The most common injuries reported with EAIs are unintentional injections. The incidence of accidental injection, mostly involving EpiPen devices, which are the most commonly used on the market, injected into the thumb, is estimated at 1 in 50,000 EpiPen units , and up to 16% of doctors who read the instructions on the autoinjector used the EpiPen® trainer in a manner that would have self-injected into their thumb . There were over 15,000 unintentional EpiPen injections reported to U.S. Poison Control Centers over 14 years of study . Of 105 unintentional injections from EAIs reported to the Food and Drug Administration Adverse Event Reporting System , more than one-third of the individuals injected were health care professionals.
Lacerations and injuries from epinephrine auto-injector needles are less common but pose a risk. Brown et al. reported 25 cases of EpiPen-associated laceration and embedded needle injuries  including 20 with thigh lacerations, a nurse with a digit laceration, and four children with stuck needles. The mean age in injured children was 3 years. Operators included parents, educators and a child, but also involved healthcare providers. The authors suggested that the 10-second hold of the EpiPen may have contributed to these injuries and may be excessive, considering evidence that EpiPens deliver epinephrine in less than 3 s [8, 9]. The hold time for EpiPen was subsequently reduced to 3 s in the United States. The hold time varies in other countries between 3 s (England and Australia), several seconds (Canada), 5 s (Sweden) and 10 s (many European, African and Asian countries).
Brown and Tuuri reported an additional case of laceration and provided guidance for providers on how to educate families regarding appropriate child restraint during injection . In the United States, patient information now includes instructions to “hold the child’s leg firmly in place and limit movement prior to and during injection” (accessdata.fda.gov/drugsatfda_docs/label/2017/019430s067lbl.pdf), although these simple instructions may fail to convey the degree of restraint needed to prevent these injuries in a combative child.
While bent needles are often reported in connection with leg laceration injuries, hooked needles are a less frequent complication of EpiPen use. Two of the current cases of hooked EpiPens were reported previously , however, the cause of and management of hooked EAI needles has not been discussed elsewhere. We propose three explanations for the hooked EpiPen needles observed in this study. The first is that the needle could hit a hard structure such as bone during injection and curve rather than penetrate further. This may explain the first 2 cases described here, where the EpiPen was injected in an area with a short skin-to-bone distance. This type of hooking might similarly happen if the needle were to hit a very stiff seam of clothing, although we are unaware of any reports of this occurring. It is unlikely that a plain film would identify the location of impact, so x-rays of the bone are unlikely to offer proof of this proposed explanation. Instructions indicate that users should avoid injecting at closing seams . Secondly, the needle could bend if the patient moves during injection. In most cases, patient movement results in needles bending in one straight line or with a simple curve, rather than a true hook . However, we describe one case where the tip of the curved needle appeared and behaved “like a hook”. Thirdly, our experience with testing many EpiPen devices suggests that EpiPens needles frequently don’t eject from the device perfectly straight. If they are sufficiently off-center to hit the cartridge carrier, this could hook the needle prior to injection.
In the fourth case we present, it is most likely that the needle became hooked prior to injection. The needle pierced the side rather than the center of the rubber needle cover, and then appears to have made contact with, and cracked, the white carrier that houses the cartridge and stopper. It then also appears to have cracked the orange shroud that typically covers the needle upon removal from the body. These two contacts appear to have hooked the needle, which likely entered the patient that way. The hooked needle was then difficult to remove.
The ergonomics of the EpiPen has attracted some concerns in the past . Upside-down use of EpiPen devices resulting in thumb injections as well as failed drug administration has been frequently reported. Suboptimal ergonomic design was cited as a reason for about half of cases of more than 100 unintentional injections, as people were trying to self-inject or inject to another person having an allergic reaction .
It is hard to determine what role needle bending plays in the creation of laceration injuries in children in other cases, but in two of our cases no laceration was noticed and the needle insertion site healed well. Bent needles are not covered by the plastic casing, which poses a potential injury to children and EAI providers. While minimizing needle injection time may have prevented some lacerations described previously , it is hard to predict if this would have been preventing the injury in the cases we present here.
Hooked needle were only seen with EpiPen devices in this study. This may reflect device prevalence in the community. Two other EAIs are available in the United States: Auvi-Q (kaléo, Richmond, VA) and a generic for Adrenaclick (Amneal Pharmaceuticals, Bridgewater, NJ.). In these devices, the syringe does not push a rubber stopper up against a cartridge carrier during firing. The mechanisms of firing are sufficiently different that they may not have the same potential for off-center firing of the needle compared with the EpiPen device.