Evaluations of Morphine and Fentanyl for Mechanically Ventilated Patients With Respiratory Disorders in Intensive Care: A Systematic Review of Methodological Trends and Reporting Quality.

BACKGROUND: Mechanically ventilated patients with respiratory disorders may require sedatives, such as opioids.

OBJECTIVES: To define methodological trends, gaps, and the reporting quality of the comparative clinical and economic evaluations of fentanyl and morphine in ventilated patients in the intensive care unit.

METHODS: We conducted a literature review of the MEDLINE, Embase, OVID, ScienceDirect, Springer Link, and EconLit databases, comparing studies in the management of ventilated patients with respiratory disorders in the intensive care unit using either fentanyl or morphine, or both. We assessed the methodological aspects of the literature characteristics and trends of, for example, modeling, data sources, cost calculation, and data analysis, appraising the quality of reporting via the CONsolidated Standards Of Reporting Trials, STrengthening the Reporting of OBservational studies in Epidemiology, and the Consolidated Health Economic Evaluation Reporting Standards checklists.

RESULTS: Among 1327 articles, 33 (comprising 22 in adults, 8 in neonates, and 3 in pediatrics) met the inclusion criteria. No head-to-head morphine versus fentanyl evaluations explicitly confined to subjects with respiratory conditions were undertaken. Studies relied on various scales to measure the sedation level as a primary study outcome, limiting the comparability of study conclusions. Seven articles of adults were identified to be economic studies from the hospital perspective. On the basis of different endpoints, the same sedation regimen performed differently in various studies. None of the randomized controlled trials, observational cohorts, or pharmacoeconomics studies met most of the assessed reporting quality criteria.

CONCLUSIONS: Our review identified poor reporting quality and high heterogeneity of methods used, potentially limiting the degree to which studies could be interpreted, decisions could be influenced, and findings could be generalized.