Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Do Doctors know how to use adrenaline autoinjectors correctly?

Three Part Question

Do [doctors] know how to use [an adrenaline autoinjector] properly in [patients with anaphylactic symptoms]

Clinical Scenario

You read through the latest edition of a medical journal and notice a case report of a doctor who accidentally injected their own thumb while trying to use an Epipen. When you subsequently try to practice using a training device you do the same thing, this makes you wonder how many doctors do know how to use AAI correctly?

Search Strategy

Medline 1946 to July Week 5 2013

EMBASE 1980 to 2013 Week 26
[exp Physicians/ OR exp Health Personnel OR exp General Practitioners/ OR doctor$.mp. OR clinician$.mp. OR health care provider$.mp. OR] AND [exp Patient Education as Topic/ OR exp Health Education/ OR exp Teaching/ OR educat$.mp. OR teach$.mp.OR instruct$.mp. OR train$.mp.]AND [exp Epinephrine/OR exp Self AdministrationOR exp Injections, Intramuscular/ OR adrenalin$.mp. OR OR OR OR OR OR intramuscular$.mp. OR] limit to (english language and humans).

Search Outcome

Three hundred and twenty-four papers were found; however, only five were relevant.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Sicherer et al,
46 paediatrics enrolled via unscheduled visits to paediatric departments and 5 private practices in the area. Only 36 doctors took part in the demonstration task Single centre cross sectional descriptive study Used a demonstration for AAI training 17% used a training device to educate patients/parents Small sample Enrolment methods were not systematic No explanation or analysis between the doctors who did and did not complete the demonstration task
Used written information for AAI training 24% gave written material to patients/parents
Ability to demonstrate 6 key steps in AAI use 9 (25%) able to use Epipen 4 (11%) able to use EpiEZ 1 (3%) able to use Anakit
Jarvinen et al,
82 doctors enrolled through a variety of means; Staff meetings (100% response rate), conferences/gran rounds (49% RR), interoffice mailings to emergency department (36% RR) Subset invited a teaching sessions (35% RR, n=26) Double site cross sectional descriptive study Comfort in prescribing AAI 66% stated they were comfortable prescribing and providing training for AAIsSmall sample Selection bias due to the convenience method of sampling, hence no information on non-responders. Possible doctors who recognised a personal weakness in this area were more likely to respond/attending teaching.
Used demonstration for AAI training Only 32% had a training device in clinic, no data on how often this was used.
Ability to correctly use an AAI before and after organised teaching Pre-teaching – 23% completed all 6 steps Post teaching – 85% Majority felt their knowledge and confidence had also improved.
Mehr et al,
100 doctors from juniors to consultants from a variety of specialities No statistical difference between groups on familiarity Single centre cross sectional descriptive study Prescribing of AAIs45 (41%) had prescribed an AAI. 49% stated them did not provide any training
Ability to use AAI using 6 step techniqueMean score of 4 out of 6. 2 participants were able to complete all 6 steps. Those who did not read the instructions, were more senior (p=0.03) and scored significantly better than those who read the instructions. (p=0.02) 95% of doctors read the instructions of which only 41% were able to correctly inject.
Adverse effects 16% injected into thumbs
Arga M et al,
196 Paediatricians at various points through training Questionnaire competed by 196 Training attended by 158 151 reassessed 6 months later Single site cross sectional descriptive cohort studyAbility to correctly use AAI improves training 1st attempt – 23.2% of doctors were able to correctly use AAI 6 months later – 74.2% were able to correctly use the AAI (P<0.001) 7.3% injected into thumb after education
Time to administer AAI decreases after training 1st attempt – 28 seconds 2nd attempt – 20 seconds.
Grouhi et al,
122 Medical/nursing/pharmacists at a conference. Ability to deliverObservational cohortAble to demonstrate 4-step approach to autoinjection technique25.4% of participants demonstrated all four stepsSmall numbers, wide variety of staff. Not a clinical setting. Details of exact method of assessment limited in paper


All five studies used objective testing of doctors’/healthcare staff's ability to perform a successful AAI injection, and throughout results were poor. Despite different scoring techniques being used (6 vs 5 vs 4 step) on average only 23% of the doctors tested were successful on the first assessment. Arga et al and Jarvinen et al attempted to improve scores; both studies used educational sessions with demonstrations, although the uptake rate of these sessions was variable, clear improvements in doctors’ ability was noted. A number of accidental thumb injections were recorded in the studies by Arga et al and Mehr et al (7.3% and 16%, respectively); some of these accidents took place after the instructional leaflet was consulted. Both author groups believe the design of AAI is difficult to comprehend and far from intuitive. Improved labelling and a transparent cover for the needle may act as visual prompts and prevent such accidents. Unfortunately, the sample sizes were small in a number of these studies and due to the convenience method of sampling used by Sicherer et al and Jarvinen et al there is a possibility of significant selection bias. It is likely that doctors who acknowledged a personal weakness in this area were more likely to enrol and hence reduce average ability levels. Although a number of nurses are involved in the care of both paediatric and adult allergy patients, only the paper by Grouhi et al included nurses (and then in small numbers). Such research would be beneficial for comparison.

Editor Comment

AAI, adrenaline autoinjector.

Clinical Bottom Line

Despite the weaknesses of these studies, the findings are consistent throughout. It is clear the majority of doctors are unable to use AAI correctly and are therefore unable to teach patients or parents effectively. The lack of AAI technique teaching in the undergraduate curriculum means many junior doctors enter the workplace without this key skill. Therefore, the provision of educational sessions for doctors regardless of specialty or seniority is necessary and has been shown to be beneficial, improving ability, knowledge and confidence. Doctors who regularly come into contact with patients with allergies or are required to prescribe AAI should ensure their skills are competent and should utilise the wide variety of educational tools including DVDs and training devices available for both themselves and their patients.


  1. Sicherer SH, Forman JA, Noone SA. Use Assessment of Self-Administered Epinephrine Among Food-Allergic Children and Pediatricians. Pediatrics 2000;105:359–62.
  2. Jarvinen KM, Sicherer SH, Sampson HA, et al. Use of multiple doses of epinephrine in food-induced anaphylaxis in children. J Allergy Clin Immunol 2008;122:133–8.
  3. Mehr S, Robinson M, Tang M. Doctor--how do I use my EpiPen? Pediatr Allergy Immunol 2007;18:448–52.
  4. Arga M, Bakirtas A, Catal F, et al. Training of trainers on epinephrine autoinjector use. Pediatr Allergy Immunol 2011;22:590–3.
  5. Grouhi M, Alshehri M, Hummel D, et al. Anaphylaxis and epinephrine auto-injector training: who will teach the teachers? J Allergy Clin Immunol 1999;104:190–3.