ABSTRACT
OBJECTIVE: To compare two methods of assessing gland size on thyroid ultrasound in newborn infants with suspected congenital hypothyroidism (CH). METHODS: Images from infants with eutopic glands referred between 2007 and 2013 were evaluated blind by two sets of observers. Subjective gland size was categorised as small, borderline-small, normal, borderline-large and large. Objective gland volume, calculated as the sum of each lobe using the prolate ellipsoid formula (length x width x depth x π/6), was put into corresponding categories: <0.8, 0.81-1.0, 1.1- <2.2, 2.2-2.4 and >2.4 ml, derived from normative Scottish data. RESULTS: Of 36 infants, permanent CH was present in 17, transient CH in 17, status uncertain in 2. Mean (SD) intraobserver error for thyroid volume measurement was 0.11 (0.23) ml [8.3%]. Subjective assessment by two observers was discordant in only four (10.8%) infants. However, subjective vs objective evaluation was discordant in 14 (39%). Eight (three permanent, five transient CH) had large glands subjectively but normal glands objectively; and six (four transient CH) had normal glands subjectively but small glands objectively. The former infants all showed a single flattened curve to the anterior thyroid margin, giving an impression of bulkiness. Gland shape was normal in the latter infants. CONCLUSION: Neither subjective nor objective evaluation predicts permanent vs transient CH. Altered gland shape may confound both methods, and undermine use of the conventional formula for measuring lobe volume. ADVANCES IN KNOWLEDGE: Until more refined methods are available for assessing thyroid size, both subjective and objective evaluation are recommended in CH.