Prevalence and Incidence of Hypoparathyroidism in the United States Using a Large Claims Database

Authors


ABSTRACT

Hypoparathyroidism is a rare endocrine disorder whose incidence and prevalence have not been well defined. This study aimed to 1) estimate the number of insured adult patients with hypoparathyroidism in the United States and 2) obtain physician assessment of disease severity and chronicity. Prevalence was estimated through calculation of diagnoses of hypoparathyroidism in a large proprietary health plan claims database over a 12-month period from October 2007 through September 2008 and projected to the US insured population. Incidence was also calculated from the same database by determining the proportion of total neck surgeries resulting in either transient (≤6 months) or chronic (>6 months) hypoparathyroidism. A physician primary market research study was conducted to assess disease severity and determine the percentage of new nonsurgical patients with hypoparathyroidism. Incidence data were entered into an epidemiologic model to derive an estimate of prevalence. The diagnosis-based prevalence approach estimated 58,793 insured patients with chronic hypoparathyroidism in the United States. The surgical-based incidence approach yielded 117,342 relevant surgeries resulting in 8901 cases over 12 months. Overall, 7.6% of surgeries resulted in hypoparathyroidism (75% transient, 25% chronic). The prevalence of chronic hypoparathyroidism among insured patients included in the surgical database was estimated to be 58,625. The physician survey found that 75% of cases treated over the past 12 months were reported due to surgery and, among all thyroidectomies and parathyroidectomies and neck dissections performed in a year, 26% resulted in transient hypoparathyroidism and 5% progressed to a chronic state. In conclusion, the two claims-based methods yielded similar estimates of the number of insured patients with chronic hypoparathyroidism in the United States (∼58,700). The physician survey was consistent with those calculations and confirmed the burden imposed by hypoparathyroidism. © 2013 American Society for Bone and Mineral Research.

Introduction

Hypoparathyroidism is an endocrine disorder in which the parathyroid glands are either removed or produce insufficient parathyroid hormone (PTH). The resulting decrease in PTH, which controls calcium, phosphate, and vitamin D levels in the blood and bone, causes derangements in mineral homeostasis, including hypocalcemia and hyperphosphatemia.[1, 2] Hypoparathyroidism is most often acquired, caused by head and neck surgery during which the parathyroid glands are either intentionally or inadvertently removed or damaged; it can also be congenital, idiopathic, or the result of an autoimmune disorder (eg, pernicious anemia, Addison's disease).[1, 3-6]

Although a comprehensive review of hypoparathyroidism was recently published,[3] there are few prevalence estimates for this disorder. Current estimates in the United States focus on surgical cases, but less is known about patients who develop the disease through other causes.[4] A population-based study, reported only in a conference presentation to date, identified a prevalence of 37 cases per 100,000 residents, with the vast majority (78%) of cases caused by surgery and a small proportion being familial (7%) or idiopathic (6%).[7] Approximately 10% to 12% of patients undergoing total thyroidectomy have resulting hypoparathyroidism, and most of those cases are transient, resolving within 6 months.[8, 9] Only a very small proportion of postsurgical cases develop chronic disease.[8-10] A recent review of the treatment guidelines reported a range between 0.5% and 6.6% of thyroidectomies resulting in transient hypoparathyroidism, although the rate was lower at centers with extensive experience in endocrine surgery.[11] The extent of surgery and experience of the surgeon also affected the rate of postsurgical hypoparathyroidism.[11]

Formal guidelines have not yet been developed for the management of hypoparathyroidism,[1] and, to date, no PTH replacement therapy has been approved for use in this condition. Currently, long-term management includes large amounts of calcium and vitamin D to relieve the symptoms of hypocalcemia.[12] These approaches, however, can cause hypercalcemia and worsen the hypercalciuria and hyperphosphatemia.[3, 13] Both under- and overtreatment can lead to unintended outcomes that can be irreversible.

In this context of varied causes, wide-ranging duration of illness, and suboptimal management, we conducted this study to estimate the incidence and prevalence of hypoparathyroidism in the insured population in the United States and to assess its impact in clinical practice. Through these estimates, we aim to better understand the burden of this disease and its unmet needs.

Materials and Methods

This study estimated the prevalence of hypoparathyroidism, specifically the number of insured patients in the United States with hypoparathyroidism. The study used two epidemiologic approaches, one diagnosis based and the other surgical based, to enhance validity. In addition, a sample of physicians was surveyed using primary market research methods to gain insights into the impact of hypoparathyroidism and approaches to its management in current clinical practice.

For the two epidemiologic approaches, cases of hypoparathyroidism were identified using health insurance data from the IMS LifeLink Health Plan Claims Database (formerly the PharmMetrics database; IMS Health, Danbury, CT, USA; http://www.imshealth.com), which has been used for epidemiologic purposes in a variety of disease states.[14-17] This large proprietary database comprises fully adjudicated medical and pharmaceutical claims. At the time the study was conducted, it included longitudinal data for nearly 77 million unique patients from 75 health plans across the United States.

Enrollees represented in the database include employer-sponsored plans, individuals purchasing coverage in the marketplace, and government-sponsored but commercially administered Medicaid and Medicare plans. The database is representative of the commercially insured patient population on a variety of demographic measures, including geographic region, age, sex, and health plan type. Only health plans that submit data for all members are included in the database to ensure complete data capture and representative samples. All data are HIPAA compliant to protect patient privacy.

Although the average duration of member enrollment is approximately 2 years, this analysis was limited to patients with a minimum of 3 years' enrollment. This study drew from a sample of approximately 7.8 million patients participating in 23 health plans. Given the proprietary nature of the database and patient enrollment across multiple plans, it was not possible to conduct chart reviews to verify diagnoses based on billing code data. However, all data contributions in the database are subjected to a series of quality checks to ensure a standardized format and to minimize coding errors.

Diagnosis-based prevalence approach

Diagnosis-based prevalence of hypoparathyroidism was identified using patient data with the diagnostic code for hypoparathyroidism between October 2007 and September 2008. In addition, because hypoparathyroidism claims may not surface until after a contributing procedure is conducted, and even then may not be recorded in the medical chart, diagnoses coded as hypocalcemia were also examined for the same 1-year period. Within that group, we then looked for related conditions or prior surgeries reported over a 3-year period to capture data for 1 year before and 1 year after the 1-year review period that would indicate hypoparathyroidism as the underlying cause of the hypocalcemia diagnosis. These indicators included parathyroidectomy; total or partial thyroidectomy; neck dissection; radiation; and conditions such as Addison's disease, polyglandular syndrome, DiGeorge syndrome, hemochromatosis, Wilson's disease, magnesium metabolism disorder, pernicious anemia, and thalassemia. Any patients counted in the group with the diagnostic code for hypoparathyroidism were removed from the group with the code for hypocalcemia to avoid double counting. This number was then projected up to the total US insured population to obtain an estimate of prevalence. Patients whose hypoparathyroidism resolved within 6 months after a surgery were considered to have transient disease, whereas those whose hypoparathyroidism did not resolve within 6 months were considered to have chronic disease. Recently diagnosed patients with nonsurgical etiologies were considered to have chronic hypoparathyroidism.

Surgical incidence approach

Surgical incidence of hypoparathyroidism was determined by counting the total number of patients who underwent parathyroidectomy, complete or partial thyroidectomy, and neck dissection during the same 1-year period (October 2007 to September 2008). The proportion of surgeries resulting in hypoparathyroidism was determined by identifying patients with a recorded diagnosis of hypoparathyroidism or hypocalcemia after the surgery. To differentiate between chronic and transient hypoparathyroidism, we looked forward 1 year from the surgery to determine the proportion of hypoparathyroidism diagnoses that were still recorded for the patient 6 months postsurgery, the period of time in which transient postsurgical hypoparathyroidism is considered to typically resolve.[3] Hypoparathyroidism that did not resolve within 6 months after a surgery was categorized as chronic.

The total number of incident surgical cases was then entered into an epidemiologic model to estimate the prevalence of hypoparathyroidism in the insured US population, based on US census data and the distribution of surgical versus nonsurgical etiologies gathered from the primary market research analysis (methods described below). Model assumptions included an average onset age of 43 years, life span of 77 years for patients without chronic kidney disease, and constant incidence over time.

Primary market research with physicians

In recognition of the growing interest in primary market research to help evaluate the management of various conditions in real-world clinical settings, a quantitative online survey was conducted in August 2010 to supplement the epidemiologic assessments. Participants, who were paid for completing the survey, were randomly selected from an online panel assembled and managed by Toluna Group Ltd. (http://us.toluna-group.com) that comprises 100,000 physicians across all specialties.

A total of 290 practicing physicians, including endocrinologists, surgeons, nephrologists, and primary care physicians, participated. A 30-minute quantitative survey was administered over the Internet and included questions about diagnosis, disease severity, patient management, and current treatments. Survey participants were required to have treated at least five cases of adult hypoparathyroidism in the past year, except for primary care physicians who were required to have treated at least one patient. Participants were also required to have practiced medicine for at least 2 years with at least 75% of their time spent in clinical practice. Among other questions, participants were also asked to categorize their hypoparathyroidism patients by disease state (chronic, transient), severity (mild, moderate, severe), and etiology (surgical, nonsurgical).

Results

Of the hypoparathyroidism patients identified in the database, 75% were female and 74% were aged 45 years and older (Table 1). Excluding the pediatric patients (ie, <18 years of age), 20.1% were aged 45 to 54, 18.7% were 55 to 64, and 34.3% were 65 and older.

Table 1. Demographics of Patients Diagnosed With Hypoparathyroidism in the Diagnosis-Based Database
CharacteristicsNo. of index patientsPercent of index patients
Total (transient and chronic combined)65,325100
Sex  
Female48,98575.0
Male16,34025.0
Age (years)  
<1819793.0
18–34663610.2
35–44833312.8
45–5411,90818.2
55–6412,41019.0
65+24,06036.8

Using the diagnosis-based prevalence approach, we identified 48,674 patients with a hypoparathyroidism diagnosis. Additionally, there were 116,440 patients with a diagnosis of hypocalcemia, of whom 16,651 also had a concomitant diagnosis code that, together, reasonably suggested hypoparathyroidism (Table 2). Thus, this diagnosis-based approach yielded an estimated overall prevalence of 65,325 insured patients living in the United States with hypoparathyroidism based on data from October 2007 to September 2008 (Fig. 1). Of these patients, 58,793 (90%) were categorized as chronic.

Table 2. Concomitant Indicators of Hypoparathyroidism Among Patients With Hypocalcemia in the Diagnosis-Based Database
 No. of patients with diagnosis of hypocalcemiaaPercent
  • aPre-index and post-index refer to the 1 year before and 1 year after the 1-year review period spanning October 2007 to September 2008.
  • bSome patients had >1 concomitant condition.
Hypocalcemia with concomitant indicators of possible hypoparathyroidism116,440100
16,65114.3
 Counts of concomitant indicators of hypoparathyroidismbPercent
Magnesium metabolism disorder10,33160.5
Radiation361121.2
Pernicious anemia238613.9
Hemochromatosis3782.2
Thalassemia1851.1
Wilson's disease970.05
DiGeorge syndrome750.04
Figure 1.

Diagnosis-based prevalence approach using claims data with the diagnostic code for hypoparathyroidism between October 2007 and September 2008. Cases that did not resolve within 6 months were categorized as chronic. hypoPARA = hypoparathyroidism.

Using the surgical incidence approach, we identified 117,342 relevant surgeries resulting in 8901 cases of hypoparathyroidism over a 12-month period. Overall, 7.6% of surgeries resulted in hypoparathyroidism, of which 75% were categorized as transient and 25% as chronic. The largest percentage of cases, 38%, occurred after a total thyroidectomy and 21% of cases after a parathyroidectomy. Partial thyroidectomy accounted for 9% of cases, and other neck surgery accounted for 5% of the cases. When the surgical incidence data were entered into the epidemiologic model, the overall prevalence of chronic hypoparathyroidism among insured patients was estimated to be 58,625 (Fig. 2).

Figure 2.

Surgical-based incidence approach determined by counting the total number of patients who underwent parathyroidectomy, complete or partial thyroidectomy, and neck dissection between October 2007 and September 2008 and had a recorded diagnosis of hypoparathyroidism or hypocalcemia after the surgery. Cases that did not resolve within 6 months were categorized as chronic. CAGR = compound annual growth rate; hypoPARA = hypoparathyroidism.

The market research portion of this study indicated that an estimated 73% of cases managed by survey respondents were caused by surgery, mainly partial and total thyroidectomies (47%), parathyroidectomies (21%), and neck dissection (5%). Overall, the sampled physicians estimated that 26% of their patients who had undergone total thyroidectomies or parathyroidectomies developed hypoparathyroidism; 5% of these progressed to a chronic state. Most respondents characterized hypoparathyroidism as chronic if it lasts more than about 6 months. Overall, they categorized 43% of these chronic cases as mild, 39% as moderate, and 18% as severe. Although the survey did not specify any definitions for severity, the respondents indicated that about 13% of their patients had required intravenous calcium treatments at some point in time.

Responses were also assessed to determine patterns of care by specialty. Endocrinologists reported treating more hypoparathyroidism patients compared with other specialties: estimated n = 42/year for endocrinologists, n = 21/year for surgeons, n = 34/year for nephrologists, and n = 13/year for primary care physicians. Compared with the other specialties, primary care physicians treat fewer hypoparathyroidism patients, most of whom (63%) have chronic disease. More than half (59%) of the primary care physicians surveyed do not manage any patients with chronic hypoparathyroidism on their own.

Among sampled physicians, care was often provided jointly between specialists (ie, endocrinologists, nephrologists, surgeons) and primary care physicians, although endocrinologists, on average, reported managing 72% of their chronic patients on their own and other specialties reported referring to an endocrinologist when a patient did not respond to treatment. Surgeons, who manage slightly more transient than chronic cases of hypoparathyroidism (54% and 46%, respectively), reported sharing their care jointly with another physician about 50% of the time. Nephrologists reported treating more cases of chronic than transient hypoparathyroidism (62% and 38%, respectively); they often jointly manage the illness with a primary care physician or an endocrinologist.

Across all specialties, about two-thirds of physicians surveyed (n = 174, 66%) reported that they had used the diagnosis code for hypocalcemia rather than hypoparathyroidism at some time. Of this group, 64% (n = 115) reported later changing the code to the correct hypoparathyroidism code.

When asked to describe the categories of disease, the majority of physicians responded with descriptions of symptoms for each category. When asked to describe mild hypoparathyroidism, the top response was “patient with mild/minimal symptoms” and moderate cases were identified as patients having moderate symptoms. When named specifically, these symptoms included numbness and tingling. Severe cases were characterized by seizures and severe symptoms, such as severe hypocalcemic episodes accompanied by severe tetany, which is often associated with emergency room visits and hospitalization, as well as long-term issues with kidney damage.

When asked to rank the degree of unmet need in the treatment of hypoparathyroidism on a scale of 1 to 7, 69% of physicians across all specialties indicated a high level of unmet need. Overall, respondents ranked treating the underlying cause as the most important unmet need, followed by improved patient compliance, relieving symptoms, reducing the amount of calcium supplements, and improving patient quality of life. Across all specialties, physicians indicated that the primary goal of treatment was to “alleviate symptoms.”

Discussion

This study used two methods to estimate the prevalence of hypoparathyroidism among the insured US adult population. The diagnosis-based prevalence approach resulted in an estimate of 58,793 patients with chronic hypoparathyroidism, and the surgical incidence approach resulted in an estimate of 58,625 chronic patients. Surgical incidence varied by procedure performed, with most hypoparathyroidism resulting from total thyroidectomies followed by parathyroidectomies.

This study purposely used two different methodologies because of the challenges of producing reliable estimates. Deriving close estimates from two complementary methods provides more reliable prevalence estimates than would have been obtained from just one or the other, allowing stronger conclusions to be drawn. Nevertheless, this study had several limitations. One of these is that, as far as we are aware, no study has been carried out to cross-validate the use of this database by other databases, although this insurance database has been used in epidemiologic assessments of other diseases.[14-17] Another is that there is uncertainty around the generalizability of the database to the general US population. Our estimates do not account for the uninsured population, in whom the prevalence pattern may differ.

Another limitation was the difficulty of ascertaining the proportion of patients with a surgical etiology who go on to develop chronic hypoparathyroidism compared with those whose condition resolves over time. Health insurance claims do not always include a diagnosis code for hypoparathyroidism even when it may be present. The primary market research portion of this study confirmed that the hypocalcemia code was often recorded instead. Despite examining complementary codes and including them in the analysis where appropriate, a claims approach would inevitably exclude some cases of hypoparathyroidism. Conversely, the concomitant conditions we identified as potential indicators of hypoparathyroidism, such as radiation, magnesium metabolism disorder, and pernicious anemia, are not bona fide markers for the disorder. However, because these indicators were considered only in the context of patients who had a diagnosis code of hypocalcemia, we felt they provided a reasonably reliable indicator for estimating prevalence.

Another limitation is that the average enrollment of 2 years for patients included in this database may not have provided an adequate window of observation for the resolution of transient postsurgical hypoparathyroidism. To avoid this, patients were included in the analysis only if they had a minimum of 3 years' enrollment. Although most studies define hypoparathyroidism as permanent if there is still insufficient PTH to maintain normocalcemia 6 months after surgery,[2] it is possible that this methodological approach may have included some patients who had been left on replacement medications without appropriate evaluation of any recovery of parathyroid gland function. Thus, we may have overestimated the number of patients with chronic versus transient disease. At the same time, the surgical-based approach may have underestimated the percentage of patients with chronic disease because physicians may not continue to record the hypoparathyroidism diagnosis beyond the 6-month cutoff.

Although the primary research confirmed the patterns of incidence and prevalence of hypoparathyroidism revealed by the database analyses, its overall scientific validity may be viewed with skepticism given the nature of commercial market research physician panels. Although the 300 participants may reflect the demographic and practice characteristics of the larger group of 100,000 physicians, it is not known whether these physicians, who self-identify to participate, accurately represent the clinicians who typically care for this population of patients. Furthermore, such panels are subject to selection bias and participants may be prone to overstate the number of patients they manage, either overall or in a given disease category, because there is a financial incentive to participate.

Nevertheless, the results of the market research were consistent with the database analyses and the peer-reviewed literature.[1, 3, 18] One point of particular interest from this study was that the physicians reported a high rate of initial miscoding of the diagnosis of hypoparathyroidism. It is also interesting to note the percentage of specialists, particularly nephrologists, who share the care of hypoparathyroidism patients with other clinicians. Indeed, as Mitchell and colleagues reported recently, renal abnormalities are particularly common in patients with hypoparathyroidism: They found that the rates of stage 3 or higher chronic kidney disease are 2- to 17-fold higher in this population compared with age-appropriate norms.[18]

It should also be noted that, although other studies of the prevalence of hypoparathyroidism in the literature are scarce, our estimates are in line with those of other investigators. For example, a study in Germany identified at least 1200 patients developing chronic postoperative hypoparathyroidism per year.[19] Another study conducted with the population-based Rochester Epidemiology Project identified a prevalence of hypoparathyroidism of 37 per 100,000, with most cases being caused by surgery.[7] Other published estimates of hypoparathyroidism or hypocalcemia after surgery ranged from 10% to 30%.[8, 9, 20, 21] Most surgically caused hypoparathyroidism resolved within 6 months of surgery, but estimates from the literature suggest 0.5% to 6.6% of surgeries resulted in chronic disease, sometimes identified as hypocalcemia.[8, 20, 22] Results from this study were comparable with these earlier epidemiologic studies, with 7.6% of surgeries resulting in hypoparathyroidism, of which 75% of cases were deemed transient and 25% chronic.

One unexpected finding from the assessment of the surgical incidence of hypoparathyroidism was that only 5% of cases were attributed to neck surgeries other than total or partial thyroidectomies and parathyroidectomies. It is possible that this is an underestimate, given that more extensive neck surgeries (eg, for head and neck cancer, goiter) might be expected to be a more frequent cause of surgical hypoparathyroidism. Nevertheless, a similarly low incidence (1.7%) was recently reported in an evaluation of complications of central neck dissection in patients with papillary thyroid carcinoma.[22] Although this study examined the types of surgeries leading to postoperative hypoparathyroidism, it was beyond the scope of the study to assess the skill of the providers. More information is needed about the incidence of postsurgical hypoparathyroidism, particularly comparing the expertise of surgical centers and the surgical methods used to understand how they affect the incidence of hypoparathyroidism.[3, 20]

In addition to helping elucidate the prevalence of hypoparathyroidism, this study also helps define the burden of disease. It underscores the importance of increasing physician awareness of best practices in the management of hypoparathyroidism, including efforts to reduce large doses of oral calcium and active vitamin D while targeting a serum albumin-corrected total calcium level near the lower limit of normal and maintaining 24-hour urine calcium within the normal range and a calcium-phosphate product below 55.There are long-term complications associated with these agents, particularly owing to formation of calcium–phosphate salts in soft tissues, such as the kidney, lens, and basal ganglia.[1] The kidney is particularly vulnerable because the filtered load of calcium increases as a direct result of increases in the serum calcium level.[1] In the absence of PTH to promote renal calcium reabsorption, there is an increased risk of hypercalciuria, which may lead to nephrocalcinosis, nephrolithiasis, and reduced renal function.[1, 3, 18]

Conclusion

Epidemiology estimates of hypoparathyroidism in the literature are scarce. In this study, two different methods yielded similar estimates of the total number of insured hypoparathyroidism patients in the United States (∼59,000 for chronic hypoparathyroidism and ∼65,000 for transient and chronic combined). This is important for determining the burden of disease and potential impact of new hypoparathyroidism treatments. When extrapolated to the entire US population, accounting for the 15.4% estimated to be uninsured at the time this study was undertaken, the total number of individuals with either transient or chronic hypoparathyroidism is estimated at about 77,000. Because of its secondary nature and the number of related illnesses that may be coded instead of hypoparathyroidism, using claims data produces one approach for a prevalence estimate. This study differentiated between surgical and nonsurgical, transient and chronic, and three levels of severity of hypoparathyroidism and found that chronic, nonsurgical patients are the most severe and most in need of new treatments. These estimates, combined with physician market research, provide valuable data about the prevalence of hypoparathyroidism, its causes and characteristics, and the severity of disease. This study adds important evidence about the burden of hypoparathyroidism and the potential impact of new treatments.

Disclosures

JP, KJ, and AR are employees of IMS Health. HL is an employee of NPS Pharmaceuticals, Inc.

Acknowledgments

The research was funded by NPS Pharmaceuticals, Inc.

Margot Embree Fisher of NPS Pharmaceuticals, Inc., provided editorial assistance.

Authors' roles: All authors contributed to the conception and design of the study. All authors contributed to the acquisition, analysis, and interpretation of the data. All authors participated in drafting or critically revising the manuscript, and all authors approved the final version of the manuscript for submission.

Ancillary