Targan 2007.

Methods	A phase III, randomized, double‐blind, placebo‐controlled , multi‐center (114 sites), 12 week induction trial Permitted concurrent therapies for Crohn’s disease: stable doses of 5‐ASA, prednisone or an equivalent corticosteroid (20 mg/day), budesonide (6 mg/day), AZA, 6‐MP, MTX, and antibiotics, were permitted Exclusion criteria: short‐bowel syndrome, an ostomy, total colectomy, a stricture with obstructive symptoms, draining fistulas, an abdominal abscess, anti‐TNF‐α therapy within the previous 12 weeks, previous natalizumab therapy
Participants	509 patients with moderate to severely active Crohn's disease (CDAI ≥220 and ≤ 450) and evidence of active inflammation (CRP > 2.87 mg/L) Patients were randomized 1:1 to treatment with natalizumab or placebo infusions at weeks 0, 4, and 8 Efficacy and safety assessments were performed at weeks 4, 8, and 12
Interventions	Three infusions of intravenous natalizumab (300 mg, n = 259) or placebo (n = 250) at weeks 0, 4, and 8
Outcomes	The primary outcome variable was response by week eight which was defined as a reduction of > 70 points in the CDAI score from baseline that sustained through week twelve Secondary outcome variables included the proportion of patients achieving response at week twelve, a clinical remission (defined as a CDAI score of less than 150) by week eight sustained to week twelve, and the proportion of patients with remission at week 12 Tertiary outcomes included proportion achieving a > 100‐point CDAI score decrease at weeks eight and twelve, achieving clinical response or remission at weeks four and eight, response at week eight, time to clinical remission, mean change in baseline CDAI, platelets and CRP at weeks four, eight and twelve, and to evaluate the effects of natalizumab on quality of life (IBDQ and SF‐36) at week twelve
Notes	Industry support: Funded by Elan Pharmaceuticals, Inc and Biogen Idec. Various authors acted as paid consultants or employees for Elan or Biogen Idec
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Centrally stratified
Allocation concealment (selection bias)	Low risk	Centrally randomized
Blinding (performance bias and detection bias) All outcomes	Low risk	Double‐blind: the patients, site staff, and study investigators were all blinded to treatment assignment
Incomplete outcome data (attrition bias) All outcomes	Low risk	Patient withdrawals were reported as failures and were evenly distributed across treatment groups
Selective reporting (reporting bias)	Unclear risk	Most expected outcomes were reported Although the mean change in SF‐36 and its components (a secondary outcome) was not reported the authors did report on the mean change in IBDQ from baseline
Other bias	Low risk	The study appears to be free of other sources of bias

CDAI: Crohn's disease activity index

AZA: azathioprine

6‐MP: 6‐mercaptopurine

MTX: methotrexate

CYA: cyclosporine

ID: investigational drug

CRP: C‐reactive protein

IBDQ: Inflamatory Bowel Disease Questionnaire

5‐ASA: 5‐aminosalicylic acid

TAC: tacrolimus

ESR: erythrocyte sedimentation rate

anti‐TNF‐α: anti‐tumour necrosis factor‐alpha

SF‐36: 36‐Item Short Form Health Survey