In this nationwide sample study, we found that among CDI-related hospitalizations, frail individuals had significantly higher mortality (4-7-fold) and morbidity, including secondary sepsis, respiratory failure, acute heart failure, AKI, dialysis use, CVA, and DVT as compared to individuals with LFS. Higher frailty scores were associated with 2–4 day longer stays than those with LFS, and significantly increased healthcare expenditure.
Frailty is the accelerated decline of physiological functions with increased vulnerability to stressors and insult [23], independent of the patient’s age. In our study, we found that frail individuals were at 19–74% increased odds of CDI-related hospitalizations when compared to individuals with LFS. Frail patients have been shown to be at increased risk of CDI [24], likely from a combination of increased inpatient visits, medication use (e.g., antimicrobials, acid-blockers), dampened immune response, and gut microbiome dysbiosis. [25, 26] In fact, several studies demonstrating a correlation between frailty and gut microbiome reduction, increased chronic inflammation, commensal opportunism and dysmotility. [27–31]
The concept of frailty as a predictor of mortality is established, however, data in CDI remain limited. [32–35] In our study, we found that frail individuals (e.g., intermediate and high HFRS) were at a 4-7-fold increased mortality. Although a cumulative deficit frailty index for older individuals hospitalized with CDI has been developed (e.g., the Frailty Index for Clostridium difficile (FI-CDI)), it has not been validated for use in patients < 55 years of age (19). The HFRS, a recently validated score using ICD-10 codes for baseline comorbidities, provides a systematic and cost-effective screening tool for frailty in hospitalized individuals, has been shown to predict mortality [9, 36], and could ultimately could guide early intervention strategies to enhance hospital outcomes within this population. In addition, our analysis demonstrated that higher frailty scores were associated with increased morbidity, including severe sepsis, secondary respiratory failure, acute heart failure, AKI and dialysis use. Frailty has been implicated in significant immune dysregulation resulting in functional immunodeficiency [37], which are known risk factors for the development of severe or fulminant CDI. [38] Last, our analysis showed that frail individuals were at higher odds of secondary arterial and venous thrombosis, which could be explained by upregulated proinflammatory mediators, chemokines, in addition to toxin induced platelet–leukocyte aggregates. [39] Therefore, frailty (e.g., the HFRS) could serve as a prognosis tool in hospitalized patients with CDI. [23, 40] Moreover, these findings highlight that although there are several models to define the severity of CDI, there is a lack of reliable prognostic tools for CDI [34, 40], which represents a key barrier to timely management and resource allocation.
Among CDI-related hospitalizations, frailty was associated with longer hospital stays and higher financial burden. While the estimated national burden of CDI and associated hospitalizations has been declining [41], it is imperative to adequately manage patients at risk for worse prognosis to further reduce the national burden of CDI.
This study has several strengths. This is the first study that provides a comprehensive analysis of CDI inpatient outcomes using the HFRS. Moreover, using the NIS, we were able to capture 148 million hospitalizations in the US from 2016 to 2020, and obtained a large sample of CDI-related hospitalizations (e.g., 415,300). This guaranteed adequate generalizability and statistical power to our analysis. Lastly, the HFRS is a sample tool that can be implemented using ICD-10 codes already present within the electronic medical records of patients, therefore, bypassing resource intensive assessment.
There were some limitations to our study design. First, the data pertaining to the severity of disease, medication use (e.g., antibiotics for CDI), and recent admissions is not available in the NIS database. Second, one individual could account for multiple hospitalizations (e.g., recurrent CDI) over the study period. Lastly, the HFRS uses ICD-10 codes which do not reflect several components of frailty such as polypharmacy, general weakness and dependence on activities of daily living.
In summary, our study presents strong evidence in support of using the HFRS as an index to predict the mortality and morbidity associated with CDI. There is a wide scope for future research to develop a validated model for the role on frailty in CDI acquirement, progression and prognosis. A focus on prevention and early treatment and resuscitation should be considered in frail individuals at risk for increased morbidity and mortality due to CDI.