Routine chest x-rays in intensive care units: a systematic review and meta-analysis

Table 1 Characteristics and methodological quality of randomized and quasi-randomized controlled trials of restrictive versus routine CXR strategies.

Study	ICU Population	Patient selection criteria	Location	Total patients (n)	CXRs/patients, restrictive group (n/n)	CXRs/patients, routine group (n/n)	Outcomes used in meta-analysis	Method of allocation	Allocation concealment	Blinded outcomes assessment^a	Zero losses to follow up
Krivopal et al., 2003 [16]	Medical, Adult	Ventilated for 48 to 72 hours Exclusion criteria: reintubated, transferred from other centers	USA	94	226/51	293/43	ICU and hospital mortality ICU and hospital length of stay Duration of mechanical ventilation	Quasi-randomization based on last digit of medical record number	No	Not specified	Yes
Clec'h et al., 2007^b [15]	Medical-surgical, Adult	Ventilated for ≥ 48 hours Exclusion criteria: reintubated, tracheostomy, or withdrawal of life support	France	165	94/81	885/84	ICU and hospital mortality ICU and hospital length of stay	Computer-generated random number table	Not reported	Not specified	Yes
Hejblum et al., 2009 [13]	Medical-surgical, Adult	Ventilated for ≥ 2 days	France	611, 849^c	3,148/306, 425^c	4,607/305, 424^c	ICU mortality ICU length of stay Duration of mechanical ventilation	Computer-generated	Not reported ('open-label with respect to allocation concealment')	Not specified	Yes

^aRadiologists and clinicians were not blinded to the group when interpreting films; ^bIn this trial all patients had daily CXRs, but results in the restrictive group were concealed from clinicians, who were free to order CXRs as needed. ^cThe first sample size is adjusted for the design effect due to clustering [14], assuming intracluster correlation 0.01 (design effect 1.39); the second is the unadjusted sample size as reported in the study. CXR, chest X-ray; n, number.

ISSN: 1364-8535