Risk of Permanent Hypoparathyroidism Requiring Calcitriol Therapy in a Population-Based Cohort of Adults Older Than 65 Undergoing Total Thyroidectomy for Graves' Disease

Abstract

Objective:

Total thyroidectomy for Graves' disease (GD) is associated with rapid treatment of hyperthyroidism and low recurrence rates. However, it carries the risk of surgical complications including permanent hypoparathyroidism, which contributes to long-term impaired quality of life. The objective of this study was to determine the incidence of permanent hypoparathyroidism requiring calcitriol therapy among a population-based cohort of older adults undergoing total thyroidectomy for GD in the United States.

Methods:

We performed a population-based cohort study using 100% Medicare claims from beneficiaries older than 65 years with GD who underwent total thyroidectomy from 2007 to 2017. We required continuous enrollment in Medicare Parts A, B, and D for 12 months before and after surgery to ensure access to comprehensive claims data. Patients were excluded if they had a preoperative diagnosis of thyroid cancer or were on long-term preoperative calcitriol. Our primary outcome was permanent hypoparathyroidism, which was identified based on persistent use of calcitriol between 6 and 12 months following thyroidectomy. We used multivariable logistic regression to identify characteristics associated with permanent hypoparathyroidism, including patient age, sex, race/ethnicity, neighborhood disadvantage, Charlson–Deyo Comorbidity Index, urban or rural residence, and frailty.

Results:

We identified 4650 patients who underwent total thyroidectomy for GD during the study period and met the inclusion criteria (mean age = 72.8 years [standard deviation = 5.5], 86% female, and 79% white). Among this surgical cohort, 104 (2.2% [95% confidence interval, CI = 1.8–2.7%]) patients developed permanent hypoparathyroidism requiring calcitriol therapy. Patients who developed permanent hypoparathyroidism were on average older (mean age 74.1 vs. 72.8 years) than those who did not develop permanent hypoparathyroidism (p = 0.04). On multivariable regression, older age was the only patient characteristic associated with permanent hypoparathyroidism (odds ratio age ≥76 years = 1.68 [CI = 1.13–2.51] compared with age 66–75 years).

Conclusions:

The risk of permanent hypoparathyroidism requiring calcitriol therapy among this national, U.S. population-based cohort of older adults with GD treated with total thyroidectomy was low, even when considering operations performed by a heterogeneous group of surgeons. These findings suggest that the risk of hypoparathyroidism should not be a deterrent to operative management for GD in older adults who are appropriate surgical candidates.

Introduction

Graves' disease (GD) is the most common cause of hyperthyroidism and contributes to impaired quality of life and long-term adverse cardiovascular outcomes.^1,2 Treatment options for GD include antithyroid drugs, radioactive iodine ablation, and thyroidectomy. Although the majority of patients with GD in the United States are treated with antithyroid drugs and radioactive iodine ablation,³ total thyroidectomy offers rapid resolution of hyperthyroidism with low recurrence rates and the potential for symptomatic relief from the removal of large goiters. In addition, the management of hyperthyroidism with thyroidectomy has been associated with a reduced risk of long-term adverse cardiovascular events when compared with radioactive iodine ablation, suggesting that there may be additional benefits of operative management for GD.^4–6 Despite these advantages, thyroidectomy is rarely used for the management of GD in older adults in the United States due, in part, to the concern for perioperative complications, including permanent hypoparathyroidism.^3,7

Estimates of the risk of permanent hypoparathyroidism following thyroidectomy overall, and for GD specifically, vary widely based on the population studied, clinical setting, and criteria used to define it. Historical estimates suggested that permanent hypoparathyroidism occurs in ∼1% of patients undergoing total thyroidectomy by a high-volume thyroid surgeon.^8,9 However, recent studies suggest that this complication is more common than originally suspected, with rates of persistent hypoparathyroidism more than 6 months after thyroidectomy for benign and malignant conditions ranging from 2.4% to 5% in studies involving high-volume surgeons^10,11 and approaching 15% and 20% in two recent population-based analyses from Europe.^12,13

Preoperative hyperthyroidism and GD specifically have been associated with an increased risk of transient and permanent hypoparathyroidism,¹⁴ raising concerns that the management of GD with total thyroidectomy will lead to unacceptable morbidity related to permanent hypoparathyroidism.¹⁵ However, there are limited population-based data from the United States to describe the risk of this complication. The absence of generalizable estimates of the risks of thyroidectomy among older adults with GD has limited the ability of clinicians to set appropriate expectations about outcomes of surgery and make informed treatment decisions with patients.

The aim of this study was to determine the risk of permanent hypoparathyroidism in a population-based cohort of older adults with GD undergoing total thyroidectomy and to identify patient or operative characteristics associated with this complication. We also sought to quantify the risk of other complications following thyroidectomy in this population, including neck hematoma and recurrent laryngeal nerve injury. We hypothesized that the risk of permanent hypoparathyroidism would be higher than historical estimates and patient or operative factors associated with perioperative complications could be identified to inform treatment decisions.

Methods

We performed a population-based cohort study of all Medicare beneficiaries who underwent total thyroidectomy with a preoperative diagnosis of GD using 100% Medicare fee-for-service claims from 2007 to 2017, including outpatient claims, carrier claims, Medicare Provider Analysis and Review (MedPAR) files, and Master Beneficiary Summary File (MBSF) files. Data were accessed via the Virtual Research Data Center housed on a secure Centers for Medicare and Medicaid Services server. The study was approved by the institutional review board at Stanford University (protocol No. 49938), and the need for informed consent was waived due to minimal study risk. This study was reported in accordance with STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines for observational research.¹⁶

Population

We identified a cohort of patients older than 65 years who underwent total thyroidectomy for GD from 2007 to 2017 based on the International Classification of Diseases, Ninth Revision (ICD-9) or Tenth Revision (ICD-10) procedure codes (ICD-9 064, 0652; ICD-10 0GTK0ZZ, 0GTK4ZZ), and Current Procedural Terminology (CPT) codes (60212, 60225, 60240, 60270, 60271, 60260) from all settings of claims and ICD-9 (242.00, 242.01) or ICD-10 (E05.00, E05.01) diagnosis codes for GD within 1 year before surgery (Fig. 1).

FIG. 1.

Participant flowchart.

Beneficiaries were excluded if they had: (1) a history of thyroid cancer based on ICD-9 (193) or ICD-10 (C73) codes; (2) a pharmacy claim for calcitriol within 1 year before thyroidectomy (not including prescriptions in the 3 months leading up to surgery); or (3) missing demographic information (e.g., age, sex, race/ethnicity). We also required beneficiaries to have continuous enrollment in Medicare Parts A, B, and D (without Part C “Medicare Advantage” enrollment) for at least 12 months before and after thyroidectomy to allow adequate time to identify preexisting medical comorbidities and capture prolonged postoperative calcitriol use.

Covariates

Patient characteristics included age, sex, and race/ethnicity. Comorbidity was assessed using the Charlson–Deyo Comorbidity Index.^17,18 Mailing zip code was used to assess each patient's socioeconomic status based on the Area Deprivation Index, a validated measure of neighborhood disadvantage in the United States,^19,20 and to construct urban versus rural inhabitance, as previously described.^21,22 A validated claims-based frailty index, grounded in the deficit accumulation model of frailty assessment, was used to categorize patients as robust, prefrail, mildly frail, or moderately to severely frail²³ because frailty has been associated with complications following thyroidectomy.^24,25

Patients who underwent concurrent parathyroidectomy were identified based on ICD-9 (0681, 0689, 0699) and ICD-10 (0GBLxxx, 0GBMxxx, 0GBNxxx, 0GBPxxx, 0GBQxxx, 0GBRxxx) procedure codes and CPT codes (60500, 60502, and 60505) from claims within the same month as thyroidectomy. Parathyroid autotransplantation was identified based on ICD-9 (0695) and ICD-10 (0GSR0ZZ, 0GSR4ZZ) procedure codes and CPT codes (60512) from claims within the same month as thyroidectomy.

Outcomes

The primary outcome was permanent hypoparathyroidism as evidenced by one or more Part D pharmacy claim(s) for calcitriol, which is the primary treatment for patients with hypoparathyroidism and is only available via prescription, between 6 and 12 months following total thyroidectomy. We also evaluated the percentage of patients with the primary outcome who had a concurrent diagnosis code for hypoparathyroidism (ICD-9 252.1; ICD-10 E20.8, E20.9) within one year following thyroidectomy. We assessed additional harms of total thyroidectomy for GD based on the occurrence of a neck hematoma and recurrent laryngeal nerve injury.

Neck hematoma was identified based on ICD-9 (998.1x) or ICD-10 (E36.01, E36.02, E89.810, E89.811) diagnosis codes on any category of claim within 14 days following thyroidectomy and was further classified as requiring operative evacuation based on ICD-9 (0602, 0609, 0692) or ICD-10 (0G9K0ZZ, 0GCK0ZZ, 0GJS0ZZ, 0WJ60ZZ, 0W9600Z, 0W960ZZ, 03LU0ZZ, 03LV0ZZ) procedure codes or CPT codes (35800, 35201) indicating an operation for hematoma evacuation during this time period. We also determined the occurrence of recurrent laryngeal nerve injury based on ICD-9 (478.3x) and ICD-10 (J38.x) diagnosis codes within 6 or 12 months following surgery or tracheostomy procedure codes (ICD-9 31.1, 31.2x; ICD-10 0B110F4, 0B110Z4, 0B113F4, 0B113Z4, 0B114F4, 0B114Z4) within 14 days of surgery.

Statistical analyses

Descriptive statistics were used to report the incidence of permanent hypoparathyroidism in the overall population and subgroups of interest with Poisson exact 95% confidence intervals [CIs]. Univariate comparisons were performed using chi-square test and Student's t-test. A multivariable logistic regression model was calculated to determine associations between patient and operative characteristics and the occurrence of permanent hypoparathyroidism. The significance level for all comparisons was p < 0.05, and tests were two-tailed. Analyses were performed using SAS statistical software, version 9.4 (SAS Institute, Inc., Cary, NC, USA).

Results

We identified 4650 Medicare beneficiaries with GD who met the study criteria and were treated with total thyroidectomy from January 1, 2007, to December 31, 2017 (Fig. 1). Patients had a mean age of 72.8 years (standard deviation = 5.5), 85.6% were female, and 78.8% were white (Table 1). A total of 583 (12.5%) patients were mildly or moderately to severely frail and 2657 (57.1%) had Charlson–Deyo Comorbidity Index score of ≥2. The majority of patients (55.9%) were from disadvantaged neighborhoods. A total of 2410 (51.8%) thyroidectomies were performed in an outpatient setting, 1439 (31.0%) were performed in an inpatient setting, and in 801 (17.2%), the clinical setting was unknown. Concurrent parathyroidectomy was performed in 411 (8.8%) thyroidectomies. Parathyroid autotransplantation was performed in 401 (8.6%) thyroid operations.

Table 1.

Characteristics of Medicare Beneficiaries with Graves' Disease Treated with Total Thyroidectomy (2007–2017)

Characteristic	Overall cohort (N = 4650)
	n	%
Sex
Female	3981	85.6
Male	669	14.4
Age group, years
66–75	3326	71.5
76 and older	1324	28.5
Race
White	3666	78.8
Black	730	15.7
Other	254	5.5
ADI group
Most advantaged	726	15.6
Slightly advantaged	1327	28.5
Slightly disadvantaged	1791	38.5
Most disadvantaged	806	17.3
Geography
Urban	3632	78.1
Rural	1018	21.9
Frailty
Robust	1522	32.7
Prefrail	2545	54.7
Mildly frail	403	8.7
Moderately to severely frail	180	3.9
Charlson–Deyo Comorbidity Index
0	964	20.7
1	1029	22.1
≥2	2657	57.1
Concurrent operations
Concurrent parathyroidectomy	411	8.8
Parathyroid autotransplantation	401	8.6

ADI, Area Deprivation Index.

Among this surgical cohort, 104 (2.2% [CI = 1.8–2.7%]) patients developed permanent hypoparathyroidism as evidenced by prolonged use of calcitriol. Only 31 (29.8%) of these 104 patients had claims with diagnosis codes indicating hypoparathyroidism. Patients who developed permanent hypoparathyroidism were more likely to be older (mean age 74.2 vs. 72.8 years) when compared with those who did not develop permanent hypoparathyroidism (p = 0.04). A total of 61 (1.8% [CI = 1.4–2.4%]) out of 3326 patients aged 66 to 75 years had permanent hypoparathyroidism, compared with 43 (3.3% [CI = 2.4–4.4%]) out of 1324 patients older than 75 years. On multivariable analysis, older age was the only patient characteristic associated with permanent hypoparathyroidism (odds ratio age ≥76 years = 1.68 [CI = 1.13–2.51] compared with age 66–75 years) (Table 2). Concurrent parathyroidectomy and parathyroid autotransplantation were not independently associated with permanent hypoparathyroidism.

Table 2.

Patient Characteristics Associated with Permanent Hypoparathyroidism Requiring Calcitriol Therapy Following Total Thyroidectomy for Graves' Disease Based on Multivariable Logistic Regression

Characteristic	Adjusted OR	95% CI
Sex
Female	Ref	—	—
Male	0.58	0.29	1.17
Age, years
66–75	Ref	—	—
76 and older	1.68	1.13	2.51
Race
White	Ref	—	—
Black	1.18	0.70	2.01
Other	1.28	0.58	2.84
ADI group
Most advantaged	Ref	—	—
Slightly advantaged	1.04	0.58	1.86
Slightly disadvantaged	0.66	0.36	1.21
Most disadvantaged	0.93	0.47	1.83
Geography
Urban	Ref	—	—
Rural	1.54	0.95	2.48
Frailty
Robust	Ref	—	—
Prefrail	1.11	0.68	1.82
Mildly frail	1.42	0.68	2.95
Moderately to severely frail	0.95	0.32	2.87
Charlson–Deyo Comorbidity Index
0	Ref	—	—
1	1.41	0.73	2.75
≥2	1.44	0.78	2.67
Concurrent operations
Concurrent parathyroidectomy	1.53	0.86	2.72
Parathyroid autotransplantation	1.48	0.82	2.69

CI, confidence interval; OR, odds ratio.

In terms of other perioperative complications of total thyroidectomy for GD, 97 (2.1% [CI = 1.7–2.5%]) patients had claims codes indicating a neck hematoma, with 42 (0.9% [CI = 0.7–1.2%]) patients requiring return to the operating room for hematoma evacuation within 14 days of thyroidectomy. Temporary or permanent recurrent laryngeal nerve injury occurred in 182 (3.9% [CI = 3.4–4.5%]) patients, with 72 (1.6% [CI = 1.2–1.9%]) patients having a claim code for recurrent laryngeal nerve injury 6–12 months following thyroidectomy. No patients underwent tracheostomy within 14 days of surgery.

Discussion

This population-based longitudinal study suggests that the incidence of permanent hypoparathyroidism requiring calcitriol therapy among older adults with GD who are treated with total thyroidectomy is low (2.2%). The only factor that was associated with an increased risk of permanent hypoparathyroidism on multivariable analysis was patient age older than 75 years. There was poor concordance between pharmacological treatment for hypoparathyroidism using calcitriol and claim codes indicating the diagnosis, suggesting that ICD-9 and ICD-10 diagnosis codes are not sensitive in identifying this complication. The incidences of neck hematoma requiring operative intervention and permanent recurrent laryngeal nerve injury were also low. Overall, these data suggest that total thyroidectomy is a viable treatment option for older adults with GD and should not be discounted due to concerns about the potential morbidity of this operation.

Our study identified a lower rate of hypoparathyroidism requiring calcitriol therapy than other recent population-based analyses of patients undergoing total thyroidectomy for benign or malignant conditions. A nationwide population-based study from Sweden recently showed that the risk of permanent hypoparathyroidism (defined based on calcitriol or calcium use more than 6 months postoperatively) after thyroid surgery for benign disease over a 10-year period was 12.5%, with independent risk factors including older age, parathyroid autotransplantation, surgery at lower volume centers (≤100 thyroidectomies per year), and female sex.¹² Looking at more than 4000 elective thyroidectomies with benign pathology performed in Hong Kong, Lui et al showed that 460 (11.2%) patients had permanent hypoparathyroidism based on pharmacy claims for persistent calcitriol and calcium use more than 6 months postoperatively.²⁶

Similarly, a study of all total and completion thyroidectomies performed at university hospital centers in the Netherlands in 2016 found persistent hypoparathyroidism (based on calcitriol use one year postoperatively) present in 14.5% of patients.¹³ Although some of the discrepancy in our estimate of permanent hypoparathyroidism requiring calcitriol therapy and those reported in these studies may be related to differences in the patient populations or the inclusion of calcium supplementation as an indicator of post-thyroidectomy hypoparathyroidism, it is notable that the rate of permanent hypoparathyroidism based on calcitriol use is so much lower in our cohort and provides reassurance that this is not a common complication of thyroidectomy for GD in our patient population.

Given thyroidectomy for GD in the United States is less common, it is possible that the lower rate of complications is related to selective referral of these patients to more specialized centers and surgeons with experience operating on patients with GD and other complex thyroid disorders. We do not have data on surgeon volume or practice setting to include in our analyses to test this hypothesis. However, it is reassuring that the low rate of permanent hypoparathyroidism requiring calcitriol therapy we identified is closer to that seen in randomized clinical trials involving patients with GD who had surgery at high-volume centers.^11,27 Our finding that less than half of patients with pharmacy claims indicating permanent hypoparathyroidism had claim diagnosis codes documenting this complication is consistent with underreporting of this complication in other studies and may explain the discrepancy between complication rates commonly quoted in clinical settings and those documented in research publications.¹²

The only patient or operative factor that we found was associated with higher adjusted odds of permanent hypoparathyroidism requiring calcitriol therapy was older age (76 years or older) at the time of surgery. Although other studies have shown an increased incidence of perioperative complications in general and hypoparathyroidism specifically in older patients undergoing thyroid surgery,^12,25 no prior investigations of patients with GD have focused on older adults or provided detailed information on this association. Given thyroidectomy is rarely chosen as the primary treatment strategy for GD in the United States,²⁸ especially among older patients, there is likely a higher threshold for surgical referral in this patient population. Therefore, it is possible that older patients referred for surgery have more severe GD (e.g., large and symptomatic goiters, poorly controlled hyperthyroidism), which may put them at higher risk of hypoparathyroidism and other complications.

Without granular clinical information on disease severity, we are limited in determining what may be driving this association. Future studies should focus on investigating the association of age with perioperative thyroidectomy complications accounting for severity of GD and relevant disease characteristics.

With these new data suggesting a low risk of permanent hypoparathyroidism requiring calcitriol therapy among older adults with GD treated with total thyroidectomy, further consideration should be given to initial operative management for patients who are likely to benefit. Thyroidectomy has the advantage of providing rapid control of hyperthyroidism, which is likely the most important factor to reduce the risk of long-term cardiovascular morbidity and mortality in patients with this condition.^4,29 In addition, surgery is associated with very low recurrence rates (<1% in contemporary studies)^3,6 and may be less likely to exacerbate Graves' orbitopathy, which is associated with poor quality of life.³⁰

Increased uptake of adjunct near-infrared autofluorescence or indocyanine green angiography to identify parathyroids with intact blood supply during thyroidectomy also hold promise to further reduce the incidence of perioperative hypoparathyroidism,^31,32 although relevant data on long-term outcomes are lacking. With recent studies documenting the safety of long-term antithyroid drug treatment for hyperthyroidism,^33,34 this will be an appealing and appropriate treatment strategy for older adults with GD who are frail or have comorbidities associated with increased perioperative risk. However, our findings suggest that the risk profile of total thyroidectomy for GD in geriatric patients in the United States is acceptable and that this treatment option should not be discounted in older adults with GD and goiters, orbitopathy, or other disease characteristics that make them likely to benefit from operative management.

Strengths of our study include that it was population-based and included a large number of older adults with GD who were treated with thyroidectomy and had more than one year of follow-up. The use of pharmacy claims to identify patients with permanent postoperative hypoparathyroidism avoids bias related to underreporting of complications by physicians and based on diagnosis codes. In addition, to the best of our knowledge, this is the first population-based assessment of post-thyroidectomy hypoparathyroidism in the United States. Our study has limitations related to the use of administrative claims for our analyses. Medicare data do not include laboratory results, so we are unable to confirm that patients with new calcitriol use for >6 months after thyroidectomy have permanent hypoparathyroidism based on calcium and parathyroid hormone testing, which is the gold standard for defining this condition.³⁵ We also cannot check for the biochemical diagnosis of GD, assess the severity of disease or whether patients were euthyroid at the time of surgery, or determine if GD was the only indication for thyroidectomy.

Similarly, we do not have information on the size or presence of an associated goiter or other patient characteristics that would contribute to the complexity of surgical intervention and risk of hypoparathyroidism. With the data available in Medicare claims, we were unable to determine the dose of calcitriol taken or whether over-the-counter calcium was used alone or in addition to calcitriol.

In addition, the assessment of calcitriol usage only in the first year following thyroidectomy may overestimate the rate of permanent hypoparathyroidism, given some patients will recover parathyroid function more than one year following thyroidectomy. The low incidence of our outcome may limit our power to assess for patient factors associated with permanent hypoparathyroidism. Finally, to ensure adequate follow-up of pharmacy claims, we restricted our cohort to patients with continuous enrollment in Medicare Parts A, B, and D for 12 months before and after thyroidectomy, which may affect the generalizability of our results.

Conclusions

This population-based analysis of older adults with GD undergoing total thyroidectomy in the United States suggests that the risks of permanent hypoparathyroidism requiring calcitriol therapy, in addition to neck hematoma and recurrent laryngeal nerve injury, are low and should not be deterrents to surgical treatment in appropriate candidates. Older adults with GD who may benefit from operative management should be counseled on these geriatric-specific risks and benefits of total thyroidectomy so that informed treatment decisions can be made.

Authors' Contributions

C.D.S.: Conceptualization, methodology, formal analysis, article preparation and editing, and funding acquisition. T.M.: Methodology, formal analysis, and article preparation and editing. R.M.C., D.T.L., E.A.M., J.C., and M.K.T.: Article preparation and editing. A.W.T.: Methodology, biostatistics support, and article preparation and editing. E.K.: Conceptualization, and article preparation and editing.

Author Disclosure Statement

The authors have no conflicts of interest relevant to this project. C.D.S. reported prior consulting for Virtual Incision Corporation. E.A.M. is co-founder of Equilibrate Therapeutics, LLC, a consultant for Abbvie, and has US20210361921A1 patent pending. T.M., R.M.C., D.T.L., J.C., M.K.T., A.W.T., and E.K. have no conflicts to disclose.

Funding Information

The authors acknowledge funding support from an American Thyroid Association research grant (Dr. Seib and Ms. Meng), and the National Institutes of Health, National Institute on Aging by award K76AG068526 (Dr. Seib). Funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the article; or decision to submit the article for publication.