Full text of DOI:10.1111/j.1553-2712.2001.tb01278.x

ACADEMIC EMERGENCY MEDICINE • February 2001, Volume 8, Number 2
139
Selected Tricyclic Antidepressant Ingestions Involving
Children 6 Years Old or Less
ROBIN B. M
CF
EE, DO, MPH, THOMAS R. CARACCIO, PHARMD,
HOWARD C. MOFENSON, MD
Abstract. Objective: According to the annual report dren ingesting >5 mg/kg (range 5–9.4 mg/kg), only
of the American Association of Poison Control Cen- one (5.3 mg/kg) was mildly symptomatic (drowsy)
ters, tricyclic antidepressant (TCA) ingestions ac- prior to admission. Thirty-one of the 48 (65%) were
counted for 15,708 exposures in 1998, of which 70% sent to the emergency department (dose range 0.59–
(all age groups) were treated at health care facilities 9.4 mg/kg). Fourteen of the 31 were admitted for 12–
(HCFs), with an estimated 2,022 children less than 6 24-hour observation and none subsequently devel-
years of age exposed. The study objective was to eval- oped symptoms. Twenty-three (74%) received acti-
uate the manifestations, referral patterns, HCF man- vated charcoal (AC). There was no difference in out-
agement, and medical outcomes in pediatric patients come between the children who did and did not
6 years old or less with TCA ingestions reported to a receive AC. Conclusions: No case of significant tox-
regional poison control center. Methods: All TCA (am- icity occurred in the children who experienced unin-
itriptyline, clomipramine, desipramine, doxepin, tentional TCA ingestions in this study population.
imipramine, nortriptyline) ingestions from January None of the children in the study had toxicity at doses
1, 1993, to December 31, 1997, involving patients <5 mg/kg. Further study is necessary to develop clin-
aged 6 years or less managed by the poison control ical guidelines for the appropriate referral of unin-
center were evaluated for dose, symptoms, treat- tentional ingestions of TCA involving children. Key
ments, disposition, and outcome. Results: Forty-four words: tricyclic antidepressants; pediatric poisoning;
of 48 patients (92%) were asymptomatic. All were sin- acute poison management; tricyclic antidepressant
gle-drug exposures. Forty-three patients (90%) in- pharmacokinetics; pediatric drug metabolism; pedi-
gested a TCA dose that was less than the normally atric psychopharmacology ACADEMIC EMER-
prescribed pediatric dose (5 mg/kg). Of the five chil- GENCY MEDICINE 2001; 8:139–144
RICYCLIC and cyclic antidepressants (TCAs) children with a variety of psychiatric and behav-
T
are frequently involved in pharmaceutical ioral problems, such as enuresis, attention deficit
overdose. According to the American Association of disorders, anorexia nervosa, and obsessive–com-
Poison Control Centers (AAPCC) annual report, pulsive disorder (OCD).2 The recommended thera-
15,708 cases of TCA exposure were reported for peutic dosage for children 6 years of age and over
1998.1 Of these, 11,707 (75%) were treated in a is 1.5–5 mg/kg/day or 25–75 mg daily before bed-
health care facility (HCF). There were 96 deaths, time.
most of which involved adults.1
Many studies have described the therapeutic
Tricyclic antidepressants are widely used in and toxicological effects of TCAs on adults. Little
adults for a variety of chronic pain syndromes (e.g., has been published concerning TCA toxicity in chil-
diabetic neuropathies, fibromylagia rheumatica), dren, despite an increase in therapeutic use of
migraines, phobias, and endogenous depression. TCAs in children with doses that in the past were
As a drug class they have generally been consid- considered to be possibly toxic. In order to char-
ered to have a narrow margin of safety. In spite of acterize the level of risk associated with unin-
this, TCAs are increasingly being prescribed to
tended low-dose TCA ingestion involving children,
we examined the manifestations and medical out-
comes of all TCA ingestion calls received by our
regional poison control center (PCC) over a five-
year period involving children Յ6 years of age.
From the Department of Preventive Medicine (RBM), Depart-
ment of Emergency Medicine (TRC, HCM), and Department of
Pediatrics (HCM), SUNY/Stony Brook, University Medical
Center, Stony Brook, NY; and The Long Island Regional Poison
Control Center at Winthrop, University Hospital, Mineola, NY
(RBM, TRC, RBM).
Received June 19, 2000; revision received September 21, 2000;
accepted October 12, 2000.
Address for correspondence and reprints: Robin McFee, DO,
MPH, 48-151B Piedmont Drive, Port Jefferson Station, NY
11776. E-mail: rbmcfee@pol.net
METHODS
Study Design. This was a retrospective study of
pediatric TCA ingestion calls received by the Long
Island Regional Poison Control Center from Jan-

140
TCA INGESTION
McFee et al. • TRICYCLIC ANTIDEPRESSANT INGESTIONS IN SMALL CHILDREN
TABLE 1. Patient Characteristics: Pediatric Tricyclic Antidepressant Ingestions
Ave.* Dose
Range
Age
n
% of Total
(mg/kg)
(mg/kg)
0–12 months
13–24 months
25–36 months
37–48 months
49–60 months
61–72 months
1
2
21
54
13
4
2.4
2.1
3.0
1.8
3.6
2.5
10
26
6
2
3
1.0–4.6
0.6–9.4
1.2–3.2
0.6–6.7
0.8–4.0
6
Ave. 29 months
*Ave. = average.
n = 48
100%
Ave. 2.5 mg/kg
Range 0.6–9.4 mg/kg
uary 1993 through December 1997. Because of its obtained. If the certified poison information spe-
retrospective nature using existing data sources, cialist was not convinced the caller was a reliable
this study was considered exempt from informed historian, or the circumstances of the exposure
consent.
were considered suspicious, the patient was re-
ferred to an HCF for further intervention. All pa-
tients received at least one follow-up call after the
initial phone consultation, to ensure the child re-
mained symptom-free, to verify the exposure, and
to provide anticipatory guidance.
Study Setting and Population. Cases were ob-
tained by our computerized record system that is
capable of identifying all pediatric patients with a
discharge diagnosis of TCA ingestion. The TCA ex-
posure parameters were selected to included both
single- and multiple-drug ingestions. All charts
were identified and individually reviewed. In ad-
dition to information directly obtained during the
call to the PCC, all charts included information re-
garding patient management from the time the
child arrived at an HCF until discharge. Every ef-
fort was made to contact the patient post-dis-
charge. Inclusion criteria were children aged 6
years or younger who were exposed to the follow-
ing TCAs: amitriptyline, clomipramine, desip-
ramine, doxepin, imipramine, and nortriptyline.
These drugs were chosen because of their similar
pharmacokinetics and toxicokinetics, as well as the
recent increase in prescribing of them for several
pediatric indications. Exclusion criteria were in-
gestions involving the monocyclics, tetracyclics, cy-
clobenzaprine, or serotonin-selective reuptake in-
hibitors, e.g., paroxetine, sertraline, fluoxetine.
Data Analysis. Data were analyzed by dividing
cases into symptomatic and asymptomatic. Cases
were analyzed based on category of ingestion risk,
whether the patients were hospitalized, the use of
gastrointestinal decontamination (GID), and by in-
dividual drug ingested, which involved comparing
the clinical courses of children with doses above
and below several mg/kg cutoffs.
RESULTS
There were 48 pediatric patients from 1993 to 1997
who ingested TCAs. All cases identified were sin-
gle-medication exposures. The average age was 29
months (range 8–60 months) (Table 1). Twenty-
eight children (58%) were female; 20 (42%) were
male. Fifty-three percent were between 24 and 36
months old.
Study Protocol. Patient data collected included
Forty-three of the 48 children (90%) ingested
age, gender, drugs ingested, clinical manifestations less than 5 mg/kg (dose range 0.59–5 mg/kg). The
of toxicity (vital signs, mental status, anticholin- average TCA ingestion was 3.03 mg/kg. Only one
ergic signs, neurologic status, respiratory effort, of the five children who ingested more than 5
and cardiovascular status), both at the time of the mg/kg (range 5–9.4 mg/kg) had symptoms (sleepi-
call to poison control and throughout care and fol- ness). As this ingestion occurred immediately be-
low-up of the patient, lab results including plasma fore the child’s normal bedtime, the ‘‘symptoms’’
and urine toxicological screening, and cardiac mon- may have been unrelated to the ingestion. Twenty-
itoring. In addition, we reviewed the hospital man- three children (48%) received activated charcoal
agement of each patient including the emergency (AC) (average TCA ingestion 3.17 mg/kg; range
department (ED) or hospital admission (intensive 0.59–9.4 mg/kg), and 25 (52%) did not receive AC
care unit or floor room) and length of stay. During (average TCA ingestion 2.08 mg/kg, range 0.59–
the initial phone consultation, all information (pa- 6.67 mg/kg) (Fig. 1). There was no statistically
tient age, weight, the medication ingested includ- significant difference in outcome between groups.
ing pill identification if needed, number of tablets, No pathological or abnormal electrocardiograms
dose ingested, and circumstances of exposure) was (ECGs) were reported for any child.

ACADEMIC EMERGENCY MEDICINE • February 2001, Volume 8, Number 2
141
Forty-four of the 48 children (92%) were asymp- with the use of AC. Eight of the 31 hospitalized did
tomatic throughout their entire courses. Four (8%) not receive GID. Their average TCA ingestion was
were initially minimally symptomatic. All four 2.65 mg/kg (range 0.59–6.67 mg/kg) There was no
were observed less than 24 hours at an HCF, and difference in outcomes between the children re-
none had abnormal ECGs. The first two children ceiving and not receiving AC. Thirty (97%) of the
presented drowsy, but arousable and responsive, hospitalized children received cardiac monitoring.
around their normal bedtimes. They were subse- There was no ECG abnormality. Of the 17 (35%)
quently asymptomatic upon awakening. The third not hospitalized, three (6%) who were asympto-
child had a history of a seizure disorder and pre- matic at the time of the call to poison control, and
sented with drowsiness. The fourth child was ini- had low-dose ingestions (<2.5 mg/kg), refused our
tially asymptomatic after ingesting 4.0 mg/kg of advice and were lost to follow-up. We did not re-
imipramine. She was ‘‘treated’’ with 50 mg of di- ceive subsequent calls from these individuals, or
phenhydramine in the ED, and soon thereafter be- from any HCFs concerning the children. All the
came ataxic. Her symptoms resolved after a period others (14) who were observed at home remained
of observation, and she was discharged without asymptomatic and did not require further inter-
further problems.
vention.
Amitriptyline, and products containing amitrip-
Thirty-one (65%) of the 48 went to an ED for
further evaluation. Of the 31 treated at the ED, 27 tyline, was the most commonly ingested TCA (42%)
(87%) were asymptomatic. Fourteen (45%) children in our study (Table 2). This is consistent with
were admitted for observation; 13 to the intensive AAPCC data as well. The average dose was 2.3
care unit and one to a monitored floor room. mg/kg (range 0.8–6.7 mg/kg), which is well below
Twenty-three (74%) of those treated in an ED re- the currently prescribed dose (Table 2). The aver-
ceived AC for GID (Fig. 2). The average TCA in- age length of stay in the ED for the 17 nonadmit-
gestion in the children who received AC was 3.17 ted patients was 9.23 hours. Of the 14 children ad-
mg/kg (range 0.59–9.4 mg/kg). Of these, 19 (83%) mitted to the hospital, 13 (97%) went to the
received a cathartic and four (17%) were la- pediatric intensive care unit. All children were dis-
avaged. Only one patient received multiple doses charged without complications, and there was no
of AC. There was no adverse outcome associated fatality. To the best of our knowledge, no child was
Figure 1. Tricyclic antidepressant ingestions: number of patients per dose range treated/not treated with activated
charcoal (AC).

142
TCA INGESTION
McFee et al. • TRICYCLIC ANTIDEPRESSANT INGESTIONS IN SMALL CHILDREN
TABLE 2. The Tricyclic Antidepressants Ingested
overdoses. Gastrointestinal decontamination with
AC as well as supportive care is considered a stan-
dard of care in most TCA ingestions referred to the
(mg/kg) hospital. Time from exposure is a key determinant
Average
Dose
(mg/kg)
Dose
Range
No. of
% of
Drug
Patients
Patients
when considering the use of GID, including AC.
The difficulty in managing pediatric TCA inges-
tions originates from the relative lack of large-
scale data. Until recently this class of drugs had
fallen largely out of favor. But with increased in-
terest in using TCAs for pediatric behavioral ther-
apy (OCD, attention deficit hyperactivity disorder,
anorexia, and enuresis), the opportunity exists to
Amitriptyline
Etrafon
Limbitrol
17
1
1
1
2
2
2
14
8
35.4
2.1
2.1
2.1
4.2
4.2
4.2
29.1
16.6
2.3
0.9
1.3
0.8
4.7
1.7
1.7
2.9
3.2
0.8–6.7
Triavil
Clomipramine
Desipramine
Doxepin
Imipramine
Nortriptyline
2.2–7.2
1.3–2.2
0.6–9.4
1.2–6.2
0.6–5.4
better delineate both the therapeutic and toxic
0.6–9.4 doses of these medications in the adolescent and
pediatric populations.5–7
TOTAL
48
100.0
2.5
Cyclic antidepressants are highly lipophilic and
possess large/variable volumes of distribution (15–
readmitted to an HCF or had a delayed serious
outcome.
40 L/kg). Absorption may be prolonged in massive
overdose owing to the anticholinergic action. The
TCAs included in this study have similar phar-
macokinetics and mechanisms of action. As such,
their toxicities in overdose are thought to be sim-
ilar. There is substantial ‘‘first-pass’’ absorption
(30–70%), which has considerable variation among
patients.2,5,8 The pharmacokinetics of TCAs differ
with age. They have a shorter half-life in children,5
and children may eliminate TCAs more rapidly
than adults. The therapeutic doses currently being
DISCUSSION
Tricyclic antidepressant ingestions are responsible
for a disproportionate share of hospital and inten-
sive care unit admissions when compared with
other medication ingestions.3,4 This is especially
true for intentional ingestions and suicide at-
tempts. It is well known that unintentional inges-
tions tend to involve lower dosages than intentional prescribed are often the same amounts previously
Figure 2. Emergency department visits: use/nonuse of activated charcoal.

ACADEMIC EMERGENCY MEDICINE • February 2001, Volume 8, Number 2
143
considered dangerous when unintentionally in- mg/kg daily doses.6,7,14 Wilens et al. reported that
gested. despite higher weight-corrected doses of desip-
Toxic manifestations of the TCAs include anti- ramine, children and adolescents tend to have
cholinergic, cardiovascular, and neurologic signs lower circulating desipramine and 2-hydroxydesip-
and symptoms.2,3 One fatality in a child ingesting ramine levels than adults and are in fact at
250 mg has been previously reported.9 A prescrip- lower risk for ECG changes associated with these
tion bottle containing 100 tablets of 25 mg of TCA circulating substances than adults.7 According to
has been described as offering the patient a lethal Benowitz, doses less than ten times the therapeu-
weapon.10 Cardiovascular toxicity includes a wide tic daily dose may produce severe intoxication, and
range of conduction abnormalities, myocardial de- in general ingestion of 10–20 mg/kg is potentially
pression, and vasodilation, which can lead to pro- life-threatening.21 It appears that sudden cardiac
found hypotension.11 Central nervous system tox- complications do not develop unless some symp-
icities, including coma and seizures, as well as the toms have been occurring beforehand, albeit
cardiovascular derangements contribute to the subtly. In Pentel and Sioris’s study, ‘‘no patient de-
morbidity and mortality associated with TCAs. It veloped arrhythmias after being alert and having
is well recognized that as the dose of ingestion in- a normal cardiogram for one hour.’’13
creases (10–20 mg/kg range), so does the likeli-
hood of delayed toxicity, largely due to the anticho- amounts that by today’s standards are medically
linergic effects of TCAs.8,12 appropriate doses2,5,7,8,14 (Tables 1 and 2). Had
Forty-three of 48 children in our study ingested
Studies have suggested that most cardiovascu- these children suffered from enuresis, and thus
lar complications from TCA overdose occur within been prescribed 5 mg/kg or less, presumably there
one hour of admission, and that no dysrhythmias would not have been a call to poison control unless
develop once the patient is alert and/or has had a the child was symptomatic. Ninety percent of these
normal ECG for one hour.11,13 Recent studies by children were and remained asymptomatic. Only
Wilens et al. suggest that when compared with four had mild initial symptoms, which rapidly re-
adults, children do not appear to be at increased solved. One of these four children became symp-
risk for ECG changes related to desipramine treat- tomatic iatrogenically. This child received at the
ment or concentrations of its major metabolite. hospital a 50-mg dose of diphenhydramine and
They suggest that doses of 5 mg/kg of desipramine subsequently became ataxic. It is unknown why
are safe.7 Biederman and Puig-Antich et al. sug- the patient was given diphenhydramine, since it is
gest that pediatric patients can be placed on an not recommended for the treatment of TCA expo-
incremental dose schedule starting at 1.5 mg/kg sures. Iatrogenic events are not an uncommon phe-
and increased to 3 mg/kg, 4 mg/kg, and ultimately nomenon. Exposing children to increased risks in-
5 mg/kg, provided the total is less than 300 mg per herent in hospitalizations should be considered in
day.14,15 A patient who has remained asymptomatic the management of pediatric TCA ingestions, es-
for six hours is highly unlikely to develop subse- pecially when a safe alternative exists.
quent life-threatening symptoms. A recent study
demonstrated that most regional PCCs recom- bination triage guidelines to include considera-
mend six hours of observation in an HCF.16
tions of symptoms, dose in mg/kg, historian/call re-
This study suggests the need to develop com-
The standard of care in TCA overdoses involv- liability, and the PCC ability to follow up with the
ing children typically has been to admit the child patient, to aid in determining treatment recom-
to the hospital, regardless of dose. This practice is mendations. Poison control centers are increas-
based largely on rare case reports of late cardiac ingly focusing on such considerations and reex-
complications.17,18 Walsh, in describing two case re- amining their referral patterns of the practice of
ports, one involving a 14-year-old girl who ingested referring all exposures to an HCF.18 For example,
1,050 mg of amitriptyline (>20 mg/kg) and the if we retrospectively applied 5 mg/kg as a bench-
other involving a 27-month-old boy who ingested mark, below which asymptomatic patients can
>15 mg/kg of imipramine, asserted that children safely be observed at home, then only eight chil-
show toxic effects with relatively small inges- dren (five with ingestions >5 mg/kg and three with
tions.18 This was supported in earlier observations questionable symptoms) would have been referred
suggesting children were at increased risk of car- to a hospital instead of 31. This represents a 74%
diovascular side effects at daily doses of more than reduction in ED visits and hospitalizations, and a
3.5 mg/kg of imipramine.19,20 Manoguerra reported substantial cost reduction estimated between
a case in which a 3-year-old developed seizure and $16,000 and $100,000 based on data from the
cardiac dysrhythmia following a 100-mg ingestion Health Care Financing Administration (HCFA).22
of desipramine (6.7 mg/kg).9 More recent studies Also of great importance is preventing the stress
have failed to find a strong association between placed on children and families by having to go to
TCA dose and cardiovascular effects, even at 5 the hospital. At a time when the health care sys-

144
TCA INGESTION
McFee et al. • TRICYCLIC ANTIDEPRESSANT INGESTIONS IN SMALL CHILDREN
3. Hoffman JR, Votey S, Bayer M. Effect of hypertonic sodium
bicarbonate in the treatment of moderate to severe cyclic an-
tidepressant overdose. Am J Emerg Med. 1993; 11:336–41.
tem in general and PCCs specifically are increas-
ingly under significant financial pressure, the re-
duction in unnecessary hospitalizations can make 4. Frommer DA, Kulig KW, Marx JA. Tricyclic antidepressant
overdose. JAMA. 1987; 257:521–6.
a major contribution to saving precious resources.
5. Fetner HH, Geller B. Lithium and tricyclic antidepressants.
Psychiatr Clin North Am. 1992; 15:223–4.
6. Donnelly M, Zametkin AJ, Rapoport JL. Treatment of child-
LIMITATIONS AND FUTURE QUESTIONS
hood hyperactivity with desipramine; plasma drug concentra-
tion, cardiovascular effects, plasma and urinary catecholamine
levels and clinical response. Clin Pharm Ther. 1986; 39:72–81.
7. Wilens TE, Biederman J, Baldessarini RJ. Electrocardio-
graphic effects of despramine and 2-hydroxydesipramine in
children, adolescents and adults treated with desipramine. J
Am Acad Child Adolesc Psychiatry. 1993; 32:798–804.
8. Ellenhorn E. Cyclic antidepressants. In: Ellenhorn’s Medi-
cal Toxicology. Baltimore, MD: Williams & Wilkins, 1997, pp
624–50.
We acknowledge that this study was retrospective
using information from a single PCC gathered via
the telephone. Every effort was made to verify the
information, including follow-up calls to the pa-
tient and the HCF. Laboratory analyses of urine or
serum were not conducted on every patient to doc-
ument these exposures to a TCA. Our small sam-
ple size precludes making global recommenda-
tions. While our data are suggestive that home
observation for asymptomatic children who ingest
low doses of TCA (<5 mg/kg) is a safe management
strategy, larger studies involving multiple EDs and
PCCs are necessary to confirm this. Further study
is necessary to develop clinical guidelines for the
appropriate referral of unintentional TCA inges-
tions by children.
9. Manoguerra AS. Tricyclic antidepressants. Crit Care Q.
1982; 43–51.
10. Pimentel L, Trommer L. Cyclic antidepressant overdose, a
review. Emerg Med Clin North Am. 1994; 12:533–45.
11. Fasoli RA, Glauser FL. Cardiac arrhythmias and ECG ab-
normalities in tricyclic antidepressant overdose. Clin Toxicol.
1981; 18:155–63.
12. Tokarski GF, Young MJ. Predicting trouble in tricyclic
overdose. Emerg Med. 1988; 11:143–4.
13. Pentel P, Sioris L. Incidence of late arrhythmias following
tricyclic antidepressant overdose. Clin Toxicol. 1981; 18:
543–8.
14. Biederman J. Pharmacological treatment of adolescents
with affective disorders and attention deficit disorder. Psycho-
pharmacol Bull. 1988; 24:81–7.
15. Puig-Antich J, Perel JM, Lupatkin W, et al. Imipramine in
prepubertal major depressive disorders. Arch Gen Psychiatry.
1987; 44:81–9.
CONCLUSIONS
The pediatric TCA ingestions in our study were
both unintentional and low-dose. In our study,
there was no serious outcome among the 48 chil-
dren who ingested any of the selected TCAs, even
at doses between 5 and 9.4 mg/kg. None of the chil-
dren showed toxicity at doses <5 mg/kg. Our re-
sults suggest that sending all unintentional TCA
ingestions involving children to an HCF may be
unnecessary, and home observation may be a safe
alternative to HCF referral. However, a larger,
multicenter prospective study is needed to aid in
the development of clinical guidelines.
16. McFee RB, Mofenson HC, Caraccio TR. A nationwide sur-
vey comparing the 1999 to 1998 management of asymptomatic
children who ingested a tricyclic antidepressant. J Clin Toxicol.
1999; 37:632–3.
17. Ryan R, Wians F, Stigelman W. Imipramine poisoning in
a child: lack of efficacy of resin hemoperfusion. Pediatr Emerg
Care. 1985; 1:201–4.
18. Walsh DA. Cyclic antidepressant overdose in children: a
proposed treatment protocol. Pediatr Emerg Care. 1986; 2:
28–35.
19. Hayes A, Panich ML, Barker E. Imipramine dosage in chil-
dren: a comment on imipramine and electrocardiographic ab-
normalities in hyperactive children. Am J Psychiatry. 1975;
132:545–7.
20. Winsberg B, Goldstein S, Yepes L. Imipramine and elec-
trocardiographic abnormalities in hyperactive children. Am J
Psychiatry. 1975; 132:542–5.
21. Benowitz NL. Cyclic antidepressants. In: Olson K (ed). Poi-
soning & Drug Overdose, 2nd ed. A Lange Clinical Manual.
Norwalk, CT: Appleton and Lange, 1993.
References
1. Litovitz TL. 1998 AAPCC annual report. Am J Emerg Med.
1999; 17:476–574.
2. Cada DJ (ed). Cyclic antidepressants. In: Facts and Com-
parisons. St. Louis, MO: Facts and Comparisons, Inc., 1997, pp multicenter study of antidepressant drug overdose. Pharma-
262–3.
22. Revicki DA, Palmer CS, Phillips SD. Acute medical costs
of fluoxetine versus tricyclic antidepressants. A prospective
coeconomics. 1997; 11(1):48–55.