Kutrani Huda, Briggs Jim, Prytherch David, Spice Claire
Centre for Healthcare Modelling and Informatics, University of Portsmouth, Portsmouth, United Kingdom.
Queen Alexandra Hospital, Portsmouth Hospitals NHS Trust, Portsmouth, United Kingdom.
PLoS One. 2025 Jan 23;20(1):e0317234. doi: 10.1371/journal.pone.0317234. eCollection 2025.
Hospital Frailty Risk Score (HFRS) has recently been used to predict adverse health outcomes including length of stay (LOS) in hospital. LOS is an important indicator for patient quality of care, the measurement of hospital performance, efficiency and costs. Tools to predict LOS may enable earlier interventions in those identified at higher risk of a long stay. Previous work focused on patients over 75 years of age, but we explore the relationship between HFRS and LOS for all adults.
This is a retrospective cohort study using data from a large acute hospital during the period from 01/01/2010 to 30/06/2018. The study included patients aged 16 years and older. We calculated HFRS for patients who had been previously admitted to the hospital within the previous 2 years. The study developed Logistic Regression models (crude and adjusted) for nine prediction periods of LOS to assess association between (LOS and HFRS) and (LOS and Charlson Comorbidity Index-CCI), using odds ratios, and AUROC to assess model performance.
An increase in HFRS is associated with prolonged LOS. HFRS alone or combined with CCI were more important predictor of long LOS in most of periods to predict LOS. However, crude HFRS was superior to the models where HFRS was combined with any other variable for LOS in excess of 21 days, which had AUROCs ranging from 0·867 to 0·890. Regarding eight age groups, crude HFRS remained the first or second most effective predictor of long LOS. HFRS alone or combined with CCI was superior to other models for patients older than 44 years for all periods of LOS; whereas for patients younger than 44 years it was superior for all LOS except 45, 60, and 90 days.
This study has demonstrated the utility of HFRS to predict hospital LOS in patients across all ages.
医院虚弱风险评分（HFRS）最近被用于预测不良健康结局，包括住院时间（LOS）。住院时间是患者护理质量、医院绩效、效率和成本衡量的重要指标。预测住院时间的工具可能有助于对那些被确定为长期住院风险较高的患者进行早期干预。以往的研究主要集中在75岁以上的患者，但我们探讨了所有成年人中HFRS与住院时间之间的关系。
这是一项回顾性队列研究，使用了一家大型急症医院2010年1月1日至2018年6月30日期间的数据。该研究纳入了16岁及以上的患者。我们计算了在过去2年内曾入院的患者的HFRS。该研究针对九个住院时间预测期建立了逻辑回归模型（粗模型和校正模型），以使用比值比评估（住院时间与HFRS）以及（住院时间与查尔森合并症指数-CCI）之间的关联，并使用曲线下面积（AUROC）评估模型性能。
HFRS的增加与住院时间延长相关。在大多数预测住院时间的时期，单独的HFRS或与CCI结合是长期住院的更重要预测因素。然而，对于超过21天的住院时间，粗HFRS优于将HFRS与任何其他变量结合的模型，其AUROC范围为0.867至0.890。关于八个年龄组，粗HFRS仍然是长期住院的第一或第二有效预测因素。对于所有住院时间的患者，单独的HFRS或与CCI结合在44岁以上的患者中优于其他模型；而对于44岁以下的患者，除了45天、60天和90天外，在所有住院时间内都更优。
本研究证明了HFRS在预测各年龄段患者住院时间方面的效用。
