Value of sensitive thyrotropin measurement in ambulatory and hospitalized patients

To compare the frequency and causes of abnormal thyrotropin (TSH) levels in ambulatory and hospitalized patients and to assess the specificity and sensitivity of suppressed TSH for the diagnosis of hyperthyroidism in these two settings, analysis of thyroid function tests including sensitive TSH measurement was performed on 715 consecutive patients who had a thyroid panel performed in one clinical laboratory. Suppressed TSH (less than 0.3 mU/L) and elevated TSH (greater than 5.5 mU/L) were found in 35 (8.5%) and 28 (6.5%) of the 411 regular ward inpatients. The prevalence of suppressed TSH was significantly higher than that of high TSH among the 267 ambulatory patients (11.6% vs 5.6%, p less than 0.03). A total of 37 severely ill ICU patients had a significantly higher prevalence of both suppressed and elevated TSH (16% and 22% respectively) than regular ward inpatients and ambulatory patients. Hyperthyroidism and exogenous thyroid hormone administration were responsible for suppressed TSH in 65% of ambulatory patients and in 34% of regular ward and ICU patients. In contrast nonthyroidal illness was implicated in 36% of hospitalized patients and in 6% of ambulatory patients. The sensitivity and specificity of suppressed TSH for the diagnosis of hyperthyroidism was 90% and 91% for ambulatory patients and 100% and 91% for hospitalized patients respectively. The predictive value of suppressed TSH for hyperthyroidism was higher in outpatients (26%) than in hospitalized patients (7%). After patients with known causes for suppressed TSH other than thyroid disease had been excluded, the corrected predictive values of suppressed TSH for hyperthyroidism were 57% in outpatients and 21% in hospitalized patients.(ABSTRACT TRUNCATED AT 250 WORDS)