The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
Searching for Thyroid Disease
Abstract & Commentary
By Mary Elina Ferris, MD, Clinical Associate Professor, University of Southern California. Dr. Ferris reports no financial relationship to this field of study.
Synopsis: Fully 95% of serum thyrotropin (TSH) measurements were normal in screening tests ordered by primary care physicians. Half of abnormal TSH labs became normal upon repeat measurement. If the initial TSH result was normal, there was only a 2% chance it would become abnormal within 5 years.
Source: Meyerovitch J, et al. Arch Intern Med. 2007;167:1533-1538.
Data from an Israeli Health Maintenance Organization covering 2.3 million adults age 21 and over was used to find 422,242 persons who had at least one serum TSH measurement in the year 2002, which comprised 18% of the total enrollees. Only tests ordered by outpatient primary care physicians during routine clinical care (total 2,800 physicians) were considered. Exclusions were any patient without follow-up through the next four years; pregnancy; thyroid disease diagnosed before 2002 or with an abnormal TSH recorded during 2001; or those who had any thyroid treatment resulting from an abnormal TSH in 2002.
Testing rates were greater in women (65.9% of total tests, which was 1/3 of the female members over age 40 years), and increased in men from 8% in men younger than age 40 years, up to 20% of men aged 60-80 years. Results were normal for 95% of initial TSH measurements, with 3.7% elevated (>5.5 mIU/L) and 1.2% decreased (<0.35 mIU/L).
Although the Free T4 level was only measured in 12% of the total sample, it was determined in 89% of those with abnormal TSH results. For elevated TSH results, 98% of the patients had Free T4 measured and it was within normal limits in almost all cases.
Follow-up over the next 5 years analyzed a further 1.5 million tests (mean 3.73 tests/patient). Excluding 3.6% of the patients who had thyroid treatment started after an initial abnormal result, the subsequent tests had 98% normal results. For initial decreased TSH results, 51.5% had their repeat test become normal. For initial moderately elevated TSH results (>5.5 to 10mIU/L), repeat tests gave 62% normal, 35% unchanged, and 3% highly elevated over 10mIU/L. For initial highly elevated TSH results (over 10mIU/L), 35% had continued elevated results, with the rest either normal or decreasing.
Despite many patient concerns about thyroid disease, particularly to explain weight gain, diagnosis by screening blood tests has a low yield. This large community-based study gives us useful information about predictable outcomes when we order a serum TSH during standard clinical care. Namely, the result is most often completely normal, and there's not much point repeating it yearly. Furthermore, even most abnormal results return to normal after several months, so caution is suggested before any treatment for thyroid disease is initiated.
The authors were interested in the natural progression of subclinical thyroid disease, which has been estimated at 4-8% in the general population and increases with age. There has been concern that without treatment, cognitive and cardiac complications will ensue. They cite several recent studies showing this not to be a concern, and conclude from their data that the rate of progression to mild or overt hypothyroidism is very slow, with at least half of both elevated and decreased TSH results returning to normal within several months.
Different organizations have issued varying recommendations for thyroid screening in our patients, which is one of our most frequently ordered laboratory tests. While the U.S. Preventive Services Task Force has concluded that evidence is insufficient to recommend for or against routine screening for thyroid disease in adults, or for treatment of subclinical thyroid disease,1consensus guidelines from the American Thyroid Association recommend screening every 5 years.2 The authors of this study feel that their findings support no widespread population screening recommendations, and furthermore caution that treatment for thyroid disease should not be considered without at least 2 abnormal TSH measurements.
1. Helfand M, et al. Screening for subclinical thyroid dysfunction in nonpregnant adults: a summary of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med. 2004 Jan 20;140:128-141.
2. Ladenson PW, et al. American Thyroid Association Guidelines for Detection of Thyroid Dysfunction. Arch Intern Med. 2000;160:1573-1575.