# Prognostic Tools for Predicting Mortality in Older Adults With Hip Fracture: A Systematic Review

## TL;DR

This paper reviews tools used to predict mortality in older adults with hip fractures, finding that multidimensional assessments offer the best predictions.

## Contribution

The study systematically evaluates and compares the predictive accuracy of various prognostic tools for mortality in older hip fracture patients.

## Key findings

- The ASAgeCoGeCC score and Almelo Hip Fracture Score both showed strong 30-day mortality prediction (AUC = 0.82).
- A nomogram including age and lab values predicted six-month mortality with AUC = 0.83.
- The Frail-VIG index had the highest one-year mortality prediction accuracy (AUC = 0.90).

## Abstract

Hip fracture (HF) is highly prevalent in older adults and is associated with functional impairment, disability, institutionalisation, increased use of healthcare resources, and mortality. Knowing the prognostic tools and identifying a gold standard to unify criteria would allow for more personalised clinical decision-making. The aim of this study is to conduct a systematic review of prognostic tools and their ability to predict mortality in patients over 65 years of age with HF. A search was conducted in the PubMed, Scopus, and Web of Science databases for studies published up to June 2022. The eligibility criteria were single-centre or multicentre studies published in English or Spanish that assessed the predictive ability of a prognostic tool for long-term mortality in patients over 65 years of age with HF. We identified 24 eligible publications. Most were single-centre observational cohort studies; 13 were prospective. The prognostic tools that showed better predictive capacity for 30-day mortality were the ASA score, age, cognitive status, gender, and Charlson Comorbidity Index score (ASAgeCoGeCC score) and the Almelo Hip Fracture Score (age, gender, haemoglobin, cognitive status, comorbidities, malignancy, mobility score, and ASA score), both with an area under the curve (AUC) of 0.82. The best predictive capacity for six-month mortality (AUC = 0.83) was reported for a nomogram that included age; albumin, sodium, and haemoglobin values; and Charlson Comorbidity Index score. The Frail-VIG index (including five domains: clinical, nutritional, functional, cognitive, and social) had the highest predictive capacity for one-year mortality (AUC = 0.90). There is great variability in tools for predicting mortality in patients with HF. Those that have greater predictive capacity include a multidimensional assessment.