# Admission heart rate and in-hospital mortality in acute myocardial infarction: a contemporary analysis of the MIMIC-III cohort

## TL;DR

High or low heart rates at hospital admission are linked to higher death risk in heart attack patients, even with modern treatments.

## Contribution

This study identifies a U-shaped relationship between admission heart rate and mortality in AMI patients using contemporary data and adjusted for modern therapies.

## Key findings

- Heart rates ≥100 bpm are associated with a 2.45-fold increased risk of in-hospital death.
- Bradycardia (<60 bpm) is linked to a 1.58-fold higher death risk.
- The HR-mortality relationship is steeper in STEMI compared to NSTE-ACS patients.

## Abstract

To quantify the shape and strength of the association between heart rate (HR) recorded during the first 30 min of intensive-care admission and in-hospital death in contemporary acute myocardial infarction (AMI), after adjustment for modern reperfusion, pharmacotherapy, and haemodynamic variables.

We extracted 1,510 adults with a primary International Classification of Diseases, Ninth Revision (ICD-9) diagnosis of AMI (410.xx) from MIMIC-III (2008–2012). HR was defined as the mean of the first three electrocardiographic readings obtained within 30 min of ICU triage, before administration of rate-modifying drugs. We modelled HR both as clinically meaningful categories (< 60, 60–99, ≥ 100 bpm) and as a continuous exposure using restricted cubic splines (RCS). Multivariable logistic regression adjusted for age, sex, Killip class, systolic blood pressure, coronary revascularisation, β-blocker use, atrial fibrillation/flutter, hypertension, diabetes, chronic obstructive pulmonary disease, serum creatinine, haemoglobin, white blood cell count, sodium, potassium, glucose, platelet count and anion gap. Pre-specified subgroup analyses compared ST-elevation MI (STEMI) with non-ST-elevation ACS (NSTE-ACS).

Mean age was 66.7 ± 13.9 years; 33.6% were women; STEMI accounted for 42%. Overall in-hospital mortality was 10.9%. HR ≥ 100 bpm (23% of patients) was associated with higher death risk (adjusted OR 2.45, 95% CI 1.56–3.85) versus 60–99 bpm. Bradycardia < 60 bpm (15%) was also associated with excess risk (adjusted OR 1.58, 95% CI 1.02–2.45), yielding a U-shaped RCS curve (non-linearity p = 0.01). The HR–mortality gradient was steeper in STEMI than in NSTE-ACS (interaction p = 0.04). Findings were robust after including the 46 patients who died within 24 h of admission.

Admission HR exhibits a U-shaped, independent relation with early mortality in modern AMI care; values outside 60–99 bpm identify high-risk patients despite urgent reperfusion and optimal medical therapy.