Sant'Anna 2014 (INTRACELL).

Study characteristics
Methods	Study design: RCT Parallel groups (2) Open label Single centre Country: Brazil Duration: 12 months
Participants	30 participants randomized in a 2:1 ratio (20 SCT and 10 control) Period (recruitment): recruitment started in 2005 Inclusion criteria: people with HF; LVEF < 35% by echocardiogram; NYHA functional class III or IV, despite full medical treatment; aged 20–65 years; diagnosis of NIDCM for ≥ 12 months before enrolment. Baseline characteristics Sex (male): control 50% vs SCT 65% Age (years), mean: control 51.6 (SD 7.79) vs SCT 48.3 (SD 8.71) LVEF (%), mean: control 24.76 (SD 4.64) vs SCT 21.75 (SD 41.19) LVEDD (mm): control 69.38 (SD 7.81) vs SCT 69.80 (SD 4.41)
Interventions	Intervention group: STC Details of SCT Cell origin: autologous Cell collection location: BMA (volume of about 80 mL from the anterior iliac crest) Type of cells infused: mononuclear cells Mean volume/number administered: 1.06 × 108 mononuclear cells per participant Cell mobilization: apparently not Delivery route: intramyocardial (cells infused through a left mini‐thoracotomy, consisting of an approximately 5 cm incision in the antero­lateral portion of the fifth left intercostal space to expose the pericardium. A T‐shaped pericardial incision was made to access the free wall of the LV. Coronary arteries were identified and the cell suspension was directly injected, using a 21F butterfly needle introduced about 5 mm intramyocardially and connected to an extension managed by the surgical assistant. 20 × 0.25 mL injections were given in the myocardium and in the anterior, lateral, posterior, and apical faces of the left ventricular free wall) Number of cell infusions: apparently single Control group Not specified Study included in Comparison 1 (STC vs control)
Outcomes	Outcomes included in review Adverse events (including procedural safety) All‐cause mortality Change from baseline in NYHA functional class Change from baseline in 6MWT Change from baseline in MLHFQ Change from baseline in LVEF (primary outcome) Other outcomes reported Change from baseline in LVESD Change from baseline in LVEDD
Notes	Registered in ClinicalTrials.gov: NCT00743639 Funding: financial support from Brazilian government agencies National Council for Scientific and Technological Development, e Committee for Postgraduate Courses in Higher Education, Ministry of Health, and FAPERGS
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants randomized using computer software for simple randomization.
Allocation concealment (selection bias)	Unclear risk	No details provided.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Open‐label trial.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Open‐label trial.
Incomplete outcome data (attrition bias) All outcomes	High risk	Quote: "In the first three months of follow‐up, 25% (5 out of 20) of the patients from the BMMC group had either died or been withdrawn from the study, resulting in a decrease in the number of treated cases available for late follow‐up. Since those patients had a lower mean ejection fraction than the whole group (18.26% vs 21.75%), we excluded them from comparative analysis, in order to avoid overestimation of treatment effect. In other words, outcome analysis was performed in as‐treated basis, not as intention‐to‐treat." Contrary to what the authors commented, we consider that the exclusion of these participants from the SCT group poses a risk of bias in favour of treatment.
Selective reporting (reporting bias)	High risk	ClinicalTrials.gov register only specifies the primary endpoint (increase of the ejection function of the LV) but not the secondary outcomes.