Clinical Characteristics, Frailty, and Prognostic Predictors in Patients Aged 75 and Older With Acute Coronary Syndrome

Incidence of major bleeding, defined as clinically overt bleeding resulting in hemodynamic compromise, intracranial hemorrhage, bleeding requiring transfusion, or bleeding requiring surgical or endovascular hemostatic intervention; includes gastrointestinal and access-site bleeding. Unit of Measure: Incidence (%)

Proportion of patients with ≥1 rehospitalization for any cause and total number of rehospitalizations per patient during follow-up; comparison between PCI and conservative therapy. Unit of Measure: % of patients with ≥1 rehospitalization; count per patient.

Patient age and sex as predictors of the decision to pursue conservative medical therapy instead of percutaneous coronary intervention (PCI). Independent predictors will be identified using multivariable logistic regression and reported as odds ratios with 95% confidence intervals.

Measure: Age (years); Sex (male/female).

Hemoglobin concentration measured at admission and recorded in the dataset. Evaluated as a predictor of conservative therapy versus PCI using multivariable logistic regression. Unit of Measure: g/L.

Estimated glomerular filtration rate (CKD-EPI) measured at admission and recorded in the dataset. Evaluated as a predictor of conservative therapy versus PCI using multivariable logistic regression. Unit of Measure: mL/min/1.73 m².

High-sensitivity troponin I measured at admission; assay name and 99th percentile documented. Evaluated as a predictor of conservative therapy versus PCI using multivariable logistic regression. Unit of Measure: ng/L.

C-reactive protein measured at admission and recorded in the dataset. Evaluated as a predictor of conservative therapy versus PCI using multivariable logistic regression. Unit of Measure: mg/L.

BNP or NT-proBNP measured at admission and recorded in the dataset. Evaluated as a predictor of conservative therapy versus PCI using multivariable logistic regression. Unit of Measure: pg/mL.

Glycated hemoglobin (HbA1c) measured at admission and recorded in the dataset. Evaluated as a predictor of conservative therapy versus PCI using multivariable logistic regression. Unit of Measure: %.

Composite score derived from admission laboratory parameters not otherwise specified as individual outcomes. Included analytes: WBC with differential, RBC, HCT, MCV, MCH, MCHC, PLT, Glucose, urea, creatinine, total proteins, albumin, Total cholesterol, HDL, LDL, triglycerides, Na, K, Cl, Ca, AST, ALT, GGT, LDH, CK-MB, TSH.

Each analyte will be coded as 1 if outside the central hospital laboratory reference range and 0 if within range (assay name, manufacturer, units, and reference intervals documented). The composite score equals the total number of abnormal parameters per patient and will be entered as a single predictor variable in multivariable logistic regression to evaluate its association with the decision for conservative therapy versus PCI. Missing data will be handled by multiple imputation; sensitivity analyses will assess robustness. Unit of Measure: Count (number of abnormal parameters per patient).

Binary composite coded as 1 if any of the following are present on the admission ECG: ST elevation, ST depression, T-wave inversion, or pathologic q/Q/qS complexes; coded as 0 if none are present. Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Unit of measure: Presence/absence (yes/no).

Binary composite coded as 1 if any of the following are present on the admission ECG: atrial fibrillation (any), atrial flutter, supraventricular tachycardia, premature atrial complexes, premature ventricular complexes, non-sustained VT, sustained VT, ventricular fibrillation, or accelerated idioventricular rhythm; coded as 0 if none are present. Evaluated as a predictor of the decision for conservative therapy versus PCI. Unit of measure: Presence/absence (yes/no).

Binary composite coded as 1 if any of the following are present on the admission ECG: sinoatrial block, sick sinus syndrome, atrioventricular block (any degree), right bundle branch block, left bundle branch block, left anterior hemiblock, or left posterior hemiblock; coded as 0 if none are present. Evaluated as a predictor of the decision for conservative therapy versus PCI.

Binary composite coded as 1 if any of the following are present on coronary angiography prior to the treatment decision: left main coronary lesion (≥50% luminal stenosis), multivessel coronary disease (≥2 epicardial vessels with significant stenosis), bifurcation lesion, or ≥70% stenosis in any vessel; coded as 0 if none are present. Evaluated as a predictor of the decision to pursue conservative medical therapy versus PCI in multivariable logistic regression. Unit of measure: Presence/absence (yes/no).

Categorical coded as 1 if LVEF ≤40%, coded as 2 if 41-49%, coded as 3 if ≥50%, based on Teichholz, Simpson biplane, or visual estimation if biplane unavailable. Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Examinations will be performed using the GE Healthcare Vivid 6 ultrasound system (General Electric) in our central echocardiography laboratory, following standardized protocols.

Binary composite coded as 1 if average E/e' ≥16, coded as 0 if <16. Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Examinations will be performed using the GE Healthcare Vivid 6 ultrasound system (General Electric) in our central echocardiography laboratory, following standardized protocols.

Binary coded as 1 if effusion present, coded as 0 if absent. Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Examinations will be performed using the GE Healthcare Vivid 6 ultrasound system (General Electric) in our central echocardiography laboratory, following standardized protocols.

Categorical coded as 1 if LAVI ≥34 mL/m², coded as 0 if <34 mL/m². Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Examinations will be performed using the GE Healthcare Vivid 6 ultrasound system (General Electric) in our central echocardiography laboratory, following standardized protocols.

Categorical coded as 1 if WMSI >1.0, coded as 0 if ≤1.0. Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Examinations will be performed using the GE Healthcare Vivid 6 ultrasound system (General Electric) in our central echocardiography laboratory, following standardized protocols.

Binary composite coded as 1 if any of the following are present: aortic regurgitation, mitral regurgitation, pulmonary regurgitation, or tricuspid regurgitation; coded as 0 if none are present. Evaluated as a predictor of conservative therapy versus PCI in multivariable logistic regression. Examinations will be performed using the GE Healthcare Vivid 6 ultrasound system (General Electric) in our central echocardiography laboratory, following standardized protocols.

Presence of predefined comorbid conditions recorded at admission and used as predictors of the decision for conservative therapy versus PCI. Each comorbidity will be recorded using standard clinical definitions and ICD codes where available. Comorbidities will be coded as binary variables (yes/no); chronic kidney disease will be classified by KDIGO stage (CKD stage 1-5). Candidate comorbid predictors will be evaluated in univariate analyses and included in multivariable logistic regression based on clinical relevance and univariate p<0.10. Missing data will be handled by multiple imputation; sensitivity analyses will be performed.

Frailty assessed at admission using the Rockwood Clinical Frailty Scale (CFS) and the Hopkins frailty instrument; assessor initials and assessment timing recorded. Scores modeled primarily as continuous and secondarily as prespecified categorical variables; inclusion in multivariable logistic regression guided by univariate screening and clinical relevance; missing data handled by multiple imputation. Measure: Rockwood CFS coded as: 0 (score 1-3); 1 (score 4-6); 2 (score 7-9). Hopkins frailty coded as: 0 (score 0); 1 (score 1-2); 2 (score 3-5). Raw Rockwood CFS (1-9) and raw Hopkins score (numeric) will be recorded.

Composite endpoint defined as cardiovascular death, myocardial infarction, or stroke. Participants will be coded as positive if any of these events occur, and negative if none occur. Unit of measure: Presence/absence (yes/no).

Presence/absence of rehospitalization due to cardiovascular causes (including acute coronary syndrome, heart failure, arrhythmia, stroke, or repeat revascularization). Unit of measure: Presence/absence (yes/no).

Presence/absence of death from any cause. Unit of measure: Presence/absence (yes/no).

The outcomes are in-hospital major adverse cardiovascular events (MACE: cardiovascular death, myocardial infarction, stroke) and all-cause mortality at 6 months. Each outcome will be reported separately. Candidate predictors include demographic characteristics, prespecified comorbidities, admission laboratory values (including renal function and eGFR), echocardiographic measures (e.g., LVEF, LV volumes, LAVI, TAPSE), ECG parameters (e.g., ST elevation/depression, T-wave inversion, rhythm abnormalities, conduction blocks), and angiographic findings (culprit vessel, multivessel disease, left main disease, percent stenosis). Independent predictors will be identified using multivariable logistic regression for in-hospital MACE and Cox proportional hazards models for 6-month mortality. Full variable definitions, units, assay names, and reference ranges are provided in the SAP.

Frailty and cognition will be assessed at admission using the Rockwood Clinical Frailty Scale (CFS, range 1-9; higher = greater frailty), the Hopkins frailty instrument (range 0-5; higher = greater frailty), and the Mini-Mental State Examination (MMSE, range 0-30; higher = better cognition). Clinical outcomes are in-hospital MACE (composite of cardiovascular death, myocardial infarction, and stroke) and all-cause mortality at 6 months. Results will be reported by the prespecified score categories (CFS 1-3 / 4-6 / 7-9; Hopkins 0 / 1-2 / 3-5; MMSE <24 vs ≥24) and expressed as the percentage of patients with each outcome within categories. Statistical analysis will include correlation coefficients (Spearman or Pearson as appropriate) and multivariable models comparing categories (logistic regression for in-hospital MACE; Cox proportional hazards for 6-month mortality); reference groups, covariates and all cutoffs are specified in the SAP.

Incidence of periprocedural complications, gastrointestinal bleeding, neurological complications, and psychiatric disorders (e.g., psycho-organic syndrome).