5197Correlates and prognostic significance of nuisance bleeding after acute myocardial infarction. The FAST-MI programme

Background

Following acute myocardial infarction (AMI), most patients receive potent antithrombotic medications, which may promote nuisance bleedings (ecchymoses, minor nose or dental bleeds etc.). Little information is available on the factors related to nuisance bleedings, nor on their prognostic significance in post-AMI patients.

Aims and methods

A health status questionnaire was sent to all patients participating in the nationwide French FAST-MI cohorts (2010 and 2015) one year after the acute episode, with a specific question on the presence of nuisance bleedings. Overall, 3968 patients answered the question on the presence nuisance bleedings. In the 2010 cohorts, we also analysed whether the presence of nuisance bleedings at one year was an independent correlate of 5-year mortality.

Results

54% of the patients reported the presence of nuisance bleedings (59% in 2010 and 51% in 2015). In univariate analyses, nuisance bleedings were more frequently found in younger patients, women, patients with STEMI, current smokers, patients treated with PCI, those receiving newer P2Y12 inhibitors or ACE-inhibitors but less frequent in patients with diabetes, hypertension, or those receiving ARBs or direct oral anticoagulants. Using logistic regression analysis, however, the only independent correlates of nuisance bleedings were: female gender (OR 1.45, 1.25–1.68), age ≤60 years (OR 1.22, 1.06–1.41), VKAs (OR 1.72, 1.28–2.31), clopidogrel (OR 1.62, 1.29–2.03), prasugrel (OR 3.16, 2.43–4.09), ticagrelor (OR 2.61, 2.04–3.35) at discharge, diabetes (OR 0.74, 0.63–0.88) and year 2015 vs 2010 (OR 0.62, 0.53–0.73).

In the 2010 cohort, the presence of nuisance bleeding at one year was not a predictor of mortality at 5 years (90% survival in both patients with or without nuisance bleedings; adjusted HR 0.96, 95% CI 0.69–1.33) (Figure).

Figure 1. 5-yr survival by nuisance bleed

Conclusion

Nuisance bleedings one year after AMI are extremely frequent. They are more common in women, younger patients, in patients receiving P2Y12 inhibitors, especially newer P2Y12-i, and in those receiving VKAs; in contrast, diabetic patients report less nuisance bleedings. The presence of nuisance bleeding at one year does not appear to impact 5-year mortality.

Acknowledgement/Funding

Amgen, AstraZeneca, Bayer, Daiichi-Sankyo, Eli-Lilly, GSK, MSD, Novartis, Pfizer, Sanofi, Servier

P648Screening and treatment of familial hypercholesterolemia in a French sample of ambulatory care: a retrospective longitudinal cohort study

Background

Familial hypercholesterolemia (FH) is largely underdiagnosed as there are typically no clinical symptoms prior to the first cardiovascular (CV) event. We conducted a study which utilised ambulatory care electronic medical record (EMRs) to alert physicians to possible cases of FH. Specifically, physicians were alerted to LDL-C levels >190 mg/dL (suggesting a risk of FH) and invited to complete the Dutch Lipid Clinic Network score (DLCN).

Purpose

Describe characteristics, comorbidities and clinical management of patients diagnosed with definite or probable FH in an ambulatory care setting.

Methods

All patients with a DLCN score of definite/probable FH (score higher or equal than 6; index event) between January 2016 and September 2018 were identified in the THIN® database (The Health Improvement Network; an anonymized EMR powered by GERSDATA, a Cegedim Health Data Division). These fully anonymized data were collected by 2000 General Practitioners (GP), 130 cardiologists and 40 endocrinologists, receiving 5.5 million patients regularly in their office. Sociodemographic, laboratory measurements, comorbidities, lipid-lowering therapies (LLT), visits to specialists, LDL-C and hospitalizations were collected and analysed at baseline, and 1, 2, 3, 6 months, and 1 year thereafter.

Results

From 999 anonymous patients with a DLCN score, 98 (10%) FH patients were identified (38 [39%] definite FH, 60 [61%] probable FH) while remaining fully anonymous, 9 (9%) of whom already had genetic testing. Mean (SD) age was 57.4 (14.3) years; 56 (57%) patients were female, half (51/98 [52%]) were diagnosed with pure hypercholesterolemia (ICD-10 code: E78.0) and 9 (9%) had a personal history of CV event. 93 patients (95%) had a LDL-C measurement prior to DLCN assessment (definite FH, 36/38 [95%]; probable FH, 57/60 [95%]). Among screened FH patients, 61.2% had LDL-C between 190 to 250 mg/dL and 16.3% had LDL-C higher than 250 mg/dL. At the time of DLCN assessment, one third (30/98 [31%]) of patients were not receiving any LLT, one third ([35%] 34/98) were receiving statins alone, 19% (19/98) receiving LLT combination with statin, and 15% (15) other LLTs. Moderate statin intensity was prescribed in 20% (20/98) of patients; high intensity statin, 17%, (17/98); low intensity, 10% (10/98). No improvement on LLT use (including use of high statin intensity) was observed over the 12-month follow-up.

Conclusion

This is the first study in France that use EMR to screen possible FH patients and support GPs in identifying patients that need to be treated. Our data highlight the need to screen, diagnosis and treat potential FH patients in ambulatory care settings. Longer follow-up is needed to evaluate the impact of FH assessment on referral to specialists, LLT and clinical outcomes.

Acknowledgement/Funding

Amgen

A Role for Behavior in the Relationships Between Depression and Hostility and Cardiovascular Disease Incidence, Mortality, and All-Cause Mortality: the Prime Study

Background

Behavioral factors are important in disease incidence and mortality and may explain associations between mortality and various psychological traits.

Purpose

These analyses investigated the impact of behavioral factors on the associations between depression, hostility and cardiovascular disease(CVD) incidence, CVD mortality, and all-cause mortality.

Methods

Data from the PRIME Study (N = 6953 men) were analyzed using Cox proportional hazards models, following adjustment for demographic and biological CVD risk factors, and other psychological traits, including social support.

Results

Following initial adjustment, both depression and hostility were significantly associated with both mortality outcomes (smallest SHR = 1.24, p < 0.001). Following adjustment for behavioral factors, all relationships were attenuated both when accounting for and not accounting for other psychological variables. Associations with all-cause mortality remained significant (smallest SHR = 1.14, p = 0.04). Of the behaviors included, the most significant contribution to outcomes was found for smoking, but a role was also found for fruit and vegetable intakes and high alcohol consumption.

Conclusions

These findings demonstrate well-known associations between depression, hostility, and mortality and suggest the potential importance of behaviors in explaining these relationships.