Ten RCTs (n~1200 patients, the exact total was unclear) were included (six of which had a cross-over design). The authors reported that two other trials were identified, but subsequently excluded because standard errors were not reported (and could not be obtained from investigators). Trial quality (overall) was described as being adequate; all trials reported on losses to follow-up, all were reported as having used adequate methods for randomisation and blinding (further details on these two criteria were not given), but only three trials reported using an intention-to-treat analysis. Only one of the crossover trials had a washout phase.
The authors stated there were statistically significant reductions of serum-free and total thyroxine in patients treated with thyroxine plus triiodothyronine, but the confidence intervals reported could not clearly confirm this.
Thyroxine alone showed significant benefits for psychological or physical well-being for the following outcomes: general health (WMD -0.38,; 95% CI -0.71 to -0.05; four RCTs), physical component summary (WMD 0.7, 95% CI 0.53 to 0.87; one RCT), mental component summary (WMD 0.58, 95% CI 0.25 to 0.75; two RCTs); physical functioning (WMD 1.60, 95% CI 1.29 to 1.90; two RCTs), bodily pain (WMD 2.50, 95% CI 2.11 to 2.88; two trials), and role-emotional (WMD 2.08, 95% CI 1.17 to 2.99; two RCTs), and mental health (WMD 1.30, 95% CI 0.97 to 1.64; two RCTs) in items of the Short Form-36 Health Survey.
Thyroxine alone also showed significant benefits for general well-being for the following: the Thyroid Symptom Questionnaire (WMD -1.90, 95% CI -2.48 to -1.32; one RCT); Letter Number Sequencing-working memory test cognitive performance scores (WMD 1.10, 95% CI 0.08 to 2.13; two RCTs), blurred vision (WMD -4.66, 95% CI -5.34 to -4.00; three RCTs) and aches and pain (WMD -0.80, 95% CI -1.34 to -0.26; one RCT).
Thyroxine plus triiodothyronine replacement therapy showed significant benefit in improving cognitive performance (WMD -0.49, 95% CI -0.90 to -0.08, two RCTs).
There were no significant differences in adverse effects.
Further results were reported.