eTable 1.ICD-9 Codes Used in Defining the Study Variables
eTable 2.ICD-9 Codes Used in Defining Postthyroidectomy Complications
eTable 3. Distribution of Each Surgeon Volume by Specialty for Thyroidectomy Indicated for Pediatric Thyroid Cancer
eFigure 1. Distribution of Surgeon Volume by Specialty for Thyroidectomy Indicated for Pediatric Thyroid Cancer
eFigure 2. Postoperative Complications in Children With Thyroid Cancer
eFigure 3. Risk for Postthyroidectomy Complications and Length of Stay of >1 Day for Pediatric Thyroid Cancer by Surgeon Volume and Specialty
Customize your JAMA Network experience by selecting one or more topics from the list below.
Al-Qurayshi Z, Hauch A, Srivastav S, Aslam R, Friedlander P, Kandil E. A National Perspective of the Risk, Presentation, and Outcomes of Pediatric Thyroid Cancer. JAMA Otolaryngol Head Neck Surg. 2016;142(5):472–478. doi:10.1001/jamaoto.2016.0104
Thyroid cancer is the most common endocrine malignant neoplasm in children and adolescents. Despite a more advanced presentation of thyroid cancer, younger patients tend to have a more favorable prognosis and a lower mortality rate than adults with thyroid cancer.
To examine the presentation and outcomes of thyroid cancer in pediatric patients.
Design, Setting, and Participants
A cross-sectional weighted analysis was performed using data from the Nationwide Inpatient Sample from January 1, 2003, to December 31, 2010. Patient data were derived from a sample of 20% of community hospitals in the United States. Six hundred forty-four children and adolescents (age, <18 years; hereinafter referred to as children) with thyroid cancer were compared with 43 536 adults (age, ≥18 years) with thyroid cancer. Data were analyzed from December 7, 2014, to November 19, 2015.
Thyroid cancer and thyroidectomy.
Main Outcomes and Measures
Presentation and management characteristics of thyroid cancer and thyroidectomy outcomes in relation to surgeon volume and specialty.
A total of 644 cases of pediatric thyroid cancer were included (female, 77.3%; mean [SEM] age, 13.8 [0.2] years), corresponding to a weighted sample of 32 563. Compared with adults with thyroid cancer, children were more likely to present with cervical lymph node involvement (31.5% vs 14.7%; odds ratio [OR], 2.29; 95% CI, 1.76-2.97; P < .001) and lung metastases (5.7% vs 2.2%; OR, 2.79; 95% CI, 1.82-4.28; P < .001), whereas bone metastases were more frequent in adults (0.3% vs 1.1%; OR, 0.23; 95% CI, 0.06-0.90; P = .04). Children were more likely to be treated by a low-volume surgeon (26.9% vs 16.0%; OR, 2.09; 95% CI, 1.26-3.48; P = .005) or a pediatric surgeon (14.5% vs 9.6%; OR, 1.66; 95% CI, 1.04-2.67; P = .04) and in a low-volume hospital (20.5% vs 15.2%; OR, 2.97; 95% CI, 1.60-5.54; P < .001) or a teaching hospital (81.7% vs 63.1%; OR, 3.61; 95% CI, 2.33-5.60; P < .001). Compared with those treated by low-volume surgeons, children treated by high-volume surgeons were less likely to experience postoperative complications (14.3% vs 35.9%; OR, 0.16; 95% CI, 0.05-0.51; P = .002) or a hospital stay of more than 1 day (49.8% vs 67.9%; OR, 0.36; 95% CI, 0.15-0.90; P = .03). Management by pediatric surgeons did not significantly alter the risk for postoperative complications compared with other specialties (21.3% vs 18.5%; OR, 1.71; 95% CI, 0.64-4.53, P = .28). Management of thyroid cancer in children was significantly more costly (>$10 067.08/case) compared with adults (P = .04).
Conclusions and Relevance
Compared with thyroid cancer in adults, pediatric thyroid cancer is more likely to present as advanced disease and to be managed by low-volume or pediatric surgeons. In addition, within the United States, surgeon volume appears to be more crucial in determining thyroidectomy outcomes than the surgeon’s field of specialization.
The risk for thyroid cancer in children and adolescents is recognizably low when compared with that of older populations.1 In 2011 in the United States, the incidence of newly diagnosed thyroid cancer was 0.89 per 100 000 population for patients younger than 20 years, whereas the rate was 15.8 per 100 000 population in those 20 to 49 years of age.2 Thyroid cancer is, however, the most common pediatric endocrine neoplasm and the third most common solid tumor for this age group.3-6 Thyroid cancer represents 3% of all pediatric malignant neoplasms and 5% of malignant neoplasms of the head and neck.4-6 The most common age for pediatric thyroid carcinoma is 11 to 15 years, with less than 4% to 5% found in preschool-aged children.4,7
Thyroid cancer in children usually manifests as an asymptomatic neck mass5-9 and more commonly presents with cervical lymph node involvement and distant metastasis compared with thyroid cancer in adults.5,7-10 Despite the more advanced presentation of thyroid cancer, younger patients tend to have a more favorable prognosis and a lower mortality rate compared with adults.1,3,7,9,11
As a consequence of the low incidence of thyroid cancer in a younger population, little evidence is available regarding the definitive management and outcomes of pediatric thyroid cancer.5,7,12 Surgeon volume and specialty have been suggested to play a crucial role in the management and outcomes of thyroid cancer in children.5,7 This suggestion is supported in part by evidence of associations between surgeon volume and advantageous outcomes for other types of surgical interventions in children.13-16 However, these studies did not identify significant differences in the risk profile based on surgeon specialization.13,17,18 In 2 studies that looked at outcomes for thyroidectomy and parathyroidectomy in children with benign or malignant conditions, higher surgeon volume was associated with better outcomes, whereas surgeon specialty did not demonstrate a significant effect.19,20 Similar conclusions were reached by 3 separate literature reviews.5,7,21 In this study, we aimed to examine the presentation of thyroid cancer in children and to assess the effect of surgeon volume and field of specialization on the management and outcomes of pediatric thyroid cancer in the United States.
Question What is the presentation of pediatric thyroid cancer, and what is the effect of surgeon volume and field of specialization on thyroidectomy outcomes?
Findings In this cross-sectional study comparing children and adolescents with adults with thyroid cancer, pediatric thyroid cancer presented at an advanced stage. Higher surgeon volume was associated with favorable outcomes, whereas specialization (pediatric surgery, otolaryngology, or general surgery) did not affect the outcomes significantly.
Meaning Surgeon volume is more crucial than field of specialization in determining thyroidectomy outcomes in patients with pediatric thyroid cancer.
This study is a cross-sectional analysis using the Nationwide Inpatient Sample (NIS) database from January 1, 2003, to December 31, 2010. The NIS is part of the Healthcare Cost and Utilization Project and is sponsored by the Agency for Healthcare Research and Quality. The NIS is the largest all-payer inpatient care database publicly available in the United States and contains data from approximately 8 million hospital stays from about 1000 hospitals to approximate a 20% stratified sample of US community hospitals.22 The NIS database contains publicly available deidentified data that are exempt from institutional review board approval and informed consent.
We used the International Classification of Diseases, Ninth Revision to define the diagnoses and procedures of interest (eTables 1 and 2 in the Supplement). The study population consisted of children and adolescents (age <18 years; hereinafter referred to as children) who were compared with adults (age ≥18 years) admitted with the primary diagnosis of thyroid cancer (code 193).
The main risk factors assessed for the comparison of pediatric and adult thyroid cancer were race or ethnicity (white, black, Hispanic, Asian/Pacific Islander, Native American, and other), body weight (overweight vs not overweight), cervical lymph node involvement at the time of admission, the presence of lung metastasis at the time of admission, the presence of bone metastasis at the time of admission, the type of thyroidectomy (none, partial or unilateral, total, and substernal), whether the patient received radioisotope therapy or chemotherapy during the hospital stay, hospital teaching status (teaching vs nonteaching), hospital volume (percentile classification based on the annual volume of thyroid cancer admission includes ≤10th percentile [low], 1-10 cases/y; >10th to ≤90th percentile [intermediate], 11-198 cases/y; and >90th percentile [high], ≥199 cases/y), and the cost of health services for patients who underwent thyroidectomy. The cost was adjusted for the inflation rate to reflect 2015 US dollar values and categorized based on the 75th percentile ($10 067.08) into high vs low.
A subset of the children and adults with thyroid cancer was created that included patients who underwent thyroidectomy during their hospital stay. We examined the distribution of surgeon volumes and specialty based on the age of the patients. Surgeon volume groups were defined based on the annual number of thyroidectomies performed by each surgeon using the unique surgeon identifier number provided in the database. Low volume was defined as 1 to 2 thyroidectomies/y; intermediate volume, 3 to 30 thyroidectomies/y; and high volume, 31 or more thyroidectomies/y. Pediatric surgeons were identified by surveying the NIS database for the years 2003 to 2009. If more than 90% of the patients treated by that specific surgeon were younger than 18 years, the surgeon was considered a pediatric surgeon. Otolaryngologists were defined as surgeons who performed thyroidectomy and tympanoplasty, myringotomy, tonsillectomy, and/or adenoidectomy in the NIS database. General and other surgeon groups consisted of surgeons who were not pediatric surgeons or otolaryngologists. Analyses that involved surgeon volumes and specialties were for the period of 2003 to 2009 only because the surgeon identifier codes used to define these variables were not available for 2010.
Last, surgeon volumes and fields of specialization were assessed for their association with the outcomes of thyroidectomy in children in terms of postoperative complications and length of stay (≤1 vs >1 day). Complications were dichotomized based on the presence or the absence of 1 or more of the general and/or specific complications related to thyroid surgery (eTable 2 in the Supplement).
Statistical analyses used weighted data reflecting the national estimate. The records’ weights were available in the NIS data and were calculated based on the stratification variables that were used in sampling methods. These variables include hospital geographic region, urban or rural location, teaching status, bed size, and ownership.
We used cross-tabulation and the χ2 test to examine the association between each of the independent factors and the outcomes of interest. Factors with significant association were considered confounders and were included in the multivariate logistic regression models that were used in calculating the odds ratios (ORs) and 95% CIs. Significance level was set as α = .05, and all analyses were conducted using SAS for Windows (version 9.3; SAS Institute Inc).
A total of 644 pediatric thyroid cancer cases (age <18 years) was extracted from the NIS database (weighted sample size, 32 563 cases) (Table 1). The mean (SEM) patient age was 13.8 (0.2) years, and most of the pediatric patients were female (77.3%) and white (57.7%). At presentation, 31.0% of children with thyroid cancer had cervical lymph node involvement, whereas lung and bone metastases were reported in 5.7% and 0.3% of the study sample, respectively. Thyroidectomy (91.8%) was the predominant method of treatment received during the hospital stay, whereas radioisotope therapy was administrated in 7.8% (36 cases). A total of 107 patients (16.5%) underwent cervical lymph node dissection.
A total of 43 536 cases of adult thyroid cancer were identified and compared with the pediatric thyroid cancer cases (weighted sample size, 2 193 900) (Table 2). Hispanic patients were at least twice as likely to present with thyroid cancer at younger than 18 years (29.6% vs 12.8%) compared with white patients (57.7% vs 70.8%; OR, 2.74; 95% CI, 2.06-3.65; P < .001). Black patients were less likely than white patients to have pediatric thyroid cancer (2.4% vs 6.7%; OR, 0.40; 95% CI, 0.20-0.81; P = .01).
With regard to presentation, adult patients with thyroid cancer were more likely to be overweight compared with children (6.2% vs 2.4%; OR, 0.45; 95% CI, 0.26-0.79; P = .005) (Table 2). Cervical lymph node involvement was significantly more common in children than in adults (31.0% vs 14.7%; OR, 2.29; 95% CI, 1.76-2.97; P < .001). Similarly, metastasis to the lung was also more frequently found in children (5.7% vs 2.2%; OR, 2.79; 95% CI, 1.82-4.28; P < .001), but bone metastases were less likely to be found in children with thyroid cancer (0.3% vs 1.1%; OR, 0.23; 95% CI, 0.06-0.90; P = .04).
In terms of management, the performance of total or partial thyroidectomy was not significantly different between children and adults; however, substernal thyroidectomy was more likely to be performed in young patients (3.9% vs 2.0%; OR, 1.99; 95% CI, 1.08-3.67; P = .03). Children with thyroid cancer were more likely than adults to be treated at low-volume hospitals (20.5% vs 15.2%; OR, 2.97; 95% CI, 1.60-5.54; P < .001) and/or teaching hospitals (81.7% vs 63.1%; OR, 3.61; 95% CI, 2.33-5.60; P < .001). The cost of health services related to performing thyroidectomy was assessed in a model that controlled for the confounding effect of demographic and clinical factors and for the length of stay and presence of complications. The model demonstrated that the management of pediatric thyroid cancer cost significantly more (>$10 067.08/case) compared with adults (P = .04).
We examined the association between surgeon volume and field of specialization (eTable 3 and eFigure 1 in the Supplement). Pediatric surgeons were more likely to be low-volume surgeons (48.9% vs 16.2%; P < .001) and less likely to be high-volume surgeons (5.0% vs 37.9%; P < .001). Otolaryngologists and general surgeons did not exhibit a significant difference in relation to surgeon volume compared with other specialties.
Children were more likely than adults in the study sample to be treated by low-volume surgeons (26.9% vs 16.0%; OR, 2.09; 95% CI, 1.26-3.48; P = .005) and/or pediatric surgeons (14.5% vs 9.6%; OR, 1.66; 95% CI, 1.04-2.67; P = .04). High- and intermediate volume surgeons, otolaryngologists, and general surgeons performed thyroidectomies in children and adults at the same rates and without a significant difference (Table 3).
The most common reported postoperative complications in children were hypocalcemia (74.2%) and vocal cord paralysis (8.6%) (eFigure 2 in the Supplement). Compared with low-volume surgeons, children treated by high-volume surgeons were at a lower risk for postoperative complications (14.3% vs 35.9%; OR, 0.16; 95% CI, 0.05-0.51; P = .002) (Table 4 and eFigure 3 in the Supplement); similar results were found for intermediate-volume surgeons compared with low-volume surgeons (13.3%; OR, 0.17; 95% CI, 0.07-0.42; P < .001). Patients treated by low-volume surgeons were more likely to have a hospital stay of more than 1 day compared with those treated by high-volume surgeons (67.9% vs 49.8%; OR, 0.36; 95% CI, 0.15-0.90; P = .03) (Table 5 and eFigure 3 in the Supplement). We found no significant difference in the risk for postoperative complications or in length of stay based on the surgeon’s field of specialization.
This study has demonstrated that Hispanic children were at a higher risk for thyroid cancer compared with white children, whereas children of a black racial background were at a lower risk. In regard to presentation, children were more likely to have cervical lymph node involvement and lung metastases at the time of hospital admission compared with adults. In their 1988 retrospective cohort study of 58 children and 981 adults, Zimmerman et al11 reported similar findings regarding the higher incidence of cervical lymph node involvement and distant metastases in children at the time of presentation. Other studies that included pediatric patients with thyroid cancer only and lacked comparison with adults12,23,24 reported similar presentations.
Children with thyroid cancer were more likely to be treated by low-volume surgeons or pediatric surgeons, and pediatric surgeons were less likely to have a high volume of thyroid cases. This finding is very likely a result of the low incidence of diseases that require thyroid operations in young patients. Despite the fact that pediatric surgeons are more experienced with younger groups, thyroidectomies performed by pediatric surgeons were not associated with a different risk for complications or length of stay compared with those performed by other specialties. We found that the higher surgeon volume was distinctly associated with a lower risk for complications, regardless of the specialization. In an analysis investigating outcomes associated with pediatric surgeons from different disciplines, Rhee et al13 found that highly specialized pediatric surgeons in the field of general and cardiothoracic surgery were associated with a lower risk of in-hospital mortality, whereas no significant association was detected in the fields of otolaryngology, neurosurgery, orthopedic surgery, or urology. Although the group examined by Rhee et al13 focused only on pediatric surgeons and did not mention which field was treating thyroid cancer specifically, their study reflects the fact that pediatric training in surgery does not necessarily equate to more favorable outcomes for all operations in pediatric patients.
In their literature review of more than 1800 patients, Thompson and Hay7 found that postoperative complications were lowest for high-volume surgeons. Similarly, Rapkin and Pashankar5 shared comparable observations in their review; however, they placed more emphasis on the surgeon’s field of specialization (ie, endocrine and pediatric) than experience. Sosa et al19 also observed a lower risk for complications for high-volume surgeons who managed benign and malignant conditions of the thyroid and parathyroid glands, and Tuggle et al20 found that surgeon volume was more influential than surgeon specialty in determining outcomes in the same study population as the study by Sosa et al. However, in both studies,19,20 surgeon volume was divided into 2 groups only with a cutoff value of 30 procedures/y.
In recognition of the differences in thyroid diseases between adults and children, the American Thyroid Association in 2015 released the inaugural guidelines for the management of thyroid diseases in pediatric patients.25 We believe that the development of such guidelines is imperative and could result in favorable outcomes of thyroid surgery in this age group. Our analysis predates the release of those guidelines, such as those concerning management of pediatric calcium levels and evaluation of vocal fold paralysis. Evaluation of the impact of those guidelines in the future will be important.
Limitations of this study include the cross-sectional design that hinders establishing causality, the administrative nature of the database, and the lack of data for posthospitalization follow-up. The administrative data lack details concerning some relevant factors, such as histopathologic reports, laboratory values, and specifics of management. Also, given that the analysis predates the release of management guidelines for pediatric thyroid cancer, a difference in practice pattern in the management of cancer in adults and children may exist. The practice pattern may vary based on the surgeon’s field of specialization or the nature of their approach, which may not have been reflected accurately using the present data. In addition, the database does not have information regarding outpatient thyroidectomy; missing such information may alter the overall results and prevent the presentation of a comprehensive picture because thyroidectomy can be performed in an outpatient setting. However, considering the very low incidence of pediatric thyroid cancer, which has been studied mainly through case reports and case series, this analysis extrapolates findings to the national level and establishes a basis for further investigation using more sophisticated resources.
The strength of the study is the large sample size that has helped to confirm several observations previously reported in a variety of descriptive and lesser-scale analytic studies. The weighted statistical analysis applied in this study, in addition to the wide array of variables available within the NIS database that has been assessed for their confounding effect, adds to the strength of this study and leads to a highly representative sample of the general population found in the United States.
Thyroid cancer among children and adults exhibits essential differences in terms of presentation, management, and even cost of health services to hospitals. Outcomes surrounding the management of thyroid cancer are optimized by characteristics such as surgeon volume rather than surgical field of specialization. Our results provide an educational tool and open a discussion platform that can improve our understanding of the pediatric thyroid cancer profile. They also support attempts at improving our outcomes via centers of higher volume.
Corresponding Author: Emad Kandil, MD, MBA, Department of Surgery, Tulane University School of Medicine, 1430 Tulane Ave, Mail Stop SL-22, New Orleans, LA 70112 (firstname.lastname@example.org).
Accepted for Publication: January 7, 2016.
Published Online: March 31, 2016. doi:10.1001/jamaoto.2016.0104.
Author Contributions: Dr Kandil had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Al-Qurayshi, Aslam, Friedlander, Kandil.
Acquisition, analysis, or interpretation of data: Al-Qurayshi, Hauch, Srivastav, Kandil.
Drafting of the manuscript: Al-Qurayshi, Hauch, Srivastav.
Critical revision of the manuscript for important intellectual content: Al-Qurayshi, Hauch, Aslam, Friedlander, Kandil.
Statistical analysis: Al-Qurayshi, Hauch, Srivastav.
Administrative, technical, or material support: Friedlander.
Study supervision: Aslam, Kandil.
Conflict of Interest Disclosures: None reported.
Previous Presentation: This paper was presented as a poster at the 36th Annual Meeting of the American Association of Endocrine Surgeons; May 17-19, 2015; Nashville, Tennessee.
Create a personal account or sign in to: