Table 3

Outcome Differences Between RCTs and OSs

Brief Title (outcome)
Number of Studies
Average Failure Rate (Number of Patients)





Control**
Treatment

Anticoagulants (MI) [13, 14, 64]



     RCT
6
17% (1748)
14% (2106)
     OS
9
31%(3615)
16%(2598)
Quinidine (Afib) [13, 14, 28] 3 months



     RCT
6
54% (390)
36% (413)
     OS
5
61%* (200)
53% (342)
Trial of labor (Breech) [13, 65]



     RCT
2
2% (128)
3% (182)
     OS
6
5% (1043)
4%(1552)
Colposuspension 1 (Incontinence) [13, 66]



     RCT
2
33% (134)
12% (139)
     OS
6
37%(508)
26% (374)
Colposuspension 2 (Incontinence) [13, 66]



     RCT
2
31% (132)
12% (139)
     OS
4
32%(190)
23%(349)
TENS (Pain) [13, 14, 67]



     RCT
2
18% (34)
3% (34)
     OS
4
76%(172)
56%(136)
Early Discharge (Childbirth) [13, 68]



     RCT
1
8% (38)
5% (93)
     OS
4
21%(379)
19%(402)
Hip Screw (Hip Fx) [13, 69]



     RCT
1
50% (26)
12% (33)
     OS
5
35%(290)
8% (560)
Local Anesthesia (CEA) [11, 13, 70]



     RCT
3
5% (79)
5% (75)
     OS
11
5%* (1509)
2% (1713)
HSG (Infertility) [11, 13, 71]



     RCT
5
81% (527)
69% (302)
     OS
6
74%(734)
58%(1072)

RCT = Randomized Controlled Trial; OS = Observational Study

MI = Myocardial Infarction

Afib = Atrial fibrillation

TENS = Transcutaneous electrical nerve stimulation

Hip Fx = Hip Fracture

CEA = Carotid Endarterectomy

HSG = Hysterosalpingography

** The treatment group is listed in the row title. As seen in Table 1, the control (i.e. comparison) groups are the negative of the listed treatment except for the following: Control group for colposuspension 1 is colporrhaphy, colposuspension 2 is needle suspension, hip fracture is fixed nail plates, CEA is general anesthesia, infertility is water soluble medium.

Significance testing was only done to test heterogeneity among the failure rates for the observational studies

* P < .05 for test of heterogeneity of failure rates combined to create the average

† P < .001 for test of heterogeneity of failure rates combined to create the average

Hartz et al. Emerging Themes in Epidemiology 2005 2:8   doi:10.1186/1742-7622-2-8