Frailty: Facts, Fables, and Future

“The key finding is that a frailty tool enhanced the predictive ability of conventional preoperative risk assessments for noncardiac surgical patients. However, this effect was small.”Image: Adobe Stock.A frail individual is particularly vulnerable to stressors, with a diminished capacity to maintain or regain homeostasis after a destabilizing event.1 If surgery is such an event, frailty may contribute to postoperative outcomes. In this issue of Anesthesiology, Xie et al.2 investigate whether their novel frailty index predicts postoperative cardiovascular outcomes. The authors enrolled 1,808 patients 65 yr or older (mean, 73 ± 6 yr) who underwent elective major noncardiac surgery across four hospitals in China between 2019 and 2021. Preoperatively, all patients underwent cardiac risk assessment, which included the Revised Cardiac Risk Index,3 the Gupta perioperative myocardial infarction score,4 as well as measurements of N-terminal pro-B-type natriuretic peptide. They assessed frailty using their newly developed Frailty Index during Preoperative Geriatric Assessment (FI-PGA), by incorporating 30 individual parameters related to comorbidities, level of social support, functional status, physical activity, nutritional status, cognition, and mood. Its range is 0 to 1 and included four of Revised Cardiac Risk Index parameters (diabetes, cerebrovascular disease, ischemic cardiac disease, and heart failure). Additionally, the established Clinical Frailty Scale5 was calculated for comparison (1 to 3 = not frail, 4 = prefrail, and 5 or greater = frail). The most common surgeries were abdominal (43%) and orthopedic (20%) procedures, while vascular surgeries were rare (1.1%). The study did not report details on anesthetic techniques or perioperative management. The patients were followed for up to 180 days. At 30 days (99.7% follow-up), 316 patients (17.5%) developed the primary outcome, i.e., any major cardiovascular event, including death. The most common complications were myocardial injury, followed by congestive heart failure and new-onset atrial fibrillation. Cardiovascular death occurred in 16 patients (0.9%). By 180 days (92.4% follow-up), 312 patients (19%) had either died or developed new disability. The FI-PGA (mean, 0.61; median, 0.14) was associated with the primary outcome: for each 0.1 increment, there was a 1.5-fold increase in postoperative cardiovascular complications. FI-PGA improved the prediction, when added to baseline risk models (age, sex, American Society of Anesthesiologists physical status, and Revised Cardiac Risk Index or Gupta, alone or in combination with N-terminal pro-B-type natriuretic peptide) by increasing the area under the curve (AUC) by 0.008 to 0.022. However, even at its best, the AUC value remained less than 0.718. Notably, the FI-PGA reclassified patients only into lower predicted risk categories. The other frailty index, the Clinical Frailty Scale (median 3) was also associated with postoperative cardiovascular complications: for each 1-point increase, the Clinical Frailty Scale was associated with a 1.2-fold increase in postoperative cardiovascular events. Like FI-PGA, the Clinical Frailty Scale added little prognostic value beyond baseline risk models, with AUC values remaining less than 0.716. The key finding is that a frailty tool enhanced the predictive ability of conventional preoperative risk assessments for noncardiac surgical patients. However, this effect was small. Although a direct statistical comparison was not sought, the simpler Clinical Frailty Scale, which is easily assessed during a preoperative evaluation by inquiring about a patient’s physical fitness, had a similar performance to the more detailed FI-PGA. Neither frailty tool improved the AUC beyond 0.75. The authors are to be commended for investigating whether frailty tools improve the prediction of important postoperative complications after elective, high-risk noncardiac surgery. Their study provides a wealth of supplemental material and details on methodology and serves as an educational resource for anesthesiologists on frailty assessment tools. How should we translate the findings of this investigation into clinical practice? Frailty is a multifaceted condition with many distinct forms. Unlike comorbidities such as advanced age, hypertension, or diabetes—which are well-defined entities detectable through standardized measurements and laboratory tests—frailty requires a dedicated assessment tool for identification. Originally conceptualized to examine long-term survival in community-dwelling populations, frailty was not designed to predict short-term postoperative complications. Fried et al.6 characterized frailty as an age-related decline in lean body mass, strength, endurance, walking performance, and physical activity. Their model, the Frailty Index, quantitated these phenotypes (presence = 1, absence = 0), categorizing patients as prefrail (one or two criteria) or frail (three or more criteria). They found that frail individuals had a 2.5-fold lower likelihood of survival. The foundational frailty tool, the Clinical Frailty Scale,5 showed that mild to severe frailty (Clinical Frailty Scale score 5 to 7) had an increased risk of mid-term mortality or institutionalization. Overall, up to 22 different instruments have been employed to define preoperative frailty, each yielding different incidence rates. Some tools can be applied retrospectively, and the majority of frailty tools can only be applied prospectively,7 while few can be self-reported.8 The incidence of frailty varies depending on the assessment tool used and do not assess the same domains. As highlighted in the current study, the Clinical Frailty Scale is a much simpler, physical fitness–based model, whereas the FI-PGA is a more detailed, “global” tool. Even within the same instrument, frailty classifications may differ. For example, in the current study, the authors used the Clinical Frailty Scale as a continuous variable, whereas it was originally reported and validated as a categorical measure: 1 to 3 (not frail), 4 (prefrail), and 5 or greater (frail). A limitation is the absence of a standardized reference to benchmark the presence, severity, and predictive performance of the FI-PGA tool. As a comparator tool, the Clinical Frailty Scale does not examine cardiac comorbidities, social support, nutritional status, mood, and cognition in detail; instead, it evaluates a “global” state of the patient. However, there is lack of a widely deployed clinical frailty tool particularly for surgical patients to use as a reference. Tools perform variably; even the widely known Revised Cardiac Risk Index is variable in its predictive ability; it has demonstrated limited predictive performance in noncardiac vascular surgery patients.9 The translation of frailty research into clinical practice remains challenged by the use of different frailty tools across different patient cohorts. In a retrospective analysis of a similar population (primarily abdominal surgery, with 20% vascular procedures), a comparable incidence of cardiovascular mortality was reported. They found that the alternative frailty instruments significantly improved risk discrimination and reclassification when incorporated into standard preoperative risk assessments.10 Similarly, in an analysis of a large, nationwide U.S. database of surgical hospitalizations, nearly one fifth of all patients undergoing noncardiac surgery were classified as frail. Higher frailty risk scores were associated with an increased risk of perioperative major adverse cardiovascular events, including myocardial infarction, cardiac arrest, and all-cause mortality. Frail patients were less likely to be discharged home and more likely to require postoperative care in a skilled nursing or intermediate care facility.11 These findings contradict the negligible contribution of the frailty tools observed in the current study. Equally important, the different types of surgery—each serving as a unique physiologic stressor—and differences in anesthetic techniques, particularly in the modern era of fast-track and early recovery approaches, further complicate effects to distinguish the impact of preexisting frailty on postoperative outcomes. Despite these challenges, frailty assessment holds promise for the future. In a quality improvement study of 50,463 patients undergoing major surgery, the implementation of a Frailty Screening Initiative using the Risk Analysis Index—integrated with an electronic health record alert—was associated with a significant reduction in systemwide 1-yr postoperative mortality, from 3.9 to 3.3%. Among frail patients specifically, 1-yr mortality decreased significantly from 20.2 to 16.0%.12 For those considering clinical implementation in the field of frailty, the trade-off between accuracy versus practicality will be crucial. While more sophisticated assessment tools, such as the FI-PGA—which has yet to be tested against other established frailty instruments—may offer greater accuracy, their implementation and sustainability in routine clinical practice presents significant challenges. This applies not only to the FI-PGA but also to other resource-intensive frailty assessments, such as the Frailty Index or the 6-min walk test—any tool that requires more than electronic health record screening and a brief preoperative evaluation. Beyond a clinical trial setting, implementing such tools effectively remains a major hurdle. Striking the right balance between accuracy and feasibility will be key moving forward. A meta-analysis of studies that prospectively assessed frailty in preoperative clinical practice identified the Clinical Frailty Scale as the instrument with the strongest association with in-hospital or 30-day mortality and the largest effect size among frailty measures. The Clinical Frailty Scale was also the most user friendly, with fewer logistical and environmental barriers and a faster administration time.13 Even though the Clinical Frailty Scale does not assess cognitive or socioeconomic domains, its simplicity makes it more feasible for widespread implementation in clinical settings. This study focused on cardiovascular events, which are undeniably robust and clinically relevant endpoints. However, incorporating additional outcomes, such as intensive care unit or hospital length of stay, as well as patient-centered measures could provide valuable insights and enhance the clinical impact of frailty assessment in preoperative patients. Frailty, as recognized, is largely independent of chronological age. A shift in focus from age to frailty—and a deeper exploration of their interaction both pre- and postoperatively—is essential. This distinction would be further validated if frailty, rather than age, emerged as the stronger predictor of observed outcomes. Moving forward, we should re-evaluate the trajectory of frailty from community settings to the perioperative environment, selecting a practical assessment tool that balances comprehensiveness with feasibility. Standardizing the surgical stressor—through comparable procedures and perioperative anesthetic protocols—would further refine frailty-related predictions. Additionally, extending the scope of frailty research beyond the immediate postoperative period will be crucial in fully understanding its long-term implications. Competing Interests The authors are not supported by, nor maintain any financial interest in, any commercial activity that may be associated with the topic of this article.