This review concluded that compared to alternative treatments, cardiac resynchronisation therapy combined with implantable cardioverter defibrillators reduced all-cause mortality in people with chronic heart failure. Questions about the quality of included data and the appropriateness of pooling data from studies that used very different comparators mean that the authors' conclusions may be unreliable.

To determine whether cardiac resynchronisation therapy with implantable cardioverter defibrillators (CRT-D) reduced mortality in people with advanced chronic heart failure.

MEDLINE, EMBASE and Cochrane Central Register of Controlled Trials (CENTRAL) were searched from 1994 to October 2008. Search terms were reported. No language or publication restrictions were applied. ClinicalTrials.gov, the United States Food and Drug Administration website and bibliographies of identified papers were checked.

Randomised controlled trials (RCTs) that lasted at least one month and compared CRT-D to no CRT-D in people with chronic heart failure due to left ventricular dysfunction (ejection fraction <35%) and reported on all cause mortality were eligible for inclusion.

In the included studies, mean age ranged from 63 to 69 years and between 67% and 89% of participants were men. Between 43% and 70% of participants had ischaemic heart disease. Left ventricular ejection fraction ranged from 20% to 25%. Duration of QRS ranged from 16 to 168 milliseconds. Concomitant treatments in some participants included angiotensin converting enzyme inhibitors/angiotensin receptor blockers and beta blockers. Comparator groups received either cardiac resynchronisation therapy (CRT), implantable cardioverter defibrillator (ICD) or medical therapy. Follow-up ranged from 4.7 to 69 months.

The authors did not state how studies were selected for the review.

Study quality was assessed using the Jadad scale of randomisation, blinding and withdrawals to a maximum score of 5.

Two authors independently assessed study quality and a third author checked information. Disagreements were resolved by consensus.

Data were extracted in order to calculate odds ratio (OR) and 95% confidence intervals (CI).

Two authors independently extracted data and a third author checked these data. Disagreements were resolved by consensus.

Pooled odds ratios and 95% CI were calculated using a random-effects model. Heterogeneity was assessed with Χ² and I². Sensitivity analyses investigated the effect of individual studies. Subgroup analyses were based on type of control (ICD, CRT and medical therapy)

Seven RCTs (4,536 participants) were included. Study size ranged from 126 to 1,520 participants. Thee studies (CRT-D to ICDs) scored 5 for quality, three scored 3 and one scored 2 on the Jadad scale.

Compared to controls, CRT-D therapy was associated with a statistically significant reduction in all-cause mortality (OR 0.55, 95% CI 0.40 to 0.76, I²=60%; seven studies). Sensitivity analysis suggested that one study was responsible for heterogeneity; results that excluded this study were not reported.

In subgroup analyses, CRT-D was associated with a statistically significant reduction in all-cause mortality when compared to CRT (OR 0.48, 95% CI 0.31 to 0.75, I²=79%; four studies) and medical therapy (OR 0.64, 95% CI 0.46 to 0.90; one study). There was no statistically significant difference in mortality between CRT-D and ICD therapy (I²=0%, three trials).

CRT-D was effective in reducing all-cause mortality in people with chronic heart failure.

The aims of the review were clearly stated in terms of inclusion criteria. Several relevant sources were searched. No language or publication restrictions were applied, which was likely to have reduced the effects of language or publication bias. Methods of data extraction and quality assessment aimed to reduce reviewer error and bias; methods used for study selection were unclear. Quality was assessed using a scoring system, but no details of the assessment were presented and this made it impossible to identify which aspects of quality individual studies failed to score on. Statistical heterogeneity between studies was assessed. Details of the investigation of heterogeneity were unclear and not all relevant results were presented. Pooling data from studies that used very different control treatments may not have been appropriate; relevant subgroup analyses were presented. Data came from studies of variable quality. In particular three of the four studies that compared CRT-D with CRT appear to be observational studies and not RCTs.

Questions about the quality of included data and the appropriateness of pooling data from studies that used very different comparators mean that the authors' conclusions may be unreliable.

Practice: The authors stated that CRT-D should be considered for people who met enrolment criteria for the included trials, but the decision to treat should take into account patient preferences, quality of life and economic considerations.

Research: The authors stated that large RCTs were needed to compare treatment with CRT-D to ICDs in people with chronic heart failure. Longer-term follow-up was needed to assess the effects of CRT-D.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.