Key result
Omitting cancer stage from surgical risk models alters mortality rates by at most ~11%.
Why the study?
A comprehensive national assessment of surgical quality across multiple cancer sites using Medicare claims data is lacking, and it is unclear if cancer stage information is needed for hospital performance evaluation.
Does the lack of cancer stage information in Medicare claims significantly modify assessments of hospital surgical performance compared to models including stage?
Observational
Yes
Does the lack of cancer stage information in Medicare claims significantly modify assessments of hospital surgical performance compared to models including stage?
Fee-for-service Medicare claims without cancer stage information are suitable for hospital-level analyses of short-term outcomes following cancer surgery, as performance measures remain stable.
Stage omission minimally alters hospital metrics in cancer surgery; extends claims-based analyses but leaves generalizability open.
RATIONAL, AIMS, AND OBJECTIVES: While public reports of hospital-level surgical quality measures are becoming increasingly common in health care, a comprehensive national assessment of surgical quality across multiple cancer sites has yet to be developed. Fee-for-service (FFS) Medicare claims present a potential resource from which to measure outcomes following cancer surgery given the national scope of patients and providers. However, due to the administrative nature of the data, clinical cancer information such as stage is not recorded. Leveraging the Surveillance, Epidemiology, and End Results (SEER) registry linked to FFS Medicare claims to analyse outcomes for patients whom we ultimately know stage information, we determined whether Medicare claims are suitable for measuring provider quality following cancer surgery by assessing the extent to which the lack of stage information modifies assessments of provider performance. METHODS: We identified patients aged 66 and older undergoing cancer surgery between 2011 and 2013 from SEER-Medicare. We compared the changes in the risk-standardized rates (RSRs), decile rankings, and c-statistics with and without risk adjustment for cancer stage for three measures of hospital performance: 30-day mortality, surgical complications, and unplanned readmissions. RESULTS: The RSR changed by at most 11.4% for mortality and by less than 4% for complications and readmissions, indicating that measures of hospital performance were stable with and without adjustment for stage. The relative performance of hospitals was also stable, as demonstrated by fewer than 20% of hospitals changing decile rank. The c-statistic declined by less than 2% across all measures, indicating that model fit was not substantially worsened without this information. CONCLUSION: These findings support the use of FFS Medicare claims for hospital-level analyses of short-term outcomes following cancer surgery. Quality reporting based on these analyses can be used to help patients choose among hospitals and for evaluating policies to improve surgical cancer care.
No takes yet. Share an insight, caveat, or question.
Lavery et al. (2019) conducted an observational in Cancer surgery. Risk adjustment without cancer stage vs. Risk adjustment with cancer stage was evaluated on Changes in risk-standardized rates, decile rankings, and c-statistics for 30-day mortality, surgical complications, and unplanned readmissions. Omitting cancer stage from risk adjustment models for hospital surgical performance changed risk-standardized rates by at most 11.4% for mortality and <4% for complications and readmissions.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: