There has been much discussion regarding appropriate management of patients in the era of widespread screening, as the majority of patients are asymptomatic at diagnosis [27, 35]. Certainly, definitive therapy for PHPT is surgical [30, 36]. There is evidence that even in patients who continue to be asymptomatic for years of active monitoring, decrements in overall health are observed by 15 years following diagnosis [15]. As an example, loss of bone mineral density has been observed even in patients without overt symptoms [20, 23]. In an analysis of an employer claims database and the Medicare Standard Analytic files, costs related to osteoporosis and the consequent non-vertebral fractures are not only strains on the health care system but also a significant cause of lost productivity [37]. Additionally, quality of life after parathyroidectomy has been shown to substantially improve, even among patients initially thought to be asymptomatic by current diagnostic guidelines, with corresponding improvement in workplace absenteeism [38–40]. Hence, definitive treatment, when indicated and patient-appropriate, appears to enable significant cost savings over time.

Using our MarketScan cohort, we studied the current status of PHPT therapy in the USA. Surgical management, as indicated by claims for CPT codes 60500, 60502, 60505, and 60512, was documented for 24.2 % (N = 7687) within a year of the date in which the first diagnostic claim for PHPT was made. In this subset of PHPT patients, the median interval from first documentation of PHPT in the claims database to parathyroid surgery was 44 days, with 83.5 % (N = 6418) of procedures performed in an outpatient setting. As such, most individuals had a length of stay less than 24 h. A search through outpatient pharmaceutical claims was undertaken to attempt to identify symptomatic patients (i.e., with indication for surgical intervention) who were otherwise unfit for surgery. No statistically significant differences in use of calcimimetics, bisphosphonates, or other medications used for PHPT treatment were found between nonsurgical and surgical patients. Specifically, cinacalcet hydrochloride was used by 0.3 % of nonsurgical patients and 0.5 % of surgical patients (p = 0.079). Alendronate sodium was used by 3.8 and 3.5 %, respectively (p = 0.325). Zoledronic acid was used by 0.1 and 0 %, respectively (p = 0.094). No other medication comparison reached statistical significance when comparing nonsurgical PHPT patients with PHPT patients who underwent parathyroidectomy. This finding underscores the emphasis placed on surgical management, as well as the advances in technique that have made parathyroidectomy accessible to a more morbid population, making pharmacologic nonsurgical management an uncommon practice in PHPT patients with a surgical indication.

Several analyses comparing costs related to management of PHPT have substantiated the economic advantage of parathyroidectomy relative to medical observation and therapy. When considering a short time frame, surveillance with or without pharmacologic treatment appears to be less costly then surgical intervention given the greater up-front expense of surgery. However, because PHPT is a lifelong disorder without surgery, the long-term costs of medical therapy and follow-up make parathyroidectomy the more cost-effective option when considering a longer time frame. It deserves mention that financial considerations are not only dependent on time frame but also on perspective, a nuance recognized particularly among those familiar with cost-effectiveness analyses (CEA) [41]. Discrete perspectives include, but are not limited to, those of the patient, the provider, the third-party payer, and society in general. While these perspectives may not all be mutually exclusive, they significantly change the costs considered in the management of a disease process.

Despite variation in the design of contemporary studies addressing costs of PHPT, surgical intervention is consistently shown to be more cost-effective than nonoperative management, even for asymptomatic patients [15, 16, 21]. Sejean et al. [24] analyzed the cost-effectiveness of three surgical approaches, as well as nonoperative surveillance, for a hypothetical 55-year-old asymptomatic female, using a health care delivery perspective. In their Markov model, three different surgical strategies and a strategy of medical management were compared over a lifetime horizon with a 1-month cycle. Across a reasonable range of associated costs, surgery remained more cost-effective than monitoring, except when there was no decrement in a medically managed patient’s quality of life [16, 24]. Zanocco et al. [15] found that in a hypothetical 60-year-old asymptomatic patient fit for surgery, using a third-party payer perspective, the incremental cost-effectiveness ratio (ICER) for parathyroidectomy was $4778 (2005 USD) per quality-adjusted life-year (QALY) gained, well below the commonly used threshold of $50,000–100,000 USD per QALY [16, 41, 42]. Pharmacologic therapy was not cost-effective, with an ICER significantly greater than $50,000 (2005 USD) per QALY, when calculating costs over the lifetime of an asymptomatic patient. Therefore, in an asymptomatic population, given a life expectancy greater than 5 years, even without costs of medications, annual follow-up may be less costly but is less effective than surgical treatment [13, 15, 24].

Current recommendations suggest that patients younger than 50 years of age, regardless of presence of symptoms, should always be referred for surgery on the basis of their expected remaining lifespan, provided they are otherwise fit for surgery [36, 43]. Zanocco and Sturgeon [13] compared cost-effectiveness of parathyroidectomy, observation, and pharmacologic management in an asymptomatic patient, varying the patient life expectancy from 0.5 to 75 years. Using a third-party payer perspective, they found that observation was preferable to inpatient parathyroidectomy if life expectancy was less than 6 years. Otherwise, parathyroidectomy should be pursued as the most cost-effective option. This threshold changed, however, when comparing surgery in an outpatient setting to observation and pharmacologic therapy. Observation was more cost-effective than parathyroidectomy only if life-expectancy was less than 4.5 years. Outpatient surgery became the preferable strategy with anticipated life-expectancy greater than 4.5 years. Pharmacologic therapy, with an ICER ranging from $2.5 to $10.5 million (2005 USD), was dominated by any other strategy [13].

Therefore, in an asymptomatic patient, the available evidence suggests that observation is less costly but less effective, and thereby less cost-effective, than parathyroidectomy. This changes only for patients with an anticipated life expectancy less than 4.5 years since this was the threshold determined for surgery performed in an outpatient setting, which is the most common practice currently [44–46]. Pharmacologic therapy without surgery is less effective and more costly for patients with asymptomatic PHPT.

In patients with symptoms who are fit for surgery, parathyroidectomy is invariably indicated and cost-effective. While there is no debate regarding use of surgical intervention in this patient population, there is considerable variation in surgical approach and use of preoperative localization studies, with a correspondingly large variation in studies addressing costs associated with these practices [14, 16, 17, 24, 47–51].

A number of surgical approaches to parathyroidectomy are employed depending on disease localization, confidence in preoperative imaging, and surgeon preference. Bilateral, or four-gland, neck exploration (BNE) was standard of care prior to the implementation of many of the now-common localization studies and minimally invasive surgical techniques. With these newer technologies has come a shift toward limited exploration (LE), which encompasses one-gland focused exploration, two-gland unilateral exploration (UNE) , and minimally invasive parathyroidectomy (MIP) approaches [44, 52]. A recent survey of surgeons performing parathyroidectomies revealed that only 10 % are still using BNE as their initial approach [44]. The practice of initial BNE is supported by its 95 % success rate consistently observed in the hands of experienced parathyroid surgeons, thereby limiting costs associated with recurrence or persistence of disease [11, 49, 53]. However, initial BNE without preoperative localization studies has greater probability of incurring costs related to complications, given the extent of surgical exposure placing more anatomical structures at risk of injury. Additionally, LE tends to have fewer costs associated with operating room time and postoperative length of stay, as many patients undergoing LE are able to avoid inpatient admission [14, 44]. Ultimately, the arguments pertaining to surgical approach and need for localization studies stem from concern regarding occult multi-glandular disease and failure of surgery to achieve a postoperative normo-hormonal and normo-calcemic state [11, 54].

Review of cost-effectiveness analyses studying the surgical approaches to parathyroidectomy reveals a lack of consistency in findings regarding the most successful, and therefore the most cost-effective, technique. Using a third-party payer perspective and probabilities based on success and complication rates of a cohort of patients at a tertiary care hospital, Baliski et al. [47] performed a decision analysis of UNE with preoperative localization sestamibi scan, BNE without preoperative localization, and MIP with preoperative localization scan, comparing cost-effectiveness in terms of avoiding complications. They found that UNE was more costly and less effective than BNE with an ICER of $5065 (2006 Canadian) per complication avoided. They found that MIP was slightly more effective than BNE in terms of avoiding complications but cost nearly $30,000 (2006 Canadian) per complication avoided, a sum they deemed prohibitive. BNE remained the most cost-effective option up to a societal willingness-to-pay threshold of $5000 (2006 Canadian) per complication avoided [47]. Because LE necessitates preoperative localization imaging and its associated costs, Baliski et al. found that BNE is the most cost-effective surgical approach when effectiveness is measured in terms of avoidance of surgical complications. However, the validity of this conclusion for a US population is uncertain since the authors note that a short time horizon was used, with probabilities based on small sample sizes, and complications only including symptomatic hypocalcemia and paresthesia . Hence, questions of cost-effectiveness among surgical approaches for PHPT remained unresolved without further corroborative studies.

Using an asymptomatic PHPT patient population and a health care delivery perspective, Sejean et al. [24] also compared the cost-effectiveness of three surgical approaches, as well as nonoperative surveillance as previously noted. In this Markov model, comparisons were made between BNE under general anesthesia without preoperative localization study performed, UNE under local anesthesia with preoperative ultrasound and sestamibi scan in addition to intraoperative PTH measurement, and video-assisted endoscopic parathyroidectomy (VAP) under general anesthesia with preoperative ultrasound and sestamibi scan in addition to intraoperative PTH measurement. Their model measured effectiveness in terms of QALYs, using a 1-month cycle length and a lifetime horizon. With these parameters, they found that a strategy of VAP was slightly more cost-effective than both UNE and BNE with an ICER of €17,250 (2002 Euros) per QALY. The ICER of UNE relative to BNE was €2688 (2002 Euros) per QALY. However, UNE became more cost-effective than VAP when the patient was older than 71 years of age, owing to the mortality risk associated with general anesthesia at this age threshold, or when only ultrasound without further imaging was used preoperatively to localize the pathologic gland. Provided that localization techniques are conclusive and concordant, they show that a focused approach is generally more cost-effective than BNE when specifically attempting to maximize QALYs, rather than to minimize complications as in Baliski et al. [24, 47].

Ruda et al. [51] also undertook a cost-effectiveness comparison of LE and BNE. Unlike Sejean et al. [24] however, they used the perspective of the health care provider and defined effectiveness as postoperative normocalcemia. Ruda, Stack, and Hollenbeak designed a decision analysis comparing BNE without preoperative localization to minimally invasive radioguided parathyroidectomy (MIRP) following initial sestamibi scan, without further imaging as long as results from this initial scan were conclusive. If initial scan results were inconclusive, a subsequent ultrasound was performed, followed by UNE. Base-case analysis revealed that pursuing initial sestamibi scan was not only less costly but more effective than BNE without localization, with an ICER of −$28,505 (2001 USD) per additional patient effectively treated. The sestamibi strategy was dominant over a reasonable range of probabilities and costs for scanning and operative strategies [51].

Because of the variation in perspectives, definitions of effectiveness, and costs assigned to different therapeutic strategies, uniformity is lacking between the available cost-effectiveness analyses addressing surgical approach, and conclusions are not easily generalizable. One’s definition of success, either achievement of surgical cure or avoidance of surgical complications, would suggest different strategies in the surgical management of a PHPT patient according to the above analyses. Further study is needed to corroborate these findings and guide surgeons in the efficient and high value care of their patients.

Approximately 85 % of patients with PHPT are found to have a solitary adenoma precipitating PTH excess. In these cases, resection of just one gland potentially provides surgical cure. Therefore, LE has become an increasingly common practice [44]. Patient evaluation prior to LE often includes one or more localization study to direct focused exploration, with high-resolution ultrasound and radionuclide Tc-99m sestamibi scans being the most common studies obtained [12, 25, 44, 51]. When localization techniques are conclusive and concordant preoperatively, focused parathyroidectomy has cure rates comparable to BNE in contemporary surgical series [14, 49, 53]. Yet even among surgeons practicing a focused approach, there has been no standardization of practice in selection of localization studies, intraoperative adjuncts, and specific surgical technique [52, 55].