Introduction
Materials and methods
Search strategy
Study selection
Data extraction and meta-analysis
Results
First author and year | Country | Setting/Population | Arms of RCT (n) | Duration of study | Type of outcome | Result |
---|---|---|---|---|---|---|
Boers, 1997 [15] | The Netherlands and Belgium | 155 early RA patients from 8 centers | Combination therapy - step-down prednisolone from 60 mg, step-down MTX and SSZ (76) vs SSZ monotherapy (79) | 28 weeks | Infections treated as outpatient | 12 infections in combination arm, 6 in SSZ monotherapy arm |
Chamberlain, 1976 [16] | UK | 49 adult RA patients from single center | 5 mg prednisolone (20) vs 3 mg prednisolone (10) vs 0 mg prednisolone (19) Allowed concomitant gold | 2- 3.5 years | n/a | No infections |
Choy, 2005 [17] | UK | 91 patients with established RA with incomplete response to DMARDs. Multicenter study | Monthly 120-mg intramuscular depomedrone (48) vs placebo (43) Allowed usual DMARDs | 2 years | n/a | No infections either arm |
Choy, 2008 [18] | UK | 467 patients within 2 years of diagnosis from 42 centers | MTX (117) MTX + cyclosporin (119) MTX + step-down prednisolone (115) MTX + cyclosporin + prednisolone (116) | 2 years | a) All-site serious infections b) Respiratory tract infections | a) 7, 3, 4, and 2 serious infections in the four respective arms b) 54, 51, 49, and 55 respiratory tract infections in the four respective arms |
Durez, 2007 [19] | Belgium | 44 patients with early RA | MTX monotherapy (14) MTX + 1 g iv methylprednisolonea (15) MTX + infliximaba (15) Infusions weeks 0, 2, 6; then 8 weekly | 46 weeks | a) Serious infection b) 'benign' infection | a) No serious infections in any arm b) 14, 12, and 12 benign infections in the three arms, respectively |
Durez, 2004 [20] | Belgium | 27 patients with active RA despite MTX | MTX + 1 g iv MP week 0 (15) MTX + infliximab weeks 0, 2, and 6 (12) | 14 weeks | Serious infections | None in either arm |
Gerlag, 2004 [21] | The Netherlands | 21 patients with active RA despite DMARDs | 60 mg prednisolone week 1, then 40 mg prednisolone week 2 (10) Placebo (11) | 2 weeks | n/a | 1 skin infection in placebo arm only |
Heytman, 1994 [22] | Australia | 60 patients with active RA previously treated with NSAIDs | Gold plus either 1 g iv methylprednisolone weeks 0, 4, and 8 (30) or placebo (30) | 24 weeks | All patient-reported side effects | No infections reported |
Jasani, 1968 [23] | UK | 9 patients with erosive RA | 4 × 1-week crossover study of ibuprofen 750 mg, aspirin 5 g, prednisolone 15 mg, and lactose as placebo | 4 weeks | n/a | No infections reported |
Kirwan, 2004 [24] | Belgium, Sweden, UK | 143 patients with active RA | Budesonide, 3 mg (37), budesonide, 9 mg (36), prednisolone, 7.5 mg (39), placebo (31) | 12 weeks | a) Respiratory infections b) Viral infections | a) 7, 4, 6, and 1 respiratory infections in the 4 groups, respectively. b) 4, 1, 0, and 0 viral infections in the four groups, respectively |
Liebling, 1981 [25] | US | 10 patients with active RA | Crossover trial of monthly 1-g iv methylprednisolone vs placebo | 12 months (6 months per arm) | n/a | 4 infections on placebo, 2 on GC |
Murthy, 1978 [26] | UK | 24 patients with > 30 minutes morning stiffness | Indomethacin, 25 mg × 4 (12), prednisolone, 5 mg (12) | 2 weeks | n/a | No infections reported |
Sheldon, 2003 [27] | UK | 26 patients with active RA | Budesonide (14) or placebo (12) plus usual DMARDs | 4 weeks | n/a | 2 cases of influenza (one from each group). |
Van Everdingen, 2002 [28] | The Netherlands | 81 patients with active, previously untreated RA | 10-mg prednisolone (40), placebo (41) | 2 years | Data reported on infections treated with antibiotics | 17 infections in 40 patients in GC arm, 22 infections in 41 patients in placebo arm |
Wassenberg, 2005 [29] | Germany/Austria/Switzerland | 192 patients with active RA, disease duration < 2 years | Gold or MTX plus either 5 mg prednisolone (93) or placebo (96) | 2 years | All adverse events collected, reported only if occurred in 3 or more patients | Total 4/93 and 3/96 (Bronchitis in 3/93 prednisolone group, 0/96 placebo group. Influenza in 1/93 prednisolone group, 3/96 placebo) |
Williams, 1982 [30] | UK | 20 patients with active RA | 1-g iv methylpredisonolone (10) or placebo (10) | 6 weeks | "Serious side effects" | None reported |
Wong, 1990 [31] | Australia | 40 patients with active RA previously treated with NSAIDs | Gold plus either three pulses of 1 g intravenous methylprednisolone weeks 0, 4, + 8 (20) or placebo (20) | 24 weeks | Patients interviewed for all possible side effects | 1 injection-site infection in placebo group |
Capell, 2004 [32] | UK | 167 patients with active RA on no DMARD therapy | SSZ plus either 7 mg prednisolone (84) or placebo (83) | 2 years | Withdrawals due to side effects | No discontinuations due to infection in either group |
Svensson, 2005 [33] | Sweden | 250 patients with active disease on DMARD therapy | DMARD + prednisolone, 7.5 mg (119), DMARD alone, open, no placebo (131) | 2 years | Adverse events leading to withdrawal | 1 abscess in non-prednisolone group. No infections leading to discontinuation in prednisolone group |
Van der Veen, 1993 [34] | The Netherlands | 30 patients with active RA | Oral MTX plus either placebo (10) or 100 mg oral prednisolone days 1, 3, and 5 (10) or 1 g iv MP days 1, 3, and 5 (10) | 1 year | Adverse events leading to discontinuation of MTX | 1 pneumonia in placebo group (at week 12) |
van Schaardenburg, 1995 [35] | The Netherlands | 56 patients with active RA aged > 60 previously treated with NSAIDs | Chloroquine, 100 mg/day (28) (rescue with gold, then SSZ allowed) vs prednisolone 15 mg/day, tapered after 1 month (28) | 2 years | Withdrawal due to adverse advents | No discontinuations due to infections in either group |
Main results
RCTs
Observational studies
Number of studies | Mean RR | I2 statistic | Ratio of RR | |
---|---|---|---|---|
Study design | ||||
Cohort | 33 | 1.55 (1.35, 1.79) | 71.3% | 1.00 (referent) |
Case-control | 9 | 1.95 (1.61, 2.36) | 79.4% | 1.26 |
Definition of exposure | ||||
Baseline | 5 | 1.46 (0.87, 2.45) | 79.7% | 1.00 (referent) |
Current (within 3/12) | 22 | 1.70 (1.47, 1.97) | 58.9% | 1.16 |
Recent (within 6/12) | 7 | 1.56 (1.24, 1.96) | 79.5% | 1.07 |
Ever | 2 | 1.80 (1.29, 2.51) | 52.5% | 1.23 |
Unclear | 6 | 2.35 (1.27, 4.36) | 36.5% | 1.61 |
Adjusted for age and sex | ||||
No | 22 | 1.32 (0.97, 1.80) | 67.6% | 1.00 (referent) |
Yes | 19 | 1.78 (1.58, 2.01) | 82.3% | 1.35 |
Adjusted for disease severity | ||||
No | 24 | 1.41 (1.14, 1.75) | 71.3% | 1.00 (referent) |
Adjusted for surrogate | 10 | 1.98 (1.68, 2.34) | 78.5% | 1.40 |
Adjusted for direct measurement | 6 | 1.52 (1.17, 1.97) | 77.0% | 1.08 |
Adjusted for disease duration | ||||
No | 33 | 1.63 (1.41, 1.89) | 76.8% | 1.00(referent) |
Yes | 6 | 1.55 (1.20, 2.01) | 83.5% | 0.95 |
Adjusted for comorbidity | ||||
No | 22 | 1.30 (0.97, 1.74) | 64.2% | 1.00 (referent) |
Yes | 17 | 1.74 (1.55, 1.96) | 75.1% | 1.34 |
Adjusted for other RA therapies | ||||
No | 22 | 1.28 (0.98, 1.67) | 61.1% | 1.00 (referent) |
Yes | 18 | 1.84 (1.62, 2.08) | 82.8% | 1.44 |