Routine vs. on-demand CXR in ICU patients

From Wiki Journal Club
Jump to navigation Jump to search
Hejblum G, et al. "Comparison of routine and on-demand prescription of chest radiographs in mechanically ventilated adults: a multicentre, cluster-randomised, two-period crossover study". The Lancet. 2009. 374(9702):1687-1693.
PubMedFull textPDF

Clinical Question

For mechanically ventilated patients, are daily chest radiographs necessary or should they be ordered only when a patient's clinical status changes?

Bottom Line

For mechanically ventilated adult ICU patients, using an on-demand strategy for CXRs reduces the number of CXRs overall by 32% compared to routine daily CXRs, without having any deleterious effect on patient safety or clinical outcome measures in the ICU.

Major Points

Chest radiographs are common in critically ill, intubated, mechanically ventilated patients. Professional society recommendations had supported the practice of routine daily radiographs, though some questioned the utility in detecting changes like displaced endotracheal tubes compared to bedside examination. Radiographs have significant consequences of cost, radiation, and results of unexpected findings which may or may not have clinical significance.

Earlier single-center studies had favoured on-demand strategies, but only two, with a total of 259 patients, focused on mechanically ventilated patients (Graat et al., Crit Care 2006; Clec'h et al., Intensive Care Med 2008). This study was much larger, including nearly 850 patients across 21 ICUs, using a cluster-randomized crossover design to control for variation in practice between ICUs. Patients in an ICU using the routine strategy received daily CXRs during morning rounds, while those in the on-demand strategy received them only when clinical examination suggested the need for one.

This study showed a reduction of daily CXRs by an average of 32% (from 1.09 to 0.75 CXRs per day), consistent with findings in the smaller studies. The number of images leading to diagnostic or therapeutic interventions were similar between both strategies, indicating that few if any important abnormalities were missed. Further, key ICU clinical outcomes of days on mechanical ventilation, ICU length of stay, and ICU mortality were similar between the groups, suggesting that the on-demand strategy is both safe and effective.


  • Multicenter, cluster-randomized, crossover trial
  • N=849
    • On-Demand Strategy (n=425)
    • Routine Strategy (n=424)
  • Setting: 21 closed, adult ICUs (13 medical, 2 surgical, 6 mixed ICUs) of a research network in Paris, France.
  • Enrollment: 2006-2007, median duration of study at each ICU was 131 days.
  • Blindings: Open-label
  • Follow-up: Until discharge from ICU or 30 days of mechanical ventilation.
  • Analysis: Intention to treat.
  • Primary outcome: Mean number of chest radiographs per patient-day of mechanical ventilation.
  • Secondary outcomes: Days of mechanical ventilation, length of stay in the ICU, and mortality of patients during ICU stay.


Inclusion Criteria

  • Adult patients in a study ICU.
  • Receiving mechanical ventilation during morning rounds.

Exclusion Criteria

  • Mechanically ventilated for less than 2 days.

Baseline Characteristics

From the Routine strategy group:

  • Demographics: Age 61 years, male 61%
  • SAPS II score (predicted hospital mortality): 52 (51%)
  • Reason for starting mechanical ventilation:
    • Thoracic diseases 40%
      • ARDS or ALI: 32%
      • Pneumonia: 25%
      • Acute on chronic respiratory failure: 17%
    • Extrathoracic diseases 49%
      • Shock: 50%
      • Coma (not intoxication): 31%
    • Postoperative care: 10%


  • ICUs randomized to one of two strategies:
    • Routine Strategy: ventilated patients received daily chest radiographs.
    • On-Demand Strategy: ventilated patients receive on-demand chest radiographs.
  • Number of chest radiographs, a reason for doing each radiograph, new findings leading to diagnostic procedures or therapeutic interventions, days of mechanical ventilation, length of stay in the ICU, and mortality of patients were recorded for each patient.
  • First Treatment Period ended after approximately 20 patients were enrolled at each ICU and the last mechanically ventilated subject was extubated, or up to 30 days after enrolment
  • After first treatment period ended, each ICU observed a 1-week washout period in which the ICU providers were free to choose any strategy to order chest radiographs
  • After the washout period, each ICU began the other strategy for the Second Treatment period with an enrollment goal of 20 patients.


Comparisons are Routine Strategy vs. On-Demand Strategy.

Primary Outcomes

Chest radiographs per patient-day of mechanical ventilation, mean
Overall: 1.09 vs. 0.75 (p<0.0001)
At morning rounds: 0.90 vs. 0.54 (p<0.0001)
Unscheduled: 0.18 vs. 0.20 (p=0.24)

Secondary Outcomes

Days of mechanical ventilation, mean[SD]
9.8 [8.2] vs. 9.9 [8.6] (p=0.90)
Length of ICU stay in days, mean[SD]
14.0 [11.6] vs. 13.2 [11.0] (p=0.28)
ICU Mortality, %
31% vs. 32% (p=0.79)

Subgroup Analysis

  • On-demand strategy associated with fewer chest radiographs than routine strategy in every ICU.
  • Both strategies had a similar number of diagnostic procedures or therapeutic interventions driven by new findings on chest radiograph, both overall (824 vs. 834 procedures or interventions (p=0.77)), and the types of interventions were similar between groups except for an increased number of changes in vent settings in the on-demand strategy.


  • All ICUs were closed with patients managed by dedicated intensivists, limiting its generalizability to ICUs with other staffing models.
  • Workflow and efficiency differences between the two strategies for clinicians and technicians were not assessed.
  • No data is provided to guide how long patients can be safely managed without a CXR if no clinical indications are present.
  • Actions based on negative CXR findings were not documented.
  • No subgroup analysis by patient type was done.


  • No conflicts of interest declared.
  • Study funded by a grant for research on standard care procedures from Assistance Publique-Hopitaux de Paris.

Further Reading

A 2012 systematic review and meta-analysis pooled 3 RCTs and 6 before-after observational studies. The authors found no evidence of harm, but noted limitations in assessing any consequences of missed findings on routine CXRs: Ganapathy A, Adhikari N, Spiegelman J, Scales DC. "Routine Chest x-rays in intensive care units: a systematic review and meta-analysis." Crit Care. 2012; 16(2): R68.