a b s t r a c t

Background: Given signs and symptoms of hyperthyroidism are non-specific, thyroid studies may be completed to evaluate patients with vague complaints such as tachycardia in the emergency department (ED). We sought to determine how often a new diagnosis of hyperthyroidism was made in patients who underwent thyroid labora- tory testing in the ED.

Objectives: The primary outcome of this study was a new diagnosis in the ED, or in the following 30 days, of hyperthyroidism or thyrotoxicosis following ED thyroid laboratory testing.

Methods: This was a retrospective chart review study in a single ED of all patients who had a TSH or free T4 ordered from 2007 to 2018, or a TSH value below, or a free T4 higher than the local reference ranges. Patients with a diagnosis of hypothyroidism were excluded.

Results: 12,366 patients underwent thyroid laboratory testing, and 12,244 patients were included. Of included patients, 655(5.4%) had abnormal thyroid studies, and 95(0.8%) patients received a new diagnosis of hyperthy- roidism or thyrotoxicosis. The sensitivity and specificity of tachycardia for the diagnosis of hyperthyroidism was 70%(95% CI 62% to 77%) and 62%(95% CI 61% to 63%), respectively; and the positive and negative likelihood ratios were 1.8(95% CI 1.7 to 2.0) and 0.5(95% CI 0.4 to 0.6), respectively.

Conclusion: This study suggests that thyroid laboratory studies may be best used for a focused assessment for thyroid disorders rather than a screening tool for patients with one symptom of hyperthyroidism.

(C) 2021

1. Introduction

Hyperthyroidism affects approximately 1.3% of people in the United States [1] and is most commonly caused by Graves’ disease and nodular thyroid disease, and less commonly by painless and subacute thyroid- itis. [2] Hyperthyroidism is commonly diagnosed by measuring serum Thyroid stimulating hormone , a pituitary-derived hormone that prompts the thyroid gland to release thyroid hormone into the blood- stream in the form of thyroxine (T4) and triiodothyronine (T3). [3] TSH, because of its high sensitivity and specificity of thyroid gland disor- ders, is commonly used to screen for hyperthyroidism. [4] In one study, 1.3% of emergency department (ED) patients were identified as possibly having hyperthyroidism, as evidenced by a low TSH value, which is sim- ilar to the prevalence of hyperthyroidism in the general population. [5] When severe, hyperthyroidism is life-threatening and requires immediate diagnosis and treatment. When less severe, recognition

* Corresponding author at: Department of Emergency Medicine, Hennepin County Medical Center, 701 Park Ave S, Mail Stop R2, Minneapolis, MN 55415, USA.

E-mail address: [email protected] (A.M. Arens).

remains important to arrange for outpatient follow-up. Diagnosing hyperthyroidism in the ED, however, can be difficult because the signs and symptoms of excess thyroid hormone are non-specific; a partial list of symptoms includes palpitations, hypertension, tremor, anxiety, fatigue, nausea, vomiting, diarrhea, heat intolerance, and diaphoresis.

[6] Many other conditions treated in the ED present with similar symp- toms, such as sepsis syndrome, acute coronary syndrome, substance use, anxiety, pain, among many others. [3] Therefore, the challenge to the emergency medicine physician is to decide when an ED patient requires thyroid hormone testing. Existing ED data examining this question have small sample sizes (<500 patients in the largest), and knowledge is limited for this important clinical question. [5,7,8]

Some physicians may choose to order thyroid hormone studies when a patient has a single symptom consistent with hyperthyroidism (eg, tachycardia or palpitations); others reserve testing for more classic presentations of excess thyroid hormone. Given the rarity of this illness, ordering thyroid studies in scenarios without High clinical suspicion can increase the number of false-positive results, with the potential for harm from misdiagnosis, while adding cost and time to an emergency department visit. [9]

https://doi.org/10.1016/j.ajem.2021.12.038

0735-6757/(C) 2021

To improve utilization of thyroid studies in the emergency depart- ment, we sought to determine the proportion of patients who under- went ED thyroid hormone testing who were newly diagnosed with hyperthyroidism or thyrotoxicosis, and to describe the presenting characteristics and symptoms of these patients.

1. Methods
   1. Study design and setting

This was a retrospective chart review study in a single, urban, aca- demic ED with an associated emergency medicine residency with an an- nual census of approximately 100,000 patients. Thyroid hormone studies are available 24 h per day, including TSH, free T4 and T3, and may be ordered by any clinician at their discretion. There is no institu- tional guideline nor protocol to inform indications for ordering thyroid testing. The local institutional review board approved this protocol.

• 1. Selection of participants

We included all patients who had a TSH or free T4 ordered in the ED from January 1, 2007 through December 31, 2018, excluding those diag- nosed with hypothyroidism in the ED, as defined by any ED diagnosis containing the following phrases in any part of the diagnosis name: hy- pothyroidism, hypothyroid, or Hashimoto’s. This yielded the base popu- lation.

We then performed additional manual chart review for patients with a TSH value below the local reference range (<0.27 mIU/L) or an elevated free T4 higher than the local reference range (>1.6 ng/dL).

• 1. Methods of measurement

For all patients, a blinded hospital data analyst extracted the follow- ing from the medical record: age; gender; chief complaint; disposition (admit versus discharge); vital signs at ED arrival; highest heart rate in the ED; laboratory results for TSH, free T4, and all T3 studies; and pri- mary final diagnosis based upon ICD-9 or ICD-10 codes. This determina- tion was made by the individual reviewer.

For patients eligible for additional manual chart review (abnormally low TSH or abnormally high free T4), trained abstractors used a stan- dardized data collection form in REDCap [10] to record whether the pa- tient had specific symptoms of hyperthyroidism (including recent weight loss, insomnia, diarrhea, heat intolerance, anxiety or irritability, increased sweating, tremors, thinning skin or brittle hair, fatigue or muscle weakness, palpitations, decreased menstrual cycles, light- headedness, nausea or vomiting, or seizure) or diagnoses possibly re- lated to hyperthyroidism (including coma or obtundation, congestive heart failure, atrial fibrillation, jaundice or hepatic failure). The abstrac- tor also documented whether there was prior history of thyroid disease, ongoing medication for thyroid disease, the rationale for ordering the thyroid laboratory studies in the ED (if documented), final ED diagnosis, whether there was subsequent diagnosis of hyperthyroidism in the 30 days following the index ED encounter, and whether treatment for hy- perthyroidism was provided in the ED or in the following 30 days. To determine whether the patient had hyperthyroidism, the abstractor re- viewed the notes from the emergency department, inpatient stay (if ap- plicable), and outpatient clinics and affiliated outpatient clinics. A patient was considered to have a diagnosis of hyperthyroidism if the treating physician in the ED or in the following 30 days formally diag- nosed the patient with a hyperthyroid state in the patient note or final diagnoses based upon ICD-9 or ICD-10 codes. This included the follow- ing diagnoses: Graves’ disease, toxic thyroid nodule(s), iatrogenic hyperthyroidism from excess levothyroxine, subacute thyroiditis, sub- clinical hyperthyroidism, and hyperthyroidism or thyrotoxicosis, not otherwise specified.

• 1. Outcomes

The primary outcome of this study was a new diagnosis in the ED, or in the following 30 days, of hyperthyroidism or thyrotoxicosis. Patients who were formally diagnosed with hyperthyroidism or thyrotoxicosis by a treating physician at the index ED visit or in the following 30 days met this outcome.

• 1. Primary data analysis

We used descriptive techniques to analyze the data, presenting counts and proportions for baseline characteristics stratified by whether the thyroid testing was abnormal and whether there was a new diagno- sis of hyperthyroidism. We describe in detail the characteristics of pa- tients who had a diagnosis of hyperthyroidism, separating those with known and new diagnoses. There was no formal statistical testing. Be- cause tachycardia is a symptom that commonly prompts a clinician to order thyroid studies, we calculated test characteristics of this symptom as it related to the primary outcome. A second independent reviewer (YS, emergency medicine) recorded outcomes for 10% of patients to cal- culate Interobserver agreement for the primary outcome (87.5% agree- ment, ? = 0.59).

1. Results
   1. Main results

Between January 1, 2007 and December 31, 2018, 12,366 patients underwent thyroid laboratory testing. After excluding 122 patients diagnosed with hypothyroidism, the base population included 12,244 patients. Of these, 11,589 (94.8%) patients had normal thyroid studies (see Fig. 1), while 655 (5.4%) had abnormal thyroid studies. A diagnosis of hyperthyroidism or thyrotoxicosis was made in 163 (1.3%, 95% CI 1.1% to 1.6%) patients, including 95 (0.8%) patients receiving a new

Fig. 1. Flow of study patients, stratified by laboratory results and final diagnosis.

diagnosis, and 68 (0.6%) patients with preexisting thyroid disease. Pa- tient characteristics, vital signs, laboratory results, chief complaints, and disposition are shown in Table 1, stratified by laboratory results and final diagnosis.

Tachycardia was present in 38% of patients with normal thyroid studies, and in 70% of patients with hyperthyroidism. HAs thyroid stud- ies are often ordered for tachycardia, we calculated test characteristics for this symptom: the sensitivity and specificity of tachycardia for the diagnosis of hyperthyroidism was 70% (95% CI 62% to 77%) and 62% (95% CI 61% to 63%), respectively; and the positive and negative likeli- hood ratios were 1.8 (95% CI 1.7 to 2.0) and 0.5 (95% CI 0.4 to 0.6), re- spectively.

Details for patients diagnosed with hyperthyroidism or thyrotoxico- sis, including medication use, symptoms of excess thyroid hormone, specific diagnoses, disposition, and treatment received, are detailed in Table 2. Tachycardia, without any additional symptoms of hyperthy- roidism, was identified in 15 (9.2%) patients diagnosed with hyperthy- roidism, which represents 0.1% of all patients who underwent thyroid study testing during the study period.

Reasons for ordering thyroid studies in the ED are shown in Table 3. For patients without a known diagnosis of hyperthyroidism, only 64% had a reason documented.

• 1. Sensitivity analysis

We performed a sensitivity analysis that included all patients, in- cluding those diagnosed with hypothyroidism. This was meant to be more inclusive, as some thyroid studies were ordered for suspected hy- pothyroidism rather than for suspected hyperthyroidism — it was not possible to specifically determine the cause of ordering for most pa- tients. In this analysis that included 12,366 patients, 777 patients (6.3%, 95% CI 5.9% to 6.7%) had abnormal thyroid studies. A diagnosis of hyperthyroidism or thyrotoxicosis was made in 163 (1.3%, 95% CI

1.1% to 1.5%) patients, including 95 (0.8%) patients receiving a new diag- nosis, and 68 (0.5%) patients with preexisting thyroid disease.

1. Discussion

In this large single-center study, there was a new diagnosis of hyper- thyroidism made in 0.8% of emergency department patients with thy- roid studies ordered, suggesting a low overall diagnostic utility. There are limited data on best practices for ordering thyroid studies in the ED. The largest prior study, with 464 patients, implemented targeted physician education on appropriate reasons to order thyroid studies and found that ED thyroid studies were suggestive of hyperthyroidism in 7% of patients, with 2% having a new diagnosis of hyperthyroidism.

[5] The authors concluded that routine ED TSH testing can be valuable for ruling out thyroid pathology, but rarely does it lead to the discovery of a new diagnosis of hyperthyroidism. A smaller study of 104 patients showed 12% of patients to have laboratory findings consistent with hy- perthyroidism, with about 80% of tests being approved as appropriate by an endocrinologist. [7] In comparison, the current study, with a much larger study size, identified a diagnosis of hyperthyroidism in 1.3% of patients and a new diagnosis in just 0.8% of patients. We suspect the larger size of our current might be more indicative of general emer- gency medicine practice, compared to studies with smaller sizes, though the exact generalizability of these results is unknown.

The symptoms of hyperthyroidism are non-specific and have signif- icant overlap with myriad maladies more commonly encountered in the ED. For example, in our study, the most common presenting signs and symptoms of patients with a new diagnosis of hyperthyroidism included: tachycardia (78%), palpitations (31%), unintentional weight loss (28%), fatigue (26%), and nausea or vomiting (25%). Patients with a diagnosis of hyperthyroidism in the setting of known thyroid dis- ease presented with similar chief complaints and symptomatology. Given these signs and symptoms may be indicative of any number of

Table 1

Baseline characteristics

Characteristic	Normal thyroid studies	Abnormal thyroid studies	Hyperthyroidism in the setting of known thyroid diseasea	New diagnosis of hyperthyroidisma
	N = 11,589	N = 655	N = 68	N = 95
Age, median (IQR)- years	48 (33-62)	48 (32-59)	53 (37-64)	41 (31-55)
Female gender - no. (%)	6876 (59)	453 (69)	61 (90)	60 (63)
Initial heart rate, median (IQR) - beats per min	88 (75-106)	100 (83-117)	103 (80-118)	114 (100-130)
Highest heart rate, median (IQR) - bpm	93 (79-112)	105 (86-123)	108 (89-121)	122 (105-141)
Highest heart rate > 100 bpm - no. (%)	4432 (38)	356 (54)	40 (59)	74 (78)
Systolic blood pressure, median (IQR) - mm Hg	134 (119-149)	132 (119-147)	130 (118-150)	138 (122-152)
Temperature, median (IQR) - Celsius	98.1 (97.7-98.4)	98.1 (97.7-98.6)	98.1 (97.5-98.6)	98.2 (97.7-98.6)
Thyroid stimulating hormone ordered - no. (%)	11,516 (99)	644 (99)	67 (99)	93 (98)
Value, median (IQR) - mIU/L	1.8 (1.1-3.1)	<0.1 (<0.1 to 0.2)	<0.1 (<0.1 to 0.1)	<0.1 (<0.1 to <0.1)
Free thyroxine ordered - no. (%)	1100 (9)	381 (58)	41 (60)	72 (76)
Value, median (IQR) - ng/dL	1.1 (1-1.3)	1.8 (1.5-3.9)	2.4 (1.8-5.2)	4.5 (2.2-7.7)
Top 10 chief complaintsb - no. (%)
Confusion / Altered mental status	1256 (11)	72 (11)	15 (22)	6 (6)
Palpitations / fast heart rate	1222 (11)	62 (9)	24 (35)	13 (14)
Chest pain	1077 (9)	75 (11)	14 (21)	8 (8)
Weakness / fatigue	746 (7)	30 (5)	8 (12)	3 (3)
Dizziness	631 (5)	19 (3)	5 (7)	2 (2)
Shortness of breath	462 (4)	29 (4)	9 (13)	4 (4)
Abdominal pain	372 (3)	30 (5)	8 (12)	4 (4)
Headache	310 (3)	16 (2)	2 (3)	2 (2)
Fall	313 (3)	12 (2)	2 (3)	1 (1)
Syncope	216 (2)	7 (1)	0	0
Vomiting	163 (1)	13 (2)	3 (4)	3 (3)
upper respiratory infection	52 (<1)	4 (1)	8 (12)	8 (8)
Sore throat	72 (1)	12 (2)	3 (4)	3 (3)
Diarrhea	47 (<1)	8 (1)	3 (4)	1 (1)
Disposition - no. (%) Discharged from the ED	6620 (57)	336 (51)	23 (34)	41 (43)
bpm = beats per minute.

^a These columns show patients that are also accounted for in the column titled “Abnormal thyroid studies”

^b This lists the top 10 chief complaints for each group, including counts for chief complaints that were top in one group but not the other.

Table 2

Details for patients with a final diagnosis of hyperthyroidism

Table 3

Reasons for ordering thyroid studies in the Emergency Department

Unintentional weight loss 3 (4) 27 (28)

Parameter - no. (%)	Hyperthyroidism with known thyroid disease	New diagnosis of hyperthyroidism		Reasona - no. (%)	History of or active thyroid	No prior diagnosis of thyroid disease
	N = 68	N = 95			disease	(N = 270)
Taking levothyroxine before ED visit	41 (60)	0			(N = 385)
Taking T3 before ED visit	0	0		History of hypothyroidism, as sole reason	120 (31)	NA
Taking methimazole or	11 (16)	0		History of hyperthyroidism, as sole reason	106 (28)	NA
propylthiouracil before ED visit				Active thyroid disease, with medication	54 (14)	NA
Taking beta-blocker before ED visit Symptoms of hyperthyroidism present Tachycardia in the ED	8 (12) 40 (59)	2 (2) 74 (78)		non-adherence Evaluate for hyperthyroidism or thyrotoxicosis, without more specific reason	66 (17)	59 (15)
Tachycardia as only ED finding,	2 (3)	13 (14)		Tachycardia	17 (4)	33 (9)
without other symptoms				Goiter	15 (4)	27 (7)
Any symptom, excluding tachycardia Palpitations	63 (93) 20 (29)	78 (82) 29 (31)		Concern for thyroid gland dysfunction, without more specific reason Thyroid nodule	18 (5) 0	20 (5) 12 (3)
Atrial fibrillation, new diagnosis	5 (56)	13 (14)		Other	105 (27)	47 (12)
Signs of congestive heart failure	8 (12)	9 (9)		No reason listed	7 (2)	97 (25)

^a Patients could have more than one reason, hence the total exceeds the number of

Frequent bowel movements or diarrhea

13 (19) 13 (14)

patients

Insomnia 4 (6) 7 (7)

Anxiety, nervousness, or irritability

15 (22) 12 (13)

patients with multiple symptoms consistent with hyperthyroidism,

Seizure 4 (6) 2 (2)

Coma 1 (2) 0

Heat intolerance 5 (7) 5 (5)

Increased sweating 3 (4) 11 (12)

Tremor 5 (7) 9 (9)

Thinning skin or brittle hair 1 (1) 4 (4)

Fatigue or muscle weakness 23 (34) 25 (26)

Nausea or vomiting 16 (24) 24 (25)

and, for patients with tachycardia, first consider more common diagno- ses such as hypovolemia, sepsis, alcohol withdrawal, and others.

Our study suggests that Vital sign abnormalities including tachy- cardia with additional symptoms of hyperthyroidism were seen much more frequently in patients with a new diagnosis of hyperthy- roidism. Thyroid studies may be useful tools in evaluating patients

with signs and symptoms suggestive of hyperthyroidism in the ED,

Decreased menstrual cycle (for women)

Specific diagnosis of hyperthyroidism Hyperthyroidism or thyrotoxicosis, not otherwise specified

Iatrogenic hyperthyroidism from excess levothyroxine

History of hyperthyroidism, not

taking anti-thyroid medications

0/61 0/60

23 (33) 68 (72)

28 (41) 0

10 (15) 0

but are of limited utility based upon chief complaint or isolated vital sign abnormalities.

1. Limitations

This study is subject to multiple limitations. As a retrospective re- view, this study is susceptible to bias in recording outcomes, which we attempted to mitigate by choosing objective clinical data whenever

Graves’ disease 3 (4) 13 (14)

Subclinical hyperthyroidism 3 (4) 9 (9)

Toxic nodule(s) 1 (1) 2 (2)

Subacute thyroiditis 0 3 (3)

Discharged from the ED 23 (34) 41 (43)

possible and relying on final physician diagnoses. In addition, the clini- cal information gathered, including chief complaint and discharge diag- nosis, are dependent upon completeness of the existing medical record and subject to error. Furthermore, relying on medical record review and

discharge diagnoses may inadvertently exclude patients. Given that our

Treatment for hyperthyroidism in the ED

Treatment for hyperthyroidism in the following 30 days

ED = emergency department.

9 (13) 20 (21)

63 (93) 70 (74)

study population was from a single center the findings might not be generalizable to centers with different ordering practices. This study was designed to identify patients with evidence of hyperthyroidism or thyrotoxicosis, and did not investigate the utility of testing to identify hypothyroidism or any other specific thyroid disorders.

diagnoses, considering a patient’s entire presentation, rather than iso- lated symptoms, as a basis to decide whether thyroid testing is indicated might improve ordering practices.

Similarly, the most common chief complaints of patients ultimately determined to have a new diagnosis of hyperthyroidism or thyrotoxico- sis were nonspecific and included: palpitations or fast heart rate in 13 (14.0%; 0.1% of total patients), chest pain in 8 (8.0%; 0.07% of total), and upper respiratory infection symptoms in 8 (8.0%; 0.07% of total) pa- tients. Considering that a new diagnosis of hyperthyroidism was made in only 0.1% patients presenting with vague symptoms of palpitations or fast heart rate who underwent thyroid testing, thyroid studies may have limited utility in patients with these chief complaints alone.

In this study, tachycardia was one of the most commonly reported reasons to order thyroid studies (9%). e The positive and negative likeli- hood ratios for this sign to diagnose hyperthyroidism were marginal (1.8 and 0.5, respectively) and therefore do not alter pre-test probability significantly. Thus, it may be prudent to reserve thyroid testing for

1. Conclusion

In this ED population, a new diagnosis of hyperthyroidism or thyro- toxicosis was made in less than 1% of all patients who underwent thy- roid hormone testing. Furthermore, less than 0.2% of patients with non-specific signs and symptoms such as isolated tachycardia, palpita- tions, or anxiety received a new diagnosis of hyperthyroidism or thyro- toxicosis after thyroid studies were completed in the ED. Our study suggests that thyroid laboratory studies may be best used for a focused assessment for thyroid disorders rather than a screening tool for pa- tients with one symptom of hyperthyroidism.

Previous presentation

Accepted for abstract presentation at Society for Academic Emer- gency Medicine 2020, Denver CO. Meeting cancelled secondary to COVID-19.

Financial disclosures

This was an unfunded investigation. No authors have any conflicts of interest to report.

Author contributions

AA, BD conceived and designed the study. MB, EM, and YSR contributed to data collection and monitoring. BD performed the data analysis. AA, MB, and BD drafted the initial manuscript; all authors contributed substantially to revisions; BD and AA made final editorial decisions.

Declaration of Competing Interest

None.

Acknowledgements

None.

References

1. Hollowell JG, Staehling NW, Flanders WD, Hannon WH, Gunter EW, Spencer CA, et al. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and nutrition examination survey (NHANES III). J Clin Endocrinol Metab. 2002;87:489-99.
2. Ross DS, Burch HB, Cooper DS, Greenlee MC, Laurberg P, Maia AL, et al. 2016 American thyroid association guidelines for diagnosis and management of hyper- thyroidism and other causes of thyrotoxicosis. Thyroid. 2016;26:1343-421.
3. Devereaux D, Tewelde SZ. Hyperthyroidism and thyrotoxicosis. Emerg Med Clin North Am. 2014;32:277-92.
4. de los Santos ET, Starich GH, Mazzaferri EL. Sensitivity, specificity, and cost- effectiveness of the sensitive thyrotropin assay in the diagnosis of thyroid disease in ambulatory patients. Arch Intern Med. 1989;149:526-32.
5. Giacomini A, Chiesa M, Carraro P. Urgent thyroid-stimulating hormone testing in emergency medicine: a useful tool? J Emerg Med. 2015;49:481-7.
6. De Leo S, Lee SY, Braverman LE. Hyperthyroidism. Lancet. 2016;388:906-18.
7. Koroscil MA, Marco CA. Utility of thyroid testing in the ED. Am J Emerg Med. 2016; 34:2029-30.
8. Apostol M, Keeran M, Klempf N, McCoskey V, Ernst AA, Weiss SJ, et al. Thyroid stim- ulating hormone testing in ED evaluation of patients with atrial fibrillation and var- ious Psychiatric diagnoses. Am J Emerg Med. 2019;37:1114-7.
9. Pimentel L, Hansen KN. Thyroid disease in the emergency department: a clinical and laboratory review. J Emerg Med. 2005;28:201-9.
10. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)-A metadata-driven methodology and workflow process for provid- ing translational research informatics support. J Biomed Inform. 2009;42(4):377-81.