proactive safety measures can be undertaken. Most importantly, establish a system of independent double checks when high alert medications are administered. For example, have one practitioner program a pump and have another independently check the drug, concentration, and rate of infusion/other pump settings. Include verification of the patient and attachment to the appropriate line/route. Making practitioners aware of poor device design and the potential for user error can improve compliance with independent verification before administration. If patients or family members will be using medication delivery devices, provide them with clear verbal and written instructions. Require return demonstrations with sufficient practice to assure competency. Alert them to potential user error with the devices. Finally, report drug-related device failure and user error to USP-ISMP and FDA so that we can advocate timely action and use the valuable information to alert others to problems.

References: 1) Weinger, MB, et al. Incorporating human factors into the design of medical devices. Letter. JAMA 1998; 280(17); 1484. 2) ECRI. Hazard Report. Abbott PCA Plus II - Patient controlled analgesic pumps prone to misprogramming, resulting in narcotic overinfusions. Health Devices 1997;26:389-91. 3). Huang S. A survey of Epi-PEN use in patients with a history of anaphylaxis. J Allergy Clin Immunol 1998;102:525-6.

Acute Care Main Page
Current Issue
Past Issues
Highlighted articles
Action Agendas - Free CEs
Special Error Alerts
Newsletter Editions
Acute Care
Long Term Care
Home | Contact UsEmployment  | Legal Notices | Privacy Policy | Help Support ISMP
Med-ERRS Med-ERRS | MSOMedication Safety Officers Society | Consumer Medication SafetyFor consumers
 ISMP Canada ISMP Canada | ISMP Spain ISMP Spain | ISMP Brasil ISMP Brasil | International Group | Pennsylvania Patient Safety Authority

200 Lakeside Drive, Suite 200, Horsham, PA 19044, Phone: (215) 947-7797,  Fax: (215) 914-1492
© 2018 Institute for Safe Medication Practices. All rights reserved