Methodology in meta–analysis: a study from Critical Care meta–analytic practice

Risk related therapy in meta-analyses of critical care interventions: Bayesian meta-regression analysis

PURPOSE

The relationship between treatment efficacy and patient risk is explored in a series of meta-analyses from the critical care domain, focusing on mortality outcome.

METHODS

Systematic reviews of randomized controlled trials were identified by electronic search over the period 2002 to July 2018. A Bayesian meta-regression model was employed, using the risk difference metric to estimate the relationship between mortality difference and control arm risk, and estimate the mortality difference with and without adjusting for control arm risk.

RESULTS

Of 780 initially identified published systematic reviews, 113 had appropriate mortality data comprising 123 analysable groups. The 123 meta-analyses were pharmaceutical therapeutic (59.3%), non-pharmaceutical therapeutic (24.4%) and nutritional (16.3%), with a 25% overall average control arm mortality. In 25/123 (20%) analyses, meta-regression indicated significant baseline risk (Bayesian 95% credible intervals excluding zero). In all analyses, the relationship between risk-difference and control arm risk was negative indicating a positive treatment effect with increasing control arm risk. Adjusted estimates identified six studies with significant positive treatment effects, not evident until after adjustment for control arm risk.

CONCLUSION

Underlying risk-related therapy is apparent in meta-analyses of the critically-ill and identification is of importance to both the conduct and interpretation of these meta-analyses.

Heparin therapy reduces 28-day mortality in adult severe sepsis patients: a systematic review and meta-analysis

INTRODUCTION

There are approximately 19 million new cases of sepsis worldwide each year. Among them, more than one quarter of patients die. We aimed to assess the effects of heparin on short-term mortality in adult patients with sepsis and severe sepsis.

METHODS

We searched electronic databases (Medline, Embase, and Cochrane Library databases; the Cochrane Controlled Trials Register) and conference proceedings (Web of Knowledge (Conference Proceedings Citation Index - Science, Conference Proceedings Citation Index - Social Sciences & Humanities)) from inception to July 2014, expert contacts and relevant websites. Controlled trials of heparin versus placebo in sepsis or severe sepsis were identified. In total two reviewers independently assessed eligibility, and four authors independently extracted data; consensus was reached by conference. We used the chi-square test and I2 to assess statistical heterogeneity (P <0.05). The primary analysis was based on the fixed-effect model to produce pooled odds ratios with 95% confidence intervals.

RESULTS

A total of nine publications were included in the meta-analysis. Heparin decreased 28-day mortality (n = 3,482, OR = 0.656, 95% CI = 0.562 to 0.765, P <0.0001). According to the meta-analysis of 28-day mortality, heterogeneity was not found among the eight randomized clinical trials (RCTs) (I2 = 0.0%). Heparin had no effect on bleeding events in sepsis (seven RCTs, n = 2,726; OR = 1.063; 95% CI = 0.834 to 1.355; P = 0.623; and I2 = 20.9%). Subgroup analysis demonstrated that the sample size may be a source of heterogeneity, but experimental design was not.

CONCLUSIONS

Heparin may reduce 28-day mortality in patients with severe sepsis, at the same time, there was no increase in the risk of bleeding in the heparin group. We recommend the use of heparin for sepsis and severe sepsis.

The efficacy of loop diuretics in acute renal failure: assessment using Bayesian evidence synthesis techniques

OBJECTIVE

To quantify the therapeutic efficacy of loop diuretics in acute renal failure using Bayesian evidence synthesis, because despite widespread use, the role of diuretics is controversial.

DATA SOURCE

Randomized controlled trials or nonrandomized studies, 1966 to January 2007, identified from MEDLINE and EMBASE databases and manual bibliographic search.

STUDY SELECTION

Studies with assessable predefined end points, exclusive of those pertaining to acute renal failure prophylaxis or chronic renal failure.

DATA EXTRACTION

Data extraction was performed jointly by the first two authors; independent study assessment was via standard checklist, unblinded.

DATA SYNTHESIS

The primary outcome was mortality; secondary outcomes were time to renal function normalization and total number of dialyses. Bayesian hierarchical random effects estimates of treatment effects were determined as risk ratio for mortality, incidence rate ratio for dialysis number, and mean difference for continuous measures. Bayesian outcome probabilities were calculated as probability (P) that risk ratio or incidence rate ratio of loop diuretics >1 and probability that mean difference >0. Five randomized controlled trials and eight nonrandomized studies were identified. Loop diuretics were not associated with decreased mortality in either randomized controlled trials or nonrandomized studies: overall risk ratio 1.10; 95% credible interval 0.85, 1.42; P(risk ratio >1) > 83.8%. The oliguric period was decreased by loop diuretics: overall mean difference -7.70 days; 95% credible interval -12.51, -2.08; P (mean difference >0) = 0.7%. Although the dialysis rate credible interval, loop diuretics vs. control, spanned unity (incidence rate ratio 0.71; 95% credible interval 0.47, 1.06), the probability that the incidence rate ratio exceeded unity indicated a substantial benefit: P (incidence rate ratio >1 = 4.1%. Uremic duration was not substantially different, loop diuretics vs. control: overall mean difference -1.54 days; 95% credible interval -5.62, 2.46; P [mean difference >0] = 17.8%).

CONCLUSIONS

Loop diuretics were not associated with improved survival benefit in acute renal failure, despite reduction in oliguric period and high probability of a significant reduction in dialysis numbers. Further studies to clarify this dichotomy appear mandated.