Smith 2008.

Methods	STUDY DESIGN: Parallel group trial LOCATION, NUMBER OF CENTRES: Two large teaching hospitals in Brisbane, Australia. DURATION OF STUDY: 4 months COMPLIANCE: Not reported CONFOUNDERS: Just under one half of the participants attended outpatient clinics within 4 to 6 weeks and received additional educational intervention.
Participants	N SCREENED: 148 N RANDOMISED: 146 N COMPLETED: 142 M = 64/F = 82 MEAN AGE: 34 years CHARACTERISTICS: Mean age of 34 (SD 13,8 years), groups did not differ in age, education, gender, income, and previous ED attendances, however the PCE group was more likely to have seen a GP in previous 7 days. Improvement in Peak flow monitoring (post intervention). Reduction of re‐attendance at 4 months post (PCE group had fewer re‐attendance). INCLUSION CRITERIA: Diagnosis of asthma prior to presentation; aged >18 yrs; able to read and write English; no other concurrent respiratory medical condition. EXCLUSION CRITERIA: Too ill to participate (as deemed by medical officer) and/or required intensive care medical treatment. Case notes were used to confirm the participant’s eligibility and to exclude patients with other respiratory diseases and conditions that cause dyspnoea.
Interventions	Education group Patient‐centered education (PCE) utilizing an Asthma Foundation Leaflet and comprising a commonly used standard education curriculum for both groups. Control group Standard patient education. FOLLOW‐UP PERIOD: 4 & 12 months
Outcomes	Re‐attendance; re‐admission; symptom control (measured with Asthma Control Questionnaire; withdrawals
Notes	S Smith supplied data and study details in comment posted on the review.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	'Patients were randomised by day of their birth month with odd days randomised to the intervention group of PCE and even dates to the SPE group'
Allocation concealment (selection bias)	Low risk	'Treating staff at both EDs were blind to the randomisation process as there was no acknowledgement or notation of group assignment in the patient’s chart or elsewhere and they were not present when the education was given' Treating staff were likely to have been unaware of treatment group assignment.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Participants were informed that they would receive education but were blind to the exact nature/content of delivery. Researcher would have been aware of treatment group assignment.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Re‐attendance data verified from patient records for all participants. Secondary outcome data were collected from outpatient appointments but there was more evidence of high withdrawal rates.
Selective reporting (reporting bias)	Low risk	Following information supplied from comments posted on this review we have revised our judgement from high risk of reporting bias to low risk of bias.
Other bias	Low risk	No issues were detected.