Postoperative HypocalcemiaAfter Thyroidectomy for Graves’ Disease

Catherine E. Pesce,1 Zita Shiue,1 Hua-Ling Tsai,2 Christopher B. Umbricht,1 Ralph P. Tufano,3

Alan P.B. Dackiw,1 Jeanne Kowalski,2 and Martha A. Zeiger1

Background: It is believed that patients who undergo thyroidectomy for Graves’ disease are more likely toexperience postoperative hypocalcemia than patients undergoing total thyroidectomy for other indications.However, no study has directly compared these two groups of patients. The aim of this study was to determinewhether there was an increased incidence or severity of postoperative hypocalcemia in patients who underwentthyroidectomy for Graves’ disease.Methods: An institutional review board-approved database was created of all patients who underwent thy-roidectomy from 1998 to 2009 at the Johns Hopkins Hospital. There were a total of 68 patients with Graves’ diseasewho underwent surgery. Fifty-five patients who underwent total thyroidectomy were randomly selected andserved as control subjects. An analysis was conducted that examined potential covariates for postoperativehypocalcemia, including age, gender, ethnicity, preoperative alkaline phosphatase level, size of goiter, whetherparathyroid tissue or glands were present in the specimen, and the reason the patient underwent surgery. Specificoutcomes examined were calcium levels on postoperative day 1, whether or not patients experienced symptoms ofhypocalcemia, whether or not Rocaltrol was required, the number of calcium tablets prescribed upon discharge,whether or not postoperative tetany occurred, and calcium levels 1 month after discharge.Results: Each outcome was analyzed using a logistic regression. Graves’ disease patients had a significantly ( p-value< 0.001) higher odds of greater number of calcium tablets prescribed upon discharge. Further, 6 of 68patients with Graves’ disease and no patient in the control group were readmitted with tetany ( p¼ 0.033). Therewas a trend, though not significant, toward patients with Graves’ disease having a higher prevalence of hypo-calcemia the day after thyroidectomy and 1 month later.Conclusions: Patients with Graves’ disease are more likely to require increased dosages of calcium as well asexperience tetany postoperatively than patients undergoing total thyroidectomy for other indications. This sug-gests that patients operated upon for Graves’ disease warrant close followup as both inpatients and outpatientsfor signs and symptoms of hypocalcemia.

Introduction

Graves’ disease is the most common cause of hyper-thyroidism and the most common autoimmune disorder

in the United States. As a first-line treatment antithyroidmedications or radioactive iodine ablation are generally used,with total thyroidectomy serving as a third-tier approach.However, in pregnant patients or patients with thyroid nod-ules, total thyroidectomy can serve as a first- or second-lineapproach. Surgical procedures generally include subtotalthyroidectomy, a near-total thyroidectomy, or total thyroid-ectomy (1). Risks of permanent recurrent laryngeal nerve

palsy and permanent hypoparathyroidism are similar in pa-tients who undergo total thyroidectomy versus those whoundergo subtotal thyroidectomy. Total thyroidectomy, how-ever, also allows for elimination of all abnormal tissue in theneck and lower recurrence rates for Graves’ disease (2).

Estimates of the percentage of patients who experiencetransient hypoparathyroidism after all thyroid surgery rangefrom 6.9% (3) to 46% (4,5). It is believed that patients under-going thyroidectomy for Graves’ disease are more likely toexperience postoperative hypocalcemia than patients under-going total thyroidectomy for other indications. However, nostudy has directly compared these two groups of patients.

Departments of 1Surgery, 2Biostatistics, and 3Otolaryngology/Head and Neck Surgery, The Johns Hopkins University School of Medicine,Baltimore, Maryland.

THYROIDVolume 20, Number 11, 2010ª Mary Ann Liebert, Inc.DOI: 10.1089/thy.2010.0047

1279

The exact mechanisms of postoperative hypocalcemia areunknown and are yet to be elucidated; however, there appearto be significant factors that are associated with this compli-cation: (i) Graves’ disease; (ii) female gender; (iii) the extent ofmanipulation of the parathyroid glands at the time of surgery;and (iv) hungry bone syndrome with rapid skeletal uptake ofcalcium in patients with osteodystrophy (6). Two surgicalgroups have investigated hypocalcemia associated with thy-roidectomy and their findings are summarized in Table 1.

We sought to determine whether Graves’ disease patientsare indeed more likely to experience postoperative hypocal-cemia and whether the hypocalcemia was more severe thanpatients who underwent total thyroidectomy for other benigndiseases.

Materials and Methods

An institutional review board-approved database wascreated of all patients who underwent total thyroidectomyfrom 1998 to 2009 at the Johns Hopkins Hospital. A total of68 patients with Graves’ disease underwent surgery duringthis period. An additional 55 patients who underwent totalthyroidectomy for benign disease during this same periodwere randomly selected from the database and served ascontrol subjects. Total thyroidectomy was performed by oneof three surgeons.

On the basis of literature review of potential covariatesinvolved in the development of postoperative hypocalcemiaafter thyroid surgery (7–12), a retrospective review of patientcharts was conducted to identify such factors. These includedage (greater or less than 50 years), gender, ethnicity, preop-erative alkaline phosphatase level, size of the goiter, whetheror not parathyroid tissue or glands were present in the spec-imen, and the season the patient underwent surgery. Specificoutcomes examined were calcium levels on postoperative day1, whether or not patients experienced symptoms of hypo-calcemia, whether or not Rocaltrol was started upon dis-charge due to hypocalcemia, the number of calcium tabletsprescribed upon discharge, whether or not postoperativetetany occurred, and calcium levels 1 month after discharge.Number of calcium tablets and not exact dosage was recordedsince a range of tablet sizes is available to patients over thecounter (200–600 mg calcium per tablet). Consequently, spe-cific dosages were unavailable for this study.

Total calcium or ionized calcium was measured dependingupon the attending surgeon’s choice. Hypocalcemia wasdefined as a total calcium level <8.5 mg/dL (normal8.5–10.5 mg/dL) or an ionized calcium level <1.13 mmol/L(normal 1.13–1.32 mmol/L) on postoperative day 1 and1 month after discharge. Symptoms of hypocalcemia among

patients included perioral numbness, extremity paresthesiasor myalgias, and/or positive Chvostek (if documented to benegative preoperatively) or Trousseau sign.

The routine practice of all three surgeons involved checkingeach patient’s calcium level the morning after thyroidectomy.If a patient was normocalcemic, no supplementation wasgiven as an inpatient or upon discharge. If the patient washypocalcemic, they were administered calcium tablets as aninpatient and given instructions for calcium supplementationupon discharge. If the calcium level after calcium tablet sup-plementation was not easily stabilized, Rocaltrol was addedand prescribed upon discharge. Graves’ disease patients andcontrol patients were supplemented identically. After dis-charge, hypocalcemic patients were followed by a nursepractitioner by phone to clarify quantity and length of calciumsupplementation required and to inquire about any addi-tional ongoing symptoms of hypocalcemia. Both normo-calcemic and hypocalcemic patients were dischargedpostoperative day 1 unless calcium levels could not be stabi-lized. If the patient was symptomatic as an outpatient and thesymptoms could not be resolved with additional calcium,serum calcium was checked.

Each outcome (‘‘calcium levels on postoperative day 1,’’‘‘whether or not patients experienced symptoms of hypocal-cemia,’’ ‘‘whether or not Rocaltrol was started upon dischargedue to hypocalcemia,’’ and ‘‘calcium levels one month afterdischarge’’) was analyzed using a logistic regression to obtainodds ratios (OR), 95% confidence intervals (CI), and p-valuesbased on a Chi-square statistic. A Poisson regression was usedto model count data for the outcome ‘‘number of calcium tabletsprescribed upon discharge.’’ The outcome of ‘‘whether or notpostoperative tetany occurred’’ required use of a Fisher exacttest. For each potential risk factor, both adjusted for Graves’disease and unadjusted OR are presented where feasible.

Results

An analyses summary of each outcome is provided (Table 2).

Calcium on postoperative day 1

Of 68 patients with Graves’ disease, 44 patients werehypocalcemic on postoperative day 1, and 24 were normo-calcemic. Of 55 control subjects, 32 were hypocalcemic, and23 were normocalcemic. The Graves’ disease patients had ahigher OR of hypocalcemia (OR¼ 1.32, 95% CI 0.63–2.74))than the control group, though not significant ( p-value¼ 0.46).After adjusting for each risk factor, the direction and signifi-cance of the association did not change. Among all character-istics examined, the presence of parathyroid tissue in the

Table 1. Factors Associated with Hypocalcemia

Study Findings

Thomusch et al. (7,8). Graves’ disease, female gender, and larger hospital operative volume are risk factors forpostoperative hypoparathyroidism.

Graves’ disease, compared to uninodular goiter, multinodular goiter, and thyroiditiscombined, leads more often to postoperative hypoparathyroidism.

Yamashita et al. (9–12). Higher incidence of postoperative tetany among women with Graves’ disease.Calcium and vitamin D supplements may be recommended before and/or after surgery

for patients with Graves’ disease.

1280 PESCE ET AL.

specimen showed a significantly higher probability of hypo-calcemia compared to the lack of tissue present, when adjustingfor Graves’ disease (OR¼ 3.18, 95% CI 0.98–10.34, p-val-ue¼ 0.05).

Whether or not patients experienced symptomsof hypocalcemia

Positive symptoms of hypocalcemia among patients weredefined as perioral numbness, extremity paresthesias or my-algias, and/or positive Chvostek’s sign (if documented to benegative preoperatively) or Trousseau’s sign. Of the 68 pa-tients with Graves’ disease, 14 experienced symptoms of hy-pocalcemia. Ten patients experienced extremity paresthesias,3 patients experienced perioral numbness, and 1 patient had apositive Chvostek’s sign. Of the 55 control patients, 5 expe-rienced symptoms of hypocalcemia. Four patients experi-enced extremity paresthesias, and 1 patient experiencedperioral numbness. Graves’ disease patients had a higher ORof experiencing symptoms of hypocalcemia (OR¼ 2.59, 95%CI 0.87–7.72) than the control group, though not significant( p-value¼ 0.09). Older age patients had a lower odds of ex-periencing symptoms of hypocalcemia (OR¼ 0.96, 95% CI0.93–1.00) than younger patients when adjusting for Graves’disease ( p-value¼ 0.04).

Whether or not Rocaltrol was required upon discharge

Of 68 patients with Graves’ disease, 12 patients requiredRocaltrol, 56 patients did not. Of 55 control subjects, 9 patientswere started on Rocaltrol, 46 patients were not. Patients withGraves’ disease presented with a higher OR of startingRocaltrol compared to the control group (OR¼ 1.1, 95% CI0.42–2.83), though not significant ( p-value¼ 0.85). Among allcharacteristics examined, the presence of parathyroid tissue inthe specimen showed a significantly higher probability ofrequiring Rocaltrol compared to the lack of tissue present,when adjusting for Graves’ disease (OR¼ 7.23, 95% CI 2.8–21.99, p-value< 0.01). Also, older age showed significantly( p-value¼ 0.04) lower odds of starting Rocaltrol than youngerage (OR¼ 0.96, 95% CI 0.93–1.00) after adjusting for Graves’disease.

Number of calcium tablets prescribed upon discharge

Among 68 patients with Graves’ disease, the mediannumber of calcium tablets prescribed upon discharge was 2(range: 0–9). Among 55 control subjects, the median numberof tablets was 0 (range: 0–6). When examining the numberof tablets, patients with Graves’ disease had a significantly( p-value< 0.001) higher odds of greater number of calcium

tablets prescribed upon discharge (OR¼ 1.85, 95% CI 1.40–2.46) than the control group. After adjusting for other char-acteristics, the association did not change.

Whether or not postoperative tetany occurred

Of the 68 patients with Graves’ disease, 6 patients experi-enced tetany postoperatively requiring visits to local emer-gency departments. All 6 patients experienced tetany within1 week of discharge. Of the 55 control group patients, noneexperienced tetany. A statistically significant association wasfound between Graves’ disease patients and the incidence oftetany ( p-value¼ 0.033).

Calcium levels 1 month after discharge

Of the 68 patients with Graves’ disease, only 38 patients hadcalcium levels 1 month after discharge available for analysis.Twenty-nine patients had a normal calcium level, and 9 werehypocalcemic. Of the 55 control group patients, 25 wereavailable for analysis. Nineteen had a normal calcium level,and 6 were hypocalcemic. Graves’ disease patients had a lowerOR of hypocalcemia 1 month after discharge (OR¼ 0.98, 95%CI 0.30–3.21) than the control group, though not significanteither alone ( p-value¼ 0.98) or after adjusting for other risk.

Discussion

Prospective multicenter studies have shown that indepen-dent risk factors for predicting the development of transienthypocalcemia after thyroidectomy include female gender, theextent of manipulation of the parathyroid glands at the time ofsurgery, and, more recently, the presence of Graves’ disease.The mechanisms suggested include hungry bone syndromewith rapid skeletal uptake of calcium in patients with osteo-dystrophy, as well as abnormal release of calcitonin by ma-nipulation of the thyroid gland (6). In addition, Graves’disease may be associated with more profuse bleeding andadhesions between the thyroid capsule and parathyroidglands. Adhesions and reduction of visibility during diffusebleeding from dissected thyroid tissue can more frequentlylead to injuries of the parathyroid glands (8). Some studiessuggest that the extent of resection and surgical techniquehave the greatest impact on the incidence of permanentpostoperative hypoparathyroidism (7).

In comparison to control patients there was a trend towardGraves’ disease patients having a higher prevalence of hy-pocalcemia on postoperative day 1, symptoms of hypocalce-mia as inpatients, having Rocaltrol started upon discharge,and having a higher prevalence of hypocalcemia 1 monthafter thyroid surgery. The OR was significant for Graves’

Table 2. Univariate Results Examining the Association Between Graves’ Disease Patients (n¼ 68)

and Control Patients (n¼ 55) with Hypocalcemia Outcomes

Calcium POD1Experiencedsymptoms

StartedRocaltrol

No. of calcium tabletsprescribed

Calcium level1 month postoperative

Control (referent) 1.00 1.00 1.00 1.00 1.00Graves’ disease 1.32 2.59 1.10 1.85 0.9895% CI (0.63–2.74) (0.87–7.72) (0.42–2.83) (1.40–2.46) (0.30–3.21)p-Value 0.46 0.09 0.85 <0.001 0.98

POD1, postoperative day 1; CI, confidence intervals.

POSTOPERATIVE HYPOCALCEMIA AFTER THYROIDECTOMY FOR GRAVES’ 1281

disease patients and the number of calcium tablets prescribedupon discharge. Further, 6 of 68 patients with Graves’ diseaseand no patient in the control group were readmitted withtetany ( p¼ 0.033). This supports our hypothesis that patientswith Graves’ disease are at increased odds of signs andsymptoms of hypocalcemia after thyroidectomy.

Not surprisingly, the presence of parathyroid gland tissuein the thyroid specimen was predictive of postoperative hy-pocalcemia. A statistically significant association was foundbetween presence of parathyroid tissue and both incidence ofhypocalcemia on postoperative day 1 as well as incidence ofstarting the patient on Rocaltrol postoperatively. These find-ings support the work of others in demonstrating the im-portance of surgical technique in preserving the parathyroidglands during surgery (5). Thomusch et al. found that removalof a single parathyroid gland was not associated with post-operative hypocalcemia, whereas resection of at least twoparathyroid glands increased the risk of transient and per-manent hypoparathyroidism. We also agree that preservationof the parathyroid glands and their blood supply duringthyroidectomy is a recommended surgical strategy to de-crease the rate of postoperative hypocalcemia. Future studiesare needed to determine whether there is decreased incidenceof hypocalcemia after parathyroid reimplantation duringthyroid surgery. In our study we did not find that incidentalparathyroidectomy occurred more often in our patients withGraves’ disease versus control patients undergoing thyroid-ectomy.

We controlled for multiple patient characteristics that oth-ers and we have hypothesized to be potential risk factors forhypocalcemia after thyroidectomy. Yamashita et al. describedfemale sex to be the most important risk factor for tetany,related to the fact that women are more susceptible to calciumand vitamin D deficiency than men (9). Another studyshowed that seasonal variations in calcium homeostasis hadan effect on incidence of postoperative hypocalcemia. Patientswith Graves’ disease were found to be more susceptible tocalcium and vitamin D deficiency during the winter monthsand thus had a higher incidence of postoperative tetany whenthyroidectomy was performed in the winter (11). We did notfind this in our study, however. Another study showed thatpatients with Graves’ disease who had vitamin D deficiencywith high serum alkaline phosphatase levels were at highestrisk for operative tetany (12). They concluded that serum vi-tamin D and alkaline phosphatase levels should be monitoredin patients with Graves’ disease undergoing thyroidectomy.Although we were able to examine alkaline phosphataselevels on our patients, vitamin D levels were not available forall patients in our retrospective review. We did not, however,find a significant association between preoperative alkalinephosphatase levels and postoperative hypocalcemia in eithercontrol or Graves’ disease patients. After completion of thisstudy and reviewing the results of others, we now collectpreoperative vitamin D levels on all patients scheduled forthyroidectomy.

Interestingly, our results demonstrated that older age pa-tients had significantly lower odds of experiencing symptomsof hypocalcemia than younger patients when adjustingfor Graves’ disease ( p-value¼ 0.04). Also, older age patientshad significantly lower odds of starting Rocaltrol thanyounger age patients after adjusting for Graves’ disease( p-value¼ 0.04). Bhattacharyya et al. analyzed data from 517

total thyroidectomy cases and similarly found that youngerage was a significant risk factor for hypocalcemia, whereaspatient sex, indication for surgery, and parathyroid implan-tation were not (13). Others, however, refute this associationnoting that advanced age is a known major risk factor forvitamin D deficiency. Erbil et al. found that age>50 years wasassociated with an increased risk of postoperative hypocal-cemia after total thyroidectomy, thought largely due to lowvitamin D concentrations associated with aging in that pop-ulation (14).

The limitations of this study include the retrospective na-ture of data analysis. In addition, certain data were notavailable on all patients, including vitamin D levels. As statedpreviously, we now collect preoperative vitamin D levels onall patients scheduled for thyroidectomy. Although resultswere adjusted when possible, we do not believe clinicallysignificant differences in the covariates varied between thetwo groups.

This study has changed our clinical practice significantly.Each Graves’ disease patient who is scheduled for total thy-roidectomy is now prescribed calcium supplementation pre-operatively until the day of surgery. Calcium levels are nowchecked both the night of surgery as well as the morning after.Finally, Graves’ disease patients are prescribed calcium sup-plementation empirically upon discharge, even if normo-calcemic as an inpatient. Future studies are planned toexamine whether this practice will alter the occurrence ofsigns and symptoms of hypocalcemia in Graves’ disease pa-tients after thyroidectomy.

Conclusion

Patients with Graves’ disease are more likely to requireincreased dosages of calcium as well as experience tetanypostoperatively than patients undergoing total thyroidec-tomy for other indications, probably due to alterations incalcium homeostasis as there was a trend for their serumcalcium to be lower than patients having thyroidectomy forconditions other than Graves’ disease. This suggests that pa-tients operated upon for Graves’ disease warrant close follow-up, as both inpatients and outpatients, for signs and symp-toms of hypocalcemia.

Disclosure Statement

The authors declare that no competing financial interestsexist.

References

1. Wilhelm S, McHenry C 2010 Total thyroidectomy is superiorto subtotal thyroidectomy for management of Graves’ dis-ease in the United States. World J Surg 34:1261–1264.

2. Efremidou E, Papageorgiou M, Liratzopoulos N, Manolas K2009 The efficacy and safety of total thyroidectomy in themanagement of benign thyroid disease: a review of 932cases. Can J Surg 52:39–44.

3. Falk S, Birken E, Baran D 1998 Temporary post-thyroidectomy hypocalcemia. Arch Otolaryngol Head NeckSurg 114:168–174.

4. Percival R, Hargreaves A, Kanis J 1985 The mechanisn ofhypocalcemia following thyroidectomy. Acta Endocrinol109:220–226.

1282 PESCE ET AL.

5. See A, Soo K 1997 Hypocalcemia following thyroidectomyfor thyrotoxicosis. Br J Surg 84:95–97.

6. Lal G, Ituarte P, Kebebew E, Siperstein A, Duh Q, Clark O2005 Should total thyroidectomy become the preferredprocedure for surgical management of Graves’ disease?Thyroid 15:569–574.

7. Thomusch O, Machens A, Sekulla C, Ukkat J, Brauckhoff M,Dralle H 2003 The impact of surgical technique on postop-erative hypoparathyroidism in bilateral thyroid surgery: amultivariate analysis of 5846 consecutive patients. Surgery133:180–185.

8. Thomusch O, Machens A, Sekulla C, Ukkat J, Lippert H,Gastinger I, Dralle H 2000 Multivariate analysis of risk fac-tors for postoperative complication in benign goiter surgery:prospective multicenter study in Germany. World J Surg24:1335–1341.

9. Yamashita H, Noguchi S, Murakami T, Uchino S, WatanabeS, Ohshima A, Kawamoto H, Toda M, Yamashita H 2000Calcium and its regulating hormones in patients withGraves’ disease: sex differences and relation to postoperativetetany. Eur J Surg 1666:924–928.

10. Yamashita H, Murakami T, Noguchi S, Shiiba M, WatanabeS, Uchino S, Kawamoto H, Toda M, Murakami N 1999Postoperative tetany in Graves’ disease: important role ofvitamin D metabolites. Ann Surg 229:237–245.

11. Yamashita H, Noguchi S, Murakami T, Watanabe S, UchinoS, Oshima A, Kawamoto H, Toda M, Yamashita H 2000Seasonal changes in calcium homeostasis affect the incidence

of postoperative tetany in patients with Graves’ disease.Surgery 127:377–382.

12. Yamashita H, Noguchi S, Murakami I, Uchino S, Wata-nabe S, Ohshima A, Toda M, Yamashita H, Kawamoto H2001 Predictive risk factors for postoperative tetany infemale patients with Graves’ disease. J Am Coll Surg192:465–468.

13. Bhattacharyya N, Fried M 2002 Assessment of the morbidityand complications of total thyroidectomy. Arch OtolaryngolHead Neck Surg 128:389–392.

14. Erbil Y, Bozbora A, Ozbey N, Sever H, Aral F, Ozarmaan S,Tezelman S 2007 Predictive value of age and serum para-thormone and vitamin D3 levels for postoperative hypo-calcemia after total thyroidectomy for nontoxic multinodulargoiter. Arch Surg 142:1182–1187.

Address correspondence to:Martha A. Zeiger, M.D., FACS, FACE

Chief of Endocrine SurgeryProfessor of Surgery, Oncology, Cellular

and Molecular MedicineThe Johns Hopkins University School of Medicine

600 North Wolfe St.Blalock 606

Baltimore, MD 21287

E-mail: mzeiger@jhmi.edu

POSTOPERATIVE HYPOCALCEMIA AFTER THYROIDECTOMY FOR GRAVES’ 1283