Objectives. To perform a systematic investigation of medications associated
with adverse sedation events in pediatric patients using critical incident
analysis of case reports.
Methods. One hundred eighteen case reports from the adverse drug reporting
system of the Food and Drug Administration, the US Pharmacopoeia, and the r
esults of a survey of pediatric specialists were used. Outcome measures wer
e death, permanent neurologic injury, prolonged hospitalization without inj
ury, and no harm. The overall results of the critical incident analysis are
reported elsewhere. The current investigation specifically examined the re
lationship between outcome and medications: individual and classes of drugs
, routes of administration, drug combinations and interactions, medication
errors and overdoses, patterns of drug use, practitioners, and venues of se
dation.
Results. Ninety-five incidents fulfilled study criteria and all 4 reviewers
agreed on causation; 60 resulted in death or permanent neurologic injury.
Review of adverse sedation events indicated that there was no relationship
between outcome and drug class (opioids; benzodiazepines; barbiturates; sed
atives; antihistamines; and local, intravenous, or inhalation anesthetics)
or route of administration (oral, rectal, nasal, intramuscular, intravenous
, local infiltration, and inhalation). Negative outcomes (death and permane
nt neurologic injury) were often associated with drug overdose (n = 28). So
me drug overdoses were attributable to prescription/transcription errors, a
lthough none of 39 overdoses in 34 patients seemed to be a decimal point er
ror. Negative outcomes were also associated with drug combinations and inte
ractions. The use of 3 or more sedating medications compared with 1 or 2 me
dications was strongly associated with adverse outcomes (18/20 vs 7/70). Ni
trous oxide in combination with any other class of sedating medication was
frequently associated with adverse outcomes (9/10). Dental specialists had
the greatest frequency of negative outcomes associated with the use of 3 or
more sedating medications. Adverse events occurred despite drugs being adm
inistered within acceptable dosing limits. Negative outcomes were also asso
ciated with drugs administered by nonmedically trained personnel and drugs
administered at home. Some injuries occurred on the way to a facility after
administration of sedatives at home; some took place in automobiles or at
home after discharge from medical supervision. Deaths and injuries after di
scharge from medical supervision were associated with the use of medication
s with long half-lives (chloral hydrate, pentobarbital, promazine, prometha
zine, and chlorpromazine).
Conclusions. Adverse sedation events were frequently associated with drug o
verdoses and drug interactions, particularly when 3 or more drugs were used
. Adverse outcome was associated with all routes of drug administration and
all classes of medication, even those (such as chloral hydrate) thought to
have minimal effect on respiration. Patients receiving medications with lo
ng plasma half-lives may benefit from a prolonged period of postsedation ob
servation. Adverse events occurred when sedative medications were administe
red outside the safety net of medical supervision. Uniform monitoring and t
raining standards should be instituted regardless of the subspecialty or ve
nue of practice. Standards of care, scope of practice, resource management,
and reimbursement for sedation should be based on the depth of sedation ac
hieved (ie, the degree of vigilance and resuscitation skills required) rath
er than on the drug class, route of drug administration, practitioner, or v
enue.