Response to Letter to the Editor Titled "Risk Modeling for Unplanned Intensive Care Unit (ICU) Admission".

To the Editor in reply

We are grateful to the letter authors for their interest in our study regarding risk factors for unplanned intensive care unit (ICU) admission after elective total joint arthroplasty (TJA), which was recently published in the Journal of Arthroplasty [1].

We agreed with the authors that in this arduous situation of the COVID-19 pandemic nationwide and around the world, issues related to medical resource utilization have become paramount for health care providers. Therefore, the American College of Surgeons and the Centers for Medicare and Medicaid Services have recommended postponing or canceling elective procedures, including TJA [2,3]. The questions from the letter writers are timely and appropriate regarding triage of urgent procedures related to TJA, including periprosthetic joint infection, periprosthetic fracture, and prosthetic dislocation. Our ultimate goal is to have an arthroplasty procedure based calculator that can cover these urgent procedures, similar to the current American College of Surgeons NSQIP risk calculator [4].

As we hopefully return to normalization during this critical period and restart elective cases, we hope our findings in the published study can help surgeons reduce the risk of ICU admission. We report significantly increased risks with bilateral versus unilateral hip [OR 13.2 (95% confidence interval, 4.9-31.4)] and knee arthroplasty [OR 10.8 (5.8-19.9)]. Moreover, we demonstrated cutoff values for age, BMI, estimated blood loss, and surgical duration, which are clinically relevant factors for ICU admission for both hip and knee arthroplasty, which may help orthopedics surgeons for preoperative decision-making and management.

There is previously published work on predictors and risk-stratified model development, which was created from stratified preoperative and intraoperative factors to predict unplanned ICU admission after total hip arthroplasty [[5], [6], [7]]. We think this model is interesting and important for clinical use. Therefore, as suggested by the letter writers, we are developing a risk calculator tool from our database and extending it to evaluate multiple preoperative factors that influence and weight the risk for ICU admission.

We want to thank the authors for their invaluable comments and suggestions. We hope that our future work can further help plan appropriately for health care resource management, especially in this time of crisis for both surgeons and patients.