Table 1 Assessment of Risk of Bias of Randomized Controlled Trials
From: Erythropoiesis stimulating agents and reno-protection: a meta-analysis
Reference | Trial features | Randomized sequence | Allocation concealment | Blinding of outcome assessors | ITT analysis | Reports on Lost patients | All patients treated in assigned group |
|---|---|---|---|---|---|---|---|
Dardashti 2014 [24] | AKI: DB, SS | Low risk: patients were randomly allocated. | Low risk: sequentially numbered, sealed, & opaque envelopes. Independent nurses prepared the study drug & syringes were delivered blinded | High risk | High risk: 5 patients that received study drug were discontinued and excluded from analysis | Low risk: lost patients reported | Low risk: all patients treated |
deSeigneux 2012 [76] | AKI: DB, SS | Low risk: a randomization code was generated by computer | Low risk: envelopes with allocation were prepared by the quality of care unit. A nurse opened the envelopes and prepared the syringes for injection. Investigators and patients were blinded to the treatment | High risk | Low risk: AKI data on all patients | Low risk: lost patients reported | Low risk: all patients treated |
Endre 2010 [26] | AKI: DB, MS (2 centers) | Low risk: allocation by a predefined computer-generated randomization sequence | Low risk: concealment was by a pharmacist; pairs of identical syringes. Patients, all medical staff, & investigators were blinded to treatment | Low risk: Data Safety Monitoring Board with unmasking followed recording of the final AEs of the patient last enrolled | Low risk | Low risk: lost patients reported | Low risk: but 1 patient withdrew |
Kim 2013 [27] | AKI: DB, SS | Low risk: computer-generated random code | Low risk: medications were prepared by a nurse who knew the patient’s group assignment but was not involved in the study | Unclear risk | Low risk: No dropouts | Low risk: lost patients reported | Low risk: all patients treated |
Oh 2012 [16] | AKI: DB, SS | Low risk: A randomization code list with a block size of two was generated. Treatments were allocated to patients through the Internet in accordance with the predefined randomization list | Low risk: a research coordinator performed randomization and prepared the study drugs | Unclear risk | Low risk | Low risk: all patients completed the trial | Low risk: all patients completed the trial |
Tasanarong 2013 [28] | AKI: DB, SS | Low risk: treatment assignment by blocked randomization. Sealed envelopes containing the allocation group were opened by nurses who did not participate in the study | Low risk: treatments were blindly given to the research coordinator. Patients and investigators were blinded to group assignment. Pairs of identical syringes containing either rHuEPO or saline were prepared | High risk | Low risk: No dropouts | Low risk: no dropouts | Low risk: no dropouts |
Yoo 2011 [29] | AKI: OL(single blinded), SS | Low risk: patients were allocated by computer-generated random numbers | Unclear risk: medications were prepared and administered by a ward physician recognizing the patient’s group but not involved in the current study, the surgeon and anesthesiologist involved were blinded | Low risk: the surgeon and anesthesiologist involved in the study and patient management were blinded to the patients’ groups until the end of the study | Low risk: complete data sets from the 74 patients were analyzed without any missing data | Low risk: no dropouts | Low risk: complete data sets from the 74 patients were analyzed without any missing data |
Aydin 2012 [31] | Transplant: DB, SS | Low risk: Patients were randomized by an independent hospital pharmacist. The randomization allocation sequence was generated by a random-number table | Low risk: patients, physicians, data managers and investigators were kept blinded throughout the study | Low risk: data managers and investigators were kept blinded throughout the study | Low risk: No dropouts | Low risk: No dropouts | Low risk: No dropouts |
Coupes 2015 [30] | Transplant: DB, SS | Low risk: patients were randomly assigned by the trial pharmacy by computer | Low risk: all study participants and the study team were blinded to the trial drug | Unclear risk | Low risk: 1 patient withdrew but was included in the analysis | Low risk: lost patients reported | Low risk |
Hafer 2012 [32] | Transplant: DB, SS | Unclear risk: randomization methodology not disclosed | Low risk: vials containing ESA and placebo had identical appearance | Unclear risk | Low risk for DGF. High risk for graft loss (3 patients died 1 in ESA group and 2 in placebo group) | Low risk: lost patients reported | High risk: 2 untreated patients (not included in analysis) and 3 patients died |
Martinez 2010 [33] | Transplant: OL, MC | Unclear risk: randomization method not disclosed | High risk: comparator arm was untreated | Low risk: Blinded evaluation of end-points | Unclear risk: 1 died in ESA group | Low risk: lost patients reported | Low risk |
Sureshkumar 2012 [34] | Transplant: DB, SS | Low risk: the hospital pharmacy created a schedule using random assignments to a series of patient study numbers | Low risk: ESA and placebo were both 1 ml syringes. The medications were administered in a double-blinded manner | Unclear risk | Low risk | Low risk: no dropouts | Low risk |
Van Biesen 2005 [35] | Transplant: OL, SS | Unclear risk: randomization method not disclosed | High risk: open label | High risk | Unclear risk | High risk | Unclear risk |
Van Loo 1996 [36] | Transplant: OL, SS | Unclear risk: randomization method not disclosed | High risk: open label | High risk | Low risk: no deaths or withdrawals | Low risk: no deaths or withdrawal | Low risk: no deaths or withdrawals |
Abraham 1990 [38] | Anemia correction: DB then OL, Anemia correction: SS | Unclear risk: randomization method not disclosed | Unclear risk: unspecified | High risk | Low risk: no dropouts | Low risk: no dropouts | Low risk |
Clyne 1992 [39] | Anemia correction: OL, 2 center | Unclear risk | High risk | High risk | Low risk: for RRT | Low risk: lost patients reported | Low risk |
Kleinman 1989 [40] | Anemia correction: DB, MC | Unclear risk: randomization method not specified | Unclear risk: unspecified | High risk | Unclear risk: no dropouts reported | Unclear risk: no dropouts reported | Low risk |
Kuriyama 1997 [41] | Anemia correction: OL, SS | Unclear risk | High risk | High risk | Low risk | Low risk: lost patients reported | Low risk |
Lim 1989 [42] | Anemia correction: DB, SS | Low risk: randomization by third party | Unclear risk | Unclear risk | High risk | Low risk: lost patients reported | Low risk |
Lim 1990 [43] | Anemia correction: OL, SS | Unclear risk | High risk | High risk | Low risk: no dropouts | Low risk: no dropouts | Low risk |
Revicki 1995 [18] | Anemia correction: OL, MC | High risk | High risk | High risk | Low risk: for RRT endpoint | Low risk: lost patients reported | Unclear risk |
Cianciaruso 2008 [45] | Anemia correction: OL, MC | Low risk: randomization by computer at a separate site | Low risk: allocation was concealed from investigators, sequences were sequentially numbered in opaque envelopes opened in sequence | High risk | Low risk | Low risk: lost patient reports | High risk: 1 patient in the treatment group did not receive ESA, study terminated early |
Gouva 2004 [47] | Anemia correction: OL, MC | Low risk: computer generated sequence | Unclear risk | High risk | Low risk | Low risk: lost patients reported | High risk: study prematurely terminated |
Levin 2005 [48] | Anemia correction: OL, MC | Low risk: computer generated sequence | Low risk: allocation was in sealed sequentially numbered opaque envelopes. Designated personnel opened the next number in sequence | High risk | Low risk | Low risk: lost patient reports | High risk: only 77/85 in the high Hb group received ESA |
MacDougall 2007 [49] | Anemia correction: OL, MC | Low risk: randomized using central randomization procedures (ClinPhone) | Unclear risk | High risk | Low risk | Low risk: lost patients reported | High risk: patients in the high Hb group received ESA on day 1 but study was prematurely terminated |
Pfeffer 2009 [50] | Anemia correction: DB, MC | Low risk: DB, and patients were randomly assigned with the use of a computer-generated, permuted-block design | Unclear risk | High risk | High risk: 9 patients were excluded prior to unblinding | Low risk: lost patient reports | High risk: 93.9% of the patients in the darbepoetin alfa group were receiving the assigned treatment at 6 months” |
Ritz 2007 [51] | Anemia correction: OL, MC | Low risk: randomization was performed centrally into treatment groups by using a block-size randomization procedure stratified by country | Unclear risk | High risk | Low risk | Low risk: lost patient reports | Unclear risk: patients in group 1 were started immediately ESA but 3 patients withdrew |
Roger 2004 [52] | Anemia correction: OL, MC | Low risk: patients were randomized according to computer-generated stratification tables | Low risk: order concealment was maintained until the intervention was assigned | High risk | Low risk | Low risk: lost patient reports | Low risk |
Rossert 2006 [53] | Anemia correction: OL, MC | Low risk: patients were randomized according to computer-generated stratification schedule | Unclear risk | High risk | Low risk | Low risk: lost patient reports | High risk: study was terminated prematurely. Many subjects did not enter maintenance or withdrew |
Villar 2011 [55] | Anemia correction: OL, MC | Low risk: block-size randomization was used | Unclear risk | High risk | Low risk | Low risk: lost patients reported | Unclear risk: most patients likely received ESA but 6 patients died or withdrew |
Akizawa 2011 [44] | Anemia correction: OL, MC | Low risk: patients were assigned by a computer according to a minimization method | Unclear risk | High risk | Low risk | Low risk: lost patients reported | High risk: after 1 administration, 43 withdrew. |
Drueke 2006 [46] | Anemia correction: OL, MC | Low risk: randomization was performed centrally with the use of a dynamic randomization method | Unclear risk | High risk | Low risk | Low risk: lost patients reported | High risk: 75 in the high Hb group withdrew |
Singh 2006 [54] | Anemia correction: OL, MC | Low risk: patients were assigned by computer-generated per-muted-block randomization | Unclear risk | High risk | Low risk | Low risk: lost patients reported | High risk: study was terminated early at the second interim analysis because power to demonstrate benefit was less than 5%, and there was a high withdrawal rate |