CORR Insights®: The Charlson and Elixhauser Scores Outperform the American Society of Anesthesiologists Score in Assessing 1-year Mortality Risk After Hip Fracture Surgery

Where Are We Now?

A hip fracture diagnosis for a patient 65 years of age or older can be devastating not only for the patient but also for his or her family. In my experience, family members want to know what to expect following this injury, and such questions often lead to discussions about the likelihood of death in the first year; we know that men, patients whose health is poorer, older patients, and patients with dementia all are at greater risk for death, and that those factors (with a few exceptions related to health conditions that can be brought quickly into better control) are not modifiable [8].

But some risk factors are modifiable and in the surgeon’s control, including the time between admission and surgery and certain details related to the procedure itself. Decreasing the time to surgery can be a consideration, but a recent multinational hip fracture accelerated surgical treatment and care track (known as HIP ATTACK) study found that among patients with a hip fracture, accelerated surgery (6 hours versus 24 hours) did not lower the risk of mortality [4]. Thus, as long as surgery is done reasonably promptly, there is no benefit to hurrying; it usually is possible to get to surgery within 24 hours.

Still, patients and their families want quantitative data on mortality, and there are tools that can help us to provide this. The overall condition of a patient can be scored using any of several scales: Modified Fried Criteria, Reported Edmonton Frail Scale, and the Frailty Index to name a few [5]. And, of course, when a patient presents with a hip fracture, the preoperative assessment typically includes the American Society of Anesthesiologists (ASA) scale. Risk-adjustment measures are valuable tools for measuring and predicting outcomes, including mortality. That score, along with two other well-known measures—the Elixhauser Comorbidity Index (ECI) and the Charlson Comorbidity Index (CCI)—are discussed in the current study by Varady and colleagues [10]. The ECI categorizes the comorbidities of patients based on the International Classification of Diseases codes found in administrative data [2]. In that index, the classification is dichotomous—patients either have a comorbidity or they do not; this tool has been used principally to anticipate the utilization of hospital resources and in-hospital mortality. The CCI is a weighted index used to predict the risk of death within 1 year for patients with specific comorbid conditions [1]. However, both outcomes are studied as predicting inpatient death in general and after orthopaedic surgery [7, 9].

In their retrospective study, Varady and colleagues [10] assess the predictive accuracy of the ASA, ECI, and CCI for 1-year mortality after hip fracture surgery. They found that the ECI and CCI scores were both more predictive of 1-year mortality when compared to ASA score. Based on these findings, the authors recommend using ECI and CCI scores for orthopaedic research on outcomes extending beyond the 90-day postoperative period.

These discoveries are helpful in providing objective data points in discussion of outcomes and research. I believe the CCI especially would be valuable to use as a predictive tool for mortality after hip fracture surgery because it is being used in the manner it was intended: prediction of mortality at 1 year after surgery. Another advantage to the CCI is that it is already calculated in the American Academy of Orthopaedic Surgery (AAOS) hip fracture registry. That facilitates ease of use in research.

Where Do We Need To Go?

Of the three scoring systems used in the current study, the ASA score has the most subjectivity [6]. The objective scoring systems, like the ECI and CCI scores, are calculated from comorbidity categories, and they reflect acute and chronic comorbidities. By including these, an overall assessment of patient vitality is made. Patient age is indirectly assumed in this instance to have an effect because with advanced age, there may be more comorbidities. But this is not always the case. There are patients who could be described as “appearing older than stated age.” Will these patients have an ECI or CCI score based on their acute and chronic health conditions that is more predictive of mortality after hip fractures? When speaking with families and the patient, I generally use the mortality rates based on the diagnosis of a hip fracture, which is often augmented with the association between outcomes and delays to surgery. Using a score calculated from definitive information readily available in the electronic medical record as a predictor of 1-year mortality after hip fracture surgery takes noncontrolled factors such as time to surgery and complications after surgery out of the discussion when talking to families and perhaps is a more desirable reality of this diagnosis.

In their study, Varady and colleagues [10] found that CCI and ECI were more accurate than ASA for 1-year mortality after hip fracture and should be used for institutional orthopaedic research involving outcomes at 90 days and beyond. Their study, though robust, was from one institution. I believe that validation studies should be performed that involve more institutions and perhaps national registries.

How Do We Get There?

But if the ultimate goal is to adopt a scoring system that can be used in research for predicting mortality after hip fracture surgery, we will need to use more than just scores based on comorbidities. That is where registry data can be helpful. The American Joint Replacement Registry is one of the registries under the AAOS Registry Program that provides data accumulated from multiple centers in the hopes of driving quality care [3]. The AAOS announced the launching of a Fracture & Trauma Registry for spring of 2021. This will include data on hip fractures. Data accumulated in these registries can be useful for surgeons to evaluate their patient outcomes and quality improvement in their practice. The Fracture & Trauma Registry includes the CCI score for hip fractures.

With patient- and procedural-specific data collected, information on outcomes will be available in the future. Because this registry is in its infancy, discussions with patients and their families can include the use of objective scoring systems, such as the CCI or ECI, which were proven in this study to be valuable in predicting mortality. This is a good starting point for providing 1-year mortality rates to our patients.