J Clin Ultrasound 2 0 : 3 7 4 2 , January 1992 0 1992 by John Wiley & Sons, Inc. CCC 0091-2751/92/01037-06 $04.00
Clinical versus Ultrasound Examination of the Thyroid Gland in Common Clinical Practice A. Brander, MD,* P. Viikinkoski, MD,* J. Tuuhea, MD,* L. Voutilainen, MD,* and L. Kivisaari. MDt
Abstract: In a prospective series of 72 patients, clinical and ultrasonographic examination of the thyroid gland were compared in detail. Normal-sized lobes were differentiated from enlarged ones both by inspection and by palpation. When lobar size was assessed by palpation, the estimate was most clearly influenced by increase in width. The correlation between two examiners in lobe size assessment was significant. In the classification of thyroid disease as diffuse, solitary, or multinodular, clinical examination and ultrasonography correlated significantly. However, only one third of the clinically solitary nodules proved to be solitary by ultrasound examination. Of 77 separate nodules, 43 escaped detection on clinical examination. Of these 43, 14 nodules exceeded 2 cm in diameter. It is concluded that the use of ultrasonography frequently alters the primary evaluation of thyroid nodularity based on palpation. Indexing Words: Thyroid, US studies Thyroid, neoplasms
-
Clinical examination constitutes the primary diagnostic test for the detection of goiter and evaluation of nodular thyroid disease. The approach to an individual nodule still depends largely on whether it is clinically solitary or multinodular, even though it is well known that the sensitivity of palpation is poor in detecting thyroid nodules. The sensitivity of high-resolution real-time ultrasonography (7.5- or 10-MHz transducers) in detecting thyroid nodules is well established, although reliable criteria for differentiating malignant nodules from benign ones are still lacking. The aim of our study was to evaluate the usefulness of ultrasonography in common clinical practice by comparing in detail clinical and ultrasound examinations of the thyroid in a n outpatient series of a district hospital. Clinical examination and thyroid ultrasonography have been correlated previously but mostly in retrospective studies in which palpatory findings have not been recorded in detail. Our prospective series consists of patients in a district hospital where thyroid ultrasonography From the *Hyvinkaa District Hospital Hyvinkaa, and tDepartment of Radiology, Helsinki University Central Hospital, Helsinki, Finland. For reprints contact A. Brander, MD. Hyvinkaa District Hospital, Sairaalank. 1, 05850 Hyvinkaa, Finland.
has been used as a n adjunct to clinical examination since 1985. Thyroid ultrasound examinations are performed daily, and most fine-needle biopsies of the thyroid are performed with ultrasound guidance. The series represents those thyroid problems that are common in clinical practice. There has been no deliberate selection, although this study primarily is concerned with suspected nodular thyroid disease. Even though our series lacks the gold standard of surgical exploration, the number of cytologically verified nodules is high.
PATIENTS AND METHODS
The series consists of 72 outpatients and patients at a medical department of a district hospital with 200 beds in southern Finland. Goiter is not considered endemic in the area. There were 64 women and 8 men, aged 17-80 (mean 49). The patients were referred to the outpatient department by general practitioners because of a suspected goiter or thyroid nodule, or disorders of thyroid function. In 8 patients thyroid abnormality was first suspected by the examining clinician at the hospital. The series was collected between 1987 and 1989. In all cases, clinical examination included in37
38
BRANDER ET AL.
spection and palpation of the thyroid. The size of each lobe was classified as normal, or slightly or clearly increased, and the presence or absence of separate nodules was recorded. The consistency of the thyroid was classified as soft or abnormally hard. A standard form was used for scoring the clinical findings. Individual nodules were located on a three-dimensional thyroid map included in the form. For assessment of interobserver variation, 27 patients were reexamined by another physician without any knowledge of previous findings. A total of 12 clinicians participated in the clinical examinations. Of 96 palpatory examinations, 64 (66.7%) were performed by internists and 32 by residents. Fifty-seven percent of the examinations were performed by three internists who were all experienced clinicians. All ultrasound examinations were performed by the same radiologist with the same equipment, a Toshiba SAL-77A dynamic image scanner with 7.5-MHz linear transducer for small parts. During the examination, the patient was supine with the neck hyperextended. An anechoic stand-off pad (Kitecko) was used in all examinations. The volume of each lobe was calculated by the method described by Brunn et al. (length x width x thickness of the lobe x 0.479).l Focal lesions were measured with electronic calipers, characterized for echogenicity, and located on a similar map as was used on palpatory examination. Blood tests for thyroid function and autoantibodies were obtained for all but 5 patients. Fineneedle aspiration biopsy was performed on 54 lesions in 53 patients, in all cases except 1with ultrasound guidance. The statistical analysis was performed employing BMDP Statistical Software Package version 1988 designed for IBM PC-DOS-computer. In comparing the ultrasonically measured volumes of lobes that were normal vs enlarged on inspection, analysis of variance was used. When three or more groups were compared, as in comparing ultrasonically measured volumes of palpatory different lobe size groups, in estimating the impact of each dimension of a lobe on the clinical size assessment, and in comparing the volumes of clinically hard vs soft lobes, the Newman-Keuls multiple range test was used. The correlation between observers in clinical examination and the correlation between ultrasound and clinical examination in the classification of thyroid nodularity was tested by calculated kappa-values.
RESULTS
Thyroid Volume The mean volume of the thyroid (ultrasound measured) in the series was 26.2 mL (range 4.2 mL-137.4 mL). The mean volume was 13.3 mL for the right lobe and 12.9 mL for the left lobe.
Clinical Estimation of Lobe Size Only the assessment of the first examiner was used in the analysis. On both inspection and palpation, the right and the left lobe were analyzed separately. The ultrasound-measured volumes of lobes estimated as clearly enlarged on palpation differed significantly from the volumes of lobes estimated as normal ( p < 0.01 for both lobes) and slightly enlarged ( p < 0.01 for right lobe, p < 0.05 for left lobe). Between palpatory slightly enlarged and normal lobes, no significant difference in ultrasound-measured volumes was found. The mean volumes and SDs are given in Table 1. On inspection there was, with respect to both lobes, a significant difference between the ultrasound-measured volumes of clinically normal versus enlarged lobes (right lobe p < 0.005, left lobe p < 0.01). The mean volumes and SDs are given in Table 2. To estimate the effect of a single lobe dimension on the clinical assessment of lobe size, the various clinical size groups were compared with regard to ultrasonically measured thickness, length, and width. Clinically normal and clearly enlarged lobes differed significantly from each other in each diameter ( p < 0.01 for each diameter for both lobes). Clearly enlarged lobes also differed significantly from slightly enlarged ones in width (p
Clinical versus Ultrasound Examination of the Thyroid Gland in Common Clinical Practice A. Brander, MD,* P. Viikinkoski, MD,* J. Tuuhea, MD,* L. Voutilainen, MD,* and L. Kivisaari. MDt
Abstract: In a prospective series of 72 patients, clinical and ultrasonographic examination of the thyroid gland were compared in detail. Normal-sized lobes were differentiated from enlarged ones both by inspection and by palpation. When lobar size was assessed by palpation, the estimate was most clearly influenced by increase in width. The correlation between two examiners in lobe size assessment was significant. In the classification of thyroid disease as diffuse, solitary, or multinodular, clinical examination and ultrasonography correlated significantly. However, only one third of the clinically solitary nodules proved to be solitary by ultrasound examination. Of 77 separate nodules, 43 escaped detection on clinical examination. Of these 43, 14 nodules exceeded 2 cm in diameter. It is concluded that the use of ultrasonography frequently alters the primary evaluation of thyroid nodularity based on palpation. Indexing Words: Thyroid, US studies Thyroid, neoplasms
-
Clinical examination constitutes the primary diagnostic test for the detection of goiter and evaluation of nodular thyroid disease. The approach to an individual nodule still depends largely on whether it is clinically solitary or multinodular, even though it is well known that the sensitivity of palpation is poor in detecting thyroid nodules. The sensitivity of high-resolution real-time ultrasonography (7.5- or 10-MHz transducers) in detecting thyroid nodules is well established, although reliable criteria for differentiating malignant nodules from benign ones are still lacking. The aim of our study was to evaluate the usefulness of ultrasonography in common clinical practice by comparing in detail clinical and ultrasound examinations of the thyroid in a n outpatient series of a district hospital. Clinical examination and thyroid ultrasonography have been correlated previously but mostly in retrospective studies in which palpatory findings have not been recorded in detail. Our prospective series consists of patients in a district hospital where thyroid ultrasonography From the *Hyvinkaa District Hospital Hyvinkaa, and tDepartment of Radiology, Helsinki University Central Hospital, Helsinki, Finland. For reprints contact A. Brander, MD. Hyvinkaa District Hospital, Sairaalank. 1, 05850 Hyvinkaa, Finland.
has been used as a n adjunct to clinical examination since 1985. Thyroid ultrasound examinations are performed daily, and most fine-needle biopsies of the thyroid are performed with ultrasound guidance. The series represents those thyroid problems that are common in clinical practice. There has been no deliberate selection, although this study primarily is concerned with suspected nodular thyroid disease. Even though our series lacks the gold standard of surgical exploration, the number of cytologically verified nodules is high.
PATIENTS AND METHODS
The series consists of 72 outpatients and patients at a medical department of a district hospital with 200 beds in southern Finland. Goiter is not considered endemic in the area. There were 64 women and 8 men, aged 17-80 (mean 49). The patients were referred to the outpatient department by general practitioners because of a suspected goiter or thyroid nodule, or disorders of thyroid function. In 8 patients thyroid abnormality was first suspected by the examining clinician at the hospital. The series was collected between 1987 and 1989. In all cases, clinical examination included in37
38
BRANDER ET AL.
spection and palpation of the thyroid. The size of each lobe was classified as normal, or slightly or clearly increased, and the presence or absence of separate nodules was recorded. The consistency of the thyroid was classified as soft or abnormally hard. A standard form was used for scoring the clinical findings. Individual nodules were located on a three-dimensional thyroid map included in the form. For assessment of interobserver variation, 27 patients were reexamined by another physician without any knowledge of previous findings. A total of 12 clinicians participated in the clinical examinations. Of 96 palpatory examinations, 64 (66.7%) were performed by internists and 32 by residents. Fifty-seven percent of the examinations were performed by three internists who were all experienced clinicians. All ultrasound examinations were performed by the same radiologist with the same equipment, a Toshiba SAL-77A dynamic image scanner with 7.5-MHz linear transducer for small parts. During the examination, the patient was supine with the neck hyperextended. An anechoic stand-off pad (Kitecko) was used in all examinations. The volume of each lobe was calculated by the method described by Brunn et al. (length x width x thickness of the lobe x 0.479).l Focal lesions were measured with electronic calipers, characterized for echogenicity, and located on a similar map as was used on palpatory examination. Blood tests for thyroid function and autoantibodies were obtained for all but 5 patients. Fineneedle aspiration biopsy was performed on 54 lesions in 53 patients, in all cases except 1with ultrasound guidance. The statistical analysis was performed employing BMDP Statistical Software Package version 1988 designed for IBM PC-DOS-computer. In comparing the ultrasonically measured volumes of lobes that were normal vs enlarged on inspection, analysis of variance was used. When three or more groups were compared, as in comparing ultrasonically measured volumes of palpatory different lobe size groups, in estimating the impact of each dimension of a lobe on the clinical size assessment, and in comparing the volumes of clinically hard vs soft lobes, the Newman-Keuls multiple range test was used. The correlation between observers in clinical examination and the correlation between ultrasound and clinical examination in the classification of thyroid nodularity was tested by calculated kappa-values.
RESULTS
Thyroid Volume The mean volume of the thyroid (ultrasound measured) in the series was 26.2 mL (range 4.2 mL-137.4 mL). The mean volume was 13.3 mL for the right lobe and 12.9 mL for the left lobe.
Clinical Estimation of Lobe Size Only the assessment of the first examiner was used in the analysis. On both inspection and palpation, the right and the left lobe were analyzed separately. The ultrasound-measured volumes of lobes estimated as clearly enlarged on palpation differed significantly from the volumes of lobes estimated as normal ( p < 0.01 for both lobes) and slightly enlarged ( p < 0.01 for right lobe, p < 0.05 for left lobe). Between palpatory slightly enlarged and normal lobes, no significant difference in ultrasound-measured volumes was found. The mean volumes and SDs are given in Table 1. On inspection there was, with respect to both lobes, a significant difference between the ultrasound-measured volumes of clinically normal versus enlarged lobes (right lobe p < 0.005, left lobe p < 0.01). The mean volumes and SDs are given in Table 2. To estimate the effect of a single lobe dimension on the clinical assessment of lobe size, the various clinical size groups were compared with regard to ultrasonically measured thickness, length, and width. Clinically normal and clearly enlarged lobes differed significantly from each other in each diameter ( p < 0.01 for each diameter for both lobes). Clearly enlarged lobes also differed significantly from slightly enlarged ones in width (p
