Author Affiliations: Medicina III, Ospedale San Paolo (Drs Ferrari and Podda), and Dipartimento di Medicina, Chirurgia e Odontoiatria (Drs Ferrari and Podda), and Unità Operativa di Medicina Interna II, Dipartimento di Scienze Cliniche “L. Sacco,” Ospedale L. Sacco (Dr Casella), Università degli Studi di Milano, Milan, Italy.
In reporting the results of the systematic review “Procalcitonin Algorithms for Antibiotic Therapy Decisions: A Systematic Review of Randomized Controlled Trials and Recommendations for Clinical Algorithms,” Schuetz et al1 state that measurement of procalcitonin (PCT) levels for antibiotic decisions in patients with respiratory tract infections and sepsis does not worsen the mortality rate. We are doubtful about their conclusions. We believe that the meta-analysis is underpowered to provide evidence that PCT can affect mortality. Mortality was the primary outcome in only 3 of 14 randomized controlled trials included in the study, and in the primary setting studies (approximately 25% of the subjects) death was an extremely rare event (1 event). A second issue to be addressed is that the meta-analysis involved trials with very different clinical settings (clinical heterogeneity); thus, we wonder if it should be more appropriate to perform a random effect analysis, irrespectively of the statistical significance of the heterogeneity. Finally, regarding the methodological quality of original randomized controlled trials, we ask the authors if the global adherence to the protocols (that was low in almost all studies) can be a bias affecting the results of the meta-analysis.
Ferrari L, Casella F, Podda GM, Gruppo di Autoformazione Metodologica (GRAM) FT. Do Procalcitonin Algorithms Modify Mortality?. Arch Intern Med. 2012;172(2):197–198. doi:10.1001/archinternmed.2011.764