Use of Population-Based Risk Score Might Fail to Detect Frailty in Hospitalized Patients With COPD

In a cross-sectional study published in JAMA Network Open, poor detection of frailty was found among patients with chronic obstructive pulmonary disease (COPD) in the hospital using the population-based Hospital Frailty Risk Score (HFRS) compared with the bedside Clinical Frailty Scale (CFS), suggesting that using only the HFRS might cause frailty to go undetected and therefore untreated.

This study was conducted because frailty is linked to severe morbidity and mortality in patients with COPD, and though it can be treated and reversed, it is not regularly assessed in pulmonary clinical practice. Prior to this study, it wasn’t known how population-based tools compared to standard bedside assessments for hospitalized patients with COPD.

The HFRS and CFS tools are easier to use than a comprehensive geriatric assessment in the hospital setting, and researchers aimed to determine if the HFRS and CFS tools overlapped in producing the same results for patients with COPD, and to measure the sensitivity and specificity of the HFRS compared with the CFS in frailty detection.

The benefit of the CFS measurement is that it is based on clinical data directly collected from the individual or caregiver and easy to use, but results can be inadvertently influenced by the person who is administering the instrument, and it requires that they are knowledgeable about and understand the frailty paradigm. Conversely, the HFRS is an administrative data measure that does not require additional clinical resources to use, but it also does not incorporate a clinician’s judgment of an individual patient.

Before starting the study, researchers hypothesized that the HFRS would not accurately identify vulnerable individuals who were prefrail as determined by the bedside CFS assessment.

A cross-sectional study of 99 hospitalized patients with COPD exacerbation was conducted in the respiratory ward of a tertiary care academic hospital (The Ottawa Hospital, Ottawa, Ontario, Canada). Participants admitted were adult inpatients with a diagnosis of acute COPD exacerbation from December 2016 to June 2019 and used a clinical care pathway for COPD.

Primary outcomes consisted of HFRS sensitivity and specificity to identify frail and nonfrail individuals in line with CFS assessments of frailty, and the secondary outcome was the optimal probability threshold of the HFRS to discern frail and nonfrail individuals.

The results of the HFRS specifically showed poor sensitivity (27%) but higher specificity (93%) in frailty detection. When measuring the degrees of functional impairment according to the CFS, 86% of analyzed patients were considered frail. However, most patients (73%) detected as frail by the CFS measurement were categorized as low risk for frailty when measured by the HFRS.

Using the optimal probability threshold of the HFRS (≥ 1.4 points) showed performance improvement with sensitivity increased to 69% but a lower specificity of 57% in frailty detection based on the CFS classification. Even with this threshold, 27% of patients considered vulnerable by CFS (or prefrail) were not considered at risk for frailty by the HFRS. Of those already classified as mildly frail for the CFS, 33% were considered low risk for frailty by the HFRS.

Most individuals (73.5%) found to be moderately or severely frail according to the CFS measurement were women, and researchers suggest more investigation be done studying frailty between sexes and their implications for people with COPD to inform better care.

These results led researchers to determine that the HFRS assessment cannot replace the CFS assessment, especially in hospitalized patients with COPD exacerbation.

Without these results, hospitalized patients with COPD exacerbation might have been inadequately assessed for frailty. Researchers recommended that future study should focus on adapting current approved bedside frailty assessment tools to improve frailty detection in younger populations, see how they perform in male vs female patients, and study if early detection of frailty can improve clinical outcomes.

Some limitations stated by the researchers include possible selection biases. Most of the patients in the cohort were younger than 75 years when the HFRS tool was used, but it is validated as a tool that should be used in patients 75 years or older. Because the CFS tool incorporates patients’ self-disclosure about their functional capacity prior to hospitalization, patients might have had a recall bias in trying to identify their baseline state while being in their illness state.

Reference

Chin M, Kendzerska T, Inoue J, et al. Comparing the Hospital Frailty Risk Score and the Clinical Frailty Scale among older adults with chronic obstructive pulmonary disease exacerbation. JAMA Netw Open. 2023;6(2):e2253692. doi:10.1001/jamanetworkopen.2022.53692