TY - JOUR
T1 - Optimizing selection of in vitro tests for diagnosing thyroid disorders
AU - Zwas, S. Tzila
AU - Rosenblum, Yossef
AU - Boruchowsky, Sabina
PY - 1987/9
Y1 - 1987/9
N2 - The optimal utilization of the thyroid related radioimmunoassays T3, T4, and TSH-RIA is derived from analysing the clinical and laboratory data for 974 patients with functional thyroid disorders. A statistical computer analysis of the contribution of each of the three tests, and in combination, to the final diagnoses of hypothyroid, euthyroid, and hyper thyroid states was designed. The best contributing test for hypothyroidism and euthyroidism was TSH-RIA (98.5 and 93%, respectively). T4/T3+TSH-RIAs were the optimal dual combination for diagnosing euthyroidism (98.0%). For diagnosing hyperthyroidism T4-RIA was the best single test (82.5%) followed by T3+T4 as an optimal dual combination (95%). Using all three tests was of no significant additional value over dual combinations. It is concluded that the work and cost of randomly performing three tests routinely is not justified without clinical basis. An algorithm is proposed to guide thyroid studies based on computer analyses of the above- mentioned single or dual-test combinations to establish accurate diagnosis at the lowest laboratory cost.
AB - The optimal utilization of the thyroid related radioimmunoassays T3, T4, and TSH-RIA is derived from analysing the clinical and laboratory data for 974 patients with functional thyroid disorders. A statistical computer analysis of the contribution of each of the three tests, and in combination, to the final diagnoses of hypothyroid, euthyroid, and hyper thyroid states was designed. The best contributing test for hypothyroidism and euthyroidism was TSH-RIA (98.5 and 93%, respectively). T4/T3+TSH-RIAs were the optimal dual combination for diagnosing euthyroidism (98.0%). For diagnosing hyperthyroidism T4-RIA was the best single test (82.5%) followed by T3+T4 as an optimal dual combination (95%). Using all three tests was of no significant additional value over dual combinations. It is concluded that the work and cost of randomly performing three tests routinely is not justified without clinical basis. An algorithm is proposed to guide thyroid studies based on computer analyses of the above- mentioned single or dual-test combinations to establish accurate diagnosis at the lowest laboratory cost.
UR - http://www.scopus.com/inward/record.url?scp=0023406350&partnerID=8YFLogxK
U2 - 10.1097/00006231-198709000-00009
DO - 10.1097/00006231-198709000-00009
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:0023406350
SN - 0143-3636
VL - 8
SP - 759
EP - 765
JO - Nuclear Medicine Communications
JF - Nuclear Medicine Communications
IS - 9
ER -