Association between distance to tertiary hospital and cardiovascular outcomes in coronary artery disease patients

Abstract Background The impact of distance from residence to Tertiary Referral Hospital and cardiovascular (CV) outcomes in patients with coronary artery disease (CAD) is unknow. Despite longer travel distances hinder access to healthcare and may worsen CV outcomes, we hypothesize that Mediterranean lifestyle and behaviors in distant rural areas may be associated with a reduced risk of CV death and events. Purpose To investigate the association between travel distance to Tertiary Hospital and mid-term cardiovascular outcomes in a population of CAD patients in Southern Spain. Methods Retrospective study including all patients discharged after percutaneous coronary intervention (PCI) at a high-volume center in Southern Spain during 2018. Those belonging to another healthcare area were excluded. One-way driving distances from residence to hospital were computed using Google Maps Distance Matrix API with R package “gmapsdistance”. Patients were stratified into tertiles according to travel distance (short, STD; intermediate, ITD; and long, LTD). Kaplan-Meier (KM) and Multivariable Cox regression (adjusted for age, sex, atrial fibrillation, cancer history, prior revascularization and clinical presentation) were used to assess the impact of travel distance on CV death and a composite outcome of MACE (Myocardial Infarction, unplanned PCI and CV death). Results Of 1005 patients discharged after PCI during the study period, 966 met the selection criteria. Flowchart and baseline characteristics by distance groups are presented in Figure 1. Median travel distance tertiles were 6.1 (STD), 41.7 (ITD) and 78.4 (LTD). During a median follow-up of 31 (IQR 28–35) months, 50 cardiovascular deaths [STD 27 (8.4%), ITD 13 (4%), LTD 10 (3.1%), p=0.006)] and 63 MACE occurred [STD 45 (13.9%), ITD 37 (11.5%), LTD 26 (8.1%), p=0.060)]. KM curves for the three distance groups are shown in Figure 2. In univariable and multivariable Cox models, longer travel distances were associated with better outcomes, as for every 10 Km increase, there was a 11% and 7% decrease in the hazards of CV death (HR adj: 0.89, CI 0.82–0.98, p=0.029) and of MACE (HR adj: 0.93, CI 0.87–0.99, p=0.025), respectively. Conclusion Travel distance was inversely associated with CV events in a population of CAD patients in Southern Spain. Patients in the first tertile of distance had a higher rate of CV death. Multicenter studies involving other Mediterranean regions are needed to confirm these findings and to look