SIU 2017: Variation in Surgical Quality Affects Outcome of RCC Treatment

Lisbon, Portugal ( In this session, Dr. Finelli covered surgical quality control in the setting of non-metastatic RCC, specifically focusing on data driven quality of care outcomes. While there are many studies expanding the horizon of what can be done, it is important to continue to assess what is already being done and if it can be improved.

He began by reviewing some of the work of his own group in assessing quality indicators within the Canadian healthcare system, and the deficiencies they found – specifically, low-volume surgeons at low-volume centers with high levels of complications, lack of adherence to established guidelines, surgeon learning curves affecting quality.

Unfortunately, this can be a difficult topic to broach and is often not received well by surgeons. But understanding our limitations is an important step to improve the field.

Volume Outcome Relationships in the Treatment of RCC

As with other malignancies, volume of surgery is associated with oncologic, functional and perioperative outcomes. 

  • In-hospital mortality was more common in the practice of surgeons with low-volume
  • High-volume surgeons were more likely to do a NSS, had less in-hospital complications and lower in-hospital mortality
Establishing Quality Indicators – in order to improve on the above findings, a quality indicator is needed. However, what is the right quality indicator? How do we measure? What do we change based on the results?

  • Donabedian model involves 3 different QI’s - structure QI’s, process QI’s, and outcomes QI’s – to help improve quality
Dr. Finelli then reviewed his work with RCC quality indicators in the Canadian healthcare setting. In a multiple specialty working group, they were able to identify 23 quality indicators along the entire spectrum of RCC diagnosis, staging, treatment, and survival outcomes. This was based on data and expert opinion.

What makes a good QI?
  • Valid
  • Reliable
  • Feasible
  • Useable
It also needs to have inter-hospital variance so it can actually discriminate performance. Only if variance exists is there a room for improvement!

The goal is to identify low-quality outliers – though all involved should continue to try and improve!

Case-mix adjustment (accounting for case complexity) is important – case mix can affect quality measures.

His first attempt at benchmarking for RCC was to utilize the NCDB (National Cancer Database) to benchmark Hospital-level quality performance for RCC
- Quality Indicators (Case-mIx adjusted)
    o Partial Nx Proportion (T1a)
    o MIS proportion (T1-2)
    o Positive margin rate (T1)
    o Length of stay (T1-4)
    o Readmission rate (T1-4)
- While not perfect, well suited for its intended purpose
- Validated in the NCDB first to ensure variations
- Composite scores allow better stratify patients – narrows down the number that are high outliers or low outliers on all indicators
    o Learn from the high outliers
    o Help the low outliers
- Composite score correlated with mortality and other adverse events – indicating validity
Once created within the NCDB external database, they went on to validate: assess Ontario hospitals with more granular data. Data pending.

Take-home message:

1) Quality is much more than just surgical care and volume-outcomes relationships
2) Ideally, quality should also reflect non-surgical management as well

Presented by: Antonio Finelli

Written by: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, Twitter: @tchandra_uromd at the 37th Congress of Société Internationale d’Urologie - October 19-22, 2017- Lisbon, Portugal