Preoperative physiotherapy for the prevention of respiratory complications after upper abdominal surgery

Boden. BMJ 2018; 360:j5916 doi:

Clinical Question

  • In patients undergoing upper abdominal surgery (UAS), does the addition of a single pre-op physiotherapy session compared to standard care reduce the rate of post-operative pulmonary complications (PPC)?


  • PPC is the most common complication following UAS with an incidence of between 10 and 50%
  • PPC is strongly associated with increased mortality, morbidity, and healthcare costs


  • Multicentre, randomised controlled trial
  • Ethics approval in accordance with Helsinki Declaration
  • Trial registered with Australian New Zealand Clinical Trials Registry: ACTRN126130006644741
  • Consecutive participants
  • Randomised using sequentially numbered sealed opaque envelopes
  • Computer-generated allocation sequence
  • Healthcare professionals and trial assessors in the post-op phase blinded to allocation
  • PPC diagnosed using standardised validated tool – Melbourne group score
  • Power calculation based on previous studies of absolute risk reduction and institutions own unpublished data in the area
    • Powered at 80%
    • Baseline PPC incidence of 20%, 10% absolute risk reduction
    • Significance set at 0.05
    • 441 patients required, which includes 11% inflation for drop-outs, non-compliance and uncertainty of baseline
  • Intention-to-treat analysis
  • Authors conducted various exploratory post-hoc analyses


  • 2 hospitals in Australia, 1 in New Zealand
  • June 2013 – August 2015


  • Inclusion:
    • Over 18 years of age
    • Elective surgery
    • Surgery requiring abdominal incision >5cm in length, that will be above, or extending above, the umbilicus
    • Minimum single overnight hospital stay
  • Exclusion:
    • Unable to comprehend verbal instructions in English
    • Unable to participate in single pre-admission session with physiotherapist (PT)
    • Emergency surgery
    • Current hospital admission for a different indication
    • Organ transplant
    • Open abdominal hernia repairs
    • Unable to stand and ambulate for at least 1 minute
  • 504 pts listed for UAS; 441 met inclusion criteria and randomised; 222 intervention and 219 control
  • Baseline characteristics: intervention vs control (selected)
    • Age: 63.4 vs 67.5 years
    • ASA physical health status 1 or 2: 69% vs 58%
    • Comorbidities – respiratory: 19% vs 26%
    • Current smoker – 21% vs 27%
    • Surgical category – colorectal: 50% vs 47%
    • Surgical category – hepatobiliary/upper GI: 22% vs 28%
    • Surgical category – renal/urology: 28% vs 24%
    • Post-op location – ICU: 43% vs 46%


  • Standard Care + Extended Physiotherapy Session
    • Seen in multidisciplinary pre-operative outpatient clinic
      • Standard care included assessment by “nurse, anaesthetist, doctor ± stomal therapist”
      • Basic physiotherapy session included assessment and provision of information booklet
        • Booklet instructed:
          • 2x 10 deep breaths
          • Followed by 3x coughs
          • To be performed hourly
      • Extended physiotherapy session immediately followed with a single 30 minute education and breathing exercise coaching session
        • Educated about:
          • Individual risk of PPCs
          • Impact of anaesthesia and surgery on lung volumes and sputum clearance
          • Prevention strategies
        • Training provided for:
          • Breathing exercises as detailed in booklet
        • Also included memory cues for patients to remember hourly exercises in the post-op phase


  • Standard Care Only
    • Seen in multidisciplinary pre-operative outpatient clinic
      • Standard care included assessment by “nurse, anaesthetist, doctor ± stomal therapist”
      • Basic physiotherapy session included assessment and provision of information booklet
        • Booklet instructed:
          • 2x 10 deep breaths
          • Followed by 3x coughs
          • To be performed hourly

Management common to both groups

  • Post-op physiotherapy care in both groups standardised
  • If PPC identified, patient flagged and subsequent physiotherapy delivered at attending physiotherapists discretion
  • All others aspects of peri-operative clinical care were left to discretion of medical and nursing teams


  • Primary outcome: the number of patients diagnosed with PPC was significantly lower in the intervention group
    • Extended Physiotherapy Session: 12%
    • Standard Care: 27%
    • Absolute Risk Reduction (ARR): 15% (95% CI 7 to 22%, p<0.001)
    • Unadjusted hazard ratio (HR): 0.43 (95% CI 0.27 to 0.67).
    • Incidence remains halved (adjusted HR 0.48; 95% CI 0.30-0.75) when adjusted for baseline imbalances in 3 prespecified covariates – age, respiratory comorbidities and surgical procedure
    • NNT 7
    • Fragility Index 14
  • Secondary outcome:
    • Incidence of hospital acquired pneumonia halved in intervention group
      • 8% vs 20% (OR 0.45, CI 0.26 – 0.78, p<0.001)
    • No difference in –
      • Mortality
      • Hospital length of stay
      • Readiness for hospital discharge
      • Unplanned readmissions
      • ICU length of stay
      • Hospital readmissions at 6 weeks
      • Ambulation attainment measures
    • No adverse events attributable to intervention

Authors’ Conclusions

  • In a general population of patients listed for elective upper abdominal surgery, a 30 minute preoperative physiotherapy session provided within existing hospital multidisciplinary preadmission clinics halves the incidence of PPCs and specifically hospital acquired pneumonia.


  • Study protocol and analysis plan pre-published
  • Comprehensive intervention arm
  • Variability in physiotherapy experience reduced by AV education video to standardised delivery of intervention
  • High screening and recruitment rate
  • High level of masking
  • Rate of PPC in study as per study power


  • Inability to attend PT session as an exclusion criteria
  • Although described as an international, multi-centre study, the majority of patients were recruited from a single-centre in Australia
    • Significant between site differences, with the NZ site showing no benefit of the intervention, albeit with small numbers
  • Baseline difference between groups
    • Age, respiratory co-morbidities and surgical type corrected for in analysis and authors report similar conclusion
  • Compliance of patients in intervention arm unmeasured (authors’ justified this as assessment could have Hawthorne effect)
  • The primary outcome of the study was based on a surrogate outcome measure, scoring four or more out of eight on the Melbourne Group Score (MGS)
    • A patient-centred outcome, such as hospital length of stay, may have been more valid and relevant
  • Pre-op clinic extensively staffed which may not be standard of care in other healthcare settings
  • Cost and health economics analysis awaited

The Bottom Line

  • Despite the lower rates of PPC (and HAP) in the intervention arm, this did not translate into improved clinical outcomes as measured in this trial
  • The introduction of such a comprehensive pre-op physiotherapy programme whilst probably feasible and safe, has significant cost and health economics implications
  • I will await further studies before preparing a business plan to take to the hospital funders

External Links


Summary author: Adrian Wong
Summary date: 31 January 2018
Peer-review editor: Fraser Magee
Additional editing: 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.