Cost Effectiveness of Intraoperative PathologyExamination during Diagnostic Hemithyroidectomyfor Unilateral Follicular Thyroid Neoplasms

Kyle Zanocco, MD, Michael Heller, BA, Dina Elaraj, MD, FACS, Cord Sturgeon, MD, MS, FACS

BACKGROUND: The use of intraoperative pathology examination (IPE) during diagnostic hemithyroidectomyfor a follicular neoplasm is controversial. Although this service rarely alters intraoperativedecision making, it does provide patients with the possibility of avoiding reoperation forcompletion thyroidectomy if malignancy is detected. We hypothesized diagnostic hemithyr-oidectomy with IPE for a unilateral follicular thyroid neoplasm diagnosed on fine-needleaspiration is not cost effective compared with diagnostic hemithyroidectomy alone.

STUDY DESIGN: Cost-effectiveness analysis with a Markov decision model was performed comparing diag-nostic hemithyroidectomy without IPE, diagnostic hemithyroidectomy with IPE, and totalthyroidectomy. Treatment outcomes and their probabilities were identified based on litera-ture review. Costs were estimated using data from Medicare, the US Bureau of Labor Statis-tics, and the Nationwide Inpatient Sample. Sensitivity analysis and a 1,000-iteration MonteCarlo simulation were used to examine the uncertainty of cost, probability, and utilityestimates in the model.

RESULTS: Diagnostic hemithyroidectomy without IPE had an expected cost of US$7,665 and an effec-tiveness of 23.95 quality-adjusted life years and dominated both the IPE and total thyroid-ectomy strategies. Intraoperative pathology examination became cost effective during one-waysensitivity analysis if the sensitivity of IPE increased from 14.3% to 34.4%, the specificityincreased from 98.6% to 99.8%, or the pretest probability of malignancy increased from 25%to 43%. Monte Carlo simulation demonstrated that the intraoperative pathology strategy wasnot cost effective in 92.7% of iterations.

CONCLUSIONS: Intraoperative pathology examination is not cost effective in the diagnosis of follicular thyroidneoplasms during diagnostic hemithyroidectomy. Improvements in both the sensitivity andspecificity of this service would be needed to justify its use. (J Am Coll Surg 2013;217:702e710. � 2013 by the American College of Surgeons)

Palpable thyroid nodules are common in the United States’adult population, with an estimated prevalence of 10 to18 million.1,2 Approximately 16% of biopsied thyroidnodules are found to be follicular neoplasms that requireexcisional biopsy.3,4 Guidelines from the 2007 BethesdaNational Cancer Institute Thyroid FNAState of the ScienceConference recommend diagnostic hemithyroidectomyor total thyroidectomy for follicular neoplasms becausefine-needle aspiration (FNA) is unable to differentiatebenign follicular adenomas from follicular carcinomas.5,6

The use of intraoperative pathology examination (IPE),including frozen section (FS) during diagnostic hemithyr-oidectomy for a follicular neoplasm is controversial.7-16

Intraoperative pathology examination offers the potentialto save patients from undergoing reoperation for comple-tion thyroidectomy by making an intraoperative diagnosisof malignancy. However, IPE has low sensitivity and

CME questions for this article available athttp://jacscme.facs.org

Disclosure Information:Authors havenothing to disclose.Timothy JEberlein,Editor-in-Chief, has nothing to disclose.

Dr Zanocco’s participation in this study was supported in part by theNorthwestern University Feinberg School of Medicine Center for Health-care Studies under an institutional award from the Agency for HealthcareResearch and Quality, T-32 HS 000078 (Principal Investigator: Jane LHoll, MD, MPH).

Presented at the Third Annual Association for Academic Surgery andSociety of University Surgeons Academic Surgical Congress, HuntingtonBeach, CA, February 2008.

Received March 27, 2013; Revised May 7, 2013; Accepted May 7, 2013.From the Department of Surgery, Section of Endocrine Surgery (Zanocco,Heller, Elaraj, Sturgeon) and Center for Healthcare Studies(Zanocco), Northwestern University Feinberg School of Medicine,Chicago, IL.Correspondence address: Cord Sturgeon, MD, MS, FACS, Department ofSurgery, Northwestern University Feinberg School of Medicine, 676 N StClair St, Ste 650, Chicago, IL 60611. email: csturgeo@nmh.org

702ª 2013 by the American College of Surgeons ISSN 1072-7515/13/$36.00

Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jamcollsurg.2013.05.008

rarely changes operative management; it also adds to costsand increases operative time. In addition, intraoperativepathology examination introduces the possibility of falsepositives, which can lead to the unnecessary removal ofthe contralateral thyroid lobe in cases of benign follicularneoplasms.Total thyroidectomy has been proposed as an alternative

to diagnostic hemithyroidectomy in patients with follicularneoplasms.17 Total thyroidectomy eliminates the need forIPE or the possibility of reoperation, but costs more thandiagnostic hemithyroidectomy. This procedure also obli-gates all patients to receive lifelong thyroid hormonereplacement and places the contralateral recurrent laryngealnerve (RLN) and parathyroid glands at risk for injury.Formal cost-effectiveness analysis has never been

performed to compare the following treatment strategiesfor follicular thyroid neoplasms: diagnostic hemithyroi-dectomy alone followed by possible completion thy-roidectomy, hemithyroidectomy with FS, and totalthyroidectomy. We hypothesized that the addition ofIPE to diagnostic hemithyroidectomy for a unilateralfollicular thyroid neoplasm is not cost effective. In addi-tion, we hypothesized that total thyroidectomy wouldnot be cost effective compared with diagnostic hemithyr-oidectomy with or without IPE.

METHODS

Reference case scenario

A reference case was defined based on recommendations ofthe Panel on Cost-Effectiveness in Health and Medicine.18

The reference case was a patient with a unilateral follicularthyroid neoplasm identified on FNA. This patient wasassumed to be an appropriate candidate for thyroidectomyvia a cervical incision and without previous neck surgery orintraoperative findings necessitating the use of FS (eg,suspicious lymphadenopathy or the need to query parathy-roid tissue). The time horizon for the analysis was the lifeexpectancy of the reference case patient, which was set at40 years based on the age of patients diagnosed with follic-ular thyroid neoplasms.8,15,19

Decision model

A Markov decision model of the diagnosis and treatmentof the reference case was constructed using decision

software (TreeAge Pro, TreeAge Software, Inc., 2012).The following diagnostic decision alternatives werecreated: diagnostic hemithyroidectomy without IPE,diagnostic hemithyroidectomy with IPE, and totalthyroidectomy without IPE. Figures 1, 2 and 3 diagramthe chance and terminal nodes of the three diagnosticstrategies. The event pathways and treatment outcomesprobabilities resulting from each decision were estimatedbased on literature review and are shown in Table 1. Thepretest probability that an FNA-identified follicularneoplasm was malignant on a definitive pathology exam-ination after surgical excision was assumed to be 25%.The model was given a cycle length of 1 year and wasterminated when the number of completed cycles wasgreater than life expectancy. The optimal decision wasdefined as the strategy producing the greatest quality-adjusted life expectancy that did not exceed an incre-mental cost-effectiveness ratio of US$100,000 perquality-adjusted life year (QALY).

Costs

The societal costs of resources consumed for each strategywere estimated using a combination of micro and grosscosting techniques. Only costs that differed between the3 diagnostic strategies were included.18,37 All costs in themodel were measured in 2010 US dollars. An inflationrate of 4.1% was used to adjust older costs to their2010 amount when necessary.38 This rate was calculatedfrom the mean annual changes in the Consumer PriceIndex for Medical Care from 2000 to 2010. All futurecosts in the model were subjected to a standard discountrate of 3%.18

Cost estimates for physician-provided services werebased on 2010 national Medicare charge limits for thesurgery, anesthesiology, and pathology services renderedby each strategy.27 Hospital costs were estimated by calcu-lating an average of 2009 Medicare cost-to-charge ratiosfor DRG 625 (Thyroid, Parathyroid, and ThyroglossalProcedures with Major Complications and Comorbid-ities), DRG 626 (Thyroid, Parathyroid, and ThyroglossalProcedures with Complications and Comorbidities), andDRG 627 (Thyroid, Parathyroid, and ThyroglossalProcedures without Complications and Comorbidities)weighted by total amount reimbursed.28 The computedratio of 0.189 was then multiplied by the national medianinpatient charge estimates for hemi- and total thyroidec-tomy reported by the 2009 Nationwide InpatientSample.29 Thyroid hormone medication costs were basedon average US wholesale prices.31 The costs for thethyroidectomy-related complications of permanent RLNinjury and permanent hypoparathyroidism have been esti-mated elsewhere and were adjusted to 2010 costs.30 The

Abbreviations and Acronyms

FNA ¼ fine-needle aspirationFS ¼ frozen sectionIPE ¼ intraoperative pathology examinationQALY ¼ quality-adjusted life yearRLN ¼ recurrent laryngeal nerve

Vol. 217, No. 4, October 2013 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms 703

cost of missed patient labor due to surgery was based on2010 annual US hourly earnings.32 All of the costs used inthe model are shown in Table 1.

Effectiveness

Effectiveness was measured as the quality-adjusted lifeexpectancy produced by each strategy. All of the model’soutcomes were assigned a corresponding quality-adjustment factor based on literature review. The timespent by a patient with a particular result was multipliedby that result’s quality-adjustment factor to yield QALYs.Patients requiring reoperation for completion thyroidec-tomy were assigned a one-time reduction in quality-adjusted life expectancy. A 3% discount rate was appliedto all QALYs that were realized in the future. The quality-adjustment factors used in the model are displayed inTable 1.

Sensitivity analysis

Threshold analysis was performed on each referencecase variable to identify values where the incremental

cost-effectiveness ratio exceeded US$100,000/QALYduring one-way sensitivity analysis. Two-way sensitivityanalysis was performed on the sensitivity and specificityof intraoperative pathologic examination for malignancyin a follicular neoplasm. A Monte Carlo simulation wasperformed, where triangular frequency distributions foreach variable were simultaneously sampled during 1,000consecutive iterations. In this probabilistic sensitivityanalysis, all variables were allowed to vary �50% of thereference case estimate or, if not possible in the case ofprobability and utility estimates, to the greatest extentthat allowed the reference case value to remain at thecenter of a range containing the boundary 0 or 1.

RESULTS

Reference case

The diagnostic hemithyroidectomy without IPE was theleast costly, with an expected cost of US$7,665 and aneffectiveness of 23.95173 QALYs. The addition of anIPE to this procedure resulted in a loss of 0.00152

Figure 1. Chance and terminal node diagram of strategy for diagnostic hemithyroidectomywithout intraoperative pathology examination. RLN, recurrent laryngeal nerve.

print&web4C/FPO

704 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms J Am Coll Surg

QALYs at an additional cost of US$72.17. The IPEstrategy was dominated because it was both more costlyand less effective than diagnostic hemithyroidectomyalone. The total thyroidectomy strategy was more costlyand less effective than the other strategies and was there-fore also dominated by diagnostic hemithyroidectomyalone. The reference case results are summarized inTable 2.

Sensitivity analysis

The threshold values necessary to achieve a cost effective-ness of US$100,000/QALY for adding intraoperativepathology to diagnostic hemithyroidectomy are listed

Figure 2. Chance and terminal node diagram of strategy for diagnostic hemithyroidectomy with intraoperative pathology examination.RLN, recurrent laryngeal nerve.

Figure 3. Chance and terminal node diagram of strategy for totalthyroidectomy. RLN, recurrent laryngeal nerve.

print&web4C/FPO

print&web4C/FPO

Vol. 217, No. 4, October 2013 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms 705

in Table 3. The optimal decision was most sensitive to thesensitivity and specificity of IPE, pretest probability ofmalignancy in the reference case, and the costs of surgery.If the IPE specificity increased from 98.6% to 99.8%, orif the sensitivity increased from 14.3% to 34.4%, IPEbecame cost effective. The IPE strategy also becamecost effective if the probability of malignancy increasedfrom 25% to 43%, or if the cost of completion thyroid-ectomy doubled. The model was not sensitive to the rateof hypothyroidism after hemithyroidectomy or the treat-ment costs or quality-adjustment factors for hypoparathy-roidism or RLN injury.

Two-way sensitivity analysis demonstrated that thedecision model was sensitive to simultaneous variationof the test characteristics (ie, sensitivity and specificity)of IPE. Figure 4 demonstrates that intraoperativepathology became cost effective for combinations ofsensitivity and specificity that were sufficiently high, butthe reference case values for these parameters didnot produce an incremental cost-effectiveness ratio<US$100,000/QALY.Monte Carlo simulation demonstrated that the intrao-

perative pathology strategy was not cost effective whencompared with diagnostic hemithyroidectomy without

Table 1. Model Assumptions

VariablesReferencecase value

Range of valuesin Monte Carlodistribution References

Probabilities, %

Malignancy in follicular neoplasm identified on fine-needle aspiration 25.0 12.5e37.5 5,20

True-positive identification of malignancy on intraoperative pathologyexamination (sensitivity) 14.3 7.2e21.5 8,15

True-negative identification of benign neoplasm on intraoperativepathology examination (specificity) 98.6 97.2e100 8,15

Hypothyroidism after hemithyroidectomy 35.0 17.5e52.5 21,22

Hypoparathyroidism after completion or total thyroidectomy 1.0 0.5e1.5 21,23-25

Recurrent laryngeal nerve injury (per lobe dissected) 1.0 0.5e1.5 26

Costs in US$

Diagnostic thyroid hemithyroidectomy 4,141.26 2,071e6,212 27-29

Completion thyroidectomy 5,513.61 2,757e8,270 27-29

Total thyroidectomy 5,500.05 2,750e8,250 28,29

Additional 15 min of anesthesiologist work time 23.57 11.79e35.36 27

Intraoperative pathology examination 159.52 79.76e239.28 27

Treatment of recurrent laryngeal nerve injury 10,366.55 5,183e15,550 30

Annual cost of permanent hypoparathyroidism, US$ 863.02 432e1,295 30

Annual cost of thyroid hormone replacement, US$ 151.51 75.76e227.27 31

Hourly cost of missed patient work, US$ 22.61 11.31e33.92 32

Quality-of-life adjustment factors

Euthyroid 1 NA* 33,34

Hypothyroid on thyroid hormone replacement 0.990 0.980e1 35

Euthyroid with recurrent laryngeal nerve injury 0.891 0.782e1 36

Hypothyroid with recurrent laryngeal nerve injury 0.881 0.762e1 35,36

Hypothyroid with hypoparathyroidism 0.893 0.786e1 35,36

Hypothyroid with hypoparathyroidism and recurrent laryngeal nerve injury 0.775 0.550e1 35,36

1-time reduction in quality-adjusted life expectancy for undergoingreoperation for completion thyroidectomy, d 1 0.5e1.5 Expert opinion

Time

Additional anesthesia time due to intraoperative pathology examination, min 20 10e30 8

Additional work missed to undergo completion thyroidectomy, h 16 8e24 Expert opinion

Life expectancy, y 40 20e60 8,15,19

Discount rate, % 3.0 1.5e4.5 18

*Euthyroidism was assigned a quality adjustment factor of 1 and was not varied on probabilistic sensitivity analysis to serve as a baseline for comparison by allother quality adjustment factors.

706 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms J Am Coll Surg

intraoperative pathology in 92.7% of the iterations. Intra-operative pathology examination was dominant in 1.8%of iterations compared with 82.1% in which hemithyroi-dectomy without IPE was dominant (Fig. 5).

DISCUSSIONThis study demonstrates that diagnostic hemithyroidec-tomy with IPE for unilateral follicular neoplasms

diagnosed on FNA produces decreased effectivenessat an increased cost and operative time compared withdiagnostic hemithyroidectomy without IPE. However,increases in the pretest probability of malignancy in follic-ular neoplasms or improvement of the sensitivity or spec-ificity of IPE would make the use of this service costeffective. Total thyroidectomy for follicular neoplasmsis more costly and less effective than diagnostic hemithyr-oidectomy with or without IPE.

Table 3. The US$100,000/Quality-Adjusted Life-Years Threshold Values for Model Variables That Make IntraoperativePathology Cost Effective

Variable Threshold

Probabilities, %

Malignancy in follicular neoplasm identified on fine-needle aspiration, % >42.8

True-positive identification of malignancy on intraoperative pathology examination (sensitivity), % >34.4

True-negative identification of benign neoplasm on intraoperative pathology examination (specificity), % >99.8

Hypothyroidism after hemithyroidectomy None

Hypoparathyroidism after completion or total thyroidectomy None

Recurrent laryngeal nerve injury (per lobe dissected), % >20.9

Costs in US$

Diagnostic thyroid hemithyroidectomy >8,990

Completion thyroidectomy >11,787

Total thyroidectomy <600

Additional 15 min of anesthesiologist work time None

Intraoperative pathology examination None

Treatment of recurrent laryngeal nerve injury None

Annual cost of permanent hypoparathyroidism None

Annual cost of thyroid hormone replacement None

Hourly cost of missed patient work, US$ >410

Quality-of-life adjustment factors

Hypothyroid on thyroid hormone replacement None

Euthyroid with recurrent laryngeal nerve injury None

Hypothyroid with recurrent laryngeal nerve injury None

Hypothyroid with hypoparathyroidism None

Hypothyroid with hypoparathyroidism and recurrent laryngeal nerve injury None

One-time reduction in quality-adjusted life expectancy for undergoing reoperation for completion thyroidectomy, d >4.3

Time

Additional anesthesia time due to intraoperative pathology exam, min None

Additional work missed to undergo completion thyroidectomy, h >293

Life expectancy, y <2.0

Discount rate, % >77

Table 2. Reference Case Results

Strategy Cost, US$Incrementalcost, US$ Effectiveness*

Incrementaleffectiveness* ICER

Diagnostic hemithyroidectomy 7,664.71 d 23.95173 d d

Diagnostic hemithyroidectomy with intraoperativepathology examination 7,736.88 72.17 23.95021 �0.00152 Dominated

Total thyroidectomy 9,575.36 1,910.65 23.79924 �0.15248 Dominated

*Reported in quality-adjusted life years.ICER, incremental cost-effectiveness ratio.

Vol. 217, No. 4, October 2013 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms 707

This is the first formal cost-effectiveness analysis of IPEfor follicular thyroid neoplasms that uses a decisionmodel. Several previous studies using different method-ology have arrived at opposing conclusions about whetherIPE is a cost-saving or cost-losing service. In a randomizedprospective evaluation of FS analysis for follicular thyroidneoplasms, the charge per informative FS was estimated

at US$12,470.15 The cost per useful operation-alteringFS has been estimated elsewhere on retrospective reviewto range from US$6,000 to US$26,040.7,8,11 Althoughthese studies argue that FS is not worthwhile becausethe cost of FS exceeded the cost of reoperation fora completion thyroidectomy, it is difficult to interpretthe cost effectiveness of FS because there was no

Figure 4. Two-way sensitivity analysis demonstrates the combined effects of changes in thesensitivity and specificity of intraoperative pathology examination for follicular neoplasms on theoverall cost effectiveness of this service during diagnostic hemithyroidectomy. The threshold linerepresents combinations of these test characteristics that produce an incremental cost-effectiveness ratio (ICER) of US$100,000 per quality-adjusted life year (QALY). The areaabove the threshold line represents values that produce an ICER of <US$100,000 per QALY andwould make intraoperative pathology a cost-effective strategy. The reference case test charac-teristics intersect below the threshold line in the area representing values that exceed thethreshold for cost effectiveness.

Figure 5. A 1,000-iteration Monte Carlo simulation where all model assumptions weresimultaneously varied across their frequency distributions. The incremental cost andincremental effectiveness of intraoperative pathology examination strategy compared with thediagnostic hemithyroidectomy alone are plotted for each iteration. Of 1,000 iterations, 927(92.7%) exceeded the US$100,000/quality-adjusted life-year threshold for cost effectiveness.ICER, incremental cost-effectiveness ratio.

708 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms J Am Coll Surg

consideration given to possible improvement in quality oflife by avoiding a second procedure.A cost-minimization analysis of FS for all benign and

indeterminate thyroid lesions in the Canadian healthcare system calculated a cost savings of C$22.01 perpatient.9 However, the sensitivity and specificity of FS(30.2% and 100%, respectively) were higher in thatstudy’s population of patients when compared with thetest characteristics of FS in subsets of patients with follic-ular lesions (14% and 99%, respectively).8,15 A retrospec-tive review of patients at the Mayo Clinic undergoingthyroidectomy for suspected follicular and Hurthle cellneoplasms reported FS sensitivity and specificity of78.3% and 99.3% and savings of >US$400,000 incharges during a 10-year time period due to avoidanceof a second operation.13 Although this study populationis also not directly comparable with a subset limited topatients with follicular lesions, it is important to notethat the relatively high sensitivity and specificity reportedat Mayo is attributed by the authors to an abundance ofsurgical pathology resources available in the performanceof intraoperative examination. A tradeoff is evidentbetween the cost of increased availability of pathologyresources and the improvement in test characteristics ofIPE. Measuring the impact on saved patient charges alonedoes not capture the costs of operating a surgicalpathology laboratory with the extra equipment andpersonnel necessary to achieve these results.Cost-effectiveness analysis with decision modeling

enables a more complete economic evaluation of theaddition of IPE to diagnostic lobectomy than has previ-ously been studied. This technique allows the value ofthe increased costs for IPE to be examined in the contextof potential effects on quality of life. In addition, sensi-tivity analysis identifies thresholds for cost effectivenessin outcomes probabilities, utility measurements, and coststhat can be applied to unique diagnostic situations withdifferent values than those that appear in the referencecase scenario.Several limitations are evident in this study. The deci-

sion model omitted several possible complications,including wound infection, hematoma formation, andoperative mortality. Because all diagnostic strategiesincluded surgery, these omissions probably have littleimpact on the results. However, some of these eventsmight be more likely during reoperation and wouldmake completion thyroidectomy less favorable, thereforeimproving the cost effectiveness of the IPE strategy. Theextra cost of transportation to undergo completionthyroidectomy was also omitted, which would have addedto the overall cost of the diagnostic hemithyroidectomywithout IPE strategy, although by only a small amount.

The overall cost of completion thyroidectomy couldhave increased by US$6,273 before IPE became cost effec-tive, so it is unlikely that this omission would change themodel’s conclusion. The model assumed that all patientswho became hypothyroid after thyroid lobectomyrequired full-dose thyroid hormone replacement. Theoutcomes of partial hypothyroidism, which would haverequired a lower level of thyroid hormone replacement,were not considered. This omission did not alter theconclusions of this study because the model was not sensi-tive to lower costs of thyroid hormone replacement or thequality-adjustment factor for treated hypothyroidism. Thesurgical complication of RLN injury has a wide spectrumof severity and treatment costs, including the need fortemporary voice therapy, vocal cord medialization, and,at worst, permanent tracheostomy, but this array of possi-bilities was simplified into a single gross cost and qualityof life estimates. We believe our sensitivity analysisadequately addresses the possible impact of these simplifi-cations. The option of outpatient thyroidectomy was notdirectly considered in this model. Diagnostic hemithyroi-dectomy and completion thyroidectomy in the outpatientsetting would have lowered the costs of these procedures,and sensitivity analysis demonstrates that this effect woulddiminish the cost effectiveness of IPE.

CONCLUSIONSThis study demonstrates that IPE is not cost effective inthe diagnosis of follicular thyroid neoplasms during diag-nostic hemithyroidectomy. Improvements in both thesensitivity and specificity of this service would be neededto justify its use.

Author Contributions

Study conception and design: Zanocco, Elaraj, SturgeonAcquisition of data: Zanocco, HellerAnalysis and interpretation of data: Zanocco, SturgeonDrafting of manuscript: Zanocco, HellerCritical revision: Zanocco, Heller, Elaraj, Sturgeon

REFERENCES

1. Gharib H, Goellner JR. Fine-needle aspiration biopsy ofthe thyroid: an appraisal. Ann Intern Med 1993;118:282e289.

2. Stojadinovic A, Peoples GE, Libutti SK, et al. Development ofa clinical decision model for thyroid nodules. BMC Surg2009;9:12.

3. Al-Hureibi KA, Al-Hureibi AA, Abdulmughni YA, et al. Thediagnostic value of fine needle aspiration cytology in thyroidswellings in a university hospital, Yemen. Saudi Med J 2003;24:499e503.

Vol. 217, No. 4, October 2013 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms 709

4. Wu HH, Jones JN, Osman J. Fine-needle aspiration cytologyof the thyroid: ten years experience in a community teachinghospital. Diagn Cytopathol 2006;34:93e96.

5. Baloch ZW, Cibas ES, Clark DP, et al. The National CancerInstitute Thyroid fine needle aspiration state of the scienceconference: a summation. Cytojournal 2008;5:6.

6. Cibas ES, Ali SZ. The Bethesda System for reporting thyroidcytopathology. Thyroid 2009;19:1159e1165.

7. Callcut RA, Selvaggi SM, Mack E, et al. The utility of frozensection evaluation for follicular thyroid lesions. Ann SurgOncol 2004;11:94e98.

8. Chen H, Nicol TL, Udelsman R. Follicular lesions of thethyroid. Does frozen section evaluation alter operative manage-ment? Ann Surg 1995;222:101e106.

9. Lai P, Segall L, de Korompay N, et al. Cost analysis of intra-operative frozen section examinations in thyroid surgery ina Canadian tertiary center. J Otolaryngol Head Neck Surg2009;38:559e563.

10. LiVolsi VA, Baloch ZW. Use and abuse of frozen section inthe diagnosis of follicular thyroid lesions. Endocr Pathol2005;16:285e293.

11. McHenry CR, Raeburn C, Strickland T, Marty JJ. The utilityof routine frozen section examination for intraoperative diag-nosis of thyroid cancer. Am J Surg 1996;172:658e661.

12. Miller MC, Rubin CJ, Cunnane M, et al. Intraoperative path-ologic examination: cost effectiveness and clinical value inpatients with cytologic diagnosis of cellular follicular thyroidlesion. Thyroid 2007;17:557e565.

13. Paphavasit A, Thompson GB, Hay ID, et al. Follicular andHurthle cell thyroid neoplasms. Is frozen-section evaluationworthwhile? Arch Surg 1997;132:674e678; discussion678e680.

14. Roach JC, Heller KS, Dubner S, Sznyter LA. The value offrozen section examinations in determining the extent ofthyroid surgery in patients with indeterminate fine-needle aspi-ration cytology. Arch Otolaryngol Head Neck Surg 2002;128:263e267.

15. Udelsman R, Westra WH, Donovan PI, et al. Randomizedprospective evaluation of frozen-section analysis for follicularneoplasms of the thyroid. Ann Surg 2001;233:716e722.

16. Wong Z, Muthu C, Craik J, et al. Role of intraoperative frozensection in the management of thyroid nodules. ANZ J Surg2004;74:1052e1055.

17. Cooper DS, Doherty GM, Haugen BR, et al. Revised Amer-ican Thyroid Association management guidelines for patientswith thyroid nodules and differentiated thyroid cancer.Thyroid 2009;19:1167e1214.

18. Gold MR. Cost-Effectiveness in Health and Medicine. NewYork: Oxford University Press; 1996.

19. Bell FC, Miller ML. Life Tables for the United States SocialSecurity Area 1900�2100, Actuarial Study No. 120 2011.Available at: http://www.socialsecurity.gov/OACT/NOTES/as120/LifeTables_Tbl_6_2010.html. Accessed November 25,2011.

20. Wang CC, Friedman L, Kennedy GC, et al. A large multi-center correlation study of thyroid nodule cytopathology andhistopathology. Thyroid 2011;21:243e251.

21. Dobrinja C, Trevisan G, Piscopello L, et al. Comparisonbetween thyroidectomy and hemithyroidectomy in treatmentof single thyroid nodules identified as indeterminate follicularlesions by fine-needle aspiration cytology. Ann Ital Chir 2010;81:403e410; discussion 410e411.

22. Su SY, Grodski S, Serpell JW. Hypothyroidism followinghemithyroidectomy: a retrospective review. Ann Surg 2009;250:991e994.

23. Page C, Strunski V. Parathyroid risk in total thyroidectomy forbilateral, benign, multinodular goitre: report of 351 surgicalcases. J Laryngol Otol 2007;121:237e241.

24. Sitges-Serra A, Ruiz S, Girvent M, et al. Outcome of pro-tracted hypoparathyroidism after total thyroidectomy. Br JSurg 2010;97:1687e1695.

25. Youngwirth L, Benavidez J, Sippel R, Chen H. Parathyroidhormone deficiency after total thyroidectomy: incidence andtime. J Surg Res 2010;163:69e71.

26. Lo CY, Kwok KF, Yuen PW. A prospective evaluation ofrecurrent laryngeal nerve paralysis during thyroidectomy.Arch Surg 2000;135:204e207.

27. Overview of the Medicare Physician Fee Schedule Search.2011. Available at: http://https//www.cms.gov/apps/physician-fee-schedule/overview.aspx. Accessed November 25, 2011.

28. MEDPAR Medicare Fee for Service for Parts A & B. 2011.Available at: http://www.cms.gov/MedicareFeeforSvcPartsAB/03_MEDPAR.asp#TopOfPage. Accessed November 25, 2011.

29. HCUPnet. A tool for identifying, tracking, and analyzingnational hospital statistics. 2011. Available at: http://hcupnet.ahrq.gov/. Accessed November 25, 2011.

30. Zanocco K, Angelos P, Sturgeon C. Cost-effectiveness analysisof parathyroidectomy for asymptomatic primary hyperparathy-roidism. Surgery 2006;140:874e881; discussion 881e882.

31. Healthcare Thomson. Red Book. Montvale, NJ: ThomsonPDR; 2010:v.

32. Employment, Hours, and Earnings from the Current Employ-ment Statistics survey (National). 2011. Available at: http://data.bls.gov/timeseries/CES0500000003?data_tool¼XGtable.Accessed November 25, 2011.

33. Elberling TV, Rasmussen AK, Feldt-Rasmussen U, et al.Impaired health-related quality of life in Graves’ disease. Aprospective study. Eur J Endocrinol 2004;151:549e555.

34. Epstein KA, Schneiderman LJ, Bush JW, Zettner A. The“abnormal” screening serum thyroxine (T4): analysis of physi-cian response, outcome, cost and health effectiveness.J Chronic Dis 1981;34:175e190.

35. Muennig P, Khan K. Designing and Conducting Cost-Effectiveness Analyses in Medicine and Health Care. 1st ed.San Francisco: Jossey-Bass; 2002.

36. Sejean K, Calmus S, Durand-Zaleski I, et al. Surgery versusmedical follow-up in patients with asymptomatic primaryhyperparathyroidism: a decision analysis. Eur J Endocrinol2005;153:915e927.

37. Muennig P. Cost-Effectiveness Analyses in Health: A PracticalApproach. 2nd ed. San Francisco: Jossey-Bass; 2008.

38. Consumer Price Index-Medical Care. 2011. Available at:http://data.bls.gov/timeseries/CUUR0000SAM?output_view¼pct_12mths. Accessed November 25, 2011.

710 Zanocco et al Frozen Section for Follicular Thyroid Neoplasms J Am Coll Surg