Paul R Carter

79 Hyperlipidaemia Reduces Mortality in Breast, Prostate, Lung and Bowel Cancer

Introduction Hyperlipidaemia is a well -established cardiovascular risk factor but the effect of hyperlipidaemia and treatment with cholesterol-lowering drugs on cancer remain equivocal. We aimed to investigate the impact of comorbid diagnosis of hyperlipidaemia amongst patients with the four most prevalent cancer types in the United Kingdom (Lung, Breast, Prostate and Bowel). We did this using a large database of patients admitted with comorbid hyperlipidaemia to hospitals in the North of England, UK between 2000–2013. Methods Anonymous information on patients with a primary diagnosis of lung, breast, prostate and bowel cancers were obtained from hospitals in North England, UK between 1st January 2000 and 31st March 2013. This data was analysed according to the ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) study protocol. ICD-10 and OPCS-4 codes were used to trace patients coded for cancer, patient demographics, prevalence of hyperlipidaemia and mortality data. The impact of hyperlipidaemia on mortality in cancer patients was analysed by cox regression adjusted for age, gender and ethnicity. P values of <0.05 were taken as statistically significant. Results 929552 patients were admitted during the study period. Of these 7997 had lung cancer, 5481 had breast cancer, 4629 had prostate cancer, and 4570 had bowel cancer. Comorbid diagnoses of hyperlipidaemia significantly reduced mortality amongst patients with all four cancer types studied. Cox regression analyses accounting for differences in age, gender and ethnicity showed that hyperlipidaemia was associated with a significantly reduced mortality rate in lung cancer (OR 0.78, 95% CI 0.70–0.87), breast cancer (OR 0.57, 95% CI 0.43–0.77), prostate cancer (OR 0.53, 95% CI 0.50–0.79) and bowel cancer (OR 0.70, 95% CI 0.58–0.84). Conclusion We have demonstrated for the first time that comorbid hyperlipidaemia has a highly protective effect on mortality amongst patients with the four most prevalent cancers in the UK. The underlying reasons for this are yet to be determined but treatment with statins may contribute. This potentially beneficial effect of lipid-lowering medications amongst cancer patients should be further investigated.Abstract 79 Table 1 Odds ratio adjusted for age, gender and ethnicity

Impact of cardiovascular risk factors and disease on length of stay and mortality in patients with acute coronary syndromes

BACKGROUND Traditional risk factors for cardiovascular disease (CVD) have been thoroughly investigated. We aimed to investigate the impact of comorbid cardiovascular risk factors and diseases on length of stay (LOS) and mortality in patients presenting with acute coronary syndromes (ACS). METHODS We examined prevalence of CVD, LOS and mortality from 25,287 consecutive admissions for ACS from seven hospitals across North West England between 2000 and 2013 using the ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) protocol using ICD-10 and OPCS-4 coding systems. RESULTS Mean LOS was 7.0days and there were 9653 (38.2%) deaths in the ACS cohort over the 13-year period. Hypertension and hyperlipidaemia were associated with decreased LOS (6.95 and 4.8days respectively, P<0.001) and mortality (36.8% and 19.4% respectively, P<0.001), as was angina pectoris (5.4days and 33.5%, P<0.001). Type 2 diabetes was associated with increased LOS and mortality (7.8days, P<0.05; 44.4%, P<0.001), whereas type 1 diabetes was associated with increased mortality only (7.0days, P=0.42; 41.3%, P<0.001). Other concomitant CVD was associated with an increased LOS and mortality: peripheral vascular disease (8.6days, P<0.05; 53%, P<0.001), atrial fibrillation (10.9days, P<0.001; 63.5%, P<0.001), cerebrovascular disease (15.9days, P<0.001; 76%, P<0.001), heart failure (11days, P<0.001; 69.9%, P<0.001), and ischaemic heart disease (6.7days, P<0.001; 38.7%, P<0.05). CONCLUSION CVD risk factors have a significant and varied impact on LOS and mortality in patients with ACS and it may be inappropriate to group them when assessing in-hospital risk. These factors should be used to identify patients at an increased risk of prolonged admissions and death post-ACS, and services should be directed accordingly.

78 Impact of Ethnicity on Mortality Amongst Tobacco Abusers in the United Kingdom

Introduction Tobacco smoking contributes significantly to the global health burden and its negative impact on cardiovascular morbidity and mortality in particular have been well documented. Smoking is affected by sociodemographic factors though with rates varying according to sex, social class and ethnicity. However, the impact of ethnicity on mortality amongst hospitalised tobacco abusers is yet to be studied. Objectives We aimed to investigate the impact of ethnicity on mortality of hospitalised patients with a comorbid diagnosis of tobacco abuse. We did this using a large database of patients admitted with comorbid tobacco abuse to hospitals in the North of England between 2000–2013. Methods Anonymous information on adult tobacco abusers was obtained for hospitals in North West England between 1st January 2000 and 31st March 2013. This data was analysed according to the ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) study protocol. ICD-10 and OPCS-4 codes were used to trace patients coded for tobacco abuse, patient demographics and mortality data. Mortality of tobacco abuse patients was compared by logistic regression. P values of <0.05 were taken as statistically significant. Results 28379 patients had comorbid tobacco abuse during the study period. Mean age of tobacco abusers was 44.9 years, 51.2% were male and 1918 (6.8%) tobacco abuse patients died. The majority were Caucasian (84.8%) with lower proportions of South Asian (4.7%), Afro-Caribbean (1.9%), Oriental (0.2%), mixed (1.0%), other (2.5%) and unknown (4.8%) ethnicities. Crude mortality was highest amongst Caucasian patients (7.3%) compared to South Asian (2.2%), Afro-Caribbean (2.7%), Oriental (1.7%), mixed (0.3%), other (3.1%) and patients with unknown ethnicities (6.1%). Of all these ethnic groups, logistic regression accounting for age and sex showed that only South Asian ethnic origin affected mortality rates amongst patients with comorbid tobacco abuse. South Asian patients were 1.90 times more likely to die (95% CI 1.21–2.97).Abstract 78 Table 1 Demonstrates the crude unadjusted and the adjusted mortality rates for tobacco users according to ethnicity Conclusion Our results demonstrate that mortality of tobacco abusers varies according to ethnicity with higher mortality amongst South Asian patients in particular. This could represent differences in smoking rates or an increased susceptibility to smoking. This is particularly important given that smoking and South Asian ethnicity are both known risk factors for developing cardiovascular disease.

35 Percutaneous coronary intervention in patients with acute coronary syndrome

Introduction Percutaneous Coronary Intervention (PCI) has changed the management of acute coronary syndrome (ACS). However, the role of PCI in patients with previous coronary artery bypass grafting (CABG) is widely debated. Lack of clear guidelines leads to subjective assessments based on clinician preferences. We sought to investigate if PCI affected all-cause mortality in ACS patients with previous CABG. Methods Completely anonymous information on patients with ACS with a background of previous CABG presenting to three multi-ethnic general hospitals in the North West of England, United Kingdom in the period 2000–2012 was obtained. Patients were traced using the ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) study protocol using ICD-10 and OPCS-4 coding systems. Information on demographics, co-morbidities and procedures were available for all patients. Predictors of mortality and survival analyses were performed using SPSS version 20.0. Results Out of 12,227 patients with ACS, 1172 (19.0%) cases had previous CAVBG. Of these 83 (7.1%) patients underwent PCI. Multi-nominal logistic regression, accounting for differences in age and co-morbidities, revealed that PCI conferred a 7.96 times improvement in mortality (2.36–26.83 95% CI) compared to not having PCI. Conclusions We have shown that PCI confers significantly improved all-cause mortality in the management of ACS in patients with previous CABG. This highlights the need for clinicians to conscientiously think about the individual benefits and risks of PCI for every patient.

108 The Impact of Marital Status on Mortality and Length of Stay in Patients Admitted with Myocardial Infarction

Introduction Ischaemic heart disease is the leading cause of mortality worldwide. The development of surgical and percutaneous interventions has improved survival rates, but the influence of sociodemographic factors on outcomes following MI and their potential use as predictors of such outcomes, are increasingly recognised. Conclusive studies show associations between marriage and lower incidences of IHD in addition to better survival prospects for married individuals suffering MI. There is however, a conflicting evidence base and a lack of literature considering the influence of marital status on LOS, which has been observed to be highly variable in MI patients. Objectives From a large patient database, we aimed to investigate the impact of marital status on the prevalence, LOS and crude mortality of MI patients admitted in Northern England, UK. Methods We compared marital status variations and associated LOS and mortality data by one way anova and cox regression respectively, using anonymous information on MI patients obtained from hospitals in North England between 1st January 2000 and 31st March 2013. This data was analysed according to the ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) study protocol, which uses ICD-10 and OPCS-4 codes to trace patients and demographics. P values <0.05 were taken as statistically significant. Results Amongst 929552 patient admissions recorded during the study period there were 25287 cases involving a new diagnosis of MI. Mean age of MI patients was 66.6 years, 64.2% of the cohort were male and 80.3% were Caucasian. 38.2% of MI patients died and mean LOS was 7.0 days. Crude mortality was highest among widowed patients (62.9%). Logistic regression accounting for age, sex and gender showed that married (OR 0.863), widowed (OR 0.959) and unmarried patients (OR 0.973) had statistically lower mortality rates when compared to single people. LOS was statistically shorter for married patients (2.12 days shorter), and unmarried patients (2.66 days shorter) compared to a mean LOS of 8.2days recorded amongst single patients. Conversely, mean LOS was 1.82 days longer for widowed patients. Conclusion Marital status has a clinically important impact on LOS and mortality of MI patients. In particular, single patients show higher mortality rates and longer LOS compared to married patients. It is reasonable to suggest that these results may be due to reduced social support at home and this should be taken into account when considering the holistic care of patients with MI.Abstract 108 Table 1 Length of stay and mortality of patients admitted with ACS stratified by marital status Marital Status Prevalencen (%) Mean LOS (Days) Difference in mean LOS compared to single patients(95% confidence intervals) Crude mortalityn (%) Odds ratio for mortality compared to single patients(95% confidence intervals) Single 2 531(10.0%) 8.2 - 752(29.7%) - Married 11 933 (47.2%) 6.1 -2.12(-1.05, -3.20)*** 4 098(34.3%) 0.863(0.798–0.933) *** Divorced 1 105(4.4%) 6.8 -1.38(-3.14, 0.39) 378(34.2%) 0.994(0.934–1.058) Widowed 4 004(15.8%) 10.0 1.82(0.58–3.07) *** 2 517(62.9%) 0.959(0.947–0.971) *** Common Law Living 5(0.02%) 7.8 -0.39(-22.34, 21.55) 0(0%) No deaths Unmarried 5 184(20.5%) 6.0 -2.66(-3.34, 0.96) *** 1 830(35.3%) 0.973(0.956–0.991) ** Separated 284(1.1%) 10.5 2.26(-0.80, 5.33) 78(27.5%) 0.983(0.945–1.022) Unknown 241(1.0%) 7.2 -1.00(-4.31, 2.30) 0(0%) No deaths * p < 0.05 ** p < 0.01 *** p < 001

4 Mortality is Higher in Heart Failure Patients Discharged from Hospital on Weekends

Introduction Heart failure is common and one of the top contributors to mortality in the United Kingdom. There is some highly controversial evidence to support a ‘‘weekend-effect’’ with mortality rates elevated for patients admitted or discharged from hospital on the weekend. The impact of weekend discharge on mortality rates of heart failure patients in the United Kingdom has not been previously studied though, and was the focus of this study. We did this using a large database of patients discharged with heart failure to hospitals in the North of England, UK between 2000–2013. Methods Anonymous information on patients with heart failure was obtained from hospitals in North England, UK between 1st January 2000 and 31st March 2013. This data was analysed according to the ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) study protocol. ICD-10 and OPCS-4 codes were used to trace patients coded for heart failure, day of discharge, patient demographics, prevalence of comorbidites and mortality data. The impact of day of discharge on mortality in heart failure patients was analysed by Kaplan-meier survival analysis and cox regression analysis adjusted for age, gender, ethnicity and the 9 top contributors to mortality (Ischaemic Heart Disease, colon cancer, breast cancer, lung cancer, pneumonia, cerebrovascular disease, Chonic Obstructive Pulmonary Disease and dementia). P values of <0.05 were taken as statistically significant. Results Of 929552 patients admitted during the study period there were 31760 patients with heart failure. 27944 (88.8%) of patients were discharged on weekdays and 3816 (11.2%) were discharged on weekends. Demographics for patients discharged on weekdays (mean age 74 ± 14 and 50.3% male) was similar to those discharged on weekends (74 ± 14 and 50.1% male). Similarly, prevalence of the 9 top contributing conditions to mortality in the UK were similar in both groups (see Table 1). Crude all-cause mortality for heart failure patients discharged on weekends (69.2%) was higher than those discharged on weekdays (66.0%). Mortality was statistically greater for heart failure patients discharged on weekends after cox regression analysis accounting for differences in age, gender, ethnicity and the top contributors to mortality (OR 1.270, 95% confidence intervals 1.219–1.323). Kaplan-Meier survival analysis demonstrated that although 5-year survival for heart failure patients discharged on Monday-Friday were similar, survival for those discharged on Saturday/Sunday was greatly reduced.Abstract 4 Table 1 Demographics of heart failure patients discharged on weekdays and weekends Conclusions We have demonstrated that long-term mortality of heart failure patients discharged on weekends is significantly higher than those discharged on weekdays. Further research is required to elucidate the reasons for these disparities and could relate to premature discharge or lack of community care for heart failure patients discharged on weekends.Abstract 4 Table 2 Prevalence of top contributors to mortality amongst heart failure patients discharged on weekdays and weekendsAbstract 4 Figure 1 5 year survival depending on day of discharge

105 Impact of CV Risk Factors/Disease on Length of Stay and Mortality in Patients Presenting with MI

Background Classical risk factors for cardiovascular disease such as hypertension and diabetes, and their association with myocardial infarction (MI), have been thoroughly investigated. However, more research is needed to investigate the correlation between these risk factors and the impact on length of stay (LOS) and mortality in patients presenting with MI, which was the aim of this study. Methods We reviewed anonymous patient data including demographics, LOS, prevalence of cardiovascular comorbidities, and mortality during 25,287 consecutive admissions for MI from seven hospitals in the North West of England between 1 January 2000 and 31 March 2013. The ACALM (Algorithm for Comorbidities, Associations, Length of stay and Mortality) protocol, using ICD-10 and OPCS-4 coding systems, was used to track patient data. LOS and mortality of MI patients with and without cardiovascular comorbidities was compared by multinomial logistic regression. P values <0.05 were taken as statistically significant. Results Of 25,287 patients admitted with MI over the study period, mean (± SD) age was 66.6 ± 14.3 and 64.2% were male. The mean (± SD) LOS was 7.0 ± 16.2 days and there were a total of 9,653 (38.2%) deaths. The classical cardiovascular risk factors hypertension and hyperlipidaemia were associated with a decreased LOS and mortality (7.0 and 4.8 days respectively, P < 0.001; 36.8% OR 0.72 [95% CI 0.67–0.77] and 19.4% OR 0.42 [95% CI 0.39–0.46] respectively, P < 0.001), whereas diabetes was associated with a longer LOS and higher mortality (7.8 days, P < 0.05; 44.4% OR 1.3 [95% CI 1.20–1.41], P < 0.001). Angina pectoris was associated with shorter LOS and reduced mortality (5.4 days; 33.5% OR 0.75 [95% CI 0.68–0.82], P < 0.001). Other concomitant cardiovascular diseases were associated with an increased LOS and mortality: PVD (8.6 days, P < 0.05; 53% OR 1.93 [95% CI 1.68–2.21], P < 0.001), AF (10.9 days; 63.5% OR 1.51 [95% CI 1.38–1.66], P < 0.001), Cerebrovascular disease (15.9 days; 76% OR 2.29 [95% CI 1.67–3.15], P < 0.001), HF (11 days; 69.9% OR 3.28 [95% CI 3.03–3.56], P < 0.001), and IHD (6.7 days, P < 0.001; 38.7% OR 1.16 [95% CI 1.06–1.26], P < 0.05). Conclusion Cardiovascular risk factors and concomitant disease have a significant impact on LOS and mortality in patients presenting with MI. The presence of these diseases should be used to identify patients at an increased risk of prolonged admissions and death post MI, and services should be directed accordingly.Abstract 105 Table 1 Characteristics of patients presenting with ACSAbstract 105 Table 2 Impact of cardiovascular risk factors and disease on LOS and death in 25,287 consecutive patients presenting with ACS

21 Impact of Combined Atrial Fibrillation and Heart Failure on Mortality: 14 Year Naturalistic Follow-Up Study

Background Atrial Fibrillation (AF) and Heart Failure (HF) frequently co-exist conferring considerable morbidity and mortality, yet current treatment options remain limited. Recent meta-analyses of patients with concomitant AF and HF have suggested no prognostic benefit of beta-blockers or digoxin, creating a paradox whereby those most in need have the fewest therapeutic choices. We sought to investigate the association between HF and AF and their impact on mortality from a large 14-year naturalistic follow-up study. Methods Anonymous data of adult patients aged ≥18 with all types of HF and AF admitted to several hospitals in the North of England between 2000 and 2013 was obtained and processed using the ACALM (Algorithm for Co-morbidity, Associations, Length of stay and Mortality) study protocol. ACALM uses the ICD-10 and OPCS-4 coding systems to identify patients and the methodology has been published widely. Analyses were performed comparing mortality between patients with HF, AF and combined HF and AF at baseline and their development during follow-up. Results At baseline, of 929,552 adult patients 29,164 (3.1%) had AF, 19,474 (2.1%) had HF, and 5,728 (0.6%) had both HF and AF. Of those with AF at baseline, 1,647 (5.6%) developed HF during follow-up, and of those with HF at baseline, 824 (4.2%) developed AF during follow-up. Demographics and crude mortality rates are shown; see Table. Patients with combined AF and HF at baseline had increased mortality than patients with AF or HF alone. Patients with AF at baseline that developed HF, and patients with HF at baseline that developed AF, experienced a greater mortality compared to those with combined HF and AF at baseline; see Figure.Abstract 21 Table 1 Characteristics of ACLM participants by atrial fibrillation and heart failure status at the start and end of the study Baseline AF Baseline HF BaselineAF + HF Baseline AF,developed HF Baseline HF,developed AF n 29164 19474 5728 1647 824 Age ± SD 74 ± 13 73 ± 14 77 ± 12 77 ± 11 77 ± 11 Male% 52.4 51.0 49.8 48.0 49.3 Caucasian% 87.6 82.9 86.8 92.2 90.2 South Asian% 1.7 4.0 1.8 1.6 2.4 Afro-Caribbean% 0.8 1.6 1.2 0.4 0.6 Other% 9.9 11.8 10.2 5.8 6.8 Crude Mortality (per 1000) 485 636 672 715 778 Mean Survival (Days) 722 631 692 880 922 AF, atrial fibrillation; HF, heart failure; SD, standard deviationAbstract 21 Figure 1 Kaplan Meier survival curves through the duration of follow-up in the ACALM database are shown based on atrial fibrillation and heart failure status at the start and end of the study. AF, atrial fibrillation; HF, heart failure Conclusion Concomitant AF and HF is associated with substantial mortality and risk of death, irrespective of which disease develops first. In light of limited current treatment for these patients, future therapies to specifically target the combined HF and AF group are required.