Premoselli 2006.

Methods	Quasi‐randomised, single‐blind controlled trial Blinded assessor Ethics approval and informed consent obtained
Participants	Total N = 50 participants (50 wrists) randomised Intervention group N = 25 wrists randomised, 18 wrists completed trial Control group N = 25 wrists randomised, 16 wrists completed trial 5 males, 45 females Mean ± SD age: Intervention group: 53.1 ± 13.3 yrs Control group: 46.5 ± 13.8 yrs Inclusion criteria: 1. Compound motor action potential (CMAP) median nerve distal latency < 4.7 ms 2. Difference between median and ulnar sensory action potential latencies > 0.4 ms Exclusion criteria: 1. Diabetes 2. "Clear CTS" (i.e., not mild recent onset CTS, as measured using electromyographic measures)
Interventions	Intervention: Neutral custom‐moulded thermoplastic resin wrist splints were worn at nighttime only, for a minimum of 6 hours per night, for six months Control: No intervention
Outcomes	Outcomes assessed at baseline, at three months, and at the end of six months of treatment: 1. Nerve conduction: sensory action potential latency (ms), sensory action potential velocity  (m/s), sensory action potential amplitude (µV), motor action potential latency (ms), motor action potential velocity (m/s), motor action potential amplitude (mV)* 2. Symptoms using carpal tunnel questionnaire (rates 11 items on ordinal scale 1: mildest pain, to 5: most severe pain)* 3. Hand function using carpal tunnel questionnaire (rates 8 items on ordinal scale 1: difficulty with activities, to 5: hindrance in performing an activity)* 4. Semeiotic testing using the Williams et al. (1992) pressure‐provocative test and the Phalen test. The time lapse between the moment of stimulation and the first manifestation of symptoms was assessed for each kind of test.*
Notes	*The results of all outcomes were reported as both endpoint and change from baseline values. There were some differences between endpoint and change from baseline values in terms of statistical significance, though to minimise selective inclusion bias, we chose to be consistent and only include endpoint values (as we included endpoint values where available for all other studies included in the review). Each participant contributed only one CTS‐affected wrist to the study, so a unit of analysis error resulting from the correlation between two wrists in bilateral CTS participants could not have occurred.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Quote: "The randomisation protocol was based on the last visit booking number (even or odd)." Comment: The trial authors used a non‐random component in the sequence generation process.
Allocation concealment (selection bias)	High risk	Quote: "The randomisation protocol was based on the last visit booking number (even or odd)." Comment: The trials authors did not adequately conceal the treatment allocation until interventions were assigned.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: Due to the nature of the interventions, it is unlikely that participants were not aware of which treatment they received (nighttime splint or no intervention).
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "The examiner was blinded to treatment status (control or treatment)." Comment: Assessment of outcomes were probably done by a blinded assessor.
Incomplete outcome data (attrition bias) 3 months or less	Unclear risk	Quote: "Fifty patients (50 hands) were enrolled, of which 36 completed the study at 6 months." Quote: "At the three‐month follow‐up visit, 24/25 case patients and 24/25 control patients were evaluated." Quote: "At the six‐month follow‐up visit, 18 case group subjects and 16 control group subjects were evaluated." Comment: The numbers in these three quotes do not add up. In the abstract it says that 36 patients were available at 6 months follow‐up, but in the text, it says that 34 (18 + 16) patients were available at 6 months follow‐up. Therefore it is not clear how many participants were lost to follow‐up and the reasons for these.
Selective reporting (reporting bias)	Low risk	Comment: All outcomes stated in the methods section of the publication were reported in their pre‐specified way.
Other bias	Low risk	Comment: No other sources of bias identified.