Postoperative Outcomes in Graves' Disease Patients: Results from the Nationwide Inpatient Sample Database

Abstract

Background:

Current surgical indications for Graves' disease include intractability to medical and/or radioablative therapy, compressive symptoms, and worsening ophthalmopathy. Total thyroidectomy for Graves' disease may be technically challenging and lead to untoward perioperative outcomes. This study examines outcomes in patients with Graves' disease who underwent total thyroidectomy and assesses its safety for this patient population.

Methods:

A retrospective cross-sectional analysis was performed using the Nationwide Inpatient Sample database from 2006 to 2011. Total thyroidectomy performed in patients with Graves' disease, benign multinodular goiter (MNG), and thyroid cancer was identified. Demographic factors, comorbidities, and postoperative complications were evaluated. Chi-square, one-way analysis of variance, and risk-adjusted multivariable logistic regression were performed.

Results:

Of 215,068 patients who underwent total thyroidectomy during the study period, 11,205 (5.2%) had Graves' disease, 110,124 (51.2%) MNG, and 93,739 (43.6%) thyroid malignancy. Patients with Graves' disease were younger than MNG and thyroid cancer patients (M _age = 42.8 years vs. 55.5 and 51.0 years; p < 0.01). The Graves' disease group included a higher proportion of women (p < 0.01) and nonwhites (p < 0.01). Postoperatively, Graves' patients had significantly higher rates of hypocalcemia (12.4% vs. 7.3% and 10.3%; p < 0.01), hematomas requiring reoperation (0.7% vs. 0.4% and 0.4%; p < 0.01), and longer mean hospital stay (2.7 days vs. 2.4 and 2.2 days; p < 0.01) compared to MNG and thyroid cancer patients, respectively. On risk-adjusted multivariate logistic regression, Graves' disease was independently associated with a higher risk of vocal-cord paralysis (odds ratio [OR] = 1.36 [confidence interval (CI) 1.08–1.69]), tracheostomy (OR = 1.35 [CI 1.1–1.67]), postoperative hypocalcemia (OR = 1.65 [CI 1.54–1.77]), and hematoma requiring reoperation (OR = 2.79 [CI 2.16–3.62]) compared to MNG patients. High-volume centers for total thyroidectomy were independently associated with lower risk of postoperative complications, including in patients with Graves' disease.

Conclusions:

Despite low overall morbidity following total thyroidectomy, Graves' disease patients are at increased risk of postoperative complications, including bleeding, vocal-cord paralysis, tracheostomy, and hypocalcemia. These risks appear to be lower when performed at high-volume centers, and thus referral to these centers should be considered. Total thyroidectomy may therefore be a safe treatment option for appropriately selected patients with Graves' disease when performed by experienced surgeons.

Introduction

Graves' disease is the most common cause of hyperthyroidism in the United States (1). Current management regimens for Graves' disease include antithyroid medications, radioactive iodine ablation, or thyroidectomy (1). Each treatment regimen differs in efficacy, safety, and cost, with no option universally superior to the other for all patients. Clinical factors that favor surgical treatment include large goiter with compressive symptoms, worsening ophthalmopathy, concurrent suspicious thyroid nodules, and/or primary hyperparathyroidism (2). Women planning to become pregnant, as well as patients with intolerance, intractability, or contraindications to medical and/or radioablative therapy, may also be referred for total thyroidectomy (3,4). Patients with severe thyrotoxicosis resulting in physiologic compromise may also require urgent thyroidectomy (5).

Total thyroidectomy offers definitive treatment for Graves' disease (6,7), and is one of the recommended first-line treatments in the 2011 American Thyroid Association (ATA)/American Association of Clinical Endocrinologists and 2016 ATA management guidelines (2,8). Although surgical management offers high rates of success and prompt resolution of hyperthyroidism, current trends still show that total thyroidectomy lags behind medical and radioactive iodine therapy as a definitive treatment for Graves' disease (1,6,9). Total thyroidectomy carries some risks, including permanent hypoparathyroidism and vocal-cord dysfunction (10 –12). Moreover, due to chronic inflammation, such operations in the setting of Graves' disease are technically challenging and may lead to increased risk of perioperative complications (13). For these reasons, endocrinologists may be discouraged from considering total thyroidectomy for treatment of Graves' disease in otherwise appropriate surgical candidates. In fact, most endocrinologists in the United States and Europe currently do not refer patients for thyroidectomy for definitive treatment of Graves' disease (3,9,14). Nonetheless, in recent years, total thyroidectomy has become the primary treatment modality for Graves' disease in some high-volume endocrine surgery centers (15 –17).

Previous studies from single institutions, predominantly tertiary referral centers, have reported a high safety profile for thyroidectomy in the setting of Graves' disease (10,11,17,18). However, complication rates following total thyroidectomy for Graves' disease have not been evaluated from large national databases that reflect outcomes from a variety of geographic and practice settings. Therefore, the aims of this population-based study were to evaluate outcomes following total thyroidectomy in patients with Graves' disease and to determine its safety as a definitive treatment for this patient population.

Materials and Methods

A retrospective cross-sectional analysis using data from the Health Care Utilization Project Nationwide Inpatient Sample (HCUP-NIS) from 2006 to 2011 was conducted. The HCUP-NIS is the largest publicly available inpatient database maintained by the United States Agency for Healthcare Research and Quality (AHQR). This database includes a 20% stratified sample of all discharges from non-federal hospitals, resulting in approximately eight million records per year. Patient cases are weighted appropriately to project national estimates for analyses.

Total thyroidectomy operations were identified using International Classification of Diseases, Ninth Revision (ICD-9) code 06.4. Adult patients with a primary diagnosis of Graves' disease were compared to patients with other indications for total thyroidectomy, including benign nontoxic multinodular goiter (MNG) or thyroid malignancy. Patient characteristics including demographic, socioeconomic, and comorbidities were recorded and compared between cohorts. Variables available from the HCUP-NIS included age, sex, race/ethnicity, primary payer type, and admission status (elective vs. non-elective). Comorbidities including history of diabetes mellitus, hypertension, congestive heart failure, chronic lung disease, renal failure, chronic liver disease, and obesity were also recorded. Hospital-level data included the HCUP-NIS-provided variable for hospital location and teaching status. Hospital total thyroidectomy volume was calculated using HCUP-NIS unique hospital identifiers assigned to each case. Hospitals were then stratified between the top 20th percentile and bottom 80th percentile, which corresponded to >47 or ≤47 total thyroidectomies per year, respectively. Outcome variables included in-hospital complications, length of stay, and total hospital charges. Perioperative complications were identified using ICD-9 diagnostic codes, with the exception of in-hospital death, which is reported in the HCUP-NIS. Perioperative factors and surgical complications analyzed included wound complications, hematoma, procedure for control of bleeding, blood transfusion, vocal-cord paralysis, need for tracheostomy, autotransplantation of parathyroid glands, and substernal thyroidectomy. Endocrine complications included in-hospital hypocalcemia and tetany. Medical complications were defined as composite of cardiovascular, renal, pulmonary, and cerebrovascular adverse events. All complications listed refer to in-hospital events, as there are no post-discharge data in the HCUP-NIS.

Statistical analyses were performed using SAS v9.4 (SAS Institute, Inc., Cary, NC). Differences between diagnoses (Graves' disease, MNG, thyroid cancer) with respect to each of the primary outcomes were tested by chi-square test, one-way analysis of variance, and risk-adjusted multivariable logistic regression models. Multivariable logistic regression models for each of the primary outcomes included primary predictor group (diagnosis; with MNG as the reference group), in addition to co-factors including age at admission (years), sex (female vs. male), race/ethnicity (white, black, Hispanic, others), insurance type (Medicare, Medicaid, private, uninsured/other), number of comorbidities, admission status (elective vs. emergent/urgent), teaching status of hospital (rural, urban non-teaching, urban teaching), hospital total thyroidectomy volume (top 20th percentile vs. bottom 80th percentile), and substernal thyroidectomy (yes vs. no). Adjusted odds ratios (OR) with corresponding confidence intervals (CI) along with p-values and area under the curve (AUC) for each model were calculated. This study was approved by the Institutional Review Board at the University of Miami Leonard M. Miller School of Medicine.

Results

There were 215,068 total thyroidectomy cases identified in adult patients using the HCUP-NIS between 2006 and 2011. Of these, 11,205 (5.2%) were performed in patients with a primary diagnosis of Graves' disease, 110,124 (51.2%) for MNG, and 93,739 (43.6%) for thyroid malignancy (Table 1). Patients with Graves' disease were significantly younger than MNG and thyroid cancer patients who underwent total thyroidectomy (M _age = 42.8 years vs. 55.5 and 51.0 years; p < 0.01), respectively. The Graves' cohort also had a higher proportion of women and nonwhites compared to the MNG and thyroid cancer groups (Table 1). Groups also differed in terms of health insurance type, with the Graves' cohort having more Medicaid (18.6% vs. 8.0% and 7.4%; p < 0.01) and uninsured patients (11.6% vs. 7.1% and 5.8%; p < 0.01) than the MNG and thyroid cancer groups, respectively. Overall, Graves' disease patients had lower rates of comorbidities compared to other total thyroidectomy groups, with the exception of heart failure, which was highest in Graves' patients (3.8%) compared to MNG (2.5%) and thyroid cancer (1.4%; p < 0.01; Table 1).

Table 1.

Characteristics of Patients who Underwent Total Thyroidectomy Based on Diagnosis

Characteristics	Overall (n = 215,068)	MNG (n = 110,124; 51.2%)	Graves' disease (n = 11,205; 5.2%)	Thyroid cancer (n = 93,739; 43.6%)	p-Value
Mean age, years (SD)	52.9 (15.3)	55.5 (14.3)	42.8 (14.4)	51.0 (15.7)	<0.01
Sex					<0.01
Female	78.3%	81.2%	83.7%	74.3%
Male	21.7%	18.8%	16.3%	25.7%
Race/ethnicity					<0.01
White	68.3%	66.1%	58.8%	71.9%
Black	13.8%	19.5%	22.3%	6.3%
Hispanic	9.6%	7.6%	10.8%	11.8%
Other	8.3%	6.8%	8.1%	10.0%
Payer type					<0.01
Medicare	26.2%	30.9%	12.9%	22.4%
Medicaid	8.3%	8.0%	18.6%	7.4%
Private insurance	58.7%	54.0%	56.9%	64.4%
Uninsured/other	6.8%	7.1%	11.6%	5.8%
Comorbidities
Diabetes mellitus	15.2%	17.5%	9.3%	13.2%	<0.01
Hypertension	40.3%	45.9%	31.8%	34.7%	<0.01
Congestive heart failure	2.1%	2.5%	3.8%	1.4%	<0.01
Chronic lung disease	12.0%	13.2%	11.9%	10.6%	<0.01
Renal failure	2.1%	2.6%	1.2%	1.7%	<0.01
Liver disease	0.7%	0.6%	0.8%	0.8%	<0.01
Obesity	10.2%	11.7%	6.9%	8.9%	<0.01

MNG, multinodular goiter; SD, standard deviation.

When perioperative factors and unadjusted analyses of clinical outcomes following total thyroidectomy between groups were performed, there were more non-elective admissions in Graves' disease patients who underwent total thyroidectomy compared to MNG and thyroid cancer patients (17.8% vs. 14.6% and 9.1%; p < 0.01), respectively (Table 2). There were also significant differences between groups with regard to hospital location and teaching status (Table 2). Furthermore, 21.8% of Graves' disease patients underwent total thyroidectomy at high-volume centers compared to 17.8% of MNG patients and 24.5% of thyroid cancer patients (p < 0.01; Table 2). Graves' patients also underwent substernal thyroidectomy less frequently than MNG patients (5.3% vs. 11.0%; p < 0.01) but at a higher frequency than patients in the thyroid cancer group (4.0%). There were also differences in rates of parathyroid autotransplantation during total thyroidectomy between cohorts (Table 2).

Table 2.

Perioperative Details and Unadjusted Outcomes Following Total Thyroidectomy Based on Diagnosis

Variable	Overall (n = 215,068)	MNG (n = 110,124; 51.2%)	Graves' disease (n = 11,205; 5.2%)	Thyroid cancer (n = 93,739; 43.6%)	p-Value
Admission status					<0.01
Elective	87.6%	85.4%	82.2%	90.9%
Emergent/urgent	12.4%	14.6%	17.8%	9.1%
Hospital location/teaching status					<0.01
Rural	5.3%	5.8%	6.5%	4.5%
Urban, non-teaching	30.6%	32.1%	29.6%	28.9%
Urban, teaching	64.2%	62.1%	63.9%	66.6%
Hospital total thyroidectomy volume					<0.01
Low (bottom 80th percentile; 0 thyroidectomy	79%	82.2%	78.2%	75.5%
High (top 20th percentile; >47 per year)	21%	17.8%	21.8%	24.5%
Substernal thyroidectomy	7.7%	11.0%	5.3%	4.0%	<0.01
Parathyroid gland autotransplantation	6.6%	5.2%	6.1%	8.4%	<0.01
Tracheostomy	1.3%	1.2%	1.0%	1.3%	<0.01
Hematoma	1.9%	2.1%	2.8%	1.5%	<0.01
Requiring surgical intervention	0.4%	0.4%	0.7%	0.4%	<0.01
Blood transfusion	1.0%	1.3%	1.4%	0.7%	<0.01
Hypocalcemia	8.8%	7.3%	12.4%	10.3%	<0.01
Tetany	0.1%	0.0%	0.3%	0.1%	<0.01
Vocal-cord paralysis	1.6%	1.1%	0.9%	2.4%	<0.01
Unilateral	1.0%	0.5%	0.3%	1.6%	<0.01
Bilateral	0.4%	0.3%	0.3%	0.5%	<0.01
Wound complications	0.3%	0.2%	0.1%	0.3%	<0.01
Venous thromboembolism	0.2%	0.2%	0.0%	0.2%	<0.01
Pulmonary embolism	0.1%	0.1%	0.0%	0.1%	<0.01
Deep venous thrombosis	0.1%	0.1%	0.0%	0.1%	0.06
Medical complication	3.4%	4.1%	2.9%	2.6%	<0.01
Cardiac	0.3%	0.4%	0.3%	0.2%	<0.01
Stroke	0.2%	0.3%	0.0%	0.0%	<0.01
Pulmonary	2.6%	3.1%	2.4%	2.1%	<0.01
Acute kidney injury	0.6%	0.8%	0.4%	0.8%	<0.01
Shock	0.1%	0.1%	0.1%	0.0%	<0.01
In-hospital death	0.3%	0.4%	0.2%	0.1%	<0.01
Length of stay, days (SD)	2.3 (4.3)	2.4 (5.0)	2.7 (4.4)	2.2 (3.4)	<0.01
Total charges (SD)	$31,312 (45,722)	$32,735 (54,422)	$29,626 (38,309)	$29,845 (33,756)	<0.01

Overall, postoperative complications were low for the entire cohort of total thyroidectomy patients (Table 2). Graves' patients experienced higher rates of postoperative hypocalcemia (12.4% vs. 7.3% and 10.3%; p < 0.01) and tetany (0.3% vs. 0% and 0.1%; p < 0.01) compared to MNG and thyroid cancer patients, respectively (Table 2). Patients with Graves' disease also had a higher rate of hematomas requiring surgical intervention compared to other groups (0.7% vs. 0.4% in MNG vs. 0.4% in thyroid cancer; p < 0.01). Graves' patients did not experience higher rates of vocal-cord paralysis, need for tracheostomy, wound complications, or major medical complications following total thyroidectomy compared to MNG or thyroid cancer patients (Table 2). The mortality rate was low overall (0.3%) but was highest in the MNG group (0.4% vs. 0.2% in Graves' and 0.1% in thyroid cancer; p < 0.01). Graves' patients also had a marginally longer mean length of hospital stay compared to MNG and thyroid cancer patients (2.7 days vs. 2.4 and 2.2 days; p < 0.01), respectively. However, mean total hospital charges were lower for Graves' patients hospitalized after total thyroidectomy compared to other indications (Table 2).

Risk-adjusted multivariable analyses for postoperative complications with diagnosis as primary predictor are presented (Table 3). Compared to MNG as a reference group, Graves' disease was an independent risk factor for hematoma requiring surgical intervention (OR = 2.79 [CI 2.16–3.62]), postoperative hypocalcemia (OR = 1.65 [CI 1.54–1.77]), vocal-cord paralysis (OR = 1.36 [CI 1.08–1.69]), and tracheostomy (OR = 1.35 [CI 1.10–1.67]; Table 3). Similarly, thyroid cancer was independently associated with increased risk of postoperative hematoma, hypocalcemia, vocal-cord paralysis, and tracheostomy (Table 3). Of note, treatment at high-volume centers was associated with lower odds of postoperative hematoma (OR = 0.77 [CI 0.70–0.84]), hypocalcemia (OR = 0.52 [CI 0.50–0.55]), vocal-cord paralysis (OR = 0.74 [CI 0.67–0.81]), tracheostomy (OR = 0.65 [CI 0.57–0.74]), and major medical complications (OR = 0.61 [CI 0.56–0.66]) compared to low-volume hospitals.

Table 3.

Risk-Adjusted Multivariate Logistic Regression Models for Outcomes Following Total Thyroidectomy Based on Diagnosis

	Diagnosis (vs. MNG)
	Thyroid cancer		Graves' disease
Outcomes	aOR [CI]	p	aOR [CI]	p	AUC
Tracheostomy	1.70 [1.55–1.87]	<0.01	1.35 [1.10–1.67]	<0.01	0.78
Hematoma	0.84 [0.78–0.91]	<0.01	1.89 [1.65–2.17]	<0.01	0.65
Reoperation hematoma	1.05 [0.89–1.23]	0.55	2.79 [2.16–3.62]	<0.01	0.65
Hypocalcemia	1.42 [1.37–1.47]	<0.01	1.65 [1.54–1.77]	<0.01	0.62
Vocal-cord paralysis	2.85 [2.63–3.10]	<0.01	1.36 [1.08–1.69]	0.01	0.71
Wound complication	1.47 [1.20–1.81]	<0.01	0.28 [0.12–0.65]	<0.01	0.74
Venous thromboembolism	1.13 [0.91–1.41]	0.28	0.33 [0.14–0.78]	0.01	0.81
Major medical complication	0.98 [0.92–1.04]	0.53	1.05 [0.91–1.20]	0.50	0.79

Each row represents a multivariable logistic regression model, with diagnosis as the main predictor variable. Models also include: age (years), sex (female, male), race/ethnicity (white, black, Hispanic, others), insurance (Medicare, Medicaid, private, uninsured/other), number of comorbidities, admission status (elective, emergent/urgent), location/teaching status of hospital (rural, urban nonteaching, urban teaching), hospital volume for thyroidectomy (top 20th percentile vs. bottom 80th percentile), and substernal thyroidectomy (yes/no).

aOR, adjusted odds ratio; CI, confidence interval; AUC, area under the curve.

Postoperative outcomes for Graves' patients treated at hospitals with a high total thyroidectomy volume compared to low-volume hospitals are described (Table 4). Graves' patients treated at high-volume centers experienced lower rates of postoperative hematoma (1.4% vs. 3.1%; p < 0.01), hypocalcemia (7.0% vs. 13.9%; p < 0.01), tracheostomy (0.2% vs. 1.3%; p < 0.01), wound complications (0.0% vs. 0.2%; p < 0.01), and major medical complications (1.2% vs. 3.4%; p < 0.01) compared to those treated at low-volume hospitals. There was no statistically significant difference in rates of reoperation for hematoma, vocal-cord paralysis, or venous thromboembolism (Table 4).

Table 4.

Selected Outcomes Following Total Thyroidectomy in Patients with Graves' Disease Based on Hospital Total Thyroidectomy Volume

	Hospital total thyroidectomy volume
Variable	Low volume (bottom 80th percentile; ≤47 per year; n = 8757; 78.2%)	High volume (top 20th percentile; >47 per year; n = 2447; 21.8%)	p
Tracheostomy	1.3%	0.2%	<0.01
Hematoma	3.1%	1.4%	<0.01
Requiring surgical intervention	0.7%	0.4%	0.07
Hypocalcemia	13.9%	7.0%	<0.01
Vocal-cord paralysis	0.9%	0.8%	0.53
Wound complications	0.2%	0.0%	0.04
Venous thromboembolism	0.1%	0.0%	0.22
Major medical complication	3.4%	1.2%	<0.01

Discussion

Total thyroidectomy is an effective and definitive treatment for Graves' disease (6,7) but remains the least commonly used treatment modality when compared to medical and radioactive iodine therapy (1). Although total thyroidectomy is a safe procedure that can be performed by experienced surgeons with low morbidity, there nonetheless remains concern that Graves' disease patients may be at increased risk of perioperative complications, and this may discourage clinicians from recommending it as a definitive choice of therapy. Recently, there have been reports supporting a high safety profile of total thyroidectomy in patients with Graves' disease treated at tertiary care centers (10,11,17). This contemporary population-based analysis of the HCUP-NIS furthermore provides an overview of in-hospital perioperative outcomes in patients with Graves' disease treated with total thyroidectomy in the United States. Results of this study suggest that despite an overall low morbidity following total thyroidectomy, Graves' disease patients are at increased risk of postoperative complications, including bleeding, vocal-cord paralysis, tracheostomy, and hypocalcemia. Rates of these complications were significantly lower when performed at high-volume centers.

Total thyroidectomy in the setting of Graves' disease may be more technically challenging due to chronic inflammation and increased vascularity (13). For these reasons, risk of hematoma following thyroidectomy for Graves' disease is believed to be greater. Vascularized remnant tissue may also contribute to increased rate of bleeding in patients undergoing subtotal thyroidectomy. Of note, it is possible that some procedures were incorrectly coded in the data set, and therefore some patients in the Graves' disease cohort may have actually undergone less than a total thyroidectomy. In this study, patients with Graves' disease had a 2.8% rate of hematoma after total thyroidectomy compared to 2.1% and 1.5% in patients with MNG and thyroid cancer, respectively. Moreover, there was a 0.7% rate of surgical intervention for control of hematoma in the Graves' cohort, which was statistically higher than the 0.4% rate in the non-Graves' groups. Similarly, the rate of blood transfusion was 1.4% in Graves' patients, which was slightly higher than in MNG (1.3%) and thyroid cancer (0.7%) patients. The aforementioned results also coincide with other studies that have reported a higher risk of postoperative hematomas in patients' with Graves' disease (19,20). In one published study that analyzed risk factors in 207 patients from 15 academic hospitals who developed a post-thyroidectomy hematoma requiring reoperation, Graves' disease was found to be an independent risk factor for neck hematoma with an odds ratio of 2.4 (20). In the authors' current report, Graves' disease patients had a 2.79 times relative risk of developing postoperative hematomas requiring surgical intervention compared to MNG patients. Of note, the rate of hematomas was significantly lower in Graves' patients treated at high-volume centers (1.4% vs. 3.1%). These results confirm a low overall rate of hematomas requiring reoperation after total thyroidectomy but nevertheless an increased risk in Graves' disease patients.

Postoperative hypocalcemia is another well-known complication of total thyroidectomy (12,21). In most cases, hypocalcemia is transient, but injury to the parathyroid glands can result in permanent hypoparathyroidism. Several studies have reported higher rates of postoperative hypocalcemia in patients with Graves' disease (12,22,23). Transient hypocalcemia may occur at higher rates in hyperthyroid patients due to increased bone turnover (21). However, permanent rates of hypoparathyroidism after total thyroidectomy for Graves' disease are reported to be between 3% and 4% (10,11). As expected, a higher incidence of hypocalcemia in Graves' disease patients than in MNG and thyroid cancer patients undergoing total thyroidectomy was appreciated in the authors' study. This complication was also lower in Graves' patients treated at high-volume hospitals. The HCUP-NIS does not include post-discharge data. Therefore, the authors were unable to discern cases of transient versus permanent hypoparathyroidism. Thus, while the rate of transient hypocalcemia appears to be higher in Graves' disease patients, it is unclear whether they are at increased risk for permanent hypoparathyroidism. Moreover, the risk of transient hypocalcemia may be mitigated in Graves' disease patients by preoperative calcium supplementation, as reported by other authors (21).

Perhaps the most feared complication of total thyroidectomy is injury to the recurrent laryngeal nerves, resulting in vocal-cord paresis and airway compromise necessitating tracheostomy. In the authors' study, the rate of vocal-cord paralysis was 0.9% in patients with Graves' disease, which was significantly lower than in the MNG (1.1%) and thyroid cancer (2.4%) groups. In comparison, rates of recurrent laryngeal nerve injury reported in the literature range from 1% to 4% in Graves' disease patients undergoing total thyroidectomy (10,11,17,24). Tracheostomy, which may be necessary in severe cases of vocal-cord paralysis, was also performed less frequently in Graves' disease patients compared to MNG and thyroid cancer groups. However, after adjusting for baseline patient differences on multivariate regression analysis, Graves' disease status was associated with higher rates of vocal-cord dysfunction and need for tracheostomy compared to MNG. These results suggest that Graves' disease is a significant risk factor for recurrent laryngeal nerve injury during total thyroidectomy.

In addition to complications believed to be directly associated with total thyroidectomy, perioperative complications including renal, pulmonary, cardiac, cerebrovascular, and venous thromboembolism were also analyzed. Thyroidectomy is generally not associated with a high risk of major medical complications, and not surprisingly, there was a low rate overall (3.4%), which was not higher in Graves' disease patients (2.9%) in this report. Most of these complications were pulmonary, which occurred in 2.4% of Graves' disease patients. Mortality is also an exceedingly rare event following total thyroidectomy (7). In-hospital death rate was 0.2% in Graves' disease patients. Rate of wound complications other than hematoma was also low overall (0.3%) and in Graves' disease patients (0.1%).

Finally, Graves' disease patients had a longer mean length of stay compared to MNG and thyroid cancer (2.7 days vs. 2.4 and 2.2 days, respectively; p < 0.01), with a median length of one day in all cohorts. Additionally, Graves' patients did not incur higher total hospital charges than other patients undergoing total thyroidectomy did. Mean hospital charges for total thyroidectomy in Graves' patients were $29,626. Of note, a previous report found that total thyroidectomy was the most expensive treatment modality for Graves' disease, but it was the most cost-effective in the long term (25).

There are certain limitations to this study. The HCUP-NIS is an administrative database that relies on diagnostic and procedural codes provided by each participating institution. There is an inherent risk of some entries having coding errors, leading to missed or inaccurate diagnoses. This can result in over- or underestimation of rates of complications analyzed in this study. In addition, certain outcomes of interest to this study, such as thyroid storm, do not have a specific corresponding ICD-9 code and therefore could not be analyzed. More importantly, there are no post-discharge data, so the current analysis is limited to in-hospital events. Therefore, complications occurring after hospital discharge and readmissions were not captured in this data set. Moreover, long-term data regarding Graves' disease status to determine treatment success are not available. Finally, certain complications such as hypocalcemia and vocal-cord paralysis leading to permanent injury and long-term morbidity could not be assessed.

In summary, this study demonstrates that total thyroidectomy in Graves' disease patients is associated with increased risk of postoperative complications, including hypocalcemia, hematoma, recurrent laryngeal nerve injury, and need for tracheostomy, compared to patients with benign MNG. However, complication rates are significantly improved when surgery is performed at high-volume centers. Therefore, total thyroidectomy may be considered as a definitive treatment option for Graves' disease that can be safely performed by experienced surgeons. Treatment recommendations for Graves' disease should be discussed with multidisciplinary input based on specific patient factors. Total thyroidectomy should be considered in the treatment algorithm and offered as an option for appropriate candidates.

Footnotes

Acknowledgments

This article was presented at the 86th annual meeting of the American Thyroid Association, Denver, CO, September 21–25, 2016.

Author Disclosure Statement

The authors declare that no competing financial interests exist.

References

Burch

, Cooper

. 2015. Management of Graves disease: a review. JAMA, 314:2544–2554.

Bahn Chair

, Burch

, Cooper

, Garber

, Greenlee

, Klein

, Laurberg

, McDougall

, Montori

, Rivkees

, Ross

, Sosa

, Stan

; American Thyroid Association; American Association of Clinical Endocrinologists. 2011. Hyperthyroidism and other causes of thyrotoxicosis: management guidelines of the American Thyroid Association and American Association of Clinical Endocrinologists. Thyroid, 21:593–646.

Burch

, Burman

, Cooper DS 2012 A 2011. survey of clinical practice patterns in the management of Graves' disease. J Clin Endocrinol Metab, 97:4549–4558.

Hookham

, Collins

, Allahabadia

, Balasubramanian

. 2017. Variation in the use of definitive treatment options in the management of Graves' disease: a UK clinician survey. Postgrad Med J, 93:198–204.

Gauger

, Guinea

, Reeve

, Delbridge

. 1999. The spectrum of emergency admissions for thyroidectomy. Am J Emerg Med, 17:591–593.

Genovese

, Noureldine

, Gleeson

, Tufano

, Kandil

. 2013. What is the best definitive treatment for Graves' disease? A systematic review of the existing literature. Ann Surg Oncol, 20:660–667.

Liu

, Masterson

, Fish

, Jani

, Chatterjee

. 2015. Thyroid surgery for Graves' disease and Graves' ophthalmopathy. Cochrane Database Syst Rev. CD010576.

Ross

, Burch

, Cooper

, Greenlee

, Laurberg

, Maia

, Rivkees

, Samuels

, Sosa

, Stan

, Walter

. 2016. 2016 American Thyroid Association guidelines for diagnosis and management of hyperthyroidism and other causes of thyrotoxicosis. Thyroid, 26:1343–1421.

Brito

, Schilz

, Singh Ospina

, Rodriguez-Gutierrez

, Maraka

, Sangaralingham

, Montori

. 2016. Antithyroid drugs—the most common treatment for Graves' disease in the United States: a nationwide population-based study. Thyroid, 26:1144–1145.

10.

Prasai

, Nix

, Aye

, Atkin

, England

. 2013. Total thyroidectomy for safe and definitive management of Graves' disease. J Laryngol Otol, 127:681–684.

11.

Querat

, Germain

, Dumollard

, Estour

, Peoc'h

, Prades

. 2015. Surgical management of hyperthyroidism. Eur Ann Otorhinolaryngol Head Neck Dis, 132:63–66.

12.

Edafe

, Antakia

, Laskar

, Uttley

, Balasubramanian

. 2014. Systematic review and meta-analysis of predictors of post-thyroidectomy hypocalcaemia. Br J Surg, 101:307–320.

13.

Mok

, Oltmann

, Chen

, Sippel

, Schneider

. 2014. Identifying predictors of a difficult thyroidectomy. J Surg Res, 190:157–163.

14.

Bartalena

, Burch

, Burman

, Kahaly GJ 2016 A 2013. European survey of clinical practice patterns in the management of Graves' disease. Clin Endocrinol (Oxf), 84:115–120.

15.

Elfenbein

, Katz

, Schneider

, Chen

, Sippel

. 2016. Thyroidectomy for Graves' disease in children: indications and complications. J Pediatr Surg, 51:1680–1683.

16.

Yip

, Lang

, Lo

. 2012. Changing trend in surgical indication and management for Graves' disease. Am J Surg, 203:162–167.

17.

Phitayakorn

, Morales-Garcia

, Wanderer

, Lubitz

, Gaz

, Stephen

, Ehrenfeld

, Daniels

, Hodin

, Parangi

. 2013. Surgery for Graves' disease: a 25-year perspective. Am J Surg, 206:669–673.

18.

Catania

, Guaitoli

, Carbotta

, Bianchini

, Di Matteo

, Carbotta

, Nardi

, Fabiani

, Grani

, D'Andrea

, Fumarola

. 2013. Total thyroidectomy for Graves' disease treatment. Clin Ter, 164:193–196.

19.

Dehal

, Abbas

, Hussain

, Johna

. 2015. Risk factors for neck hematoma after thyroid or parathyroid surgery: ten-year analysis of the nationwide inpatient sample database. Perm J, 19:22–28.

20.

Campbell

, McCoy

, Shen

, Carty

, Lubitz

, Moalem

, Nehs

, Holm

, Greenblatt

, Press

, Feng

, Siperstein

, Mitmaker

, Benay

, Tabah

, Oltmann

, Chen

, Sippel

, Brekke

, Vriens

, Lodewijk

, Stephen

, Nagar

, Angelos

, Ghanem

, Prescott

, Zeiger

, Aragon Han

, Sturgeon

, Elaraj

, Nixon

, Patel

, Bayles

, Heneghan

, Ochieng

, Guerrero

, Ruan

. 2013. A multi-institutional international study of risk factors for hematoma after thyroidectomy. Surgery, 154:1283–1289; discussion 1289–1291.

21.

Oltmann

, Brekke

, Schneider

, Schaefer

, Chen

, Sippel

. 2015. Preventing postoperative hypocalcemia in patients with Graves disease: a prospective study. Ann Surg Oncol, 22:952–958.

22.

Pesce

, Shiue

, Tsai

, Umbricht

, Tufano

, Dackiw

, Kowalski

, Zeiger

. 2010. Postoperative hypocalcemia after thyroidectomy for Graves' disease. Thyroid, 20:1279–1283.

23.

Welch

, McHenry

. 2011. Total thyroidectomy: is morbidity higher for Graves' disease than nontoxic goiter?. J Surg Res, 170:96–99.

24.

Barczynski

, Konturek

, Hubalewska-Dydejczyk

, Golkowski

, Nowak

. 2012. Randomized clinical trial of bilateral subtotal thyroidectomy versus total thyroidectomy for Graves' disease with a 5-year follow-up. Br J Surg, 99:515–522.

25.

, Pearce

, Wong

, Burgess

, McAneny

, Rosen

. 2009. Treatment options for Graves disease: a cost-effectiveness analysis. J Am Coll Surg, 209:170–179.e1–2.