Electronic searches of several databases including MEDLINE, EMBASE, and CINAHL (from the first record to
February 2010) were searched to identify randomized controlled trials of patients with cancer who were malnourished
or considered to be at risk of malnutrition and receiving oral nutritional support compared with routine
care. We performed a meta-analysis using a fixed effect model, or random effects models when statistically
significant heterogeneity was present, to calculate relative risk (mortality) or mean difference (weight, energy
intake, and QOL) with 95% confidence intervals (CIs). Heterogeneity was determined by using the x2 test and the
I2 statistic. All statistical tests were two-sided
Electronic searches of several databases including MEDLINE, EMBASE, and CINAHL (from the first record toFebruary 2010) were searched to identify randomized controlled trials of patients with cancer who were malnourishedor considered to be at risk of malnutrition and receiving oral nutritional support compared with routinecare. We performed a meta-analysis using a fixed effect model, or random effects models when statisticallysignificant heterogeneity was present, to calculate relative risk (mortality) or mean difference (weight, energyintake, and QOL) with 95% confidence intervals (CIs). Heterogeneity was determined by using the x2 test and theI2 statistic. All statistical tests were two-sided
การแปล กรุณารอสักครู่..
