Journal of Pediatric Surgery 54 (2019) 1969–1975

Contents lists available at ScienceDirect

Journal of Pediatric Surgery

journal homepage: www.elsevier.com/locate/jpedsurg

Featured Articles

Surgical management of pediatric thyroid disease: Complication rates

after thyroidectomy at the Children's Hospital of Philadelphia
high-volume Pediatric Thyroid Center☆
Heron D. Baumgarten a, Andrew J. Bauer b, Amber Isaza b, Sogol Mostoufi-Moab b,
Ken Kazahaya a,c, N. Scott Adzick a,⁎
a
Department of Surgery, Children’s Hospital of Philadelphia, and the Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA
b
Department of Pediatrics, and the Pediatric Thyroid Center, Children’s Hospital of Philadelphia, and the Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA
c
Division of Pediatric Otolaryngology, Children’s Hospital of Philadelphia, and the Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA

a r t i c l e i n f o a b s t r a c t

Article history: Background: Recent studies suggest improved outcomes for children undergoing thyroidectomy at high-volume
Received 26 September 2018 pediatric surgery centers. We present outcomes after thyroid surgery at a single center and advocate for referral
Received in revised form 29 December 2018 to high-volume centers for multidisciplinary management of these children.
Accepted 3 February 2019 Methods: Medical records were reviewed for all pediatric patients undergoing thyroid surgery at a single institu-
tion from 2009 through 2017. Routine recurrent laryngeal nerve and parathyroid hormone monitoring was used.
Key words:
Lymph node dissections were performed in appropriately selected cancer patients. Data collection focused on
Pediatric thyroidectomy
Multidisciplinary team
pathologic diagnosis, surgical technique, and surgical complications, including postoperative hematoma,
High-volume center neurapraxia, permanent nerve damage, hypocalcemia, and transient and permanent hypoparathyroidism.
Operative volume Results: From 2009 through 2017, 464 patients underwent thyroid surgery. Median age of the cohort was 15
Pediatric endocrine surgeon years (range 2–24). Thirty-three percent were diagnosed with benign nodules (n=151), 36% with papillary or
Graves’ disease follicular thyroid cancer (n = 168), 27% with Graves’ disease (n = 124), 3% with medullary thyroid cancer
Papillary thyroid cancer (n=14), and 1.5% underwent prophylactic thyroidectomy for MEN2a (n=7). Six patients required return to
Hypocalcemia the OR for hematoma evacuation including 5 patients after surgery for Graves’ disease (RR 8.7, 95% CI
Hypoparathyroidism
1.06–71.85). In sixteen cases, concern about neurapraxia resulted in laryngoscopy, revealing eleven patients
Neurapraxia
with vocal cord paresis. Two of these patients demonstrated a persistent deficit at 6 months postoperatively
Recurrent laryngeal nerve injury
Modified radical neck dissection (0.4%). Thirty-seven percent of patients had transient hypoparathyroidism (n=137), and two patients had per-
Radioactive iodine ablation sistent hypoparathyroidism 6 months after total thyroidectomy (0.6%). There was no significant difference in ei-
ther hypocalcemia or hypoparathyroidism after total thyroidectomy based on age or diagnosis.
Conclusions: Characterizing outcomes for pediatric patients based on diagnosis will assist in preoperative
counseling for patients and their families. This high-volume center reports low complication rates after pediatric
thyroid surgery, highlighting that referral to high-volume centers should be considered for children and adoles-
cents with thyroid disease requiring surgery.
Level of evidence: Level IV
© 2019 Elsevier Inc. All rights reserved.

The care of pediatric patients with thyroid disease is under review by specialists often have less experience in managing thyroid disease,
the pediatric surgical community. Thyroid disease is less common in both in the outpatient setting and in the operating room [1,2].
children when compared to adults, and as such pediatric surgical The most common indications for thyroid surgery include definitive
treatment of Graves’ disease (GD), thyroid nodules, and thyroid malig-
nancy [3–5]. Because less than 30% of pediatric GD patients achieve sta-
☆ All authors participated in study design; data acquisition, analysis, and interpretation; ble remission after discontinuation of antithyroid medications,
and manuscript drafting and critical review.Conflict of interest: The authors have no con- definitive treatment to achieve permanent hypothyroidism is com-
flict of interest to declare. monly pursued [6,7]. Both radioactive iodine (RAI) ablation and total
⁎ Corresponding author at: Department of Surgery, Children’s Hospital of Philadelphia,
34th and Civic Center Boulevard, Philadelphia, PA 19104. Tel.: +1 215 590 2727; fax: +1
thyroidectomy are effective, with the decision based on previously pub-
215 590 4875. lished clinical criteria, including patient age, presence or absence of thy-
E-mail address: [email protected] (N.S. Adzick). roid eye disease, thyroid size, presence or absence of thyroid nodules,

https://doi.org/10.1016/j.jpedsurg.2019.02.009
0022-3468/© 2019 Elsevier Inc. All rights reserved.
1970 H.D. Baumgarten et al. / Journal of Pediatric Surgery 54 (2019) 1969–1975

patient and family preference, and regional resources and practice par- surgeons (NSA and KK) each perform N30 pediatric thyroid surgeries
adigms [8–10]. Both are effective in achieving hypothyroidism with per year and the medical team is managed by one pediatric endocrinolo-
complication rates correlating with patient selection and preparation gist (AJB) and an additional physician dual-trained in pediatric endocri-
as well as the surgical and medical expertise of the treating institute. nology and oncology (SM-M). The Center provides primary endocrine
With regard to thyroid nodules, there is an increased likelihood that a care for pediatric GD patients, consultative care for GD patients referred
pediatric patient will undergo surgery based on the 2 to 5-fold increased for definitive management, diagnostic management for the initial presen-
risk of malignancy in pediatric thyroid nodules as compared to adults, tation of thyroid nodules and differentiated thyroid cancer (DTC), as well
with up to 25% of nodules in children ultimately diagnosed as cancer as definitive care for patients referred with suspected or pathology-
[11,12]. In addition, in pediatric patients with papillary thyroid cancer proven DTC. We also manage patients with familial history of thyroid
(PTC), 60% have regional lymph node metastases at the time of diagnosis cancer, including Multiple Endocrine Neoplasia (MEN) types 2A and 2B,
[13,14]. With the increasing incidence of thyroid cancer in both adults and familial tumor predisposition syndromes with an increased risk for thy-
pediatric patients over the last two decades there are an increased num- roid cancer, and prior childhood cancer treatment history with radiation
ber of pediatric patients being referred for thyroid surgery [15,16]. therapy.
Sosa et al and others have reported an increased risk of complications There is a single contact phone number for the CHOP Pediatric Thy-
for children undergoing thyroid surgery compared to adults [17]. This is roid Center, and there is one thyroidologist who is the “gate-keeper” for
hypothesized to be attributed to the low incidence of disease and subse- care, ensuring accurate and complete pre-surgical evaluation. The
quent referral to pediatric surgeons who perform less than 25 to 30 thy- thyroidologist determines the optimal, individualized management
roid surgeries per year (‘low-volume’) [13,15]. Patients managed at low- plan for the patient and refers for fine needle aspiration biopsy (FNA),
volume facilities are also more likely to be readmitted within 30 days imaging, and surgical management. Surgery directs inpatient care,
after thyroid surgery and to receive care that is discordant from national with the endocrine team consulting, and the endocrinologist resumes
practice guidelines [15,18]. This inverse association between surgical vol- primary care after discharge from the hospital. Each step of care is sup-
ume and complication rates has also been shown in the adult literature ported by specialized nursing, ensuring effective communication be-
[19]. Based on these reports, the inaugural American Thyroid Association tween teams and family. The whole team meets once a month to
management guidelines for children and adolescents recommended that review patients with new diagnoses and patients with active disease.
pediatric thyroid surgery be performed by a high-volume thyroid sur- Since the inception of the Thyroid Center, our case volume has in-
geon, defined as a surgeon who “performs at least 30 or more cervical en- creased significantly. This increase is attributable to several factors, in-
docrine procedures annually” [14]. cluding the hiring of a well-respected pediatric thyroidologist, active
In an effort to improve outcomes, there are an increasing number of marketing to regional centers, frequent presentation of research at na-
regional centers that have developed multidisciplinary teams “with a tional meetings, publications in peer-reviewed journals, increased inci-
full range of pediatric and oncologic resources,” focused on the care of dence of disease, and increased patient preference for surgical
pediatric patients with thyroid disease [13,15,17]. Reports published management.
by single hospitals pioneering this approach have shown favorable re- We routinely use FNA analysis of nodules with features concerning
sults, yet it is difficult to draw evidence-based conclusions about out- for malignant disease on ultrasound (US) [11,12]. Bethesda class III and
comes given the relatively small number of pediatric patients and IV nodules are typically managed with diagnostic thyroid lobectomy.
retrospective reports that often span a range between 20 and 30 years Total thyroidectomy is recommended for Bethesda class V and VI find-
[3,5,20]. Because of the ongoing debate regarding which metric is ings with central node dissection (CND) performed in the setting of a
most reliable and accurate in predicting acceptable outcomes, addi- preoperative malignant diagnosis — unilateral in the setting of unilateral
tional data from pediatric thyroid centers are necessary to serve as a ba- thyroid nodule and bilateral in the setting of confirmed malignancy bilat-
rometer for surgeons, a safety measure for hospital leaders, and a guide erally or in the isthmus. In a recent review of FNA cytopathology from the
for patients and families choosing where to seek care. Here we present Hospital of Pennsylvania (submitted for publication), the risk of malig-
the largest single-institution review of pediatric patients undergoing nancy in our cohort of pediatric patients is higher than that in adults by
thyroid surgery and offer a comprehensive analysis of our approach Bethesda score with category III portending a 20% risk of malignancy
and our outcomes for care delivered over the first nine years of the Chil- compared to 5%–15% for adults, category IV a 40% risk compared to
dren’s Hospital of Philadelphia (CHOP) Pediatric Thyroid Center. We be- 15%–30%, and category V a 100% risk compared to N75%, similar to previ-
lieve that our high-volume review will provide further evidence to ous pediatric-specific publications [21,22]. Modified radical neck dissec-
support a multidisciplinary team approach to pediatric thyroid disease tion (MRND) is performed for FNA-confirmed metastatic disease in the
management. lateral neck. Prophylactic thyroidectomy for MEN includes CND, with lat-
eral neck dissection performed based on the preoperative calcitonin and
1. Methods radiological imaging.
For total thyroidectomy patients, we routinely monitor intraoperative
This is an Institutional Review Board approved retrospective chart parathyroid hormone levels (ioPTH), drawing a value immediately before
review for all patients undergoing thyroid surgery beginning in January initiation of the surgery (1st), 10 min (2nd) and 30 min (3rd) after re-
2009, at the inception of the CHOP Pediatric Thyroid Center, through moval of the thyroid gland, respectively. This allows us to identify patients
December 2017. Subjects were identified using the comprehensive da- who may need more aggressive postoperative calcium and calcitriol sup-
tabase of patients cared for at this Center. Only patients who had under- plementation. If the 2nd or 3rd PTH is b10 pg/ml (normal 10–65 pg/ml),
gone thyroid surgery were included in this review. Patients were IV calcitriol is administered in the operative room or the recovery room.
excluded if their thyroid surgery was completed at another hospital. The absolute postoperative values, rather than a delta value, are used to
However, those referred for completion thyroidectomy or completion guide calcium and calcitriol (1,25-dihydroxycholecalciferol) supplemen-
lymph node dissection were included. tation. Intraoperative recurrent laryngeal nerve (RLN) monitoring
(IONM) is routinely utilized for total thyroidectomies. The Medtronic
1.1. Clinical management Trivantage® (Jacksonville, FL) endotracheal tube (ETT) for noninvasive
nerve monitoring is used, with the smallest size being a 5 French
The CHOP Pediatric Thyroid Center utilizes an integrated multidisci- cuffed ETT. For those children who require smaller ETT, direct neural
plinary team of pediatric specialists including general surgery, otolaryn- monitoring is performed by placement of hookwire electrodes [23]
gology, endocrinology, oncology, pathology, radiology, nuclear medicine, (Medtronic, Jacksonville, FL) (See Appendix for complete Clinical Practice
social work and behavioral health services at a single location. The two Guidelines).
 H.D. Baumgarten et al. / Journal of Pediatric Surgery 54 (2019) 1969–1975 1971

1.2. Data collection and analysis Table 1

Demographics.

Data collection focused on pathologic diagnosis, surgical technique, Patients n (%)

and surgical complications, including postoperative hematoma, neur- Age (mean, SD) 14.3±3.7
apraxia, permanent RLN injury, hypocalcemia, and transient and perma- Male 93 (20)
nent hypoparathyroidism. Vocal cord palsy, as confirmed by flexible Female 371 (80)
laryngoscopy, was considered permanent if it lasted for more than 6 Radiation exposure 33 (7.1)
Predisposing condition 40 (8.6)
months. Hypocalcemia was defined as an ionized calcium level b 1.1
MEN 21 (52.5)
mmol/L (normal 1.15–1.34 mmol/L) or a serum calcium b8.7 mg/dL (nor- Total thyroidectomy 332 (71.5)
mal 8.9–10.4 mg/dL). Calcium levels are not routinely collected for lobec- Graves’ disease 123 (37)
tomy patients, thus rates are reported for total thyroidectomy patients Malignancy 152 (45.8)
only. Postoperative transient hypoparathyroidism was defined as any PTC 136 (89.5)
Postoperative RAI 102 (74)
postoperative PTH level less than 10 pg/mL with return to normal
FTC 2 (1.3)
serum calcium and phosphorous off of calcium and calcitriol supplemen- MTC 14 (9.2)
tation within 6 months of surgery. Parathyroid tissue in the pathologic Central node dissection 117 (77)
specimen and parathyroid autotransplantation were included for analy- Modified radical neck dissection 43 (28.3)
Additional lymph node surgery 15 (9.9)
sis. Permanent hypoparathyroidism was defined as persistently low PTH
Lobectomy 121 (26)
and calcium values requiring calcium supplementation for more than 6 Malignant 26 (21.5)
months after surgery. Operative times, hospital lengths of stay, and the PTC 23 (88.5)
need for radioactive iodine ablation or further surgery were also collected FTC 4 (15.4)
for review. The majority of our patients continue with long-term follow- Completion 19 (15.7)
Postoperative RAI 12 (63.2)
up at the CHOP Thyroid Center and transition to adult endocrinology in
Central node dissection 14 (73.7)
their early 20s. There were no patients with missing data related to Transferred to CHOP 11 (2.4)
loss-to-follow-up within the data collection analysis for this cohort. Neck Dissection 4 (36.4)
Data analysis was performed using GraphPad Prism 7. D’Agostino Completion 7 (63.6)
Total 464
Pearson omnibus and Shapiro–Wilk tests were used to assess data nor-
mality. Standard description statistics, including mean, standard devia- Age reported in years. SD=standard deviation, MEN=multiple endocrine neo-
tion, median, range, proportions (categorical data), and 95% confidence plasia, PTC=papillary thyroid cancer, RAI=radioactive iodine, FTC=follicular
thyroid cancer, MTC=medullary thyroid cancer.
intervals were used to describe study group characteristics and out-
comes. The Student t test or Wilcoxon–Mann–Whitney tests were
used to check differences between groups. The Chi-squared and Fisher did not define the number of lymph nodes described as ‘minimal’ micro-
exact tests were employed to assess differences in variables distribu- scopic central neck disease, however, Jeon et al recently proposed ≤ 5
tion. One-way ANOVA with multiple comparisons and Kruskal–Wallis lymph nodes as a cutoff for patients at low-risk for persistent post-
tests were used to evaluate differences between multiple groups. Linear surgical disease [24]. A MRND was performed for 23% (n =43) of our
regression was used to analyze trends over time for various outcomes. PTC patients based on preoperative imaging and FNA confirmation of
Intention-to-treat was used to categorize patients in the setting of con- lateral neck lymph node metastasis. Fourteen of these patients had bi-
current diagnoses of GD and cancer. Pathologic diagnosis was used to lateral modified radical neck dissections.
categorize patients in the setting of preoperatively undiagnosed cancer. For the remainder of patients that underwent thyroid surgery, 31%
had benign disease (n = 145), 35% had papillary or follicular thyroid
2. Results cancer (PTC or FTC; n=164), 27% had GD (n=123), 3% had medullary
thyroid cancer (MTC; n = 14), and 1.5% underwent prophylactic thy-
2.1. Patient demographics roidectomy for MEN2a (n=7) (Table 1).

From January 2009 through December 2017, 464 patients underwent 2.2. Preoperative course
thyroid surgery (Table 1). The median age of patients was 15 years (range
2–24 years), and 80% of thyroid surgery patients were female. Seventy- Nearly all patients underwent US imaging to evaluate their thyroid
one percent underwent total thyroidectomy (n=332). Twenty-six per- and neck (for patients with a thyroid nodule) prior to surgery. There
cent underwent lobectomy (n=121), with 16.5% of those requiring com- were 13 patients with GD who did not complete preoperative imaging.
pletion thyroidectomy (n=20). Seven completion thyroidectomies were Sixty-two patients (33% of patients with malignant diagnosis) demon-
performed after initial lobectomy at another institution, and four modi- strated abnormal lymphadenopathy on preoperative imaging. FNA bi-
fied radical neck dissections were performed for persistent disease after opsy was performed in 285 patients with results detailed in Table 2.
thyroidectomy at another institution. Percutaneous methylene blue injec- FNA was discordant (Bethesda class II–IV) in 28 patients found to
tion under ultrasound guidance was used to localize metastatic lymph have PTC, 8 with papillary thyroid microcarcinoma (PTmC), and 4 pa-
node disease just prior to resection in 4% of patients with malignancy tients found to have FTC.
(n=7). We utilize intraoperative methylene blue labeling for patients un- Those who did not undergo FNA either had benign-appearing dis-
dergoing repeat neck dissection for persistent lymph node disease after ease based on preoperative US or had a previous diagnosis of PTC and
initial surgery. The number of cases requiring lymph node mapping is were found to have evidence of persistent disease based on postopera-
low owing to the optimization of presurgical lymph node mapping for pa- tive imaging. Of the 18 patients who were diagnosed with cancer but
tients undergoing initial surgery at CHOP as well as the use of US-guided, did not undergo FNA, 2 patients were found to have PTC and 6 were
percutaneous ethanol ablation rather than repeat surgery in patients with found to have PTmC after surgery for GD, 6 were newly diagnosed
only 1 to 2 FNA proven, persistent lymph node metastases. with PTC, and 4 were found to have PTmC (Table 2).
Of the 188 patients diagnosed with a thyroid malignancy, a CND was Preoperative laryngoscopy was performed for those patients requir-
performed in 75% (n = 140). Bilateral CND was performed for 64 pa- ing completion surgeries or for those with any concern for hoarseness or
tients. The trend for CND dissection has increased over time in an effort dysphagia preoperatively. The pediatric otolaryngologist performed
to more accurately define the invasive behavior of the PTC in keeping more routine preoperative laryngoscopic evaluations than did the gen-
with the 2015 ATA pediatric guidelines [14]. The 2015 ATA guidelines eral surgeon.
1972 H.D. Baumgarten et al. / Journal of Pediatric Surgery 54 (2019) 1969–1975

Table 2 IoPTH measurements were drawn for 55% of patients (n=257), and
Preoperative diagnostics. were routinely drawn for all patients undergoing TT since November
Preoperative diagnostics n (%) 2012. Fifteen percent of cases involved parathyroid autotransplantation,
Imaging
and 42% of pathology specimens included parathyroid tissue (n=193).
US 451 (97.2) Rates of transient hypoparathyroidism were not different between age
Scintigraphy 36 (8) groups (p=0.57).
MRI 24 (5) Surgeon experience at our center averaged 31.8 cases per surgeon
CT 36 (8)
per year for the past 6 years (64 surgeries/year). After development of
No imaging 13 (2.8)
FNA 286 (61.4) benign malignant the multidisciplinary Pediatric Thyroid Center at CHOP, surgeon case
I 4 (1.4) 4 0 number increased from 3–4 cases per surgeon in 2009, to 23–24 cases
II 73 (25.6) 66 7 per surgeon in 2011, to more than 40 cases per surgeon in 2015
III 41 (14.4) 31 10 (Figure 1). Complication rates have not changed over time (Table 5).
IV 46 (16.1) 23 23
There has been a downward trend in the number of cases with parathy-
V 13 (4.6) 1 12
VI 109 (37.9) 1 108 roid tissue in the specimen over time (slope= −2.2, R2 = 0.34).
Discordant FNA 40 (14)
No FNA 178 (38.4) 2.4. Postoperative course
GD 115 (64.6)
MEN 18 (10.1)
Benign imaging findings 45 (25.3) Average hospital stay for all patients was 1.7 days (±1.37 days, me-
Pathology diagnosed PTC 18 (10.1) dian 1 day, range 0.5–12). Of all 464 patients, 6 required return to the
US =ultrasound, MRI= magnetic resonance imaging, CT=computed topography, OR for hematoma evacuation and control. The majority of postoperative
FNA=fine needle aspiration, GD=Graves’ disease, MEN=multiple endocrine neoplasia, bleeding (n=5) occurred after surgery for GD (RR 8.7, 95% CI
PTC =papillary thyroid cancer. FNA classes I–VI as defined by the Bethesda criteria: 1.06–71.85, p=0.02). In 16 cases, intraoperative concern for nerve injury
I=inadequate sample to perform pathologic analysis, II=benign tissue, III=follicular (lost signal on the RLN monitor) and/or postoperative concern for
cells of unknown significance, IV=follicular neoplasia, V= concern for malignancy,
VI=malignant diagnosis.
neurapraxia resulted in laryngoscopy, showing 10 patients with unilat-
eral vocal cord paresis and 1 patient with bilateral paresis. Only two of
these patients had persistent unilateral RLN deficit 6 months postopera-
2.3. Intraoperative characteristics tively (0.4% overall).
Seventy-three percent of patients (TT, completion thyroidectomy, and
Nerve monitoring was used for 79% of thyroid surgeries (n=365). neck dissection patients) had at least transient biochemical hypocalcemia
One of the two surgeons has used nerve monitoring since our multidis- (one recorded ionized calcium level b 1.1 mg/dL, or one recorded serum
ciplinary approach began in 2009, and the other surgeon began using calcium level b8.7 mg/dL; n=253), and 30% of patients had at least tran-
nerve monitoring in December, 2012. There was no difference in the in- sient hypoparathyroidism (one recorded PTH b 10 pg/mL; n=137). After
cidence of neurapraxia between the two surgeons (6/198 vs. 5/266; implementation of ioPTH measurements and Clinical Pathway Guidelines,
p = 0.47) as well as no difference in neurapraxia with and without postoperative IV calcium was not required for the last 4 years of the case
nerve monitoring (1/69 vs. 4/197, pN0.99). study (2014–2017; Table 5). Thirty-three percent of TT patients were
Mean operative time across all surgeries was 150 min (± 98 min; discharged with calcium supplementation only, and thirty percent of TT
median 124 min, total range 40–666), and operative time was signifi- patients were discharged with prescriptions for both calcitriol and cal-
cantly different based on the extent of surgery (pb 0.01). The median cium supplementation secondary to clinically significant hypocalcemia.
operative time for total thyroidectomy (TT) without CND for malignant Among patients undergoing surgery for malignancy, there was no differ-
disease compared to GD was 128 vs. 130 min (pN 0.99). TT for benign ence in the need for oral calcium and calcitriol supplementation upon dis-
disease averaged 110 min. Completion thyroidectomy took longer charge based on extent of surgery — that is, patients undergoing total
than initial lobectomy alone (median 103 vs. 76 min, pb 0.01), attrib- thyroidectomy (68/160), total thyroidectomy with central node dissec-
uted to the addition of CND with most completion surgeries. Modified tion (34/77), and total thyroidectomy with modified radical neck dissec-
radical neck dissection added a significant amount of time to the surgery tion (11/43) all had similar rates of transient hypocalcemia requiring oral
(median 412 vs. 140 min for TT in the setting of malignancy, pb 0.01). TT medications upon discharge from the hospital (p=0.09). Furthermore,
for malignant disease took longer with the addition of CND compared to there was no difference in the need for calcium supplementation after
no node dissection (median 160 v 128 min, p=0.009) (Table 3). thyroidectomy for GD (47/123) vs. malignancy (63/174, p =0.21)
(Table 4). Finally, after separating the cohort by age (0–5, 6–10, 11–15,
16–20, and N20 years old), no difference in rates of hypocalcemia by
Table 3
Operative times.

Procedure Operative time (median, IQR) p value

TT for malignancy 166, 128–282 b0.01

with CND 160, 128.3–208.5 0.009⁎
without CND 128, 94–160 N0.99 ⁎⁎⁎⁎
with bilateral MRND 431.5, 353.3–520.8 N0.99⁎⁎
with unilateral MRND 398.5, 286.3–447.8 b0.001^
Completion lobectomy 103, 94–165 N0.99⁎⁎⁎
TT for benign disease 110, 88–138.5 b0.01
TT for Graves 130, 110.5–172 b0.01
Lobectomy 76, 62–90.75 b0.01

TT=total thyroidectomy, CND=central node dissection, MRND=modified radical neck

dissection.
⁎ Malignant TT with vs. without CND.
⁎⁎ Bilateral vs. unilateral MRND.
⁎⁎⁎ Completion vs. TT without CND.
⁎⁎⁎⁎ TT without CND vs. Graves.
^
TT with CND vs. unilateral MRND. Figure 1. Case volume for each surgeon over time.
 H.D. Baumgarten et al. / Journal of Pediatric Surgery 54 (2019) 1969–1975 1973

Table 4
Operative complications by diagnosis.

Diagnosis n age (SD) LOS in days Hematoma, Neurapraxia, n(%) / Hypocalcemia, n (%) Transient Hypoparathyroid, n(%) /
(m/mdn) n (%) Permanent nerve injury, Permanent Hypoparathyroid,
n(%) n(%)

Graves’ Disease 123 13.6 (3.9) 1.84/2 5 (4) 3 (2.4) / 0 47 (38.2) 27 (22) / 1 (0.8)
Papillary and Follicular Thyroid Cancer 174 15.1 (3.1) 2.2/2 1 (0.6) 4 (2.3) / 2 (1.1) 63 (36.2) 43 (24.7) / 1 (0.6)
Medullary Thyroid Cancer 14 6.6 (3.3) 1.64/1 0 0 6 (42.8) 4 (28.5) / 0
Total Thyroidectomy for Benign Disease 57 14.7 (3.7) 1.36/1 0 1 (1.8) / 0 14 (14.7) 13 (22.8) / 0
Lobectomy for Benign Disease 96 14.1 (3.6) 0.97/1 0 1 (1) / 0 0 0

LOS=length of stay reported in days, M=mean, mdn=median. Age reported in years. Hypocalcemia is defined as any patient with an ionized Calcium value b1.1 or serum Calcium level b8.7
mg/dL and discharged from the hospital with prescriptions for calcium and calcitriol supplementation. Hypoparathyroid is defined as any PTH value b10 (10–65 pg/mL). Permanent nerve injury
and permanent hypoparathyroidism are defined as unresolved at more than 6 months postoperatively. Total thyroidectomy for benign disease includes 7 prophylactic surgeries for MEN2a.

laboratory value, discharge medication, or rescue IV calcium therapy was there is a 13% decreased likelihood of developing transient hypocalce-
found (Table 6). Only 2 total thyroidectomy patients had permanent hy- mia [26]. This is not consistent with our results, and the absence of
poparathyroidism at 6 months postoperatively (0.6%). these influences on postoperative hypocalcemia has not yet been re-
Seventy percent of patients with PTC required postoperative radio- ported in any previous retrospective review of outcomes.
active iodine ablation for metastatic disease. The diagnosis of persistent
and recurrent disease is based on an elevated thyroglobulin and/or 3.2. Complications: hypoparathyroidism
antithyroglobulin and/or evidence of anatomic disease on radiological
imaging [14]. Fifteen patients (8%) whose cancer care was executed Although significant hypocalcemia can be dangerous for children, the
only at CHOP required further surgery for persistent disease in the cer- real concern in this setting is the risk of permanent hypoparathyroidism
vical lymph nodes, and 19 (15.7%) required completion thyroidectomy requiring close monitoring and calcium supplementation for life. This
for diagnosis of PTC in the setting of FNA showing Bethesda class II–IV. complication has predominantly influenced pediatric endocrinologists
One patient with GD underwent postoperative radioactive iodine abla- to recommend radioactive iodine ablation in lieu of surgery for definitive
tion for active remnant thyroid tissue (0.8%; Table 7). treatment of GD and has resulted in partial thyroidectomy or avoidance of
central neck dissection in the setting of PTC. Recently, high-volume, mul-
3. Discussion tidisciplinary surgical teams have begun to report better outcomes, and
accordingly surgery is increasingly chosen over radioactive iodine therapy
Concern over the safety of pediatric thyroidectomy has sparked dis- for GD [30]. The risk of permanent hypoparathyroidism in the hands of
cussion among pediatric surgeons, otolaryngologists, and endocrinolo- low-volume surgeons is evident in the literature, with one study
gists interested in reducing the risk of complications for children reporting a 20% permanent incidence [25]. In contrast, this complication
requiring thyroid surgery. Previous studies have shown a lower rate of is only 0%–2.6% in high-volume centers [2,22,27].
surgical complications at higher volume centers [3,20] compared to Our rates of parathyroid transplantation are comparable to other
lower volume centers [25]. Our outcomes are consistent with other re- publications [4], and our rates of parathyroid tissue removal have de-
ports from high volume centers. The very large clinical volume in our se- clined over the years as operative volumes have increased. Even with a
ries provides an opportunity to investigate the nuances of surgical high rate of patients selecting thyroidectomy over radioactive iodine ab-
pediatric thyroid disease management. lation for GD, and an increased number of thyroid cancer patients with
significant lymph node metastasis requiring lymph node resection, we
3.1. Complications: hypocalcemia report an overall rate of permanent hypoparathyroidism of 0.4% in the
entire series of 464 patients and a rate of 0.6% after total thyroidectomy.
Rates of transient hypocalcemia after thyroidectomy are reported in
several reviews, but the definition for hypocalcemia is not uniform. 3.3. Complications: recurrent laryngeal nerve injury
There is a large range of postoperative hypocalcemia results (between
7% and 69%) from different institutions and national database reviews Recurrent laryngeal nerve injury is also a serious complication of thy-
[4,20,25–30]. We selected a conservative definition for hypocalcemia in roidectomy [15]. The high volume pediatric and adult literature reports
comparison to much of the published literature, requiring only one labo- rates of permanent nerve injury from 0% [20] to 1.46% [31], and transient
ratory value below the normal range to qualify, which may not be clini- neurapraxia from 1.6% [20] to 4.8% [29,31]. Our rates fall within these
cally relevant. For this reason, we stratified our patients discharged on ranges with 0.4% permanent injury and 1.9% transient neurapraxia. We
supplemental calcium and calcitriol to better represent clinically signifi- also did not see a difference in nerve injury based on extent of lymph
cant hypocalcemia. In our cohort, the need for IV calcium “rescue” therapy node dissection, which has been previously reported in adults [32]. We
after surgery secondary to symptomatic hypocalcemia stopped after initi- routinely use intraoperative nerve monitoring (IONM) because there is
ation of intraoperative PTH monitoring and Clinical Practice Guidelines some evidence showing a trend for a lower rate of RLN palsy when
(see Appendix), allowing for identification of patients at increased risk IONM is used in second operations, in total thyroidectomy with neck dis-
of postoperative hypocalcemia and early initiation of calcitriol in the oper- section, and in cases of malignant disease [33–35]. Although the gold
ating room or the Recovery Room. We thus report a lower rate of IV cal- standard will always remain direct visualization of the nerve by the sur-
cium therapy than other high-volume centers and we recommend geon, thyroid or endocrine fellowship-trained surgeons exposed during
broader implementation of this approach [20]. We believe this approach their fellowship training to attending surgeons who either use or do not
has advantages over protocols using preoperative calcitriol 48–72 h use IONM, subsequently use IONM in nearly 100% of their own cases
prior to thyroidectomy for all patients that exposes patients to a medica- after their training is complete [33].
tion they may not need and places them at risk for hypercalcemia.
Despite the variation in definition for hypocalcemia in the literature, 3.4. Complications: bleeding
our results are comparable to outcomes reported. Interestingly, we did
not see increased rates of hypocalcemia based on patient age or extent Our overall rate of postoperative hematoma is comparable to other
of lymph node dissection, as is reported in several other studies studies. From a study comparing adult and pediatric outcomes after thy-
[18,20,26]. One study concluded that for every year increase in age, roidectomy for GD from a single institution, postoperative bleeding
1974 H.D. Baumgarten et al. / Journal of Pediatric Surgery 54 (2019) 1969–1975

Table 5
Complications each year.

Complications per year, n (%) 2009 2010 2011 2012 2013 2014 2015 2016 2017
(n=7) (n=27) (n=47) (n=48) (n=65) (n=75) (n=69) (n=67) (n=59)

Hematoma 0 0 0 (0) 0 1 (2) 1 (2) 2 (3) 2 (3) 0 (0)

Neurapraxia 1 (13) 0 0 2 (4) 1 (2) 2 (2.7) 2 (3) 0 1 (1.6)
Hypocalcemia 2 (25) 15 (56) 31 (66) 30 (63) 28 (43) 44 (59) 42 (61) 38 (57) 23 (39)
Discharged on calcium/calcitriol 1 (13) 1 (4) 15 (32) 12 (25) 17 (26) 32 (43) 31 (45) 13 (19) 15 (25)
Required IV calcium 2 (25) 5 (19) 12 (26) 10 (21) 3 (5) 0 0 0 0
Hypoparathyroid 1 (13) 2 (7) 8 (17) 14 (29) 10 (15) 13 (17) 20 (29) 14 (21) 5 (8)
Parathyroid tissue in the specimen 4 (50) 9(33) 27 (57) 21 (44) 26 (40) 34 (45) 25 (36) 14 (21) 21 (36)
Permanent nerve damage 0 0 0 0 0 1 (1.3) 0 1 (1.5) 0
Permanent hypoparathyroid 0 0 0 0 0 1 (1.3) 1 (1.5) 0 0
Surgery for recurrent disease 0 1 (4) 5 (11) 1 (2) 1 (2) 4 (5) 1 (1) 1 (1) 0

occurred in 4.8% of patients [29]. Our cohort reports an equivalent rate pediatric thyroidectomies performed by adult surgeons nationally, lead-
of postoperative bleeding (4%). Interestingly, we report a much higher ing to poorer outcomes for smaller children owing to the unfamiliarity
risk for postoperative bleeding after surgery for GD (4%) compared to of these surgeons with the small operative field and delicate tissues, as
other thyroid surgery diagnoses (0.3%), a distinction not previously well as the potential difference in intra- and perioperative care if the sur-
published. None of these patients were diagnosed with thyroid storm, gery is not performed in a tertiary pediatric center.
although one patient requiring return to the OR did have high blood Our CPG was created by the director of the thyroid center with input
pressures recorded on the first postoperative day with bleeding attrib- from the inpatient endocrine consult team and pharmacy. The CPG has
uted to this abnormality in the medical record. GD thyroid glands are been optimized over time via feedback from the surgical and inpatient
hypervascular, despite our patients being treated preoperatively with endocrine teams. Adherence to the guideline is very high owing to
potassium iodine drops (SSKI) and rendered medically euthyroid prior close communication between team members, most importantly by
to surgery [36]. Recognition of this increased risk should result in atten- our fellows and nurses. We have created an order set for postoperative
tion to ensuring a dry surgical field prior to wound closure, close atten- lab surveillance and the guidelines are published on the hospital
tion to the postoperative neck examination, and early recognition and website for easy reference.
treatment of postoperative thyroid storm in GD patients.
3.6. Limitations of this study
3.5. High-volume, multidisciplinary care
Our study is unable to show any scientific difference in outcomes
Theodor Kocher, the pioneer of safe thyroid surgery, performed 101 based on surgical volume, as we do not analyze outcomes from years
thyroidectomies during his first decade in Bern, Switzerland with a prior to formation of our multidisciplinary team. We make the assump-
mortality rate near 13%. This mortality decreased to 0.5% by 1917 by tion, based on multiple published studies focused on this variable, that
the time Kocher had performed more than 5000 thyroidectomies [1]. the routine nature of these surgeries for patients within our group is ben-
Tuggle et al found that the average surgeon operating on pediatric pa- eficial. The statistical analysis performed for this study did not include
tients for thyroidectomy performed just 1–2 cases per year, noting the multivariate regression analyses for any outcome. Univariate analyses
difficulty for individual surgeons to gain adequate experience. They rec- can blunt the ability to find differences between groups with various re-
ommended referral to high-volume thyroid surgeons and high-volume lated factors contributing to the variable of interest. Although, we do an-
centers regardless of surgeon specialty [1]. alyze each outcome from various angles, looking at the differences
Several studies have reported a higher complication rate and longer between extent of surgery, age, and diagnosis, perhaps a multivariate re-
length of hospital stay (LOS) for younger patients. In a review of national gression would bring forth statistical differences unrecognized by this
outcomes from the Kids’ Inpatient Database (KID — years 2009 and 2012), form of analysis. Specifically, with reference to our hypocalcemia out-
patients 1–5 years old stayed in the hospital an average of 7.8 days [4]. By comes, the conservative definition used and the absence of data analyzing
comparison, our average LOS for this age group was 1.58 days (SD 0.91), clinically symptomatic hypocalcemia, limit our understanding of clinically
which was lower still than the average reported for patients ages 16–20 relevant complications.
years in the national review, see Table 6 [4]. The shorter LOS reported
by our group could be attributable to our high case volumes and to the 4. Conclusion
joint management of these patients in hospital by the surgery and endo-
crine services while adhering to strict Clinical Practice Guidelines (CPG). The formation of multidisciplinary teams has improved outcomes
Perhaps these outcomes are also partially attributable to the number of for patients with thyroid disease [28,37–40]. Since inception of our

Table 6
Complication by age.

Complications by age, n (%) 0–5 years (n=20) 6–10 years (n=48) 11–15 years (n=201) 16–19 years (n=176) 20+ years (n=19) p value (Chi-Square)

Hematoma 0 1 (2) 2 (1) 3 (2) 0 0.89

Neurapraxia 1 (5) 1 (2) 4 (2) 3 (2) 0 0.84
Permanent nerve damage 0 2 (4) 0 0 0 0.002
Transient hypocalcemia 11 (55) 29 (60) 112 (56) 90 (51) 11 (58) 0.79
Discharged with calcium and calcitriol 8 (40) 15 (31) 67 (33) 41 (23) 6 (32) 0.21
IV calcium administration 0 1 (2) 19 (9) 11 (6) 1 (5) 0.24
Transient hypoparathyroidism 5 (25) 8 (17) 36 (18) 31 (18) 6 (32) 0.57
Parathyroid tissue in the specimen 8 (40) 15 (31) 84 (42) 74 (42) 12 (63) 0.007
Permanent hypoparathyroidism 0 1 (2) 1 (0.5) 0 0 0.41
Hospital stay (average (SD)) 1.6 (0.9) 2 (1.7) 1.7 (1.2) 1.7 (1.3) 1.6 (1.3) 0.43 (ANOVA)
 H.D. Baumgarten et al. / Journal of Pediatric Surgery 54 (2019) 1969–1975 1975

Table 7 [8] Kovatch KJ, Bauer AJ, Isaacoff EJ, et al. Pediatric thyroid cancer in patients with
Postoperative radiation and recurrence. Graves’ Disease: the role of ultrasound in selecting patients for definitive therapy.
Horm Res Paediatr 2015;83:408–15.
Patients with Malignancy, n 188 [9] Rivkees SA. Controversies in the management of Graves' disease in children. J
Endocrinol Investig 2016;39(11):1247–57.
Postoperative RAI, n (%) 122 (70.1)
[10] Bauer AJ. Approach to the pediatric patient with Graves' disease: when is definitive
Surgery for cervical metastasis 15 (8)
therapy warranted? J Clin Endocrinol Metab 2011;96(3):580–8.
Patients undergoing lobectomy for suspected benign disease 121 [11] Mussa A, De Andrea M, Motta M, et al. Predictors of malignancy in children with thy-
Completion thyroidectomy for diagnosis of DTC 19 (15.7) roid nodules. J Pediatr 2015;167(4):886–892.e1.
Patients with Graves’ Disease 123 [12] Martinez-Rios C, Daneman A, Bajno L, et al. Utility of adult-based ultrasound malignancy
Postoperative RAI 1 (0.8) risk stratifications in pediatric thyroid nodules. Pediatr Radiol 2018;48(1):74–84.
Preoperative RAI 3 (2.4) [13] Al-Qurayshi Z, Hauch A, Srivastav S, et al. A national perspective of the risk, presen-
tation, and outcomes of pediatric thyroid cancer. JAMA Otolaryngol Head Neck Surg
RAI=radioactive iodine ablation, DTC=differentiated thyroid cancer. Postoperative RAI
2016;142(5):472–8.
was recommended for patients with PTC and FTC only (n=174). [14] Francis GL, Waguespack SG, Bauer AJ, et al. Management guidelines for children with
thyroid nodules and differentiated thyroid cancer. Thyroid 2015;25(7):716–59.
[15] Youngwirth LM, Adam MA, Thomas SM, et al. Pediatric Thyroid cancer patients re-
Pediatric Thyroid Center, our case volume has increased markedly. The ferred to high-volume facilities have improved short-term outcomes. Surgery
care of our patients has evolved as a result of the regular interaction be- 2018;163(2):361–6.
[16] Siegel DA, King J, Tai E, et al. Cancer incidence rates and trends among children and
tween our multidisciplinary team comprised of one to two members
adolescents in the United States, 2001–2009. Pediatrics 2014;134(4):e945–55.
from endocrinology, oncology, radiology, pathology, and surgery (pedi- [17] Sosa JA, Tuggle CT, Wang TS, et al. Clinical and economic outcomes of thyroid and
atric general surgery and pediatric otolaryngology). The additional sup- parathyroid surgery in children. J Clin Endocrinol Metab 2008;93(8):3058–65.
port from our nursing, research and administrative support serves as an [18] Breuer CK, Solomon D, Donovan P, et al. Effect of patient age on surgical outcomes
for Graves' disease: a case–control study of 100 consecutive patients at a high vol-
ideal model for pediatric thyroid care. ume thyroid surgical center. Int J Pediatr Endocrinol 2013;2013(1):1–9856 [2013-1].
Our model does not include the participation of adult thyroid sur- [19] Adam MA, Thomas S, Youngwirth L, et al. Is there a minimum number of thyroidec-
geons in the operating room [38]. Several institutions have partnered tomies a surgeon should perform to optimize patient outcomes? Ann Surg 2017;
265(2):402–7.
a high volume adult thyroid surgeon and a pediatric surgeon, thus [20] Chen Y, Masiakos PT, Gaz RD, et al. Pediatric thyroidectomy in a high volume thyroid
adopting a two attending surgeon approach. We believe that institu- surgery center: risk factors for postoperative hypocalcemia. J Pediatr Surg 2015;50
tions responsible for training pediatric surgical specialists may sacrifice (8):1316–9.
[21] Monaco SE, Pantanowitz L, Khalbuss WE, et al. Cytomorphological and molecular ge-
their ability to produce adequately trained attending thyroid surgeons if netic findings in pediatric thyroid fine-needle aspiration. Cancer Cytopathol 2012;
they employ this two attending approach. However, each center has to 120(5):342–50.
decide the best avenue to optimize care based on their institutional re- [22] Lale SA, Morgenstern NN, Chiara S, et al. Fine needle aspiration of thyroid nod-
ules in the pediatric population: a 12-year cyto-histological correlation experi-
sources. We recommend that every pediatric surgical trainee be in- ence at North Shore–Long Island Jewish Health System. Diagn Cytopathol
volved in at least 8–10 thyroid surgeries per year of training and that 2015;43(8):598–604.
new pediatric thyroid surgeons collaborate with an experienced thyroid [23] Cheng J, Kazahaya K. Endolaryngeal hookwire electrodes for intraoperative recur-
rent laryngeal nerve monitoring during pediatric thyroid surgery. Otolaryngol
surgeon during the initial years of practice.
Head Neck Surg 2013;148(4):572–5.
We conclude that pediatric thyroid surgery performed at a high- [24] Jeon MJ, Kim YN, Sung TY, et al. Practical initial risk stratification based on lymph
volume center is associated with shorter LOS and very low-risk of surgi- node metastases in pediatric and adolescent differentiated thyroid cancer. Thyroid
cal complications such as permanent hypoparathyroidism and recur- 2018;28(2):193–200.
[25] Cohen RZ, Felner EI, Heiss KF, et al. Outcomes analysis of radioactive iodine and total
rent laryngeal nerve injury. We recommend referral of pediatric thyroidectomy for pediatric Graves' disease. J Pediatr Endocrinol Metab 2016;29(3):
patients with thyroid disease to pediatric thyroid centers with access 319–25.
to high-volume pediatric thyroidologist that will optimize selection [26] Yu YR, Fallon SC, Carpenter JL, et al. Perioperative determinants of transient hypocal-
cemia after pediatric total thyroidectomy. J Pediatr Surg 2017;52(5):684–8.
and preparation of patients for surgery, high-volume pediatric surgeons [27] Freire AV, Ropelato MG, Ballerini MG, et al. Predicting hypocalcemia after thyroidec-
experienced with operating in small operative fields and a high-volume tomy in children. Surgery 2014;156(1):130–6.
multidisciplinary thyroid team with experience in intraoperative and [28] Burke JF, Sippel RS, Chen H. Evolution of pediatric thyroid surgery at a tertiary med-
ical center. J Surg Res 2012;177(2):268–74.
perioperative care, in an effort to minimize the risk of complications [29] Chiapponi C, Stocker U, Mussack T, et al. The surgical treatment of Graves' disease in
and optimize short and long-term outcome. children and adolescents. World J Surg 2011;35(11):2428–31.
[30] Wu VT, Lorenzen AW, Beck AC, et al. Comparative analysis of radioactive iodine ver-
sus thyroidectomy for definitive treatment of Graves disease. Surgery 2017;161(1):
147–55.
Appendix A. Supplementary data [31] Sundaresh V, Brito JP, Thapa P, et al. Comparative effectiveness of treatment choices
for Graves' hyperthyroidism: a historical cohort study. Thyroid 2017;27(4):
Supplementary data to this article can be found online at https://doi. 497–505.
[32] Chen Q, Wei T, Wang XL, et al. The total number of prelaryngeal and pretracheal
org/10.1016/j.jpedsurg.2019.02.009 and https://www.chop.edu/ lymph node metastases: is it a reliable predictor of contralateral central lymph
clinical-pathway/hypocalcemia-surveillance-total-thyroidectomy- node metastasis in papillary thyroid carcinoma? J Surg Res 2017;214:162–7.
clinical-pathway. [33] Al-Qurayshi Z, Randolph GW, Alshehri M, et al. Analysis of variations in the use of
intraoperative nerve monitoring in thyroid surgery. JAMA Otolaryngol Head Neck
Surg 2016;142(6):584–9.
[34] Rulli F, Ambrogi V, Dionigi G, et al. Meta-analysis of recurrent laryngeal nerve injury
References in thyroid surgery with or without intraoperative nerve monitoring. Acta
Otorhinolaryngol Ital 2014;34(4):223–9.
[1] Tuggle CT, Roman SA, Wang TS, et al. Pediatric endocrine surgery: who is operating [35] Zheng S, Xu Z, Wei Y, et al. Effect of intraoperative neuromonitoring on recurrent la-
on our children? Surgery 2008;144(6):869–77 [discussion 877]. ryngeal nerve palsy rates after thyroid surgery—a meta-analysis. J Formos Med
[2] Al-Qurayshi Z, Robins R, Hauch A, et al. Association of surgeon volume with out- Assoc 2013;112(8):463–72.
comes and cost savings following thyroidectomy: a national forecast. JAMA [36] Randle RW, Bates MF, Long KL, et al. Impact of Potassium iodide on thyroidectomy for
Otolaryngol Head Neck Surg 2016;142(1):32–9. Graves' disease: implications for safety and operative difficulty. Surgery 2018;163(1):
[3] Scholz S, Smith JR, Chaignaud B, et al. Thyroid surgery at Children's Hospital Boston: 68–72.
a 35-year single-institution experience. J Pediatr Surg 2011;46(3):437–42. [37] Raval MV, Browne M, Chin AC, et al. Total thyroidectomy for benign disease in the
[4] Hanba C, Svider PF, Siegel B, et al. Pediatric thyroidectomy. Otolaryngol Head Neck pediatric patient—feasible and safe. J Pediatr Surg 2009;44(8):1529–33.
Surg 2017;156(2):360–7. [38] Wood JH, Partrick DA, Barham HP, et al. Pediatric thyroidectomy: a collaborative sur-
[5] Kundel A, Thompson GB, Richards ML, et al. Pediatric endocrine surgery: a 20-year gical approach. J Pediatr Surg 2011;46(5):823–8.
experience at the Mayo Clinic. J Clin Endocrinol Metab 2014;99(2):399–406. [39] Sherman J, Thompson GB, Lteif A, et al. Surgical management of Graves disease in
[6] Barrio R, Lopez-Capape M, Martinez-Badas I, et al. Graves' disease in children and ad- childhood and adolescence: an institutional experience. Surgery 2006;140(6):
olescents: response to long-term treatment. Acta Paediatr 2005;94(11):1583–9. 1056–61 [discussion 1061-2].
[7] Jevalikar G, Solis J, Zacharin M. Long-term outcomes of pediatric Graves' disease. J [40] Elfenbein DM, Katz M, Schneider DF, et al. Thyroidectomy for Graves' disease in chil-
Pediatr Endocrinol Metab 2014;27(11-12):1131–6. dren: indications and complications. J Pediatr Surg 2016;51(10):1680–3.