Skip to main content
. 2006 Oct 17;66(3):377–388. doi: 10.1136/ard.2006.062091

Table 4 Evidence of safety—pooled relative risk (RR) or odds ratio (OR)* and 95% confidence interval (CI).

Intervention† Adverse events RR/OR (95% CI) Evidence References
Paracetamol GI discomfort 0.80 (0.27 to 2.37) RCTs 48
GI perforation/bleed 3.60 (2.60 to 5.10) Case‐control study 49
GI bleeding 1.2 (0.8 to 1.7) Case‐control studies 50
Renal failure 2.5 (1.7 to 3.6) Case‐control study 51
Renal failure 0.83 (0.50 to 1.39) Cohort study 52
Topical NSAIDs GI events 0.81 (0.43 to 1.56) RCTs 25
GI bleed/perforation 1.45 (0.84 to 2.50) Case‐control 42
Glucosamine sulphate preparations Any 0.97 (0.88 to 1.08) RCTs 53
Diacerhein Diarrhoea 3.98 (2.90 to 5.47) RCTs 54, 55
NSAIDs GI perforation/ulcer/bleed 5.36 (1.79 to 16.10) RCTs 56
GI perforation/ulcer/bleed 2.70 (2.10 to 3.50) Cohort studies 56
GI perforation/ulcer/bleed 3.00 (2.70 to 3.70) Case‐control studies 56
GI protective strategies versus NSAID alone
H2 blocker+NSAID Serious GI complications 0.33 (0.01 to 8.14) RCTs 57
Symptomatic ulcers 1.46 (0.06 to 35.53) RCTs 57
Serious CV or renal events 0.53 (0.08 to 3.46) RCTs 57
PPI+NSAID Serious GI complications 0.46 (0.07 to 2.92) RCTs 57
Symptomatic ulcers 0.09 (0.02 to 0.47) RCTs 57
Serious CV or renal events 0.78 (0.10 to 6.26) RCTs 57
Misoprostol+NSAID Serious GI complications 0.57 (0.36 to 0.91) RCTs 57
Symptomatic ulcers 0.36 (0.20 to 0.67) RCTs 57
Serious CV or renal events 1.78 (0.26 to 12.07) RCTs 57
Diarrhoea 1.81 (1.52 to 2.61) RCTs 58
COX‐2 selective Serious GI complications 0.61 (0.34 to 1.10) RCTs 57
Symptomatic ulcers 0.41 (0.26 to 0.65) RCTs 57
Serious CV or renal events 0.95 (0.55 to 1.66) RCTs 57
COX‐2 specific (coxibs) Serious GI complications 0.55 (0.38 to 0.80 RCTs 57
Symptomatic ulcers 0.49 (0.38 to 0.62) RCTs 57
Serious CV or renal events 1.19 (0.80 to 1.75) RCTs 57
Surgery
 T+LRTI/IA v T Any 2.12 (1.24 to 3.60) RCTs 29
 TJR v T+IA Any 5.00 (0.26 to 95.02) RCTs 29

*RR was calculated for an RCT or cohort study and OR was for a case‐control study. RR (or OR)  = 1: no difference between treatment and control; RR (or OR )>1: more risky with treatment; RR <1: less risky with treatment. The results were pooled if more than one study was involved; †compared with placebo/non‐exposure unless otherwise stated.

H2‐blockers, histamine type 2 receptor antagonists; PPIs, proton pump inhibitors; GI, gastrointestinal; CV, cardiovascular; CNS, central nervous system; T, trapeziectomy; LRTI, ligament reconstruction and tendon interposition; IA, interposition arthroplasty; TJR, total joint replacement.