Sands 2007.

Methods	Multicenter, randomized, double‐blind, placebo‐controlled 10 week induction trial was conducted at 17 sites in the United States
Participants	79 adult patients with at least a 6 month history of Crohn's with active disease (CDAI ≥ 150) receiving infliximab therapy Permitted concomitant medication: oral 5‐ASA if on a stable dose for at least 4 weeks; stable dose oral antibiotics if on a stable dose for at least 4 weeks; oral corticosteroids (≤ 25 mg prednisolone or it's equivalent per day) for at least 4 weeks and at a stable dose of at least 2 weeks; azathioprine, 6‐mercaptopurine, or methotrexate if on a stable dose for at least 8 weeks Exclusion criteria: bowel surgery within 3 weeks of entry, persistent intestinal obstruction, bowel perforation, uncontrolled bleeding, abdominal abscess, infection, symptoms largely a result of fibrotic strictures, neoplastic disease of the bowel, a positive stool culture for enteric pathogens at screening, nasogastric/nasoenteric tube feeding, an elemental diet, total parenteral nutrition within 2 weeks of week 0, significantly abnormal results of liver function tests at screening, and prior natalizumab therapy
Interventions	Patients received either 300 mg natalizumab (n = 52) or placebo (n = 27) every 4 weeks at weeks 0, 4, and 8, in addition to receiving an intravenous infliximab infusion (5 mg/kg) at week 6
Outcomes	The primary outcome was the safety and tolerability of natalizumab in patients with Crohn's disease concurrently receiving infliximab Secondary outcomes included therapeutic efficacy, health‐related quality of life, the effects of natalizumab on markers of inflammation, and the effects of concurrent therapy
Notes	Industry support: Funding provided by Elan pharmaceuticals, various authors report being employees or paid consultants of Elan
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomization
Allocation concealment (selection bias)	Low risk	Centrally randomized
Blinding (performance bias and detection bias) All outcomes	Low risk	Double blind: the placebo was identical in appearance to natalizumab, contained the same buffering solution, and infused over the same length of time
Incomplete outcome data (attrition bias) All outcomes	Low risk	Patient withdrawals were reported and distributed evenly across treatment groups
Selective reporting (reporting bias)	Low risk	All expected outcomes were reported
Other bias	Low risk	The study appears to be free of other sources of bias