Sumathipala 2008.

Methods	Study design: randomised controlled trial
Participants	Diagnosis: MUS Method of diagnosis: participants (16‐65 years old) with ≥ 5 MOS and repeated consultations during the last 6 months were selected by the primary care doctors. They underwent an extensive physical examination by trial co‐ordinator and trial physician plus an independent reassessment through 2 questions: 1. What are your symptoms/problems, why are you here today? and, 2. Are there any other symptoms/problems? to elicit the number of symptoms and visits in the previous 6 months. After inclusion the standardised BSI questionnaire was filled out by all participants Exclusion criteria: dementia, psychosis or alcohol dependence current treatment for a psychiatric disorder Total number randomised: 150 Age: aged 16‐58 years, M 35 years Sex: for intervention group 79% women (n = 59), 21% men (n = 16); for control group 77% women (n = 58), 23% men (n = 17) Severity of symptoms at baseline: number of symptoms M = 8.6 (SD = 2.2) Duration of symptoms at baseline: M = 42.0 months (SD = 40) Setting: a general outpatient clinic of a general hospital where participants initiate their own visits without prior appointments (primary care) Location: Colombo, Sri Lanka Number of treatment centres: 1 Co‐morbidities: not described Adjunctive therapy: none described Adjunctive medication: none described
Interventions	Participants were randomly assigned to either 1. CBT (n = 75) Duration: 6 x 30‐minute structured sessions over a period of 3 months (of which 3 mandatory) Treatment protocol: treatment consisted of structured regular visits to 1 professional carer thereby aiming to reduce unstructured visits to different practitioners and co‐ordinating care. Treatment was based on the principles of CBT, using modifications of that described by Salkovskis and Sharpe et al. and Goldberg et al.'s reattribution technique for details Sumathipala 2000, ref Salkovskis, Sharpe and Goldberg) Treatment started with a SEMI interview by 1 of the authors: results, summary, and formulation based on findings were passed on to primary care physician, they were also trained to use this information to inform the strategy for their CBT intervention. Participants in the intervention group completed a diary, which was used during training. The CBT training was a short course consisting of 5 sessions covering the basis of MUS; the relevance of the explanatory model, elicited by the SEMI, to the CBT model of such symptoms; and the CBT treatment approach. Training was accomplished through lectures, supplemented by case vignettes and role‐play of therapeutic sessions by simulated participants based on case scenarios from the pilot trial, all with reference to the intervention manual (for details: Sumathipala 2008, ref 30) Therapist: trained primary care physicians 2. Structured care (n = 75) Duration: 6 x 30‐minute structured sessions over a period of 3 months (of which 3 mandatory) Treatment protocol: treatment started with a SEMI interview by 1 of the authors: only results were passed on to primary care physician, they were not trained to use this information to inform the strategy for their CBT intervention No CBT was given, but during the 6 sessions the physicians were free to manage the participants as they wished within the sessions. No training or supervision was provided for these doctors, and the intervention was not manualised Therapist: untrained primary care physicians
Outcomes	Time points for assessment: baseline, 3 months, 6 months, 9 months and 12 months after baseline Outcomes: 1. number of symptoms (as counted by researcher) 2. level of distress/psychiatric morbidity (GHQ‐30) 3. number and severity of symptoms (BSI + 2 open‐ended questions) 4. number of participant initiated doctor visits (diary)
Notes	Date of study: not described Funding source: The Wellcome Trust (international programme) provided a project grant Declarations of interest among the primary researchers: none reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Trial participants were first randomised to the 2 intervention groups using a random permuted block design, with a block size of 4. Next, participants were randomly allocated to 1 of the 3 doctors selected to deliver the intervention to which they had been allocated
Allocation concealment (selection bias)	Low risk	Randomisation codes were generated by a statistician in the UK and passed on to the independent epidemiologist (M.R.N.A.) in Sri Lanka, who executed the random allocation of treatment condition
Blinding of participants and personnel (performance bias) All outcomes	High risk	Neither the primary care doctors who delivered the interventions nor the participants who received them could be masked to their allocation because of the nature of the interventions
Blinding of outcome assessment (detection bias) All outcomes	High risk	Participants were not blinded, and most outcomes were participant reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Loss to follow‐up was highest after 6 and 12 months: 20% at both time points in the intervention group and 28% (6 months) and 30% (12 months) in the structured care group (> 20%)
Selective reporting (reporting bias)	Low risk	All intended outcomes are reported
Treatment fidelity	Low risk	Treatments were structured. A treatment manual was prepared to keep the therapeutic sessions uniform (Sumathipala 2000, p. 750)
Researcher allegiance	Low risk	No indication that researchers had a preference for 1 of the treatment modalities
Other bias	Low risk	No other sources of bias