Table 1.
Characteristics of included trials
| Trial | Interventions | Population | No of patients* | Follow-up (weeks) | Low dose aspirin allowed | Concealed allocation† | Blinding‡ | Events adjudicated§ | Intention to treat | Source |
|---|---|---|---|---|---|---|---|---|---|---|
| ADAPT | Placebo v naproxen (440 mg/d) v celecoxib (400 mg/d) | At risk for Alzheimer disease | 2528 | 198 | Yes | Yes | Yes | Yes | Yes | Journal article,w1 trial websitew2 |
| Aisen 2003 | Placebo v naproxen (440 mg/d) v rofecoxib (25 mg/d) | Alzheimer disease | 351 | 60 | Yes | Yes | Yes | No | Yes | Journal article,w3 personal communicationw4 |
| Geusens 2004 | Placebo v naproxen (1000 mg/d) v lumiracoxib (200, 400 mg/d) | Rheumatoid arthritis | 1124 | 26 | Yes | Yes | Yes | No | Yes | Journal article,w5 personal communicationw6 w7 |
| APC | Placebo v celecoxib (400, 800 mg/d) | Adenomatous polyps (colon) | 2035 | 160 | Yes | Yes | Yes | Yes | Yes | Journal article,w8 study registerw9 |
| GAIT | Placebo v celecoxib (200 mg/d) | Osteoarthritis | 631 | 24 | No | Yes | Yes | No | Yes | Journal articlew10 |
| IQ5-97-02-001 | Placebo v celecoxib (400 mg/d) | At risk for Alzheimer disease | 425 | 52 | Yes | Yes | Yes | No | No | FDA reports,w11 study register,w12 personal communicationw13 w14 |
| PreSAP | Placebo v celecoxib (400 mg/d) | Adenomatous polyps (colon) | 1561 | 159 | Yes | Yes | Yes | Yes | Yes | Journal article,w15 personal communicationw13 w14 |
| Lehmann 2005 | Placebo v celecoxib (200 mg/d) v lumiracoxib (100 mg/d) | Osteoarthritis | 1684 | 13 | Yes | Yes | Yes | Yes | Yes | Journal article,w16 personal communicationw6 w7 |
| APPROVe | Placebo v rofecoxib (25 mg/d) | Adenomatous polyps (colon) | 2586 | 229 | Yes | Unclear | Yes | Yes | Yes | Journal articlew17 |
| Reines 2004 | Placebo v rofecoxib (25 mg/d) | Alzheimer disease | 692 | 65 | Unclear | Yes | Yes | Yes | Yes | FDA reportsw18 |
| Thal 2005 | Placebo v rofecoxib (25 mg/d) | At risk for Alzheimer disease | 1457 | 210 | Yes | Yes | Yes | Yes | Yes | Journal articlew19 |
| VICTOR | Placebo v rofecoxib (25 mg/d) | Adjuvant (colon cancer) | 2327 | 235 | Yes | Yes | Yes | Yes | No | FDA reports,w18 journal article,w20 study register,w21 trial websitew22 |
| ViP | Placebo v rofecoxib (25 mg/d) | At risk for prostate cancer | 4741 | 75 | Yes | Yes | Yes | Yes | Yes | Study registerw23 |
| A3191152 | Naproxen (1000 mg/d) v celecoxib (200 mg/d) | Osteoarthritis | 589 | 26 | Yes | Yes | Yes | No | No | Study registerw24 |
| SUCCESS-1 (USA/Canada) | Naproxen (1000 mg/d) v celecoxib (200, 400 mg/d) | Osteoarthritis | 2736 | 12 | Yes | Yes | Yes | No | No | Journal article,w25 study register,w26 personal communicationw13 w14 |
| ADVANTAGE | Naproxen (1000 mg/d) v rofecoxib (25 mg/d) | Osteoarthritis | 5557 | 14 | Yes | Unclear | Yes | Yes | No | FDA reports,w18 w27 journal articlew28 w29 |
| VIGOR | Naproxen (1000 mg/d) v rofecoxib (50 mg/d) | Rheumatoid arthritis | 8076 | 54 | No | Yes | Yes | Yes | Yes | FDA reports,w18 w27 w30 journal articlew31 |
| TARGET (0117) | Naproxen (1000 mg/d) v lumiracoxib (400 mg/d) | Osteoarthritis | 9511 | 56 | Yes | Yes | Yes | Yes | No | FDA reports,w32 journal article,w33 personal communicationw6 w7 |
| CLASS (N49-98-02-035) | Ibuprofen (2400 mg/d) v celecoxib (800 mg/d) | Osteoarthritis and rheumatoid arthritis | 3975 | 65 | Yes | Yes | Yes | No | No | FDA reports,w11 w34 w35 personal communicationw13 w14 |
| TARGET (2332) | Ibuprofen (2400 mg/d) v lumiracoxib (400 mg/d) | Osteoarthritis | 8814 | 56 | Yes | Yes | Yes | Yes | No | FDA reports,w23 journal article,w33 personal communicationw6 w7 |
| CAESAR | Diclofenac (100 mg/d) v celecoxib (200 mg/d) | Osteoarthritis | 916 | 52 | Yes | Yes | Yes | No | No | FDA reports,w11 study register,w36 conference abstract,w37 personal communicationw13 w14 |
| CLASS (N49-98-02-102) | Diclofenac (150 mg/d) v celecoxib (800 mg/d) | Osteoarthritis and rheumatoid arthritis | 3993 | 52 | Yes | Yes | Yes | No | No | FDA reports,w11 w34 w35 personal communicationw13 w14 |
| Emery 1999 | Diclofenac (150 mg/d) v celecoxib (400 mg/d) | Rheumatoid arthritis | 655 | 24 | No | Yes | Yes | No | No | Journal article,w38 personal communicationw13 w14 |
| SUCCESS-1 (World) | Diclofenac (100 mg/d) v celecoxib (200, 400 mg/d) | Osteoarthritis | 10458 | 12 | Yes | Yes | Yes | No | No | Journal article,w25 study register,w26 personal communicationw13 w14 |
| EDGE | Diclofenac (150 mg/d) v etoricoxib (90 mg/d) | Osteoarthritis | 7111 | 73 | Yes | Yes | Yes | Yes | No | FDA reports,w39 journal articlew40 |
| EDGE II | Diclofenac (150 mg/d) v etoricoxib (90 mg/d) | Rheumatoid arthritis | 4086 | 150 | Yes | Yes | Yes | Yes | No | Journal articlew41 |
| MEDAL | Diclofenac (150 mg/d) v etoricoxib (60, 90 mg/d) | Osteoarthritis and rheumatoid arthritis | 23498 | 176 | Yes | Yes | Yes | Yes | No | Journal articlew42 |
| Cannon 2000 | Diclofenac (150 mg/d) v rofecoxib (12.5, 25 mg/d) | Osteoarthritis | 784 | 52 | No | Yes | Yes | No | No | FDA reports,w43 w44 journal articlew45 |
| Saag 2000 | Diclofenac (150 mg/d) v rofecoxib (12.5, 25 mg/d) | Osteoarthritis | 693 | 54 | No | Yes | Yes | No | No | FDA reports,w43 w44 journal articlew46 |
| Fleischmann 2003** | Celecoxib (200 mg/d) v lumiracoxib (200, 400 mg/d) | Osteoarthritis | 1376 | 13 | Yes | Yes | Yes | No | No | Journal article,w47 personal communicationw6 w7 |
| Tannenbaum 2004** | Celecoxib (200 mg/d) v lumiracoxib (200, 400 mg/d) | Osteoarthritis | 1459 | 13 | Yes | Yes | Yes | No | Yes | Journal article,w48 personal communicationw6 w7 |
FDA=US Food and Drug Administration. See web extra appendix 3 for full names of trials and references to sources.
*Number of randomised patients of included trial arms.
†Yes if investigators responsible for patient selection were unable to suspect before allocation which treatment was next in line (central randomisation, sequentially numbered, sealed, opaque assignment envelopes, coded drug packs).
‡Yes if drugs looked similar (for example, matching placebo) or double dummy was used.
§Relates to myocardial infarctions only. See web extra appendix 2 for more detailed description of external adjudication of events in individual trials.
¶Yes if all randomised patients were included in analysis.
**Placebo arm excluded (<100 patient years of follow-up).