SPC365

Internal validation of the EHMRG score for predicting in-hospital mortality and its discriminative ability for level of care

Session: Sessão de Posters 44 - Miscelânea: genética, estratificação de risco e mecânica miocárdica
Speaker: Luís Pedro Dos Santos

Congress: CPC 2026
Topic: D. Heart Failure
Theme: 11. Acute Heart Failure
Subtheme: 11.2 Acute Heart Failure – Epidemiology, Prognosis, Outcome
Session Type: Posters Eletrónicos
FP Number: ---
Authors: Luís Pedro dos Santos; Mariana Duarte Almeida; Gonçalo RM Ferreira; João Gouveia Fiúza; Oliver Kungel; Francisco Rodrigues Santos; Maria Luísa Gonçalves; António Costa

<p style="text-align:justify"><span style="font-size:medium"><span style="font-family:&quot;Times New Roman&quot;,serif"><span style="color:#000000"><strong><span style="font-family:Aptos,sans-serif"><span style="color:black">Introduction:</span></span></strong>&nbsp;<span style="font-family:Aptos,sans-serif"><span style="color:black">The Emergency Heart Failure Mortality Risk Score (EHMRG) was developed to estimate early mortality in acute heart failure (AHF) using admission data. It is unclear whether higher EHMRG values are also associated with the need for admission to a cardiac intensive care unit (CCU).</span></span></span></span></span></p>

<p style="text-align:start"><span style="font-size:medium"><span style="font-family:&quot;Times New Roman&quot;,serif"><span style="color:#000000"><strong><span style="font-family:Aptos,sans-serif"><span style="color:black">Objective:</span></span></strong>&nbsp;<span style="font-family:Aptos,sans-serif"><span style="color:black">To assess, in a cohort of patients hospitalized with AHF, the performance of EHMRG in predicting in-hospital mortality and its association with the level/setting of care during hospitalization.</span></span></span></span></span></p>

<p style="text-align:start"><span style="font-size:medium"><span style="font-family:&quot;Times New Roman&quot;,serif"><span style="color:#000000"><strong><span style="font-family:Aptos,sans-serif"><span style="color:black">Methods:</span></span></strong>&nbsp;<span style="font-family:Aptos,sans-serif"><span style="color:black">Single-center retrospective study including 260 consecutive patients admitted with AHF to a Cardiology department in 2024 (ward and CCU). Mean age was 75 &plusmn; 11 years and 62% were male. EHMRG variables (age, transport by ambulance, systolic blood pressure, heart rate, oxygen saturation, creatinine, potassium, troponin, active cancer, and prior metolazone use) were collected, the total score was calculated, and patients were classified into EHMRG risk groups (1 to 5b)&nbsp;according to the previously validated score and risk groups, by Lee et. al. 2012&nbsp;(see Figures 1 and 2 for more detail). The outcome &ldquo;in-hospital death&rdquo; was recorded. An ordinal three-level variable reflecting level of care was created: (1) ward; (2) ward with subsequent transfer to CICU; (3) direct CICU admission. Prognostic performance was assessed using ROC analysis and logistic regression; associations with level of care were evaluated with ANOVA and the chi-square test (IBM SPSS v30).</span></span></span></span></span></p>

<p style="text-align:start"><span style="font-size:medium"><span style="font-family:&quot;Times New Roman&quot;,serif"><span style="color:#000000"><strong><span style="font-family:Aptos,sans-serif"><span style="color:black">Results:</span></span></strong>&nbsp;<span style="font-family:Aptos,sans-serif"><span style="color:black">In-hospital mortality was 10.8%. EHMRG showed good discrimination for predicting mortality: AUC 0.73 (SE 0.042; 95% CI 0.652&ndash;0.816; p&lt;0.001). Mortality increased across risk groups (0%, 5%, 11.5%, 13.0%, 17.9%, and 24.2% for groups 1 to 5b; &chi;&sup2;=15.78; p=0.007). In univariable logistic regression, EHMRG remained associated with mortality (OR 1.012; 95% CI 1.006&ndash;1.018; p&lt;0.001; Hosmer&ndash;Lemeshow p=0.534), corresponding to an &asymp;12% increase in odds per 10-point increase. EHMRG increased progressively with level of care: ward &minus;3.1&plusmn;60.5; transferred 33.4&plusmn;70.0; CICU 44.1&plusmn;77.0 (ANOVA F=11.78; p&lt;0.001), with a significant post-hoc difference between ward and CICU (Tukey). In the categorical analysis, 84% of ward patients were in groups 1&ndash;4, whereas 48% of CICU patients were in groups 5a&ndash;5b (&chi;&sup2;=34.3; df=10; p&lt;0.001).</span></span></span></span></span></p>

<p style="text-align:start"><span style="font-size:medium"><span style="font-family:&quot;Times New Roman&quot;,serif"><span style="color:#000000"><strong><span style="font-family:Aptos,sans-serif"><span style="color:black">Conclusion:</span></span></strong>&nbsp;<span style="font-family:Aptos,sans-serif"><span style="color:black">EHMRG retained good performance for predicting in-hospital mortality in this AHF population and reflected clinical severity as expressed by the need for CICU care. Higher scores were observed not only in patients directly admitted to the CICU, but also in those initially admitted to the ward who later required CICU transfer, suggesting potential utility of EHMRG as an emergency department triage tool.</span></span></span></span></span></p>