Back to accessibility links

Variation in length of stay for prostatectomy in NSW


Variation in average length of stay (ALOS) in public hospitals for prostate surgery has been identified at both national and state levels.  Under the auspices of the Taskforce, the ACI Urology Network and the Health Economics and Evaluation Team (HEET) investigated possible differences in ALOS for prostatectomy in NSW public hospital using linked administrative data.


Several analyses using historical data over the period 2008/09 to 2013/14 was undertaken by Diagnosis Related Groups (DRG) and procedures. Linked hospital data1  was used to allow a 'look-back' to previous hospital admissions prior to surgery for identifying patient co-morbidity as well as linked1 death registrations to determine patient mortality.

Multi-level models were used to obtain case-mix adjusted hospital-specific estimates of ALOS with 'special cases' identified as those more than +/-3 standard deviations from the state average. Adjustments were made for patient characteristics such as demographics (e.g. age), co-morbidity and diagnoses (e.g. cancer) with stratification by age group and procedure. All analyses were repeated for those patients with zero co-morbidity this will help determine if the observed variation is unwarranted (clinically avoidable).


There were:

  • 3,400 prostatectomies per year at 40 hospitals.
  • ALOS 4.6 days – varied from 1.3 to 11.7 days.
  • 6 hospitals with statistically significant higher ALOS and several hospitals with significantly lower ALOS.


Recognising the limitations of administrative data, hospital anonymised reports were developed and provided to each site for comment. Further investigation to identify additional relevant service and clinical characteristics is necessary to understand the observed variation in ALOS and to rule out the contribution of any possible data limitations (unmeasured confounders, inconsistent coding or data collection practices).

Brackets report the ALOS for the most recent complete financial year (13/14). TURP = Transurethral Resection of the Prostate; RP =  Radical Prostatectomy; Open refers to open surgery. *Indicates that the identification of the hospital as a special case remained when the analysis was further restricted to patients with zero co-morbidity.

1  from the NSW Admitted Patient Data (Source: SAPHaRI, Centre for Epidemiology and Evidence, NSW Ministry of Health. This data included an anonymous patient level identifier generated through record linkage (Centre for Health Record Linkage) to allow linkage to fact of death data (Registry of Births, Deaths and Marriages Death Registrations).