Prognostic ability of the haemoglobin-to-red blood cell distribution width ratio in predicting in-hospital mortality: a retrospective, longitudinal, observational study among elderly patients with acute ischaemic stroke.

OBJECTIVE

To evaluate the prognostic ability of the haemoglobin-to-red blood cell distribution width ratio in predicting in-hospital mortality among elderly patients diagnosed with acute ischaemic stroke.

DESIGN

Retrospective, longitudinal, observational study using a hospital-based database.

SETTING

Claims data from 1 January 2014 until 31 January 2020 were extracted from the database of two hospitals affiliated with Nanjing University.

PARTICIPANTS

Patients aged ≥80 years who were diagnosed with acute ischaemic stroke, excluding those with cancers or other malignant diseases.

OUTCOME MEASURES

Patients were stratified into three groups based on the tertiles of the haemoglobin-to-red blood cell distribution width ratio. The dose-dependent relationship between this ratio and in-hospital mortality risk was determined with robust locally weighted regression analyses and restrictive cubic spline on continuous variables. The primary endpoint was defined as all-cause mortality during the hospital stay, and the secondary endpoint centred on the duration of the hospital stay.

RESULTS

A total of 606 patients constituted the dataset for the conclusive analysis (mean age, 84.6 ± 3.2 years; female, 40.3%). The haemoglobin-to-red blood cell distribution width ratio was categorised into three tertiles (T1, <6.52; T2, 6.53-8.33; and T3, >8.34). After adjusting for relevant demographic and clinical variables, a statistically significant inverse correlation was observed between higher ratios and lower risk of in-hospital mortality, with HRs of 0.48 (95% CI, 0.34 to 0.68) for T2:T1 and 0.14 (95% CI, 0.08 to 0.23) for T3:T1. A dose-dependent relationship was evident between the haemoglobin-to-red blood cell distribution width ratio and in-hospital mortality risk. The sensitivity analysis indicated that no attenuation was observed in the HR in both non-anaemic and anaemic cases. The results also indicated that a shorter length of hospital stay was associated with a higher haemoglobin-to-red blood cell distribution width ratio.

CONCLUSIONS

A high haemoglobin-to-red blood cell distribution ratio may be an independent protective factor for in-hospital mortality and reduced length of stay in elderly patients suffering from acute ischaemic stroke.