Dear Commenters,
Thank you for the reading and interest in our work. Many of the letter writer’s comments are shared by the primary authors and reflect our opinion that risk prediction tools, as they are currently constructed, are best thought of as population risk analyses that largely fail to accurately predict individual patient risk prior to the event of surgery. The manuscript we published was intended to mimic and study the pragmatic manner in which these tools are used in practice and not to develop a risk tool. We chose the risk assessment and prediction tool tool as a comparator because it has been internally and externally validated and is in wide spread use thus providing a practical backdrop to the newer frailty information we provided [1]. Our finding that frailty Index performs as an indicator of length of stay and discharge disposition similar to the risk assessment and prediction tool tool thus has context.
Standalone risk prediction tools currently are best at providing definition to populations being compared… think American College of Surgeon’s Risk Prediction Tool. That tool, born out of a national level data set, fails to predict individual patient risk but could reasonably be used to define an institution’s patient population risk exposure and set an expectation of performance in certain quality metrics [2].
Our senior author has spent significant effort evaluating risk prediction tools and variables in arthroplasty as well as attempting to develop a high performing tool with meaningful rea under the curve values for specific complications and outcomes [[1], [2], [3]]. Those development efforts have been informative, have helped define how continuous variables should be statistically handled, but as of yet have not accomplished the goal of producing a meaningful predictive tool that can be pragmatically applied in routine clinical practice. Machine learning and artificial intelligence may provide the tools for success in the future as the letter writer suggests. In short, there is more to do and we look forward to your contributions.
Regards
Dave Manning
Conflicts of interest
Dave Manning has received Royalties from, is on the Speakers bureau/paid presentations for, and is a Paid consultant for Medacta USA and receives research support from Medacta, Stryker, and Smith and Nephew.
The other author declares no potential conflicts of interest.
For full disclosure statements refer to https://doi.org/10.1016/j.artd.2025.101894.
CRediT authorship contribution statement
David W. Manning: Conceptualization, Writing – original draft, Writing – review & editing. Isaac Sontag-Milobsky: Project administration, Writing – review & editing.
Appendix A. . Supplementary data
References
- 1.Manning D.W., Edelstein A.I., Alvi H.M. Risk prediction tools for Hip and Knee Arthroplasty. J Am Acad Orthop Surg. 2016;24:19–27. doi: 10.5435/JAAOS-D-15-00072. [DOI] [PubMed] [Google Scholar]
- 2.Edelstein A.I., Kwasny M.J., Suleiman L.I., Khakhkhar R.H., Moore M.A., Beal M.D., et al. Can the American College of Surgeons risk calculator predict 30-day complications after Knee and Hip arthroplasty? J Arthroplasty. 2015;30(9 Suppl):5–10. doi: 10.1016/j.arth.2015.01.057. [DOI] [PubMed] [Google Scholar]
- 3.Kwasny M.J., Edelstein A.I., Manning D.W. Statistical methods dictate the estimated impact of body mass index on major and minor complications after total joint arthroplasty. Clin Orthop Relat Res. 2018;476:2418–2429. doi: 10.1097/CORR.0000000000000493. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.