Incremental prognostic value of the frailty assessment tools in hospitalized patients undergoing cardiac interventions: long-term follow-up
European Heart Journal Supplements

Abstract
Selecting appropriate frailty assessment tools remains a challenge. This study compared the incremental prognostic value of three assessment tools in predicting adverse outcomes in patients admitted with different cardiac conditions.
We conducted a prospective, single-center cohort study at King Abdulaziz Cardiac Center, King Abdulaziz Medical City for National Guard, Riyadh, Saudi Arabia, between 2015 and 2017, including patients aged ≥60. A comprehensive three-dimensional (3D) frailty assessment was performed for hospitalized cardiac patients: the short portable mental status questionnaire (SPMSQ), short physical performance battery (SPPB), and handgrip strength measurement as a preprocedural evaluation. We followed patients until April 2025 for different endpoints. The endpoints were all-cause mortality (ACM) and major adverse cardiac events (MACE) (heart failure hospitalization, reinfarction, or arrhythmia). Various predictive model metrics were used to assess the incremental prognostic value of each frailty assessment relative to the basic clinical model. SAS 9.0 was used to run statistical analyses and models.
A total of 320 patients (mean age, 72.9 ± 6.6 years; 30.3% female) underwent a 3D frailty assessment prior to undergoing various cardiac interventions, including coronary or valve interventions through surgical or minimally invasive procedures. The prevalence of cardiovascular risk factors, hypertension, diabetes, and dyslipidemia was 79.1%, 73.8%, and 60.0%, respectively. The median values for the three frailty scores were 2 (1, 3) for SPMSQ, 22 (16, 28) for handgrip, and 8 (5, 11) for SPPB. During the follow-up period, 90 (28.1%) patients died. When assessing incremental prognostic value, the SPPB trend improved prediction of 1-year ACM (adjusted OR: 0.876, 95% CI: 0.77–1.00;
The SPPB measures a high incremental prognostic value for predicting ACM and MACE. Further validation is needed to support integrating these approaches into routine practice.
Incremental prognostic value of the frailty assessment tool across different models
| AdjustedOR (95% CI) | Measure of a predictive model | |||||
|---|---|---|---|---|---|---|
| C-statistic | AIC | AICC | BIC | |||
| Prediction of 30-day ACM | ||||||
| SPMSQ value | 0.840 (0.49, 1.44) | 0.527 | 0.014 | 1.739 | 2.031 | 5.456 |
| Hand grip value | 0.949 (0.85, 1.06) | 0.348 | 0.004 | 1.072 | 1.364 | 4.789 |
| SPPB value | 0.887 (0.72, 1.10) | 0.271 | −0.006 | 0.112 | 0.404 | 3.829 |
| Prediction of 1-yr ACM | ||||||
| SPMSQ value | 1.121 (0.82, 1.54) | 0.481 | 0.005 | 0.829 | 1.121 | 4.546 |
| Hand grip value | 0.963 (0.90, 1.03) | 0.263 | 0.005 | 0.768 | 1.06 | 4.485 |
| SPPB value | 0.876 (0.77, 1.00) | 0.052 | 0.009 | −3.488 | −3.196 | 0.229 |
| Prediction of 5-year ACM | ||||||
| SPMSQ value | 1.133 (0.89, 1.44) | 0.309 | −0.001 | 0.688 | 0.980 | 4.405 |
| Hand grip value | 0.917 (0.87, 0.97) | 0.002 | 0.023 | −10.377 | −10.085 | −6.660 |
| SPPB value | 0.890 (0.80, 0.99) | 0.028 | 0.008 | −6.755 | −6.463 | −3.038 |
| Prediction of 30-day MACE | ||||||
| SPMSQ value | 0.846 (0.59, 1.21) | 0.357 | −0.004 | 1.382 | 1.675 | 5.099 |
| Hand grip value | 0.992 (0.93, 1.06) | 0.807 | 0.001 | 1.898 | 2.190 | 5.615 |
| SPPB value | 0.863 (0.75, 1.00) | 0.044 | 0.022 | −2.302 | −2.010 | 1.415 |
| Prediction of 1-year MACE | ||||||
| SPMSQ value | 0.990 (0.78, 1.26) | 0.938 | 0.000 | 1.950 | 2.242 | 5.667 |
| Hand grip value | 0.939 (0.89, 0.99) | 0.0196 | 0.017 | −4.500 | −4.208 | −0.783 |
| SPPB value | 0.860 (0.77, 0.96) | 0.0053 | 0.023 | −8.775 | −8.483 | −5.058 |
All measures of a predictive model show the difference between the model that includes the frailty tool and its basic model.
The basic model includes age, gender, body mass index (BMI), ejection fraction (EF), diabetes, dyslipidemia, hypertension, previous myocardial infarction, previous PCI, previous CABG, chronic obstructive pulmonary disease (COPD), peripheral vascular disease (PVD), cerebrovascular accident (CVA), renal failure, Dialysis, and the CHAESLON Index.
AIC, AICC, and BIC interpretation: the smaller, the better.
ACM, all-cause mortality; AIC, akaike information criterion; AICC, corrected akaike information criterion; BIC, bayesian information criterion; MACE, major adverse cardiac event; SPMSQ, short portable mental status questionnaire; SPPB, short physical performance battery.
Contributors

Hadeel S Ghazal
Author

Fawaz Almutairi
Author

Naif Alzahrani
Author

Hesham Alotaibi
Author

Naif Almutairi
Author

Nasser Alsahani
Author

Attayeb Alameen
Author

Iyad Farah
Author

Sultan Alanazi
Author

Mohamed Aziz
Author

Hussam I Ardah
Author

Amjad M A Ahmed
Author
You may be interested in




