Spotting frailty in CHF patients: how 6 tools stack up

Takeaway

• Frailty commonly accompanies chronic heart failure (CHF).
• Clinical frailty scale (CFS) performed best for frailty assessment.
• Editorial:
  • Discusses limitations of sensitivity and specificity, dichotomous outcomes.
  • “The optimal approach for [frailty] measurement continues to be a source of debate.”

Why this matters

• In this population, frailty correlates with higher mortality risk and hospitalizations, but there is no standardized approach to frailty assessment.

Study design

• Single-center cross-sectional study of consecutive CHF outpatients (n=467) and controls (n=87) who underwent frailty assessments:
  • 3 screening tools: Derby frailty index (DFI), acute frailty network criteria, CFS.
  • 3 assessment tools: Fried physical-frailty criteria, Edmonton frailty score (EFS), Deficit Index.
• Gold standard for each was combined results of other 5; if ≥3 tools diagnosed frailty, patient was deemed frail.
• Outcomes: frailty prevalence, classification performance, tools’ agreement.
• Funding: None disclosed.

Key results

• Frailty prevalence was higher in patients with heart failure with normal vs reduced ejection fraction.
• Of screening tools, CFS correlated best with assessment tools:
  • Correlation coefficient, 0.86-0.89; all P<.001.
  • Sensitivity, 87%; specificity, 89%.
• Of assessment tools:
  • Fried had highest sensitivity (93%) and false-positive rate (14%).
  • EFS had highest specificity (98%) and false-negative rate (18%).

Limitations

• External validation needed.
• Tool predictive value not assessed.