Hyperthyroidism, second to diabetes, is the second common endocrinal disorder in the general population. The most common causes of hyperthyoidism are diffuse toxic goiter (Graves' disease) (GD) (˜1%), solitary toxic nodules (STN) (˜0.1%) and toxic multinodular goiter (TMG) (˜0.53%) .
Although thyrotoxic emergencies are not common causes for admission to critical care units, thyrotoxicois is not uncommonly suspected and thoroughly searched for in many admitted patients. This is due to many reasons: first, the overlap of thyrotoxic symptoms and signs with many diseases encountered in critical care patients especially those with severe cardiac and chest diseases and patients with cardiac neurosis such as tachycardia, arrhythmias, dyspnea, sweating, weight loss, nervousness, severe anxiety and panic, staring look, collapsing pulse and hand tremors. Second, thyotoxicosis may not be clinically manifest (subclinical or masked) especially in old age and in this situation, if not properly and early diagnosed, may aggressively deteriorate many seriously ill patients like those with malignant arrhythmias, severe heart failure and acute coronary syndromes [2,3]. Third, some thyrotoxic patients may be initially clinically presenting with thyotoxic cardiac complications like atrial fibrillation, ventricular tachycardia, atrial or ventricular ectopics, thyrotoxic heart failure with absence, or even not clinically prominent, common thyrotoxic symptoms [4,5]. Fourth, thyromegaly, which is a common sign of thyrotoxicosis that is searched for to establish diagnosis may not be evident in about one half of thyotoxic patients .
In critically ill patients, a fifth conflict is added in diagnosis of thyrotoxicosis especially in mild (usually early) Graves' disease and in patients with pretoxic thyroid nodules. This conflict is caused by the inconsistent laboratory findings of thyroid hormones and TSH due to acute disruption of the normal steady state of these hormones [6–9]. Another problem in early diagnosis of thyrotoxicosis in seriously ill patients is added by the delay of assessment of thyroid hormones in most laboratories for at least one or two days.
Thyroid ultrasound (US) has been used as an adjunctive tool to assess different thyroid diseases in specialized centers and outpatient clinics. This test is rapid, safe, feasible, easily available and informative in many cases if performed by expert personnel [10–14]. However, its use in the ICU to diagnose thyrotoxicosis has not been studied.
Aim of the study
This study aimed to:
- Assess the thyrotoxic burden in critically ill patients admitted in ICU.
- Evaluate the value of thyroid ultrasound as a rapid and feasible bedside diagnostic tool in early evaluation of suspected thyroid disease in critically ill patients especially those with life threatening cardiac emergencies.
Patients and methods
We studied 854 patients who had been admitted to the general intensive care unit of Critical Care Department of El-Fayoum University Hospital and the intensive care unit of Internal Medicine Department of El-Minia University Hospital.
In all admitted patients, those suspected to have thyrotoxicosis were selected and thoroughly studied.
Patient selection was based on
- Presence of suspected thyrotoxic symptoms and signs (palpitations, insomnia, nervouseness, panics, weight loss, appetite changes, sweating, heat intolerance, antithyroid drug therapy, T4 replacement therapy, past history of thyroid disease or thyroid operation, presence of goiter or thyrotoxic eye signs or other suspecting thyrotoxic symptoms and signs),
- Presence of persistent unexplained sinus tachycardia, or presence of atrial fibrillation, PSVT, atrial or ventricular premature beats, ventricular tachycardia or ventricular fibrillation at presentation or during ICU admission.
All suspected patients were subjected to:
- Thorough clinical assessment including history, general examination including heart, chest, abdominal and neurological examination) and local thyroid examination using World Health Organization 1960 palpation criteria for determination of goiter: (group 0: thyroid gland not palpable or visible; IA: thyroid gland is only palpable but not visible; group IB: thyroid gland is palpable and visible with neck only in raised position; group II: thyroid gland is palpable and visible with neck not in raised position; group III: gross enlargement of the thyroid gland) . We opted to use WHO 1960 rather than WHO 1994 palpation criteria as the earlier was more descriptive for the clinical size of goiter and hence would compare better with the detailed ultrasound documented goiter. Thyroid palpation was done by either the classic posterior approach or by the alternative anterior approach. Thyrotoxic cardiac emergencies were considered when there was cardiac emergency suspected to be caused or compromised by thyrotoxicosis (e.g. cardiac arrhythmias, heart failure).
- Laboratory assessment including FT3, FT4, TSH had been done on the day of ICU admission and if needed on the day of discharge and then followed in the outpatient clinic in proved cases. Serum sodium and potassium, arterial blood gases, in addition to all routine and required investigations according to the patient condition had been done as needed.
- Thyroid ultrasound.
- Echocardiographic assessment.
- Continuous ECG monitoring.
- Thyroid scanning with 99Tc in selected cases after discharge.
Patients with inserted tracheostomy tubes were excluded due to difficulty in procedure of thyroid ultrasound examination.
Thyroid ultrasonography was done using a Toshiba scanner with a 10 MHZ linear transducer. All patients were examined with the neck hyperextended and a pillow under the shoulders. Thyroid volumetry was estimated by summation of both thyroid lobe volumes. The volume of each thyroid lobe was calculated according to the formula of calculation of the volume of an ellipsoid mass by multiplication of the maximum diameters of the thyroid lobe (width, depth and length) in centimeters by the correction factor 0.479 to express the volume in milliliters [16–18]. Goiter was defined by US if total thyroid volume was more than 25 ml in males and 18 ml in females) [17,18].
Thyroid echogenicity was assessed as normal (like echogenicity of parotid or submandibular salivary gland), hypoechogenic, hyperechogenic, homogenous or heterogenous hyper or hypoechogenic, or diffusely inhomogenous hypoechogenic [17,18].
Thyroid nodules (solitary or multinodular) or the postablative remnant of the nodules were assessed by ultrasound volumetry, echogenicity and invasion to surrounding structures. Regional lymph node assessment had been done by ultrasound .
Collected data were organized, calculated and tabulated. The results were expressed in numbers and percentages. For quantitative data, the mean and standard deviation (SD) were calculated. Comparison of diagnostic accuracy of US and clinical suspicion in diagnosis of goiter and thyrotoxicosis was made by measuring the diagnostic accuracy parameters: sensitivity, specificity, positive and negative predictive values, and overall accuracy.
Demographic data of the patients
We studied 854 patients aged 21–78 years (54.2 + 6.6 years; mean + SD) who had been admitted in the general intensive care units of Critical Care Department of El-Fayoum University Hospital and Internal Medicine Department of El-Minia University Hospital. Of these patients, 113 had been evaluated for clinically suspected thyrotoxcosis. Patient demographics and cause of admission are shown in Table 1.
Diagnosis of thyroid disease
Diagnosis of goiter
Goiter was clinically suspected in 61 out of 113 patients suspected to be thyrotoxic (54% of all suspected cases and 1.32% of all admitted cases ranging from grade IB to grade III according to WHO classification). Ultrasound thyroid volumetry had proved presence of goiter (defined as increased total thyroid volume more than 25 ml in males and 18 ml in females) [17,18] in 42 (out of 61 patients clinically suspected to have goiter) (68.9%) [including cases of GD with clinically evident goiter (12 cases), cases with clinically evident thyroid nodules (8 cases) and cases of diffuse simple goiter (22 cases)] and was beneficial in excluding diagnosis of goiter as an important clue of most thyrotoxic states in the other 19 patients (31.1%) (Fig. 1). In the other 52 patients who had been clinically suspected to have thyrotoxicosis without goiter suspicion, goiter was proved in 8 cases (5 cases with GD and 3 cases with toxic nodules). Comparison of diagnostic accuracy parameters of clinical versus ultrasound assessment of goiter is illustrated in Table 2.
Diagnosis of Graves' disease
Graves' disease was proved in 17 cases (15%) (12 females and 5 males) as described by minimal total thyroid volume of 13.5 ml in naive patients and characteristic echogenicity [17–20]. Twelve cases (70.5%) had clinically evident goiter (grade IA – III) and 5 cases (29.5%) were classified grade 0. Thyroid ultrasound examination showed increased thyroid volume of all cases (ranging 23.3–74.2 ml in females and 26.3–67.4 ml in males) (Table 3).
Thyroid echogenicity was characteristic and was found as 2 forms: the first is diffuse homogenous hypoechogenicity in both lobes (Fig. 2-A), the second is diffuse hetrogenous hyperechogenicity with no definite, even tinny, nodular lesions (Fig. 2-B).
In 4 cases suspected for Graves' disease, thyrotoxicosis was recurrent after subtotal thyriodectomy and remnant of the thyroid gland volumes was estimated by ultrasound and ranged from 23.3 to 29.3 ml with diffuse hypoechogenicity. Active GD was finally confirmed in 3 cases by hormonal assessment and one case proved to be in remission.
Diagnosis of toxic and pretoxic thyroid nodules (Table 4)
Diagnosis of definite toxic thyroid nodules was found in 11 cases (9.7%) (7 females and 4 males). Only 8 cases (72.7%) could be detected clinically (grade IA – II) and 3 cases (27.3%) were clinically overlooked (grade 0).
Eight cases (72.7%) had toxic nodules (defined as increased Ft3, Ft4, suppressed TSH and increased radiotracer uptake by the nodule with suppression of extranodular thyroid tissue by thyroid scintiscan: hot nodules) [21,22] and 3 cases (27.3%) had pretoxic nodules. These 3 cases had mild increase in Ft3, ft4, suppressed TSH with increased radiotracer uptake by the nodule with no suppression of the extranodular thyroid tissue by thyroid scintiscan: worm nodules [21,22].
Thyroid nodules were solitary in 7 cases (63.6) and multinodular in 4 cases (36.4%). Nodule volumes ranged 12.3–35.7 ml and most of them (N: 10) (90.9%) were hyperechogenic in character, one case was isoechogenic (9.1%) and 5 cases (45.5%) had nodules with a peripheral halo (reflecting normal compressed thyroid tissue) (Fig. 3). Nodule volumes ranged from 12.3 to 35.7 ml.
Diagnosis of thyrotoxicosis in critically ill patients including critically ill cardiac patients
Out of the 113 suspected thyrotoxic patients, 76 (67.3% of suspected patients) were admitted as critically ill cardiac patients and presented mainly by cardiac arrhythmias (N: 51), heart failure (N: 15), acute chest pain (N: 6) and malignant hypertension (N: 4). Out of these 76 patients, thyrotoxicosis was proved in 21 patients (27.6%). In the other suspected patients (N: 37 patients; 32.7% of the 113 suspected patients) who were admitted for various clinical critical presentations including respiratory failure, cerebrovascular stroke, shock, acute renal failure and diabetic ketoacidosis, thyrotoxicosis was proved in 7 patients (18.9%).
In all patients suspected for thyrotoxicosis (N: 113), most of them had tachyarrhythmias (with one or more forms, including life threatening arrhythmias) and some were on antiarrhythmic drugs having negative chronotropic agents and presented with falsely normal heart rate. Details of different cardiac arrhythmias of suspected patients and proved thyrotoxic patients are illustrated in Table 5. In addition, six proved thyrotoxic patients had heart failure (out of 15 patients with heart failure clinically suspected for thyrotoxicosis i.e. 40%): 2 of them had rheumatic heart disease, 1 had dilated cardiomyopathy, 1 had ischemic cardiomyopathy and 2 patients had no structural heart disease and diagnosed as thyrotoxic heart failure). One proved thyrotoxic patient had STEMI (out of 6 clinically suspected thyrotoxic patients with acute chest pain i.e. 16.6%).
Overall thyrotoxic burden in ICU (Table 6)
In 854 patients admitted in ICU, 113 patients were suspected to be thyrotoxic (13.2%). Of these 113 patients, 17 had Graves' disease (15% of suspected patients and 2% of all admitted patients) and 11 patients had toxic nodules (9.7% of suspected patients and 1.3% of all admitted patients). This means that 28 cases (24.7% of suspected patients and 3.3% of all admitted patients) proved to be thyrotoxic. Masked thyrotoxicosis, in which the patient had few easily overlooked thyrotoxic symptoms and signs, was proved in 16 cases (59.3%) whereas 12 cases (40.7%) had either moderate or marked thyrotoxic manifestations. The stepwise approach in our study for diagnosis of thyrotoxicosis by clinical examination, US assessment and final laboratory assessment is shown in Fig. 4. Details of overall thyrotoxic burden are illustrated in Table 6.
Diagnostic accuracy of US versus clinical assessment in diagnosis of thyrotoxicosis
Comparison of diagnostic accuracy parameters of clinical versus ultrasound assessment of thyrotoxicosis in the 113 studied patients suspected with thyrotoxicosis is shown in Table 7. It should be mentioned that true and false negative parameters of clinical assessment of thyrotoxicosis were not assessed as this would need laboratory assessment of all admitted patients. Hence, sensitivity, specificity, negative predictive value and overall accuracy of the clinical evaluation could not be assessed.
This study had 2 aims. First, was to assess the thyrotoxic burden in critically ill patients admitted in ICU. The second was to assess the role of thyroid ultrasound as a rapid and feasible bedside diagnostic tool in accurate and early evaluation of suspected thyroid disease in critically ill patients especially those with life threatening cardiac emergencies.
Initially, it is to be mentioned that diagnosis of thyrotoxicosis is a constellation of clinical, laboratory and thyroid imaging parameters (ultrasound and radioisotope scanning) which are all complementary to each others. Whereas the clinical and laboratory parameters have many fallacies especially in the ICU and the thyroid radioisotope scanning is not readily performed in the ICU in addition to its known diagnostic limitations, the value of thyroid US in bedside diagnosis of thyrotoxicosis in emergency situations should be emphasized. To our knowledge, this is the first study to assess the feasibility of thyroid US in the diagnosis of thyrotoxicosis in emergency settings.
It was very important to evaluate the global burden of thyrotoxic presentation and complications in the general ICU. This was of great importance to avoid overestimation or underdiagnosis of thyroid diseases especially in the common conflicting issue of the euthyroid sick syndrome encountered in ICU. Clinical and laboratory difficulties commonly encountered in urgent diagnosis of thyrotoxicosis in critically ill patients created a need to search for a rapid, feasible and accurate bedside tool to confirm or exclude the diagnosis and hence to allow early and proper management of thyrotoxicosis and its life threatening co-morbid conditions especially in hemodynamically unstable patients.
Thyrotoxic burden in the general ICU in our study was found in 3.3% of all admitted patients. GD was encountered in 2% and toxic nodules represented 1.3% of admitted patients. Final diagnosis of thyrotoxicosis was proved in 24.7% of suspected cases, of whom 59.3% had masked thyrotoxic symptoms and signs and could be easily overlooked. In a previous study of prevalence of thyroid disease in 2805 attendants of the outpatient clinic of Internal Medicine Department in El-Minia University Hospital, 1.06% had GD and 0.64% had toxic nodules . Thus, thyrotoxic burden in the ICU was found to be about double the burden encountered in the outpatient clinic which also came hand in hand by most reports concerning prevalence of thyrotoxicosis in the general population . This may reflect the impact of thyroid dysfunction on co-morbid conditions mandating ICU admission, and/or the impact of co-morbidity on unmasking or aggravating thyrotoxicosis.
It was well noted that most of our patients (67.3%) with suspected thyrotoxicosis had cardiac problems especially cardiac arrhythmias (45.1%). These findings highlighted the increased occurrence of thyrotoxicosis among patients with various tachyarrhythmias and emphasize meticulous search for thyrotoxicosis in such patients.
Data concerning the precise incidence of different tachyarrhythmias in thyrotoxic patients are lacking except for sinus tachycardia which occurs in nearly all patients (except for patients taking agents with negative chronotropic effect and case reports of severe thyrotoxicosis presented with complete heart block) and AF which occurs in 5–15% of thyrotoxic patients as reported by different population studies . In this study, AF was found to occur in a higher percentage than reported by different studies. The higher incidence of AF may be attributed to the type of patients in our study, the thyrotoxic critically ill patients, who may have additional contributing factors for the development of AF than patients with mere thyrotoxicosis. However, further large sampled study may be needed to evaluate the relation of thyrotoxicosis with other types of tachyarrhythmias found in our study.
Thyroid ultrasound had been used as an adjunctive tool to assess different thyroid diseases in specialized centers and outpatient clinics. However, its use in the ICU patients for rapid or complementary assessment of thyrotoxic emergencies had not been previously studied.
Thyroid ultrasound had proved to be beneficial in diagnosis of goiter. Goiter is defined as thyroid enlargement of more than 18 ml in females and 25 ml in males [17,18]. Goiter per se is an important clinical clue in diagnosis of most cases of thyroxicosis. Assessment of thyroid volumetry by bedside thyroid US was an invaluable tool in confirmation and exclusion of the fallacious palpation of goiter in our study (31.1% falsely felt and 29.5% and 27.3% missed goiter and nodule, respectively). However, it should be noted that clinical palpation of the thyroid may have been inaccurate in recumbent, especially intubated patients.
Echogenicity of the thyroid gland by US, along with increased thyroid volume, could instantly and correctly diagnose patients with previously undiagnosed or diagnosed but remitted GD, as well as those with recurrent GD after medical or ablative therapy. Thyroid volume greater than 16.5 ml is essential to diagnose GD in naive patients (GD in activity and not under treatment). In all patients with active GD, thyroid US showed either diffuse homogenous hypoechogenicity or diffuse hetergenicity with no definite nodular lesion.
One important issue that could create confusion is that about 30% of patients of old remittent cases of GD retain the same US image of diffuse hypoechogenicity for many years and subsequently resembling the US image of active GD. In such a case, the clinical presentation and hormonal assay of thyrotoxicosis would solve the dilemma. In our study, we had 4 cases suspected to have recurrence of GD; 3 cases suspected to have postoperative recurrence and one case suspected to have relapse of the disease. In these suspected cases, presence of GD per se, whether recent, recurrent or inactive had been confirmed by US (increased thyroid volume and characteristic thyroid echogenicity). This ultrasound picture along with presence of clinical thyrotoxic manifestations pointed to activity rather than remission of GD. However, 3 of those critically ill cases were hemodynamically unstable and the immediate clinically-based decision for intiating therapy was imperative while laboratory results were still pending. Antithyroid drugs were initiated considering the fact that thyrotoxic consequences in a critically ill patient are more precarious than the undue antithyroid therapy for a limited duration (2 days). Final diagnosis of GD activity (by thyroid hormone assay) was confirmed in 3 cases and disproved in one case.
Diagnosis of thyroid nodules (solitary and multiple) was perfectly confirmed by US. Three out of 11 patients had been clinically overlooked (27.3%) (grade 0) due to obesity and short neck. As shown in previous studies, solitary toxic nodule is usually slowly progressive in size and toxicity merging from pretoxic to toxic state and usually becomes toxic when attaining a diameter of about 2.5 cm [21,22]. Most toxic nodules in this study were hyperechogenic and some had peripheral halo expressing normal compressed thyroid tissue. Nodule volumes attained a size of 12.3–35.7 ml and toxicity was more prevalent in larger sized nodules. Small sized nodules were mostly pretoxic as proved later by radioisotope study.
Proper diagnosis of toxic thyroid nodules not only allowed early management of thyrotoxicosis in cardiac arrhythmias but also helped in precluding drugs with potentially deleterious effects such as amiodarone. Amiodarone is commonly used in treatment of supraventricular and ventricular arrhythmias and may induce thyrotoxicosis in patients with thyroid nodules even if the nodule is nontoxic as it is an iodine containing drug. It also increases toxicity of GD. This actually occurred in one interesting case in our study who had been referred to ICU with ventricular tachycardia which was described to be refractory to treatment with amiodarone given intravenously in a loading dose and later showed more deterioration of ventricular tachycardia with the addition of IV maintenance dose of amiodarone in the following 12 h. Thyroid nodule was not detected clinically in this case as the patient was obese and had a short neck. By US examination, a solitary hyperechogenic thyroid nodule. 2.7 cm in diameter was detected. This might explain refractoriness and deterioration of ventricular tachycardia as amiodarone was alleged to induce toxicity if the nodule was nontoxic and to cause more toxicity if the nodule was toxic. Amiodarone treatment was stopped and propylthyrouracil and propranolol were given before thyroid hormone assessment. Ft3 and Ft4 were assessed and were found to be markedly increased and TSH was normal indicating recent rather than old thyrotoxicosis as TSH takes a long time to be suppressed in response to excess thyroid hormones. Thus in this patient, amiodarone had induced toxicity in a previously present nontoxic thyroid nodule.
Thyroid US helped in proper and early diagnosis of thyrotoxicosis in seriously ill patients who had moderate and prominent thyrotoxic symptoms (40.7% of diagnosed patients) allowing early and aggressive management of thyrotoxicosis before waiting for laboratory diagnosis. This benefit was clearly manifest in the case of STEMI, malignant arrhythmias and refractory heart failure.
Thyroid ultrasound was beneficial in proper and early diagnosis of thyrotoxicosis in critically ill hemodynamically unstable patients who had no prominent thyrotoxic symptoms (masked thyrotoxicosis). This was represented in 59.3% of all diagnosed thyrotoxic patients in ICU. This high percentage of presentation of masked thyotoxicosis in ICU reflects the overlap of clinical picture of thyrotoxicosis with variable signs and symptoms commonly encountered in critically ill patients such as tachyarrhythmias, dyspnea, insomnia, weight loss, anxiety, panics, sweating, hand tremors and staring. Second, masked thyrotoxicosis is commonly encountered in older aged group who represented most of the admitted patients in our study and in most intensive care units. Third, subclinical thyrotoxicosis, although not easily diagnosed clinically in a healthy individual in the outpatient clinic, results in significant manifestation in a critically ill patient with other co-morbid serious conditions, and hence can be commonly encountered and should be searched for in the ICU. Fourth, thyromegaly was not prominent clinically in 29.5% of patients with GD and in 27.3% in toxic nodules and diagnosed perfectly by US.
The overall accuracy of US in diagnosis of thyrotoxicosis was assessed and compared to clinical assessment and both measures (clinical and US) were evaluated in relation to final diagnosis of thyrotoxicosis by hormonal profile in all suspected patients and radioisotope scanning in selected cases. The sensitivity of thyroid US in diagnosis of thyrotoxicosis was 100%, sensitivity 98.8%, negative predictive value 100% and overall accuracy 96.5% compared to a positive predictive value of 24.8% for clinical assessment addressing the invaluable diagnostic role of US. On the other hand, diagnosis of goiter per se by clinical examination had a sensitivity 84%, specificity 69.8%, positive predictive value 68.9%, negative predictive value 84.6% and overall accuracy 76.1% compared to 100% for all these accuracy parameters of US emphasizing the role of US as the gold standard for the diagnosis of goiter.
There are three limitations that need to be acknowledged and addressed regarding the present study. The first limitation concerned thyroid palpation in critically ill that could have been imprecise as there might be inter-observer variability but we tried to limit the number of examining physicians by only 3 examiners. The second limitation concerned thyroid palpation in lying position that could have limited the accuracy of palpation especially in intubated patients; however, this was only found in 9 patients. The third limitation was that the authors could not assess the sensitivity of clinical examination, nor the euthyroid sick state, since the design of the study was only to perform US and hormone assay for suspected cases rather than all admitted cases.
In conclusion, thyrotoxicosis is commonly encountered in the ICU and masked cases should be thoroughly searched for especially in cardiac patients. Bedside US is an invaluable tool in early, rapid and accurate diagnosis of thyrotoxicosis and thyrotoxic emergencies.
 Kaplan MM. Endocrinology and metabolism clinics of North America (thyrotoxicosis
), vol 27 (1). Philadelphia: W.B. Saunders; 1998.
 Cappola AR, Fried LP, Arnold AM, et al. Thyroid status, cardiovascular risk, and mortality in older adults. JAMA 2006;295(9):1033-1041.
 Olshausen KV, Bischoff S, Kahaly G, et al. Cardiac arrhythmias and heart rate in hyperthyroidism. Am J Cardiol 1989;63:930.
 Alfadhli E, Gianoukakis AG. Management of severe thyrotoxicosis
when the gastrointestinal tract is compromised. Thyroid 2011;21(3):215-220.
 Wyne Kathleen. Commentary: the role of thyroid hormone therapy in acutely ill cardiac patients. Crit Care 2005;9:333-334.
 Stathatos Nikolaos. The controversy of the treatment of critically ill patients with thyroid hormone. Best Pract Res Clin Endocrinol Metab 2001;15(4):465-478.
 Nylen ES. Endocrine changes in critical illness. J Intensive Care Med 2004;19:67-82.
 Ligtenberg JJM. Hormones in the critically ill patient: to intervene or not to intervene. Intensive Care Med 2001;27:1567-1577.
 Sakharova Olga. Endocrine assessments during critical illness. Crit Care Clin 2007;23:467-490.
 Eng CY, Quraishi MS, Bradley PJ. Management of thyroid nodules in adult patients. Head Neck Oncol 2010;2:11.
 Frates MC, Benson CB, Charboneau JW, Cibas ES, Clark OH, Coleman BG, et al. Management of thyroid nodules detected at US: Society of Radiologists in Ultrasound consensus conference statement. Radiology 2005;237(3):794-800, http://www.myoops.org/cocw/tufts/courses/14/content/D265882/C82210.jpg
 Ross DS. Nonpalpable thyroid nodules - managing an epidemic. J Clin Endocrinol Metab 2002;87(5):1938-1940, http://www.bmb.leeds.ac.uk/teaching/icu3/lecture/25/index.htm
 Tessler FN, Tublin ME. Thyroid sonography: current applications and future directions. AJR Am J Roentgenol 1999;173(2):437-443.
 Zimmermann Michael, Saad Amina, Hess Sonja, Torresani Toni, Chaouki Noureddine. Thyroid ultrasound
compared with World Health Organization 1960 and 1994 palpation criteria for determination of goiter Endemic goiter prevalence in regions of mild and severe iodine deficiency. Eur J Endocrinol 2000;143:727-731.
 Chanoine JP, Toppet V, Lagasse R, Spehl M, Delange F. Determination of thyroid volume by ultrasound from the neonatal period to late adolescence. Eur J Pediatr 1991;150(6):395-399.
 Hegedus L. Thyroid size determined by ultrasound. Influence of physiological factors and nonthyroidal disease. Dan Med Bull 1990;37(3):249-263.
 Hintze G, Windeler J, Baumert J, Stein H, Kobberling G. Thyroid volume and goiter prevalence in the elderly as determined by ultrasound and their relationships to laboratory indices. Acta Endocrinol Copenh 1991;124(91):12-18.
 Salama AS, Hamed ES, Abde-Wahab AM, Momtaz OM. The role of intranodular and intrathyroidal absolute ethanol injection (PEI) in treatment of solitary toxic nodule, toxic multinodular goiter and Graves' disease, preliminary results and 42 month follow up (1997-2000). Egypt J Intern Med 2001;11(2):253-265.
 Sakan S. The prognostic application of thyroid volume determination in patients with Graves' disease. Nippon Naibunpi Gakkai Zasshi 1990;66(5):543-556.
 Brkljacic B, Ivanac G, Cikara I, Bozikov V. Percutaneous ultrasound-guided ethanol injection treatment of thyroid toxic and autonomous adenomas: results and long-term follow-up. J Ultrasound Med 2009;28(3):S83.
 Dremier S, Coppee F, Delange F. Clinical review 84. Thyroid autonomy: mechanism and clinical effects. J Clin Endocrinol Metab 1996;81:4187-4193.
 Hamed ES, Abdel-Wahab AM, Salama AS, Momtaz OM. Prevalence of thyroid disease among attendants of El-Minia University hospital. EL-Minia Med Bull 1998;9(1):40-52.
 Mark PJ. Vanderpump: the epidemiology of thyroid disease. Br Med Bull 2011;99:39-51.
 Roffi M, Cattaneo F, Topol J. Thyrotoxicosis
and the cardiovascular system: subtle but serious effects. Clevland Clin J Med 2003;70:57-63, Number 1 Jan 2003.