Frailty Index as an Adjunctive Tool for Surgical Risk Assessment and Hospitalization Duration in Open Heart Surgery Patients

Yaron Barac ¹ Efrat Kurtzwald-Josefson ¹ Robert Klempfner ³ Nir Shlomo ³ Victor Rubchevsky ¹ Inbal Grossman ¹ Irit Glik ¹ Einat Kodesh ² Batya Engel–Yeger ² Sharony Ram Eitan Snir Dan Aravot ¹ Philip Biderman

¹The Cardiothoracic Department, Rabin Medical Center, Petach-Tikva, Israel
²The Therapy Department, Faculty of Social Welfare and Health Sciences, Haifa University, Haifa, Israel
³The Rehabilitation Department, Sheba Medical Center, Ramat-Gan, Israel

Background: Open heart surgical patients become older and frailer, and the validity of the available tolls to predict M&M is often in doubt. Different surgical disciplines have recently reported that adding a frailty index score may facilitate a more accurate risk scoring model in the elderly high-risk population. The duration of hospitalization is of great economic and clinical importance; therefore, Predicting duration of stay can facilitate appropriate preparations and optimization of care.

Aim: We aimed to compare for patients undergoing cardiac surgery predictors of hospitalization duration. Therefore, we prospectively assessed pre-operatively the EuroScore II, and the Comprehensive Assessment of Frailty scores (CAF) of 94 consecutive patients admitted to a tertiary cardiac surgery department. The frailty parameters were evaluated by fulfilling a questioner as well as requesting the patient to preform diverse tasks.

Patients operated on were mostly male (89%), median age of 67 (inter quartile range [IQR] 59-74), had a median EuroScore of 1.74 (IQR 1.1-3.56) and a hospitalization duration of 10 days (IQR 7-12). Patients presented with the following comorbidities: hypertension (79%), diabetes (48%), heart failure (25%), prior myocardial infarction (12%), renal dysfunction (14%) and underwent coronary bypass (82%) or valve (34%) surgery.

Frail patients defined as CAF > 10, had similar age (66.3 vs. 65.9; p=NS) yet longer hospitalizations (12.8 vs. 7.8 days; p<0.02) and higher EuroScore (19 vs. 6; p=0.01). Regression analysis showed that both the CAF score and Euroscore II were significantly associated with the duration of hospital stay (p=0.001). However, comparison of model fit showed that the CAF score was a more powerful predictor of hospital stay than Euroscore II (adjusted R2 0.16 vs. 0.03; p<0.001).

Conclusion: Therefore, our findings suggest that pre-surgical CAF score is an independent predictor of hospitalization stay following cardiac surgery, and may provide incremental prognostic information to EuroScore II.