The Trauma-specific Frailty Index Identifies Orthopaedic Trauma Patients at Risk for Stepdown Admission,Skilled Nursing Facility Discharge,and Prolonged Hospitalization

Abstract

Objectives

To compare the Trauma-Specific Frailty Index (TSFI) with an institutional Perioperative Geriatric Score (PGS) for predicting postoperative level of care, discharge disposition, and length of stay in older adults undergoing operative fixation of lower-extremity fractures.

Methods

Design: Retrospective cohort study. Setting: Two hospitals within a single academic health system: an urban level I and a suburban level III trauma center. Patient Selection Criteria: Patients aged ≥65 years who underwent operative fixation of lower-extremity fractures between January 1 and June 30, 2025; nonoperative fractures and encounters lacking primary outcome data were excluded. Frailty was measured with the TSFI (0–1 deficit-accumulation index). Perioperative risk was measured with an institutional PGS combining vasopressor use, blood loss, cardiac comorbidity, case duration, and delay to surgery. Outcome Measures and Comparisons: Primary outcomes were postoperative level of care, discharge disposition, and hospital length of stay; 90-day readmission and mortality were secondary. Multivariable regression adjusted for age, sex, body mass index, and fracture type.

Results

Among 248 patients (median age, 78 years; 64.1% female), higher TSFI values were associated with stepdown admission (odds ratio [OR] 2.43; 95% confidence interval [CI], 1.13–5.21), intensive care unit admission (OR 2.79; 95% CI, 1.16–6.70), discharge to a skilled nursing facility (OR 1.96; 95% CI, 1.11–3.47), and longer length of stay (OR 1.61; 95% CI, 1.19–2.17). The PGS showed weaker associations, reaching significance only for intensive care unit admission (OR 1.16; 95% CI, 1.07–1.26). Ninety-day readmission and mortality were too infrequent for reliable estimates.

Conclusions

The TSFI outperformed an institutional perioperative geriatric triage score in predicting postoperative care escalation and discharge disposition in lower-extremity fracture patients. Incorporation of frailty assessment into orthopaedic trauma workflows may improve risk stratification, counseling, and resource planning.

Keywords

frailty geriatric trauma lower-extremity fractures Level III

Introduction

Sustained losses in mobility and activities of daily living are common early after surgery and often persist, reflected by marked declines in health-related quality of life across physical, emotional, and social domains.^1,2 Mortality remains elevated for months to years after fracture, and many patients do not return to preinjury function, with a sizable proportion requiring new institutional care.^1-3 Against this backdrop, operatively treated lower-extremity fractures represent sentinel events with enduring effects on independence and survival.^1,4,5

These injuries are also major drivers of resource utilization.⁶ Particularly among older adults, who comprise a growing share of orthopaedic trauma, resource needs are amplified. Advanced age and comorbidity are associated with higher perioperative complication rates, longer hospitalization, and greater short-term mortality.^7,8 Length of stay correlates strongly with total inpatient costs, and constraints in postacute placement further extend bed-days and downstream expenditures.⁹

Accurate point-of-care risk stratification is central to improving outcomes and resource use in geriatric orthopaedic trauma.¹⁰ Traditional perioperative assessments emphasize chronological age, comorbidity burden, acute physiology, and operative factors, but these variables incompletely capture physiologic reserve. Frailty, a multidimensional construct reflecting vulnerability to stressors, incorporates functional status, cognition, mobility, and nutrition, and provides a more global measure of resilience.^11-13 Within trauma populations, the Trauma-Specific Frailty Index (TSFI) operationalizes this construct as a deficit-accumulation score that weights comorbidities, dependence in activities of daily living, mood and exhaustion, falls, mobility limitations, and nutritional markers into a 0-1 frailty index (Table 1).

Table 1.

Components and Calculation of the Trauma-specific Frailty Index

Domain/Category	TSFI item (deficit if “yes” unless otherwise specified)	Coding of deficit
Comorbidities	History of cancer	Documented cancer diagnosis
	Coronary artery disease	Prior MI, PCI/CABG, or angina
	Dementia	Documented diagnosis of dementia
Functional dependence	Needs help with grooming	Requires assistance
	Needs help managing money	Requires assistance
	Needs help doing housework	Requires assistance
	Needs help toileting	Requires assistance
	Needs help walking	Requires assistance or device
Psychological/fatigue	Feels less useful	Affirmative response on questionnaire
	Feels sad	Affirmative response
	Feels effort to do everything	Affirmative response
	Feels lonely	Affirmative response
Other health deficits	Falls	≥1 fall in the specified time frame
	Not sexually active	No current sexual activity
	Low serum albumin	Albumin below institutional lower limit of normal

Each deficit present is scored as 1 point; the TSFI is calculated as the number of deficits divided by the total number of possible deficits (0–1). Higher values indicate greater frailty.

In contrast, the Perioperative Geriatric Score (PGS) used at our institution represents a more conventional perioperative risk tool, combining intraoperative hemodynamic support, blood loss, cardiac history, anesthesia and operative duration, and delay to surgery to guide early postoperative triage decisions (Table 2).

Table 2.

Components and Calculation of the Perioperative Geriatric Score

Risk factor	Threshold/Definition	Points
Vasopressor use under anesthesia	> 0.03 mcg/kg/min	5
Blood loss	> 200 mL	4
History of cardiac disease	Yes	3
Time under anesthesia	> 120 minutes	3
Operative time	> 60 minutes	3
Time from injury to surgery	> 24 hours	1

The Perioperative Geriatric Score (PGS) is calculated by summing the points for each risk factor present (total score range 0–19 in this cohort). Higher scores indicate greater perioperative risk; in institutional practice, PGS ≥ 12 is considered high risk for unplanned ICU admission and escalated postoperative monitoring.

Prior work on frailty in orthopaedic trauma often pools elective and urgent procedures, restricts inclusion to single fracture patterns, or employs heterogeneous risk tools, limiting direct comparison between frailty-based and traditional perioperative instruments.¹⁴ Few studies have examined how a trauma-specific frailty index performs against an institutional perioperative triage score for key outcomes such as level of postoperative care, discharge disposition, and length of stay in geriatric lower-extremity fracture patients. To address this gap, we conducted a head-to-head evaluation of the TSFI and the PGS within a contemporary cohort, assessing each tool’s correlation, discrimination, and incremental prognostic value for prespecified clinical and operational outcomes. It is hypothesized that a preoperative frailty index derived from social and functional factors would predict hospital length of stay and discharge disposition in this population.

Methods

Study Design

This retrospective cohort study of consecutively treated older adults undergoing operative fixation of lower-extremity fractures was approved under an orthopaedic retrospective umbrella protocol with waiver of informed consent. Data resided on secure institutional servers and were de-identified for analysis. The sample size was determined by the number of eligible patients treated during the study period; no a priori sample size or power calculation was performed. The study is reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for observational studies.

Setting and Data Sources

Eligible encounters from January-June 2025 were identified from an institutional Trauma Registry and verified in the electronic health record (EHR) across two acute-care hospitals within a single academic health system. One site was an urban level I trauma center with postoperative admission options to either the intensive care unit (ICU) or a medical–surgical floor (n = 143, 57.7% of the cohort). The second site was a suburban level III trauma center with ICU, stepdown, and medical-surgical units (n = 105, 42.3%); the stepdown unit functioned as an intermediate level of care between the ICU and medical–surgical floor, with higher nurse-to-patient ratios and enhanced cardiorespiratory monitoring. Institutional characteristics, including trauma level, available postoperative care units, and cohort distribution by site, are summarized in Table 3. Registry fields supplied demographics, fracture pattern, procedures, and dates; the EHR provided perioperative variables, risk scores, and outcomes. The Trauma-Specific Frailty Index (TSFI) was implemented at the level III trauma center beginning January 1, 2025, and therefore was available only for that site, whereas the Perioperative Geriatric Score (PGS) was recorded at both hospitals.

Table 3.

Institutional Characteristics and Postoperative Care Capacity

Characteristic	Urban Level I Trauma Center	Suburban Level III Trauma Center
Setting	Academic, urban	Community, suburban
Trauma designation	Level I	Level III
Postoperative care units available	ICU; medical–surgical floor	ICU; stepdown; medical–surgical floor
Patients in cohort, n (%)	143 (57.7)	105 (42.3)
TSFI availability	Not implemented	Implemented January 1, 2025
PGS availability	Yes	Yes
TSFI scores documented, n	-	78
Complete covariate data for TSFI analyses, n	-	77

Abbreviations: ICU, intensive care unit; TSFI, Trauma-Specific Frailty Index; PGS, Perioperative Geriatric Score. Dash indicates not applicable.

Participants

Inclusion: patients ≥65 years undergoing operative treatment of an incident lower-extremity fracture (femoral neck/intertrochanteric/subtrochanteric, femoral shaft, distal femur, tibial plateau, tibial shaft, or ankle). Fractures were classified according to the Orthopaedic Trauma Association/AO 2018 Fracture and Dislocation Classification Compendium, corresponding to proximal femur (31; femoral neck/intertrochanteric/subtrochanteric), femoral shaft (32), distal femur (33), proximal tibia (41; tibial plateau), tibial shaft (42), and ankle (44) injuries.¹⁵ Exclusion criteria were age <65 years, nonoperative management of the index fracture, fractures outside the femur–tibia–ankle region, and encounters lacking all primary outcome data. For patients undergoing multiple procedures during a single admission, the first fixation was designated the index procedure.

Demographic Variables

Preoperative variables included age, sex, height, weight, body mass index (BMI), primary diagnosis or fracture type, overall comorbidity profile, and American Society of Anesthesiologists (ASA) physical status classification. Comorbidities were abstracted from chart review and encompassed cardiovascular, metabolic, renal, pulmonary, musculoskeletal, psychiatric, and oncologic conditions commonly encountered in older adults.

Risk Instruments

The institutional perioperative geriatric triage score (PGS) was calculated as a composite measure derived from predefined perioperative risk factors: vasopressor infusion exceeding 0.03 μg/kg/min, estimated blood loss greater than 200 mL, documented cardiac disease, anesthesia duration longer than 120 minutes, operative time greater than 60 minutes, and time from arrival to surgery exceeding 24 hours. Each variable was scored dichotomously, and the total score was generated per the institutional algorithm; in secondary analyses, scores of 12 or higher were classified as high risk.

Frailty was quantified using the TSFI, a deficit-accumulation index calculated as the proportion of predefined frailty characteristics present. These characteristics encompassed select comorbidities including cancer, coronary artery disease, and dementia, dependence in basic or instrumental activities of daily living, indicators of mood and exhaustion, history of falls, mobility limitations, reduced sexual activity, and nutritional markers including serum albumin. The TSFI yields values ranging from 0 to 1, with higher values indicating greater frailty. Consistent with prior literature and local practice, a TSFI value greater than 0.25 was considered high risk in secondary analyses.¹⁶ The TSFI was available only at the hospital where it had been implemented.

Healthcare Utilization

Healthcare utilization outcomes included (1) postoperative level of care immediately after surgery (medical-surgical floor, stepdown, ICU), (2) hospital discharge disposition (home, skilled nursing facility, inpatient rehabilitation, hospice/other), (3) length of stay (LOS) in days, (4) 90-day readmission, and (5) 90-day mortality. Weight bearing as tolerated (WBAT) at discharge and time from admission to surgery were exploratory outcomes. For 90-day endpoints, analyses were restricted to patients with ≥90 days of follow-up or administratively censored at the data-freeze date.

Statistical Analysis

Continuous variables were assessed for normality using the Shapiro-Wilk test and summarized as medians with interquartile ranges or means with standard deviations, as appropriate. Categorical variables were reported as frequencies and percentages. Correlation between the TSFI and the institutional PGS was evaluated using Spearman rank correlation, and linear regression was performed to model their relationship with regression coefficients and R² values reported.

Multivariable regression analyses evaluated associations between each risk instrument and postoperative outcomes. Multinomial logistic regression modeled postoperative level of care (medical-surgical floor, stepdown, or ICU) and discharge disposition (home, skilled nursing facility, inpatient rehabilitation, hospice/other). Binary logistic regression was used for 90-day readmission and 90-day mortality, using the same covariate structure. Robust linear regression modeled hospital length of stay to account for heteroskedasticity. All models adjusted for age, sex, BMI, and fracture type; the American Society of Anesthesiologists (ASA) classification was analyzed in parallel models as an additional predictor. Both risk instruments were evaluated as continuous measures in all analyses. Given the low frequency of 90-day mortality and readmission events and occasional model non-convergence, these 90-day analyses were considered exploratory and corresponding estimates were interpreted cautiously. All tests were two-tailed with significance defined as p < 0.05. All statistical analyses were conducted using Python (Version 3.10; Python Software Foundation, Wilmington, DE) with the stats models, SciPy, and pandas libraries.Results.

Demographic Characteristics

A total of 248 patients were included in the study cohort (Table 4). The median age was 78 years (IQR 73-85). The majority of patients were female (n =159, 64.1%), while 89 patients (35.9%) were male. The median BMI was 25.8 kg/m² (IQR 21.9-30.0), consistent with a cohort distributed across the normal to overweight range. For analyses that included the TSFI, the analytic cohort was limited to patients treated at the level III trauma center, where this instrument had been implemented.

Table 4.

Baseline Demographic and Clinical Characteristics of the Study Cohort

Characteristic	Total (N=248)
Sex, n (%)
Female	159 (64.1%)
Male	89 (35.9%)
Age, median years (IQR)	78 (73-85)
BMI, median kg/m² (IQR)	25.8 (21.9-30.0)
LOS, median days (IQR)	6 (4-9)
ASA classification, n (%)
ASA 2	37 (14.9%)
ASA 3	183 (73.8%)
ASA 4	28 (11.3%)
Fracture type, n (%)
Femoral neck	62 (25.0%)
Intertrochanteric/Subtrochanteric	99 (39.9%)
Femoral shaft	20 (8.1%)
Distal femur	15 (6.0%)
Tibial plateau	13 (5.2%)
Tibial shaft	14 (5.6%)
Ankle	23 (9.3%)
Missing	2 (0.8%)

Data are reported as median (interquartile range) or number (percentage). BMI, body mass index; IQR, interquartile range; LOS, length of stay; ASA, American Society of Anesthesiologists physical status classification.

Stepdown Unit

The TSFI demonstrated a significant association with stepdown unit admission (Odds Ratio [OR] 2.43, 95% Confidence Interval [CI] 1.13–5.21; p = 0.023). In contrast, the perioperative geriatric score (PGS) was not significantly associated with stepdown utilization (OR 0.99, 95% CI 0.92–1.06; p = 0.786).

Intensive Care Unit (ICU)

Higher TSFI scores were significantly associated with ICU admission (OR 2.79, 95% CI 1.16–6.70; p = 0.022). The PGS also demonstrated a significant but more modest association with ICU admission (OR 1.16, 95% CI 1.07–1.26; p < 0.001).

Skilled Nursing Facility (SNF)

The TSFI was significantly predictive of SNF disposition (OR 1.96, 95% CI 1.11–3.47; p = 0.021). By comparison, the PGS demonstrated only a borderline and non-significant association (OR 1.03, 95% CI 0.96–1.11; p = 0.452).

Inpatient Rehabilitation (IPR)

The TSFI was not significantly associated with IPR admission (OR 0.69, 95% CI 0.22–2.19; p = 0.529). The PGS showed a modest but non-significant trend toward increased IPR utilization (OR 1.07, 95% CI 0.98–1.17; p = 0.148).

Length of Stay (LOS)

Both frailty indices were significantly associated with increased hospital length of stay. The TSFI was associated with longer LOS (OR 1.61, 95% CI 1.19–2.17; p = 0.002), and the PGS demonstrated a similar significant association (OR 1.18, 95% CI 1.07–1.30; p = 0.001).

Weight-Bearing Status and Time to Surgery

In secondary analyses restricted to patients with complete data, higher TSFI values were not significantly associated with weight-bearing as tolerated at discharge or with time from arrival to surgery greater than 24 hours. The perioperative geriatric score similarly did not demonstrate a clinically meaningful association with either outcome; although its relationship with delayed surgery reached statistical significance, this was expected because delay to surgery is a component of the score itself. These analyses were therefore considered exploratory, and no definitive conclusions were drawn for these secondary endpoints.

Ninety-Day Readmission and Mortality

Ninety-day adverse events were infrequent in this cohort, limiting the ability to generate stable regression estimates. In the readmission model, neither frailty instrument demonstrated a significant association with 90-day readmission, but confidence intervals were extremely wide, reflecting sparse event counts and limited statistical power. For 90-day mortality, the TSFI model showed evidence of quasi-separation and did not converge, precluding meaningful interpretation of effect estimates. The perioperative geriatric score similarly showed no significant association with mortality. Taken together, the low frequency of 90-day events and convergence instability suggest that these analyses should be considered exploratory only, and no definitive conclusions can be drawn regarding the predictive performance of either frailty measure for rare postoperative outcomes.

A Summary of all Model Estimates Is Presented in Table 5

Discussion

Many patients with lower-extremity fractures experience sustained functional decline, prolonged hospitalization, and discharge to institutional care.¹⁷ Therefore, accurate risk stratification for patients who may require higher levels of postoperative monitoring or non-home discharge can guide perioperative decision-making and optimize hospital resource allocation.¹⁸ To the authors’ knowledge,(Table 6) this study is the first to date to characterize acute postoperative escalation of care using a frailty-based risk score, the TSFI. In this cohort, the TSFI consistently predicted acute escalation of postoperative care (stepdown and ICU admission), discharge to SNFs, and prolonged LOS. Compared with a PGS, the TSFI more strongly discriminated against post-acute care needs and higher-level postoperative monitoring.(Table 7) In the context of United States Medicare policy, in which patients typically must meet a three-night inpatient stay requirement to qualify for skilled nursing facility (SNF) coverage, early identification of individuals likely to require institutional discharge has direct implications for how quickly teams mobilize post-acute planning.¹⁹ Because SNF precertification by insurers and physical and occupational therapy evaluations are often limited or unavailable on weekends and after-hours, additional resources must be deployed during these periods to avoid preventable delays in discharge and extended length of stay.(Table 8) These results confirmed the prognostic utility of frailty in orthopaedic trauma and suggested that incorporation of the TSFI into routine workflows may improve early risk stratification.

Table 5.

Multivariable Model for Stepdown Unit Admission (Multinomial Logistic Regression). Reference Category: Medical-Surgical Floor. Odds Ratios for the TSFI are Reported per 0.1 unit Increment. All Tests Two-Sided; p < 0.05 Considered Significant. Events per Variable = 11.2 (56 Stepdown Events, 5 Predictors). n = 77. Stepdown Contrast From the Multinomial Level-Of-Care Model

Predictor	OR (95% CI)	P Value
TSFI (per 0.1 unit)*	2.43 (1.13-5.21)	0.023
PGS (per unit)*	0.99 (0.92-1.06)	0.786
Age (per year)	1.08 (0.96-1.22)	0.215
Sex (male vs female)	0.40 (0.08-1.93)	0.254
BMI (per unit)	1.01 (0.92-1.11)	0.844
Fracture type	0.73 (0.49-1.10)	0.130

*TSFI and PGS were evaluated in separate models, each adjusting for age, sex, BMI, and fracture type (5 predictors per model). PGS results are shown for comparison.

Table 6.

Multivariable Model for ICU Admission (Multinomial Logistic Regression). Exploratory Analysis. Reference Category: Medical-Surgical Floor. Odds Ratios for the TSFI are Reported per 0.1 unit Increment. Events per Variable = 1.8 (9 ICU Events, 5 Predictors); Estimates Should be Interpreted Cautiously. n = 77. ICU Contrast From the Multinomial Level-Of-Care Model

Predictor	OR (95% CI)	P Value
TSFI (per 0.1 unit)*	2.79 (1.16-6.70)	0.022
PGS (per unit)*	1.16 (1.07-1.26)	<0.001
Age (per year)	1.13 (0.97-1.32)	0.119
Sex (male vs female)	1.97 (0.23-16.71)	0.534
BMI (per unit)	1.04 (0.95-1.14)	0.399
Fracture type	0.98 (0.60-1.60)	0.933

*TSFI and PGS were evaluated in separate models, each adjusting for age, sex, BMI, and fracture type (5 predictors per model). PGS results are shown for comparison.

Table 7.

Multivariable Model for Skilled Nursing Facility Discharge (Binary Logistic Regression). Home Discharge (n = 16) vs Skilled Nursing Facility Discharge (n = 48). Odds Ratios for the TSFI are Reported per 0.1 unit Increment. PGS Evaluated in a Separate Model on the Same Cohort With the Same Covariates. All Tests Two-Sided; p < 0.05 Considered Significant. n = 64. Patients Discharged to Inpatient Rehabilitation (n = 7), Hospice (n = 3), or Other Destinations (n = 4) Were Excluded From This Binary Comparison

Predictor	OR (95% CI)	P Value
TSFI (per 0.1 unit)	2.28 (1.21-4.29)	0.011
Age (per year)	1.12 (1.01-1.25)	0.031
Sex (male vs female)	0.52 (0.11-2.40)	0.403
BMI (per unit)	1.07 (0.91-1.25)	0.404
Fracture type	0.75 (0.47-1.22)	0.252
PGS (per unit)*	1.05 (0.90-1.22)	0.525

*PGS evaluated in a separate binary logistic model (home vs SNF) on the same 64-patient cohort with the same covariates.

Abbreviations: TSFI, Trauma-Specific Frailty Index; PGS, Perioperative Geriatric Score; SNF, skilled nursing facility; OR, odds ratio; CI, confidence interval.

Table 8.

Multivariable Model for Hospital Length of Stay (Robust Linear Regression). Beta Coefficients Represent the Change in Length of Stay (days) per Unit Change in Each Predictor. All Tests Two-Sided; p < 0.05 Considered Significant. Observations per Predictor = 15.6 (78 Observations, 5 Predictors)

Predictor	β (95% CI)	P Value
TSFI (per unit)*	4.76 (1.76-7.75)	0.002
PGS (per unit)*	0.17 (0.07-0.27)	0.001
Age (per year)	-0.01 (-0.07-0.06)	0.837
Sex (male vs female)	-0.07 (-1.10-0.96)	0.895
BMI (per unit)	0.01 (-0.01-0.02)	0.512
Fracture type	0.04 (-0.22-0.31)	0.746

*TSFI and PGS were evaluated in separate models, each adjusting for age, sex, BMI, and fracture type (5 predictors per model). PGS results are shown for comparison.

The present findings align with the current body of literature establishing the TSFI as a robust predictor of discharge to a skilled nursing facility or rehabilitation, prolonged length of stay, major complications, and in-hospital mortality across diverse geriatric trauma populations.^16,20-24 Yet, while these outcomes are important markers of long-term recovery and survival, they generally reflect events that occur later in the hospitalization or after discharge. By contrast, the immediate perioperative period, particularly when escalation to a stepdown or ICU is required, represents a pivotal juncture in patient management and influences resource allocation, staffing, and postoperative monitoring intensity. In addition to predicting 90-day healthcare utilization, TSFI scores were also associated with escalation to stepdown and ICU levels of care. In practice, these results may afford clinicians additional predictive utility across the perioperative continuum. The clinical and policy implications of these findings were substantial. Stepdown and ICU utilization drive hospital resource consumption, and early identification of patients likely to require these services enables proactive planning.^21,24 Incorporating frailty assessment into initial trauma evaluations may support discussions with patients and families regarding realistic postoperative expectations, guide early consultation with geriatrics and rehabilitation services, and streamline discharge disposition planning. Because Medicare SNF eligibility depends on three qualifying inpatient overnights and timely physical and occupational therapy assessments - which may be difficult to obtain on weekends or after hours - an admission frailty score that predicts SNF disposition could help prioritize early therapy evaluations and insurance precertifications for high-risk patients, allowing transfer to SNF as soon as eligibility criteria are met. From a systems standpoint, embedding frailty tools such as the TSFI into perioperative workflows could help optimize bed utilization, reduce delays in postacute placement, and align with institutional efforts toward value-based care and bundled payment models.^.16,20,21,23

This study should be interpreted in the context of its limitations. It was conducted within a single academic health system, which may limit generalizability, and the TSFI was implemented at only one clinical site, introducing the possibility of site-related bias. As a retrospective analysis, causal inference cannot be established, and unmeasured confounding remains possible. The sample size, while sufficient to detect several clinically meaningful associations, was modest relative to large trauma registries, and no a priori sample size or power calculation was performed; the study was therefore powered by the available sample, which may have limited the ability to detect associations for less frequent outcomes. This limitation was most evident in the 90-day outcomes, where the low frequency of mortality and readmission events led to wide confidence intervals and model non-convergence; therefore, these analyses should be considered exploratory rather than definitive. Finally, although we compared the TSFI with an institutional perioperative geriatric triage score, future work should evaluate how the TSFI performs relative to other established frailty instruments and in larger, multicenter cohorts.

Conclusion

The TSFI reliably predicted not only discharge to skilled nursing facilities and prolonged hospitalization but also, for the first time, escalation to stepdown and ICU care. Taken together, these findings support a broader role for frailty screening that extends beyond long-term outcomes to immediate perioperative management. Incorporating TSFI into routine trauma workflows may improve risk stratification, strengthen patient counseling, and optimize resource allocation in orthopaedic trauma.

Footnotes

Acknowledgments

The authors thank the orthopaedic trauma registry staff and clinical research personnel at our institution for their assistance with case identification and data abstraction for this project. This study received no external funding.

ORCID iDs

William A. Crites

Anish Gangavaram

Ronak J. Mahatme

Ethical Considerations

This study received ethical approval from the University of Cincinnati Institutional Review Board (IRB #2020-0116) on June 23, 2025. This was an IRB-approved retrospective chart review; all patient information was de-identified, and informed consent was waived. Patient data are stored on secure institutional servers and will not be shared with third parties.

Consent to Participate

Informed consent to participate was waived by the University of Cincinnati Institutional Review Board due to the retrospective nature of the study and use of de-identified data.

Consent for Publication

Not applicable. The manuscript does not contain any individual person’s identifiable data, images, or videos.

Author Contributions

William A. Crites, BA: Conception and design of the study; data collection; data analysis and interpretation; drafting of the manuscript; critical revision of the manuscript. Benjamin E. Jevnikar, BA: Data management; assistance with analysis and interpretation; drafting of the manuscript; critical revision of the manuscript. Samuel K. Gerak, BA: Data collection; assistance with manuscript drafting and critical revision. Anish Gangavaram: Statistical support; assistance with manuscript revision. Ronak J. Mahatme, MD: Study design; clinical interpretation of findings; critical revision of the manuscript for important intellectual content. Shawn A. Moore, BS: Statistical support; assistance with data analysis and interpretation; critical revision of the manuscript. Richard T. Laughlin, MD: Study conception and supervision; interpretation of results; critical revision of the manuscript; final approval of the version to be submitted. All authors approved the final manuscript and agree to be accountable for all aspects of the work.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data that support the findings of this study are derived from institutional clinical and trauma registry databases and contain protected health information. De-identified data may be available from the corresponding author on reasonable request and with appropriate Institutional Review Board and institutional approvals.*

References

Cooper

. The crippling consequences of fractures and their impact on quality of life. Am J Med. 1997;103(2A):12S-19S.

Amarilla-Donoso

López-Espuela

Roncero-Martín

, et al. Quality of life in elderly people after a hip fracture: a prospective study. Health Qual Life Outcomes. 2020;18(1):71.

Panula

Pihlajamäki

Mattila

, et al. Mortality and cause of death in hip fracture patients aged 65 or older: a population-based study. BMC Musculoskelet Disord. 2011;12:105. doi:10.1186/1471-2474-12-105.

Cummings

Melton

. Epidemiology and outcomes of osteoporotic fractures. Lancet. 2002;359(9319):1761-1767. doi:10.1016/S0140-6736(02)08657-9.

Haentjens

Magaziner

Colón-Emeric

, et al. Meta-analysis: excess mortality after hip fracture among older women and men. Ann Intern Med. 2010;152(6):380-390. doi:10.7326/0003-4819-152-6-201003160-00008.

Burge

Dawson-Hughes

Solomon

Wong

King

Tosteson

. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025. J Bone Miner Res. 2007;22(3):465-475. doi:10.1359/jbmr.061113.

Polanczyk

Marcantonio

Goldman

, et al. Impact of age on perioperative complications and length of stay in patients undergoing noncardiac surgery. Ann Intern Med. 2001;134(8):637-643.

Healthcare Cost and Utilization Project (HCUP) . HCUP Facts and Figures, 2006: Statistics on Hospital-Based Care in the United States. Rockville, MD: Agency for Healthcare Research and Quality (US); 2008.

Rocha

Valente

Ramalho

Borri

Mattos

Bittar

. Cost and time of hospitalization for elderly people with bone fractures in a reference hospital. Einstein (Sao Paulo). 2024;22:eGS0493.

10.

Steffensmeier

Hoge

Shah

, et al. Evaluation of a novel multidisciplinary preoperative workup strategy for geriatric hip fractures. J Orthop Trauma. 2022;36(8):413-419.

11.

Makary

Segev

Pronovost

, et al. Frailty as a predictor of surgical outcomes in older patients. J Am Coll Surg. 2010;210(6):901-908.

12.

Robinson

Pointer

Dunn

Cleveland

Moss

. Simple frailty score predicts postoperative complications across surgical specialties. Am J Surg. 2013;206(4):544-550. doi:10.1016/j.amjsurg.2013.03.012.

13.

Panayi

Orkaby

Sakthivel

, et al. Impact of frailty on outcomes in surgical patients: a systematic review and meta-analysis. Am J Surg. 2019;218(2):393-400. doi:10.1016/j.amjsurg.2018.11.020.

14.

Roopsawang

Zaslavsky

Thompson

Aree-Ue

Kwan

RYC

Belza

. Frailty measurements in hospitalised orthopaedic populations age 65 and older: a scoping review. J Clin Nurs. 2022;31(9-10):1149-1163. doi:10.1111/jocn.16035.

15.

Meinberg

Agel

Roberts

Karam

Kellam

. Fracture and dislocation classification compendium—2018. J Orthop Trauma. 2018;32(suppl 1):S1-S170.

16.

Joseph

Saljuqi

Amos

, et al. Prospective validation and application of the Trauma-Specific Frailty Index: results of an American Association for the Surgery of Trauma multi-institutional observational trial. J Trauma Acute Care Surg. 2023;94(1):36-44. doi:10.1097/TA.0000000000003817.

17.

Piirtola

Löppönen

Vahlberg

Isoaho

Kivelä

Räihä

. Fractures as an independent predictor of functional decline in older people: a population-based study with an 8-year follow-up. Gerontology. 2012;58(4):296-304. doi:10.1159/000334565.

18.

Loftus

Balch

Ruppert

, et al. Aligning patient acuity with resource intensity after major surgery: a scoping review. Ann Surg. 2022;275(2):332-339. doi:10.1097/SLA.0000000000005079.

19.

Vashi

Cafardi

Powers

Ross

Shrank

. Observation encounters and subsequent nursing facility stays. Am J Manag Care. 2015;21(4):e276-e281.

20.

Hamidi

Haddadin

Zeeshan

, et al. Prospective evaluation and comparison of the predictive ability of different frailty scores to predict outcomes in geriatric trauma patients. J Trauma Acute Care Surg. 2019;87(5):1172-1180. doi:10.1097/TA.0000000000002458.

21.

Alizai

Colosimo

Hosseinpour

, et al. It is not all black and white: the effect of increasing severity of frailty on outcomes of geriatric trauma patients. J Trauma Acute Care Surg. 2024;96(3):434-442. doi:10.1097/TA.0000000000004217.

22.

Joseph

Orouji Jokar

Hassan

, et al. Redefining the association between old age and poor outcomes after trauma: the impact of frailty syndrome. J Trauma Acute Care Surg. 2017;82(3):575-581. doi:10.1097/TA.0000000000001329.

23.

Joseph

Pandit

Zangbar

, et al. Validating the Trauma-Specific Frailty Index for geriatric trauma patients: a prospective analysis. J Am Coll Surg. 2014;219(1):10-17.e1. doi:10.1016/j.jamcollsurg.2014.03.020.

24.

Joseph

Pandit

Zangbar

, et al. Superiority of frailty over age in predicting outcomes among geriatric trauma patients: a prospective analysis. JAMA Surg. 2014;149(8):766-772. doi:10.1001/jamasurg.2014.296.