What Is the Hospital Frailty Risk Score (HFRS)?
The Hospital Frailty Risk Score (HFRS) is an automated screening tool designed to identify frail older adults in acute care settings. Unlike traditional, manual frailty assessments that require a hands-on approach, the HFRS leverages a health system's administrative data to efficiently screen a large volume of patients. Developed using a cohort of hospitalized individuals aged 75 years or older, the HFRS is designed to predict a patient's risk of adverse events, including prolonged hospital stays, re-admission, and mortality, allowing for more targeted and personalized care plans.
The Foundational Framework: ICD-10 Diagnostic Codes
The calculation of the HFRS is fundamentally based on a predefined set of ICD-10 codes. The developers identified 109 specific three-character ICD-10 diagnostic codes that were over-represented in a cluster of older patients with high healthcare resource utilization. These codes represent a range of symptoms, diseases, and conditions commonly associated with frailty, such as dementia, falls, urinary incontinence, and cerebrovascular disease.
The Role of Weighted Points
Each of the 109 frailty-related ICD-10 codes is assigned a specific, weighted point value. This weighting system reflects the strength of each diagnostic code's association with frailty, as determined during the score's development via regression analysis. The point values typically range from 0.1 to 7.1, with higher values indicating a stronger correlation to frailty. This weighted approach ensures that the score is not simply a tally of conditions but a more nuanced reflection of the patient's overall vulnerability.
A Detailed Guide on How to Calculate Hospital Frailty Risk Score
Calculating the HFRS is a multi-step process that relies on access to a patient's administrative health records. For a health system, this process is automated, but it can be understood conceptually through the following steps:
- Identify the Patient and Timeframe: Focus on older adult inpatients (often aged 65 or 75+ depending on the study). The calculation requires a look-back period of up to two years, encompassing the current admission and any previous emergency or inpatient admissions within that time. For example, for a patient admitted today, the system would scan for relevant diagnoses from today and back two years.
- Extract ICD-10 Codes: From the patient's electronic medical record (EMR), extract all relevant ICD-10 diagnostic codes recorded during the specified timeframe. This includes primary, secondary, and comorbid diagnoses from all relevant admissions.
- Map and Assign Weights: Use a specialized algorithm or table that maps the 109 frailty-related ICD-10 codes to their specific weighted point values. The system checks if any of the extracted codes match a code in the HFRS list. If a match is found, the corresponding points are assigned.
- Sum the Points: The HFRS is the sum of all the weighted points for every matched ICD-10 code. The final number can range from 0 to over 170.
- Categorize the Risk: Based on the final summed score, the patient is placed into one of the established frailty risk categories:
- Low Risk: Score < 5
- Intermediate Risk: Score 5–15
- High Risk: Score > 15
Interpreting the HFRS Categories in Practice
The HFRS is primarily used for risk stratification rather than diagnosing frailty at the individual patient level. The score provides a hospital system with a valuable tool for population-level planning and resource allocation. For example, patients identified as high risk might be automatically flagged for comprehensive geriatric assessments or receive more proactive care coordination. A higher HFRS value is correlated with a higher risk of adverse outcomes like prolonged hospital stays and re-admission, even when accounting for other factors.
HFRS vs. Other Frailty Assessment Tools
To understand the unique role of the HFRS, it's helpful to compare it with other frailty assessment methods. This table outlines the key differences between the HFRS, the Clinical Frailty Scale (CFS), and Fried's Frailty Phenotype.
| Feature | Hospital Frailty Risk Score (HFRS) | Clinical Frailty Scale (CFS) | Fried's Frailty Phenotype |
|---|---|---|---|
| Data Source | Administrative data (ICD-10 codes) from EMRs | Clinical judgment based on overall health and function | Direct physical measurements and patient-reported data |
| Calculation Method | Automated algorithm sums weighted ICD-10 code values | A clinician scores a patient's condition based on a 9-point scale | Evaluates five specific criteria: unintentional weight loss, weakness, self-reported exhaustion, slow gait speed, and low physical activity |
| Use Case | Systematic screening and risk stratification across a large hospital population | Bedside assessment, suitable for immediate clinical decision-making | Research and clinical practice; provides an objective, standardized measure |
| Advantages | Low cost, automated, no extra staff burden, retrospective analysis possible | Quick, flexible, integrates a holistic view of the patient | Gold-standard for research; precise, based on objective criteria |
| Limitations | Reliant on quality of ICD-10 coding, potential for misalignment with clinical state | Subjective, prone to inter-rater variability, requires trained assessors | Time-consuming, resource-intensive, not suitable for mass screening |
Benefits and Limitations of the HFRS
Benefits
- Automated and Efficient: The greatest advantage of the HFRS is its ability to be calculated automatically using existing administrative data. This removes the need for additional staff time or specialized training.
- System-wide Application: It can be implemented across entire hospital systems to screen large volumes of patients, allowing for population-level health planning and identification of at-risk groups.
- Predictive Power: Numerous studies have validated the HFRS for predicting adverse outcomes, such as longer length of stay and higher readmission rates, particularly among non-critically ill older adults.
Limitations
- Data Dependency: The accuracy of the HFRS is highly dependent on the quality and completeness of the diagnostic coding within the EMR. Inaccurate or missing ICD-10 codes can lead to an incorrect score.
- Population Specificity: While validated for older adults in general acute care, its effectiveness may vary in specific patient populations, such as critically ill patients, where it has shown poorer predictive validity for mortality.
- Not a Diagnostic Tool: The HFRS should not be used as a standalone diagnostic tool for an individual patient. It is a risk stratification score and should inform, not replace, clinical judgment.
Implementing the HFRS in Clinical Practice
Successful integration of the HFRS requires careful planning and consideration of workflow. For instance, an intermediate or high-risk score could trigger an automatic alert in the EMR, prompting a more in-depth clinical assessment for frailty using tools like the CFS. This two-stage approach—automated screening followed by targeted clinical evaluation—optimizes efficiency and ensures at-risk individuals receive the appropriate level of attention and care. The use of the HFRS in this manner allows hospitals to allocate resources more effectively to those who need them most.
To learn more about the methodology and validation of the HFRS, see the original research published in The Lancet here.
Conclusion
The Hospital Frailty Risk Score offers a powerful, automated approach to screening for frailty in older adults by leveraging existing administrative data. By summing weighted ICD-10 codes from current and previous hospital admissions, it provides a valuable risk stratification tool for health systems. While it is not a replacement for comprehensive clinical assessment, its utility in efficiently identifying patients at a higher risk for adverse outcomes is clear. Healthcare systems can use the HFRS to enhance care coordination, inform targeted interventions, and ultimately improve outcomes for vulnerable older patients. Understanding how to calculate hospital frailty risk score is the first step toward integrating this important tool into modern geriatric care practices.
