JOA - 2026-09-23 - Journal Article
Modified 6-Factor Frailty Index Enhances Risk Stratification in Revision Total Hip Arthroplasty.
Swaminathan N, Welsh C, Wagner P, Panwalker N, Richards A, Areti A, Wochna JC, Mallender JL, Harrington MA, Martin DP
Topics
Key Takeaway
Adding hypoalbuminemia (<3.5 g/dL) to the mFI-5 improves 30-day mortality discrimination from AUC 0.687 to 0.757 in revision THA, with severe hypoalbuminemia (<2.5 g/dL) conferring up to 8.98-fold higher odds of adverse outcomes.
Summary Depth
Choose how much analysis to show on this article page.
Summary
This study asked whether adding hypoalbuminemia as a sixth item to the mFI-5 improves 30-day adverse outcome prediction in revision THA. Using ACS-NSQIP data (2015–2020), mFI-6 outperformed mFI-5 across all outcomes including any adverse event (AUC 0.618 vs. 0.570), readmission (0.588 vs. 0.563), and mortality (0.757 vs. 0.687), all P<0.001. Mild hypoalbuminemia (2.5–3.5 g/dL) conferred 2.56–3.27-fold higher odds of adverse outcomes, while severe hypoalbuminemia (<2.5 g/dL) conferred 3.56–8.98-fold higher odds across all frailty strata.
Key Limitation
AUC values for most outcomes remain modest (0.574–0.618), indicating that mFI-6, while statistically superior to mFI-5, still has limited absolute discriminative ability for outcomes other than mortality.
Original Abstract
INTRODUCTION
The 5-item modified frailty index (mFI-5) estimates perioperative risk in revision total hip arthroplasty (rTHA), but may miss nutritional vulnerability. We evaluated a novel six-item extension (mFI-6) and hypothesized that adding a point for hypoalbuminemia (less than 3.5 g/dL) would improve discrimination for 30-day adverse outcomes after rTHA.
METHODS
We performed a retrospective cohort study of adults undergoing rTHA (procedure codes 27134, 27137, and 27138) using the American College of Surgeons National Surgical Quality Improvement Program from 2015 to 2020. After excluding patients who had missing data for mFI-5-defining variables, albumin, body mass index, or key perioperative variables, 10,040 cases were analyzed. Discrimination of mFI-5 versus mFI-6 was compared using areas under the receiver operating characteristic curves. Multivariable logistic regression analyses assessed associations of mild (2.5 to less than 3.5 g/dL) and severe (less than 2.5 g/dL) hypoalbuminemia with 30-day outcomes across mFI-5 nonfrail (0), prefrail (1), and frail (≥ 2) strata.
RESULTS
The mFI-6 demonstrated greater discrimination than mFI-5 for any adverse outcome (area under the curve, 0.618 versus 0.570), reoperation (0.574 versus 0.549), readmission (0.588 versus 0.563), mortality (0.757 versus 0.687), and wound complications (0.575 versus 0.541; all adjusted P < 0.001). Mild hypoalbuminemia was associated with 2.56- to 3.27-fold higher odds of any adverse outcome across frailty strata, while severe hypoalbuminemia was associated with 3.56- to 8.98-fold higher odds. Similar patterns were observed across secondary outcomes.
CONCLUSION
The mFI-6 provided predictive value beyond mFI-5 alone in rTHA, with worsening albumin levels associated with progressively higher risk across frailty strata. Incorporating albumin into frailty assessment may improve preoperative risk stratification for rTHA.