Zwar 2016.

Study characteristics
Methods	Cluster‐RCT (36 clusters); follow‐up: 12 months, control group: usual care (copy COPD treatment guidelines)
Participants	Newly diagnosed COPD patients Eligible: 287, I: 169, C: 118 Randomised: 254, I: 144, C: 110 Completed: 222, I: 126, C: 96 Mean age: I: 67 years, C: 65 years Sex (% male): I: 62 C: 58 Inclusion criteria: current and former smokers, aged 40 to 85 years, newly identified as having COPD on post‐bronchodilator spirometry (post‐bronchodilator FEV₁/FVC < 0.7), had attended the practice at least twice with ≥ 1 visit in the preceding 12 months Major exclusions: recorded diagnosis of COPD, unable to understand English sufficiently to complete study questionnaires or procedures, cognitive impairment
Interventions	Nurses and GPs in intervention practices were educated to work in partnership to identify patients with COPD and to initiate an evidence‐based early intervention programme Intervention components: care plan, education, optimal diagnosis, management of anxiety and depression, medication, influenza and pneumococcal vaccination, referral to PR and/or dietician if necessary/appropriate, smoking cessation advice and resources if necessary. Multidisciplinary teams; professional roles Duration intervention: intervention duration not fixed; expected to be completed within 6 months Disciplines involved: GP, nurse, physiotherapist, dietician
Outcomes	SGRQ, CAT, general health status, post‐bronchodilator FEV₁, COPD knowledge score, awareness of diagnosis of COPD, smoking status, immunisation status for influenza and pneumococcus, effective inhaler use (when prescribed), attendance at pulmonary rehabilitation, healthcare utilisation, intervention uptake
Notes	Dominant component: e ducation
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “randomization and group allocation of GP practices was performed by an independent statistician using a computer‐generated randomisation program”
Allocation concealment (selection bias)	Low risk	Quote: “allocation concealment will be ensured as group allocation will be conducted at the same time as randomisation. Practices will be informed about their group allocation by fax (Bunker et al, 2012)
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: " i n this pragmatic trial , participating GPs, PNs, and patients were not blind to the aims of the study n or to their randomisation group"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: “project officers, who collected study outcome measures, and the statistician undertaking analyses were blind to group allocation”
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: “the primary analysis was by intention to treat and relied on the diagnosis of COPD assigned by the PN/GP on the basis of case‐finding spirometry” Comment: 34 and 10 patients withdrew from intervention and control groups, respectively. Reasons for withdraw not likely to be related to true outcomes
Selective reporting (reporting bias)	Low risk	Comment: all outcome measures described in the protocol, except patient satisfaction, are reported
Recruitment bias	High risk	Comment: patient inclusion following case‐finding procedures were performed after randomisation and group allocation of GP practices
Baseline imbalance between groups	Low risk	Comment: groups did not differ substantially in mean SGRQ nor in other characteristics
Loss to follow‐up of clusters	Unclear risk	Comment: 4 practices withdrew after randomisation, and 2 practices merged into 1 during the study period
Adequate analysis methods for CRT	Low risk	Quote: “intra‐cluster(practice) correlation coefficients (ICCs) were determined for all primary outcome variables” “The effect of the intervention on outcomes measured on a continuous scale (such as SGRQ score) were estimated and tested using mixed‐model analysis of variance in which time and treatment group were fixed effects and GP practice and subject nested within practice were random effects. The effect of the intervention on dichotomous variables was analysed using generalized estimating equations with a logistic link and a model structure that is analogous to that described above"

4MWT: four‐minute walking test; 6MWD: six‐minute walking distance; ADL: activities of daily living; BMI: body mass index; BTS: British Thoracic Society; COOP: Dartmouth Primary Care Co‐operative Quality of Life Questionnaire; COPD: chronic obstructive pulmonary disease; CRDQ: Chronic Respiratory Disease Questionnaire; CRQ: Chronic Respiratory Questionnaire; ED: emergency department; FEV₁: forced expiratory volume in one second; FVC: forced vital capacity; GOLD: Global Initiative for Chronic Obstructive Lung Disease; GP: general practitioner; HADS: Hospital Anxiety and Depression Scale; HCCQ: Health Care Communication Questionnaire; HRQoL: health‐related quality of life; I: intervention; MACL: Mood Adjective Check List; MBHI: Millon Behavioral Health Inventory; MCO: managed care organisation; MMSE: Mini–Mental State Examination; MRC: Medical Research Council; NYHA: New York Heart Association; PR: pulmonary rehabilitation; RCT: randomised controlled trial; SGRQ: St. George's Respiratory Questionnaire; SIP: Sickness Impact Profile; VAS: visual analogue scale; VC: vital capacity; YQLQ: York Quality of Life Questionnaire.