Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model

Top Stroke Rehabil. 2010 Jul-Aug;17(4):271-81. doi: 10.1310/tsr1704-271.


Objective: There are barriers to providing early physical medicine and rehabilitation (PM&R) in the intensive care unit (ICU). We present a specific model for undertaking quality improvement (QI) projects and a case study focused on QI for early PM&R in the ICU.

Methods: The QI project was undertaken using a 4-step model: (1) summarizing the evidence, (2) identifying barriers, (3) establishing performance measures, and (4) ensuring patients receive the intervention. To evaluate the application and outcomes of this model, we present data collected during a 4-month QI period versus an immediately preceding 3-month control period.

Results: Deep sedation was a major barrier to early PM&R that was addressed in the QI project. Compared to the control period, there was a decrease in medical ICU (MICU) days with any benzodiazepine use (73% vs 96% of days, P = .03) and narcotic use (77% vs 96%, P = .05) and improved delirium status (MICU days without delirium, 53% vs 21%, P = .003). In addition, more QI patients had physical therapy consultations (93% vs 59%, P = .004) and greater number of rehabilitation treatments with higher functional mobility (treatments involving sitting or greater mobility, 78% vs 56%, P = .03). Hospital data for the QI period demonstrated a decrease in average length of stay in the MICU (4.9 vs 7.0 days, P = .02) and hospital (14.1 vs 17.2, P = .03) compared to the prior year.

Conclusion: A structured QI model can be applied to implementation of early PM&R in the ICU resulting in markedly improved delirium status, delivery of PM&R, functional mobility, and length of stay.

MeSH terms

  • Delirium / drug therapy
  • Humans
  • Hypnotics and Sedatives / administration & dosage
  • Hypnotics and Sedatives / adverse effects
  • Intensive Care Units* / standards
  • Intensive Care Units* / statistics & numerical data
  • Models, Statistical*
  • Physical Therapy Modalities
  • Quality Improvement*
  • Stroke / drug therapy
  • Stroke Rehabilitation*
  • Substance-Related Disorders
  • Time Factors


  • Hypnotics and Sedatives