Twenty-nine trials involving 4,795 episodes of febrile neutropenia (range per trial: 27 to 958) were included.
The quality scores for the studies included in the review ranged from 0.16 to 0.7 (mean 0.42); the individual scores were listed in the review.
The pooled estimate of effect was marginally in favour of the monotherapy regimens: the OR was 0.88 (95% confidence interval, CI: 0.78, 0.99) when using a fixed-effect model or 0.87 (95% CI: 0.75, 1.01) when using a random-effects model. The test for heterogeneity was not statistically significant (p=0.20).
Subgroup analyses found similar findings for beta-lactam monotherapy versus beta-lactam plus aminoglycoside, cephalosporin monotherapy, and carbapenem monotherapy; i.e. the pooled estimate slightly favoured monotherapy but did not always quite attain statistical significance, particularly if the random-effects model was used.
When only studies of patients with severe neutropenia were included in the meta-analysis, the overall ORs were 0.91 (95% CI: 0.77, 1.09) and 0.92 (95% CI: 0.79, 1.08) when using the random-effects and fixed-effect models, respectively, and neither was statistically significant. When only studies of patients aged over 14 years were pooled, the random-effects and fixed-effect ORs were 0.82 (95% CI: 0.67, 0.99) and 0.82 (95% CI: 0.69, 0.97), respectively.
When only treatment of bacteraemic episodes were pooled, the random-effects and fixed-effect ORs were 0.72 (95% CI: 0.54, 0.95) and 0.70 (95% CI: 0.54, 0.92), respectively.
A sensitivity analysis that included only the higher quality studies reported random-effects and fixed-effect ORs of 0.94 (95% CI: 0.81, 1.09) and 0.94 (95%CI: 0.82, 1.07), respectively, i.e. the treatment difference was not statistically significant. The exclusion of various studies due to the choice of comparators, methodology, size and so on, had small effects on the findings. A cumulative analysis using a fixed-effect model also found a very slight treatment benefit of monotherapy, which achieved borderline statistical significance.