Standardized Rehabilitation and Hospital Length of Stay Among Patients With Acute Respiratory Failure

Morris. JAMA 2016; 315:2694-2702. doi:10.1001/jama.2016.7201

Clinical Question

  • In patients with acute respiratory failure, does early delivery of a standardised ICU and hospital rehabilitation programme improve physical function and reduce hospital length of stay?


  • Single centre, assessor blind randomised study
  • Computer generated variable size randomisation using blocks of 2, 4, 6 or 8
  • Intention to treat analysis
  • Physical function assessed at ICU discharge, hospital discharge, and at 2, 4, and 6 months post enrolment
  • Quality of life measured at hospital discharge and at 2, 4, and 6 months post enrolment
  • Initial plan was to recruit 326 patients to give 80% power to detect a 30% decrease in median length of stay at a 5% 2 sided level of significance, presuming 20% in hospital mortality, an exponential LOS distribution, and a loss of 5% of patients to follow up


  • Wake Forest Baptist Medical Centre, Winston, Salem, North Carolina
  • Oct 2009 to Nov 2014


  • Inclusion:
    • Admission to medical ICU
    • >18 years
    • Mechanical ventilation via ET tube or NIV via Mask
    •  P/F ratio <300 mmHg
  • Exclusion:
    • Inability to walk without assistance prior to acute illness (however canes/walkers not an exclusion)
    • Cognitive impairment prior to ICU admission described by surrogate
    • BMI >50
    • Neuromuscular disease impairing weaning from mechanical ventilation
    • Acute hip fracture
    • Unstable C-spine or pathological fracture
    • Mechanically ventilated >80 hours
    • Current hospitalisation (including transferring hospital) >7 days
    • Do not intubate order on admission
    • Considered to be moribund by primary attending
    • Enrolled in another research study
  • 4084 screened, 618 eligible, 300 randomised- recruitment stopped early due to lower than anticipated drop out rate


  • Standardised Rehabilitation Therapy (SRT)
    • 3 exercises – passive range of motion, physical therapy (bed mobility, transfer and balance training) and progressive resistance exercises
    • Done 3x/day 7 days a week
    • If patient unconscious only passive ROM done


  • Physical therapy ordered by clinical team as felt necessary. Only Monday to Friday


  • Significant differences seen in amounts of physical therapy delivered:
    • Intervention group: 87% passive ROM/54.6% physical therapy/35.7% resistance exercises
    • Control group: 11.7% received physical therapy
  • Primary outcome: No difference in median length of stay (both 10 days).
  • Secondary outcome: No significant difference in any secondary outcomes at ICU discharge, 2 months and 4 months
    • at 6 months, significant differences seen in :
      • Short Performance Physical Battery (9 vs 8, P< 0.004)
      • SF-36 PFS (31.1 vs 30.8; P< 0.04)
      • Functional performance Index (2.2 vs 2.0, P<0.02)

Authors’ Conclusions

  • Among patients hospitalised with acute respiratory failure, Standardised rehabilitation therapy compared with usual care did not decrease hospital length of stay.


  • An important question to ask. Weakness post ICU is increasingly recognised as a priority for patients
  • Registered with
  • Assessors blinded to intervention
  • Clear separation of between intervention and control group regarding delivery of therapy


  • Single centre study- this limits its external validity
  • Very narrow cohort of patients based on inclusion/exclusion criteria. While improving the internal validity of the trial, this again places limits on its external validity.
  • There was significant variability in patients in the intervention groups receiving treatment – 13% did not receive passive ROM, 45.4% did not receive physical therapy, and 74.3% did not receive resistance exercises. While the overall “dose” of therapy was higher in the intervention group, the “effective dose” may not have been that much more than the control.
  • Other aspects of patient care were not standardised – e.g. sedation, nutrition.
  • Control group physical therapy activities not specified or standardised.
  • Length of stay as an outcome measure is prone to multiple biases, and may not have detected increases in strength and balance that may have occurred.
  • Significant drop out rate over the 6 months post hospitalisation (44%)- this limits the significance of differences seen in secondary outcomes.

The Bottom Line

  • In this single centre study of patients with acute respiratory failure, a standardised hospital rehabilitation programme did not reduce hospital length of stay.
  • However, this is a study with very limited external validity due to a narrow cohort in a single centre.
  • Early mobilisation may have other benefits. I will follow my unit’s protocols regarding physiotherapy while awaiting a large multi-centre RCT to provide more information on the best “dose” and timing of physical therapy in ICU patients.

External Links


Summary author: Segun Olusanya
Summary date: 17/08/2016
Peer-review editor: Duncan Chambler

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.