By Samuel Nadler, MD, PhD
Critical Care, Pulmonary Medicine, The Polyclinic Madison Center, Seattle
Dr. Nadler reports no relationships relevant to this field of study.
SYNOPSIS: Quality improvement protocols can reduce the number of chest radiographs in the ICU without compromising care.
SOURCE: Sy E, et al. Implementation of a quality improvement initiative to reduce daily chest radiographs in the intensive care unit. BMJ Qual Saf 2015 Sept 8 [Epub ahead of print].
Routine, daily chest radiographs (CXRs) are often ordered on all patients in the ICU, particularly on mechanically ventilated patients. However, several studies have called into question the utility of this practice. In 2003, Krivopal et al randomized patients to routine vs non-routine CXR and demonstrated no difference in duration of mechanical ventilation, ICU length of stay (LOS), and total LOS despite a 36% reduction in the number of radiographs in the non-routine group.1 Graat et al performed a prospective, non-randomized study requiring a clinical indication and new order for radiographs.2 The number of studies dropped by 56% without a change in length of ICU stay, ICU readmission, and hospital mortality.2 A 2010 meta-analysis included these studies, demonstrating that elimination of routine CXRs had no effect on hospital or ICU mortality.3 More recently, the American College of Radiology changed its recommendations to state that CXRs should be ordered “for clinical indications only,” even in mechanically ventilated patients.4 However, ICUs continue ordering millions of routine CXRs.
This study is a before-after design that sought to reduce the number of routine CXRs in the ICU. Between June 2014 and 2015, data were collected as the pre-intervention baseline in a 350-bed, tertiary, teaching hospital with a 15-bed mixed medical-surgical ICU. Patients with trauma, thoracic surgery, neurosurgery, or liver/lung transplants were excluded. The interventions to change practices included education of the house staff at the beginning of each month and a prompt within the computerized order entry system that required an acceptable indication to order a CXR. Overall, 1492 patients were enrolled, 738 in the pre-intervention group and 754 in the post-intervention group, with a mean age of 58 and 59 years, respectively. The two groups were well matched for APACHE II (22 vs 21), ICU mortality (20% vs 19%), hospital mortality (31% vs 26%), ICU LOS (6 vs 5 days), and days of mechanical ventilation (1.3 vs 1.4 days). Prior to the intervention, 0.73 CXRs per patient-day were ordered compared with 0.54 afterward, a 26% reduction. There was no significant increase in urgent CXRs performed (post/pre ratio 1.03; 95% confidence interval [CI], 0.91-1.15), although there was a modest increase in stat CXRs (post/pre ratio 1.18; 95% CI, 1.05-1.30). There were no changes in the number of chest CTs ordered, lines or intubations, days of mechanical ventilation, ICU LOS, ICU mortality, or hospital mortality.
COMMENTARY
This study presents a practical, real-world protocol that was effective in reducing daily CXRs in a mixed medical-surgical ICU without obvious adverse effects. In the 1-year intervention period, the protocol prevented 1086 CXRs in this 15-bed ICU, and the institution continues this protocol. The cost savings were estimated at CA$27,150 and would likely be far greater in many clinical settings. The intervention included regular education to house staff and an easily generalizable prompt within the electronic medical record for ordering CXRs. The acceptable indications for CXRs were: unexplained new cardiopulmonary symptoms or signs, suspected new pneumonia, suspected new pneumothorax, suspected new pleural effusion, insertion of endotracheal tubes/central venous catheters/chest tubes, suspected malposition of existing tubes, and an “other” category.
This study raises several questions that deserve further consideration. There was a statistically significant increase in stat CXRs, but this increase did not affect mortality or LOS. Data were presented regarding the number of CXRs per patient-day on a weekly basis and demonstrated a significant degree of variability (0.5-1 pre-intervention and 0.3-0.8 post-intervention). This suggests that despite more standardized criteria for CXRs, large practice variations persist. As residents are learning not to order routine CXRs, the need for protocols to reduce these studies may become less frequent. Additionally, the indications mandated for CXRs in this protocol include pleural effusion and pneumothorax. Both may be better evaluated by bedside ultrasonography, which more recent trainees are using with greater frequency.
Overall, this study shows the efficacy of a simple, generalizable quality improvement project to reduce unnecessary routine CXRs in the ICU that realizes significant cost savings without compromising clinical care.
REFERENCES
-
Krivopal M, et al. Utility of daily routine portable chest radiographs in mechanically ventilated patients in the medical ICU. Chest 2003;123:1607-1614.
-
Graat M, et al. Elimination of daily routine chest radiographs in a mixed medical-surgical intensive care unit. Intensive Care Med 2007;33:639-644.
-
Oba Y, Zaza T. Abandoning daily routine chest radiography in the intensive care unit: Meta-analysis. Radiology 2010;255:386-395.
-
Amorosa JK, et al. ACR appropriateness criteria routine chest radiographs in intensive care unit patients. J Am Coll Radiol 2013;10:170-174.
