Measuring the quality of surgical care in Australia

A single robust and widely accepted measure for comparing outcomes between hospitals and systems would be a step forward

Surgical procedures account for large proportions of Australian healthcare activity and costs. Few Australians will never have surgery, and most will undergo several procedures during their lives. Surgery outcomes should be of interest to all, and ensuring that quality outcomes are consistently delivered should be a priority.

Patient‐specific factors, such as age and comorbidity, influence outcomes after surgery. However, outcomes can also be influenced by the performance of individual surgeons and by the system in which those surgeons work. As complications cannot be completely avoided, good systems take their risk into account and “rescue” patients with significant complications, whereas poorly functioning systems “fail to rescue”.1 Delayed recognition and treatment of significant complications exacerbate their impact and increase post‐procedure mortality and length of hospital stay.

Surgical outcomes typically assessed include peri‐operative mortality, complication rates, and length of stay. Mortality is easy to define, but is unlikely to drive quality improvements for procedures with low mortality risk. Complications can be difficult to compare across institutions, as different centres may apply different definitions. Length of stay can be affected by the nature of the procedure, the surgical approach (eg, laparoscopic or open surgery), age, and comorbidity. These factors vary between hospitals, and patients at higher risk are likely to be managed in larger public hospitals. As length of stay can be reduced by early death or early discharge to a step‐down or rehabilitation facility, a shorter length of stay in a single hospital may not indicate a better outcome. Clearly, patients with complications, especially serious complications that require further surgery or intensive care admission, return home later than those who have had uncomplicated stays.

To drive performance improvement, better outcome measures are needed, and the quality of surgical care should ideally be determined in a standardised manner to allow comparisons of surgeons, hospitals, and procedures. Benchmarking outcomes should encourage most surgeons to deliver even better care. The number of days at home alive in the thirty days following surgery (DAH₃₀) has been proposed as such a measure. Described in 2017 by Myles and colleagues, based on their analysis of outcomes data from several clinical trials of elective and emergency surgery at the Alfred Hospital in Melbourne,2 and validated by Bell and colleagues in Sweden in 2019,3 the DAH₃₀ offers a single measure that might better reflect surgical outcomes and the effects of complications, mortality, and patient demographic factors than length of stay, mortality, and complication rates alone. This single measure could be used for benchmarking and comparing the performance of different hospitals.

The DAH₃₀ is more likely to be of value for procedures associated with longer hospital stays. For a short stay procedure such as elective laparoscopic cholecystectomy, undertaken either as a day procedure or with an overnight stay and for which adverse events are infrequent, the DAH₃₀ could identify occasional outliers in terms of poor outcomes. However, the DAH₃₀ would be more useful for a high risk longer stay procedure, such as oesophagectomy, and should be considered as a measure for driving quality improvement. We have reported a peri‐operative mortality rate of 3.1% following oesophagectomy in Australia and New Zealand; the rate varies between hospitals, and the lowest rate (1.7%) was achieved in hospitals where at least twelve procedures were undertaken each year.4 Efforts to improve outcomes have focused on concentrating surgery in higher volume centres and implementing enhanced recovery programs. The DAH₃₀ appeals as a tool for assessing the outcomes of this type of procedure, and could be easily used for benchmarking across Australia.

In this issue of the MJA, Reilly and colleagues5 report their analysis of a large Medibank Private administrative dataset to validate the DAH₃₀ in a cohort of 130000 people who underwent elective (91%) or emergency inpatient surgery (9%) as private patients in Australian hospitals during 2016–2017. The authors concluded that higher DAH₃₀ values indeed “reflect shorter hospital stays and fewer serious complications, re‐admissions, and deaths.” Administrative datasets are limited by the fact that their data are not collected for determining clinical outcomes. Nevertheless, a single robust and widely accepted measure for comparing outcomes between hospitals and systems, such as the DAH₃₀, would be a step forward. Further work to determine the utility of this measure in public hospitals, following emergency surgery, and for specific higher risk procedures are the logical next steps. If effective in such areas, we could provide surgeons with the feedback they need to improve outcomes, and this can only benefit all who will undergo surgery in future years.