PMCC PMCC

Search tips
Search criteria

Advanced
Results 1-25 (55)
 

Clipboard (0)
None

Select a Filter Below

Year of Publication
more »
1.  Type 2 diabetes and incidence of cardiovascular diseases: a cohort study in 1·9 million people 
Summary
Background
The contemporary associations of type 2 diabetes with a wide range of incident cardiovascular diseases have not been compared. We aimed to study associations between type 2 diabetes and 12 initial manifestations of cardiovascular disease.
Methods
We used linked primary care, hospital admission, disease registry, and death certificate records from the CALIBER programme, which links data for people in England recorded in four electronic health data sources. We included people who were (or turned) 30 years or older between Jan 1, 1998, to March 25, 2010, who were free from cardiovascular disease at baseline. The primary endpoint was the first record of one of 12 cardiovascular presentations in any of the data sources. We compared cumulative incidence curves for the initial presentation of cardiovascular disease and used Cox models to estimate cause-specific hazard ratios (HRs). This study is registered at ClinicalTrials.gov (NCT01804439).
Findings
Our cohort consisted of 1 921 260 individuals, of whom 1 887 062 (98·2%) did not have diabetes and 34 198 (1·8%) had type 2 diabetes. We observed 113 638 first presentations of cardiovascular disease during a median follow-up of 5·5 years (IQR 2·1–10·1). Of people with type 2 diabetes, 6137 (17·9%) had a first cardiovascular presentation, the most common of which were peripheral arterial disease (reported in 992 [16·2%] of 6137 patients) and heart failure (866 [14·1%] of 6137 patients). Type 2 diabetes was positively associated with peripheral arterial disease (adjusted HR 2·98 [95% CI 2·76–3·22]), ischaemic stroke (1·72 [1·52–1·95]), stable angina (1·62 [1·49–1·77]), heart failure (1·56 [1·45–1·69]), and non-fatal myocardial infarction (1·54 [1·42–1·67]), but was inversely associated with abdominal aortic aneurysm (0·46 [0·35–0·59]) and subarachnoid haemorrhage (0·48 [0·26–0.89]), and not associated with arrhythmia or sudden cardiac death (0·95 [0·76–1·19]).
Interpretation
Heart failure and peripheral arterial disease are the most common initial manifestations of cardiovascular disease in type 2 diabetes. The differences between relative risks of different cardiovascular diseases in patients with type 2 diabetes have implications for clinical risk assessment and trial design.
Funding
Wellcome Trust, National Institute for Health Research, and Medical Research Council.
doi:10.1016/S2213-8587(14)70219-0
PMCID: PMC4303913  PMID: 25466521
2.  The science of clinical practice: disease diagnosis or patient prognosis? Evidence about “what is likely to happen” should shape clinical practice 
BMC Medicine  2015;13:20.
Background
Diagnosis is the traditional basis for decision-making in clinical practice. Evidence is often lacking about future benefits and harms of these decisions for patients diagnosed with and without disease. We propose that a model of clinical practice focused on patient prognosis and predicting the likelihood of future outcomes may be more useful.
Discussion
Disease diagnosis can provide crucial information for clinical decisions that influence outcome in serious acute illness. However, the central role of diagnosis in clinical practice is challenged by evidence that it does not always benefit patients and that factors other than disease are important in determining patient outcome. The concept of disease as a dichotomous ‘yes’ or ‘no’ is challenged by the frequent use of diagnostic indicators with continuous distributions, such as blood sugar, which are better understood as contributing information about the probability of a patient’s future outcome. Moreover, many illnesses, such as chronic fatigue, cannot usefully be labelled from a disease-diagnosis perspective. In such cases, a prognostic model provides an alternative framework for clinical practice that extends beyond disease and diagnosis and incorporates a wide range of information to predict future patient outcomes and to guide decisions to improve them. Such information embraces non-disease factors and genetic and other biomarkers which influence outcome.
Summary
Patient prognosis can provide the framework for modern clinical practice to integrate information from the expanding biological, social, and clinical database for more effective and efficient care.
doi:10.1186/s12916-014-0265-4
PMCID: PMC4311412  PMID: 25637245
Clinical decision-making; Contested diagnoses; Diagnosis; Evidence-based medicine; Information; Outcomes of care; Overdiagnosis; Prognosis; Stratified medicine
3.  Defining Disease Phenotypes Using National Linked Electronic Health Records: A Case Study of Atrial Fibrillation 
PLoS ONE  2014;9(11):e110900.
Background
National electronic health records (EHR) are increasingly used for research but identifying disease cases is challenging due to differences in information captured between sources (e.g. primary and secondary care). Our objective was to provide a transparent, reproducible model for integrating these data using atrial fibrillation (AF), a chronic condition diagnosed and managed in multiple ways in different healthcare settings, as a case study.
Methods
Potentially relevant codes for AF screening, diagnosis, and management were identified in four coding systems: Read (primary care diagnoses and procedures), British National Formulary (BNF; primary care prescriptions), ICD-10 (secondary care diagnoses) and OPCS-4 (secondary care procedures). From these we developed a phenotype algorithm via expert review and analysis of linked EHR data from 1998 to 2010 for a cohort of 2.14 million UK patients aged ≥30 years. The cohort was also used to evaluate the phenotype by examining associations between incident AF and known risk factors.
Results
The phenotype algorithm incorporated 286 codes: 201 Read, 63 BNF, 18 ICD-10, and four OPCS-4. Incident AF diagnoses were recorded for 72,793 patients, but only 39.6% (N = 28,795) were recorded in primary care and secondary care. An additional 7,468 potential cases were inferred from data on treatment and pre-existing conditions. The proportion of cases identified from each source differed by diagnosis age; inferred diagnoses contributed a greater proportion of younger cases (≤60 years), while older patients (≥80 years) were mainly diagnosed in SC. Associations of risk factors (hypertension, myocardial infarction, heart failure) with incident AF defined using different EHR sources were comparable in magnitude to those from traditional consented cohorts.
Conclusions
A single EHR source is not sufficient to identify all patients, nor will it provide a representative sample. Combining multiple data sources and integrating information on treatment and comorbid conditions can substantially improve case identification.
doi:10.1371/journal.pone.0110900
PMCID: PMC4219705  PMID: 25369203
4.  Socioeconomic Deprivation and the Incidence of 12 Cardiovascular Diseases in 1.9 Million Women and Men: Implications for Risk Prediction and Prevention 
PLoS ONE  2014;9(8):e104671.
Background
Recent experimental evidence suggests that socioeconomic characteristics of neighbourhoods influence cardiovascular health, but observational studies which examine deprivation across a wide range of cardiovascular diseases (CVDs) are lacking.
Methods
Record-linkage cohort study of 1.93 million people to examine the association between small-area socioeconomic deprivation and 12 CVDs. Health records covered primary care, hospital admissions, a myocardial infarction registry and cause-specific mortality in England (CALIBER). Patients were aged ≥30 years and were initially free of CVD. Cox proportional hazard models stratified by general practice were used.
Findings
During a median follow-up of 5.5 years 114,859 people had one of 12 initial CVD presentations. In women the hazards of all CVDs except abdominal aortic aneurysm increased linearly with higher small-area socioeconomic deprivation (adjusted HR for most vs. least deprived ranged from 1.05, 95%CI 0.83–1.32 for abdominal aortic aneurysm to 1.55, 95%CI 1.42–1.70 for heart failure; I2 = 81.9%, τ2 = 0.01). In men heterogeneity was higher (HR ranged from 0.89, 95%CI 0.75–1.06 for cardiac arrest to 1.85, 95%CI 1.67–2.04 for peripheral arterial disease; I2 = 96.0%, τ2 = 0.06) and no association was observed with stable angina, sudden cardiac death, subarachnoid haemorrhage, transient ischaemic attack and abdominal aortic aneurysm. Lifetime risk difference between least and most deprived quintiles was most marked for peripheral arterial disease in women (4.3% least deprived, 5.8% most deprived) and men (4.6% least deprived, 7.8% in most deprived); but it was small or negligible for sudden cardiac death, transient ischaemic attack, abdominal aortic aneurysm and ischaemic and intracerebral haemorrhage, in both women and men.
Conclusions
Associations of small-area socioeconomic deprivation with 12 types of CVDs were heterogeneous, and in men absent for several diseases. Findings suggest that policies to reduce deprivation may impact more strongly on heart failure and peripheral arterial disease, and might be more effective in women.
doi:10.1371/journal.pone.0104671
PMCID: PMC4140710  PMID: 25144739
5.  Association between clinical presentations before myocardial infarction and coronary mortality: a prospective population-based study using linked electronic records 
European Heart Journal  2014;35(35):2363-2371.
Background
Ischaemia in different arterial territories before acute myocardial infarction (AMI) may influence post-AMI outcomes. No studies have evaluated prospectively collected information on ischaemia and its effect on short- and long-term coronary mortality. The objective of this study was to compare patients with and without prospectively measured ischaemic presentations before AMI in terms of infarct characteristics and coronary mortality.
Methods and results
As part of the CALIBER programme, we linked data from primary care, hospital admissions, the national acute coronary syndrome registry and cause-specific mortality to identify patients with first AMI (n = 16,439). We analysed time from AMI to coronary mortality (n = 5283 deaths) using Cox regression (median 2.6 years follow-up), comparing patients with and without recent ischaemic presentations. Patients with ischaemic presentations in the 90 days before AMI experienced lower coronary mortality in the first 7 days after AMI compared with those with no prior ischaemic presentations, after adjusting for age, sex, smoking, diabetes, blood pressure and cardiovascular medications [HR: 0.64 (95% CI: 0.57–0.73) P < 0.001], but subsequent mortality was higher [HR: 1.42 (1.13–1.77) P = 0.001]. Patients with ischaemic presentations closer in time to AMI had the lowest seven day mortality (P-trend = 0.001).
Conclusion
In the first large prospective study of ischaemic presentations prior to AMI, we have shown that those occurring closest to AMI are associated with lower short-term coronary mortality following AMI, which could represent a natural ischaemic preconditioning effect, observed in a clinical setting.
Clinical trials registration
Clinicaltrials.gov identifier NCT01604486.
doi:10.1093/eurheartj/ehu286
PMCID: PMC4163194  PMID: 25038774
Myocardial infarction; Epidemiology; Ischaemia
6.  Improving the Transparency of Prognosis Research: The Role of Reporting, Data Sharing, Registration, and Protocols 
PLoS Medicine  2014;11(7):e1001671.
George Peat and colleagues review and discuss current approaches to transparency and published debates and concerns about efforts to standardize prognosis research practice, and make five recommendations.
Please see later in the article for the Editors' Summary
doi:10.1371/journal.pmed.1001671
PMCID: PMC4086727  PMID: 25003600
7.  Red Blood Cell Transfusion and Mortality in Trauma Patients: Risk-Stratified Analysis of an Observational Study 
PLoS Medicine  2014;11(6):e1001664.
Using a large multicentre cohort, Pablo Perel and colleagues evaluate the association of red blood cell transfusion with mortality according to the predicted risk of death for trauma patients.
Please see later in the article for the Editors' Summary
Background
Haemorrhage is a common cause of death in trauma patients. Although transfusions are extensively used in the care of bleeding trauma patients, there is uncertainty about the balance of risks and benefits and how this balance depends on the baseline risk of death. Our objective was to evaluate the association of red blood cell (RBC) transfusion with mortality according to the predicted risk of death.
Methods and Findings
A secondary analysis of the CRASH-2 trial (which originally evaluated the effect of tranexamic acid on mortality in trauma patients) was conducted. The trial included 20,127 trauma patients with significant bleeding from 274 hospitals in 40 countries. We evaluated the association of RBC transfusion with mortality in four strata of predicted risk of death: <6%, 6%–20%, 21%–50%, and >50%. For this analysis the exposure considered was RBC transfusion, and the main outcome was death from all causes at 28 days. A total of 10,227 patients (50.8%) received at least one transfusion. We found strong evidence that the association of transfusion with all-cause mortality varied according to the predicted risk of death (p-value for interaction <0.0001). Transfusion was associated with an increase in all-cause mortality among patients with <6% and 6%–20% predicted risk of death (odds ratio [OR] 5.40, 95% CI 4.08–7.13, p<0.0001, and OR 2.31, 95% CI 1.96–2.73, p<0.0001, respectively), but with a decrease in all-cause mortality in patients with >50% predicted risk of death (OR 0.59, 95% CI 0.47–0.74, p<0.0001). Transfusion was associated with an increase in fatal and non-fatal vascular events (OR 2.58, 95% CI 2.05–3.24, p<0.0001). The risk associated with RBC transfusion was significantly increased for all the predicted risk of death categories, but the relative increase was higher for those with the lowest (<6%) predicted risk of death (p-value for interaction <0.0001). As this was an observational study, the results could have been affected by different types of confounding. In addition, we could not consider haemoglobin in our analysis. In sensitivity analyses, excluding patients who died early; conducting propensity score analysis adjusting by use of platelets, fresh frozen plasma, and cryoprecipitate; and adjusting for country produced results that were similar.
Conclusions
The association of transfusion with all-cause mortality appears to vary according to the predicted risk of death. Transfusion may reduce mortality in patients at high risk of death but increase mortality in those at low risk. The effect of transfusion in low-risk patients should be further tested in a randomised trial.
Trial registration
www.ClinicalTrials.gov NCT01746953
Please see later in the article for the Editors' Summary
Editors' Summary
Background
Trauma—a serious injury to the body caused by violence or an accident—is a major global health problem. Every year, injuries caused by traffic collisions, falls, blows, and other traumatic events kill more than 5 million people (9% of annual global deaths). Indeed, for people between the ages of 5 and 44 years, injuries are among the top three causes of death in many countries. Trauma sometimes kills people through physical damage to the brain and other internal organs, but hemorrhage (serious uncontrolled bleeding) is responsible for 30%–40% of trauma-related deaths. Consequently, early trauma care focuses on minimizing hemorrhage (for example, by using compression to stop bleeding) and on restoring blood circulation after blood loss (health-care professionals refer to this as resuscitation). Red blood cell (RBC) transfusion is often used for the management of patients with trauma who are bleeding; other resuscitation products include isotonic saline and solutions of human blood proteins.
Why Was This Study Done?
Although RBC transfusion can save the lives of patients with trauma who are bleeding, there is considerable uncertainty regarding the balance of risks and benefits associated with this procedure. RBC transfusion, which is an expensive intervention, is associated with several potential adverse effects, including allergic reactions and infections. Moreover, blood supplies are limited, and the risks from transfusion are high in low- and middle-income countries, where most trauma-related deaths occur. In this study, which is a secondary analysis of data from a trial (CRASH-2) that evaluated the effect of tranexamic acid (which stops excessive bleeding) in patients with trauma, the researchers test the hypothesis that RBC transfusion may have a beneficial effect among patients at high risk of death following trauma but a harmful effect among those at low risk of death.
What Did the Researchers Do and Find?
The CRASH-2 trail included 20,127 patients with trauma and major bleeding treated in 274 hospitals in 40 countries. In their risk-stratified analysis, the researchers investigated the effect of RBC transfusion on CRASH-2 participants with a predicted risk of death (estimated using a validated model that included clinical variables such as heart rate and blood pressure) on admission to hospital of less than 6%, 6%–20%, 21%–50%, or more than 50%. That is, the researchers compared death rates among patients in each stratum of predicted risk of death who received a RBC transfusion with death rates among patients who did not receive a transfusion. Half the patients received at least one transfusion. Transfusion was associated with an increase in all-cause mortality at 28 days after trauma among patients with a predicted risk of death of less than 6% or of 6%–20%, but with a decrease in all-cause mortality among patients with a predicted risk of death of more than 50%. In absolute figures, compared to no transfusion, RBC transfusion was associated with 5.1 more deaths per 100 patients in the patient group with the lowest predicted risk of death but with 11.9 fewer deaths per 100 patients in the group with the highest predicted risk of death.
What Do These Findings Mean?
These findings show that RBC transfusion is associated with an increase in all-cause deaths among patients with trauma and major bleeding with a low predicted risk of death, but with a reduction in all-cause deaths among patients with a high predicted risk of death. In other words, these findings suggest that the effect of RBC transfusion on all-cause mortality may vary according to whether a patient with trauma has a high or low predicted risk of death. However, because the participants in the CRASH-2 trial were not randomly assigned to receive a RBC transfusion, it is not possible to conclude that receiving a RBC transfusion actually increased the death rate among patients with a low predicted risk of death. It might be that the patients with this level of predicted risk of death who received a transfusion shared other unknown characteristics (confounders) that were actually responsible for their increased death rate. Thus, to provide better guidance for clinicians caring for patients with trauma and hemorrhage, the hypothesis that RBC transfusion could be harmful among patients with trauma with a low predicted risk of death should be prospectively evaluated in a randomised controlled trial.
Additional Information
Please access these websites via the online version of this summary at http://dx.doi.org/10.1371/journal.pmed.1001664.
This study is further discussed in a PLOS Medicine Perspective by Druin Burch
The World Health Organization provides information on injuries and on violence and injury prevention (in several languages)
The US Centers for Disease Control and Prevention has information on injury and violence prevention and control
The National Trauma Institute, a US-based non-profit organization, provides information about hemorrhage after trauma and personal stories about surviving trauma
The UK National Health Service Choices website provides information about blood transfusion, including a personal story about transfusion after a serious road accident
The US National Heart, Lung, and Blood Institute also provides detailed information about blood transfusions
MedlinePlus provides links to further resources on injuries, bleeding, and blood transfusion (in English and Spanish)
More information in available about CRASH-2 (in several languages)
doi:10.1371/journal.pmed.1001664
PMCID: PMC4060995  PMID: 24937305
8.  Genetic Variants at Chromosome 9p21 and Risk of First Versus Subsequent Coronary Heart Disease Events 
Objectives
The purpose of this analysis was to compare the association between variants at the chromosome 9p21 locus (Ch9p21) and risk of first versus subsequent coronary heart disease (CHD) events through systematic review and meta-analysis.
Background
Ch9p21 is a recognized risk factor for a first CHD event. However, its association with risk of subsequent events in patients with established CHD is less clear.
Methods
We searched PubMed and EMBASE for prospective studies reporting association of Ch9p21 with incident CHD events and extracted information on cohort type (individuals without prior CHD or individuals with established CHD) and effect estimates for risk of events.
Results
We identified 31 cohorts reporting on 193,372 individuals. Among the 16 cohorts of individuals without prior CHD (n = 168,209), there were 15,664 first CHD events. Ch9p21 was associated with a pooled hazard ratio (HR) of a first event of 1.19 (95% confidence interval: 1.17 to 1.22) per risk allele. In individuals with established CHD (n = 25,163), there were 4,436 subsequent events providing >99% and 91% power to detect a per-allele HR of 1.19 or 1.10, respectively. The pooled HR for subsequent events was 1.01 (95% confidence interval: 0.97 to 1.06) per risk allele. There was strong evidence of heterogeneity between the effect estimates for first and subsequent events (p value for heterogeneity = 5.6 × 10−11). We found no evidence for biases to account for these findings.
Conclusions
Ch9p21 shows differential association with risk of first versus subsequent CHD events. This has implications for genetic risk prediction in patients with established CHD and for mechanistic understanding of how Ch9p21 influences risk of CHD.
doi:10.1016/j.jacc.2014.01.065
PMCID: PMC4035794  PMID: 24607648
coronary heart disease; genomics; incident; subsequent; 9p21; Ch9p21, chromosome 9p21 locus; CHD, coronary heart disease; CI, confidence interval; HR, hazard ratio; MI, myocardial infarction
9.  Blood pressure and incidence of twelve cardiovascular diseases: lifetime risks, healthy life-years lost, and age-specific associations in 1·25 million people 
Lancet  2014;383(9932):1899-1911.
Summary
Background
The associations of blood pressure with the different manifestations of incident cardiovascular disease in a contemporary population have not been compared. In this study, we aimed to analyse the associations of blood pressure with 12 different presentations of cardiovascular disease.
Methods
We used linked electronic health records from 1997 to 2010 in the CALIBER (CArdiovascular research using LInked Bespoke studies and Electronic health Records) programme to assemble a cohort of 1·25 million patients, 30 years of age or older and initially free from cardiovascular disease, a fifth of whom received blood pressure-lowering treatments. We studied the heterogeneity in the age-specific associations of clinically measured blood pressure with 12 acute and chronic cardiovascular diseases, and estimated the lifetime risks (up to 95 years of age) and cardiovascular disease-free life-years lost adjusted for other risk factors at index ages 30, 60, and 80 years. This study is registered at ClinicalTrials.gov, number NCT01164371.
Findings
During 5·2 years median follow-up, we recorded 83 098 initial cardiovascular disease presentations. In each age group, the lowest risk for cardiovascular disease was in people with systolic blood pressure of 90–114 mm Hg and diastolic blood pressure of 60–74 mm Hg, with no evidence of a J-shaped increased risk at lower blood pressures. The effect of high blood pressure varied by cardiovascular disease endpoint, from strongly positive to no effect. Associations with high systolic blood pressure were strongest for intracerebral haemorrhage (hazard ratio 1·44 [95% CI 1·32–1·58]), subarachnoid haemorrhage (1·43 [1·25–1·63]), and stable angina (1·41 [1·36–1·46]), and weakest for abdominal aortic aneurysm (1·08 [1·00–1·17]). Compared with diastolic blood pressure, raised systolic blood pressure had a greater effect on angina, myocardial infarction, and peripheral arterial disease, whereas raised diastolic blood pressure had a greater effect on abdominal aortic aneurysm than did raised systolic pressure. Pulse pressure associations were inverse for abdominal aortic aneurysm (HR per 10 mm Hg 0·91 [95% CI 0·86–0·98]) and strongest for peripheral arterial disease (1·23 [1·20–1·27]). People with hypertension (blood pressure ≥140/90 mm Hg or those receiving blood pressure-lowering drugs) had a lifetime risk of overall cardiovascular disease at 30 years of age of 63·3% (95% CI 62·9–63·8) compared with 46·1% (45·5–46·8) for those with normal blood pressure, and developed cardiovascular disease 5·0 years earlier (95% CI 4·8–5·2). Stable and unstable angina accounted for most (43%) of the cardiovascular disease-free years of life lost associated with hypertension from index age 30 years, whereas heart failure and stable angina accounted for the largest proportion (19% each) of years of life lost from index age 80 years.
Interpretation
The widely held assumptions that blood pressure has strong associations with the occurrence of all cardiovascular diseases across a wide age range, and that diastolic and systolic associations are concordant, are not supported by the findings of this high-resolution study. Despite modern treatments, the lifetime burden of hypertension is substantial. These findings emphasise the need for new blood pressure-lowering strategies, and will help to inform the design of randomised trials to assess them.
Funding
Medical Research Council, National Institute for Health Research, and Wellcome Trust.
doi:10.1016/S0140-6736(14)60685-1
PMCID: PMC4042017  PMID: 24881994
10.  Acute myocardial infarction: a comparison of short-term survival in national outcome registries in Sweden and the UK 
Lancet  2014;383(9925):1305-1312.
Summary
Background
International research for acute myocardial infarction lacks comparisons of whole health systems. We assessed time trends for care and outcomes in Sweden and the UK.
Methods
We used data from national registries on consecutive patients registered between 2004 and 2010 in all hospitals providing care for acute coronary syndrome in Sweden and the UK. The primary outcome was all-cause mortality 30 days after admission. We compared effectiveness of treatment by indirect casemix standardisation. This study is registered with ClinicalTrials.gov, number NCT01359033.
Findings
We assessed data for 119 786 patients in Sweden and 391 077 in the UK. 30-day mortality was 7·6% (95% CI 7·4–7·7) in Sweden and 10·5% (10·4–10·6) in the UK. Mortality was higher in the UK in clinically relevant subgroups defined by troponin concentration, ST-segment elevation, age, sex, heart rate, systolic blood pressure, diabetes mellitus status, and smoking status. In Sweden, compared with the UK, there was earlier and more extensive uptake of primary percutaneous coronary intervention (59% vs 22%) and more frequent use of β blockers at discharge (89% vs 78%). After casemix standardisation the 30-day mortality ratio for UK versus Sweden was 1·37 (95% CI 1·30–1·45), which corresponds to 11 263 (95% CI 9620–12 827) excess deaths, but did decline over time (from 1·47, 95% CI 1·38–1·58 in 2004 to 1·20, 1·12–1·29 in 2010; p=0·01).
Interpretation
We found clinically important differences between countries in acute myocardial infarction care and outcomes. International comparisons research might help to improve health systems and prevent deaths.
Funding
Seventh Framework Programme for Research, National Institute for Health Research, Wellcome Trust (UK), Swedish Association of Local Authorities and Regions, Swedish Heart-Lung Foundation.
doi:10.1016/S0140-6736(13)62070-X
PMCID: PMC4255068  PMID: 24461715
11.  Prognostic models for stable coronary artery disease based on electronic health record cohort of 102 023 patients 
European heart journal  2013;35(13):844-852.
Aims
The population with stable coronary artery disease (SCAD) is growing but validated models to guide their clinical management are lacking. We developed and validated prognostic models for all-cause mortality and non-fatal myocardial infarction (MI) or coronary death in SCAD.
Methods and results
Models were developed in a linked electronic health records cohort of 102 023 SCAD patients from the CALIBER programme, with mean follow-up of 4.4 (SD 2.8) years during which 20 817 deaths and 8856 coronary outcomes were observed. The Kaplan–Meier 5-year risk was 20.6% (95% CI, 20.3, 20.9) for mortality and 9.7% (95% CI, 9.4, 9.9) for non-fatal MI or coronary death. The predictors in the models were age, sex, CAD diagnosis, deprivation, smoking, hypertension, diabetes, lipids, heart failure, peripheral arterial disease, atrial fibrillation, stroke, chronic kidney disease, chronic pulmonary disease, liver disease, cancer, depression, anxiety, heart rate, creatinine, white cell count, and haemoglobin. The models had good calibration and discrimination in internal (external) validation with C-index 0.811 (0.735) for all-cause mortality and 0.778 (0.718) for non-fatal MI or coronary death. Using these models to identify patients at high risk (defined by guidelines as 3% annual mortality) and support a management decision associated with hazard ratio 0.8 could save an additional 13-16 life years or 15-18 coronary event-free years per 1000 patients screened, compared with models with just age, sex, and deprivation.
Conclusion
These validated prognostic models could be used in clinical practice to support risk stratification as recommended in clinical guidelines.
doi:10.1093/eurheartj/eht533
PMCID: PMC3971383  PMID: 24353280
Stable coronary artery disease; Stable angina; Prognosis; Myocardial infarction; Electronic health records; CALIBER
12.  Comparison of Random Forest and Parametric Imputation Models for Imputing Missing Data Using MICE: A CALIBER Study 
American journal of epidemiology  2014;179(6):764-774.
Multivariate imputation by chained equations (MICE) is commonly used for imputing missing data in epidemiologic research. The “true” imputation model may contain nonlinearities which are not included in default imputation models. Random forest imputation is a machine learning technique which can accommodate nonlinearities and interactions and does not require a particular regression model to be specified. We compared parametric MICE with a random forest-based MICE algorithm in 2 simulation studies. The first study used 1,000 random samples of 2,000 persons drawn from the 10,128 stable angina patients in the CALIBER database (Cardiovascular Disease Research using Linked Bespoke Studies and Electronic Records; 2001–2010) with complete data on all covariates. Variables were artificially made “missing at random,” and the bias and efficiency of parameter estimates obtained using different imputation methods were compared. Both MICE methods produced unbiased estimates of (log) hazard ratios, but random forest was more efficient and produced narrower confidence intervals. The second study used simulated data in which the partially observed variable depended on the fully observed variables in a nonlinear way. Parameter estimates were less biased using random forest MICE, and confidence interval coverage was better. This suggests that random forest imputation may be useful for imputing complex epidemiologic data sets in which some patients have missing data.
doi:10.1093/aje/kwt312
PMCID: PMC3939843  PMID: 24589914
angina, stable; imputation; missing data; missingness at random; regression trees; simulation; survival
13.  Comparison of Random Forest and Parametric Imputation Models for Imputing Missing Data Using MICE: A CALIBER Study 
American Journal of Epidemiology  2014;179(6):764-774.
Multivariate imputation by chained equations (MICE) is commonly used for imputing missing data in epidemiologic research. The “true” imputation model may contain nonlinearities which are not included in default imputation models. Random forest imputation is a machine learning technique which can accommodate nonlinearities and interactions and does not require a particular regression model to be specified. We compared parametric MICE with a random forest-based MICE algorithm in 2 simulation studies. The first study used 1,000 random samples of 2,000 persons drawn from the 10,128 stable angina patients in the CALIBER database (Cardiovascular Disease Research using Linked Bespoke Studies and Electronic Records; 2001–2010) with complete data on all covariates. Variables were artificially made “missing at random,” and the bias and efficiency of parameter estimates obtained using different imputation methods were compared. Both MICE methods produced unbiased estimates of (log) hazard ratios, but random forest was more efficient and produced narrower confidence intervals. The second study used simulated data in which the partially observed variable depended on the fully observed variables in a nonlinear way. Parameter estimates were less biased using random forest MICE, and confidence interval coverage was better. This suggests that random forest imputation may be useful for imputing complex epidemiologic data sets in which some patients have missing data.
doi:10.1093/aje/kwt312
PMCID: PMC3939843  PMID: 24589914
angina, stable; imputation; missing data; missingness at random; regression trees; simulation; survival
14.  Prognostic models for stable coronary artery disease based on electronic health record cohort of 102 023 patients 
European Heart Journal  2013;35(13):844-852.
Aims
The population with stable coronary artery disease (SCAD) is growing but validated models to guide their clinical management are lacking. We developed and validated prognostic models for all-cause mortality and non-fatal myocardial infarction (MI) or coronary death in SCAD.
Methods and results
Models were developed in a linked electronic health records cohort of 102 023 SCAD patients from the CALIBER programme, with mean follow-up of 4.4 (SD 2.8) years during which 20 817 deaths and 8856 coronary outcomes were observed. The Kaplan–Meier 5-year risk was 20.6% (95% CI, 20.3, 20.9) for mortality and 9.7% (95% CI, 9.4, 9.9) for non-fatal MI or coronary death. The predictors in the models were age, sex, CAD diagnosis, deprivation, smoking, hypertension, diabetes, lipids, heart failure, peripheral arterial disease, atrial fibrillation, stroke, chronic kidney disease, chronic pulmonary disease, liver disease, cancer, depression, anxiety, heart rate, creatinine, white cell count, and haemoglobin. The models had good calibration and discrimination in internal (external) validation with C-index 0.811 (0.735) for all-cause mortality and 0.778 (0.718) for non-fatal MI or coronary death. Using these models to identify patients at high risk (defined by guidelines as 3% annual mortality) and support a management decision associated with hazard ratio 0.8 could save an additional 13–16 life years or 15–18 coronary event-free years per 1000 patients screened, compared with models with just age, sex, and deprivation.
Conclusion
These validated prognostic models could be used in clinical practice to support risk stratification as recommended in clinical guidelines.
doi:10.1093/eurheartj/eht533
PMCID: PMC3971383  PMID: 24353280
Stable coronary artery disease; Stable angina; Prognosis; Myocardial infarction; Electronic health records; CALIBER
15.  Type and timing of heralding in ST-elevation and non-ST-elevation myocardial infarction: an analysis of prospectively collected electronic healthcare records linked to the national registry of acute coronary syndromes 
Aims:
It is widely thought that ST-elevation myocardial infarction (STEMI) is more likely to occur without warning (i.e. an unanticipated event in a previously healthy person) than non-ST-elevation myocardial infarction (NSTEMI), but no large study has evaluated this using prospectively collected data. The aim of this study was to compare the evolution of atherosclerotic disease and cardiovascular risk between people going on to experience STEMI and NSTEMI.
Methods:
We identified patients experiencing STEMI and NSTEMI in the national registry of myocardial infarction for England and Wales (Myocardial Ischaemia National Audit Project), for whom linked primary care records were available in the General Practice Research Database (as part of the CALIBER collaboration). We compared the prevalence and timing of atherosclerotic disease and major cardiovascular risk factors including smoking, hypertension, diabetes, and dyslipidaemia, between patients later experiencing STEMI to those experiencing NSTEMI.
Results:
A total of 8174 myocardial infarction patients were included (3780 STEMI, 4394 NSTEMI). Myocardial infarction without heralding by previously diagnosed atherosclerotic disease occurred in 71% STEMI (95% CI 69–72%) and 50% NSTEMI patients (95% CI 48–51%). The proportions of myocardial infarctions with no prior atherosclerotic disease, major risk factors, or chest pain was 14% (95% CI 13–16%) in STEMI and 9% (95% CI 9–10%) in NSTEMI. The rate of heralding coronary diagnoses was particularly high in the 12 months before infarct; 4.1-times higher (95% CI 3.3–5.0) in STEMI and 3.6-times higher (95% CI 3.1–4.2) in NSTEMI compared to the rate in earlier years.
Conclusions:
Acute myocardial infarction occurring without prior diagnosed coronary, cerebrovascular, or peripheral arterial disease was common, especially for STEMI. However, there was a high prevalence of risk factors or symptoms in patients without previously diagnosed disease. Better understanding of the antecedents in the year before myocardial infarction is required.
doi:10.1177/2048872613487495
PMCID: PMC3821819  PMID: 24222835
Cardiovascular diseases; epidemiology; myocardial infarction; risk factors
16.  Prognosis Research Strategy (PROGRESS) 3: Prognostic Model Research 
PLoS Medicine  2013;10(2):e1001381.
In this article, the third in the PROGRESS series on prognostic factor research, Sara Schroter and colleagues review how prognostic models are developed and validated, and then address how prognostic models are assessed for their impact on practice and patient outcomes, illustrating these ideas with examples.
doi:10.1371/journal.pmed.1001381
PMCID: PMC3564751  PMID: 23393430
17.  Prognosis Research Strategy (PROGRESS) 2: Prognostic Factor Research 
PLoS Medicine  2013;10(2):e1001380.
In the second article in the PROGRESS series on prognostic factor research, Sara Schroter and colleagues discuss the role of prognostic factors in current clinical practice, randomised trials, and developing new interventions, and explain why and how prognostic factor research should be improved.
doi:10.1371/journal.pmed.1001380
PMCID: PMC3564757  PMID: 23393429
18.  Data Resource Profile: Cardiovascular disease research using linked bespoke studies and electronic health records (CALIBER) 
The goal of cardiovascular disease (CVD) research using linked bespoke studies and electronic health records (CALIBER) is to provide evidence to inform health care and public health policy for CVDs across different stages of translation, from discovery, through evaluation in trials to implementation, where linkages to electronic health records provide new scientific opportunities. The initial approach of the CALIBER programme is characterized as follows: (i) Linkages of multiple electronic heath record sources: examples include linkages between the longitudinal primary care data from the Clinical Practice Research Datalink, the national registry of acute coronary syndromes (Myocardial Ischaemia National Audit Project), hospitalization and procedure data from Hospital Episode Statistics and cause-specific mortality and social deprivation data from the Office of National Statistics. Current cohort analyses involve a million people in initially healthy populations and disease registries with ∼105 patients. (ii) Linkages of bespoke investigator-led cohort studies (e.g. UK Biobank) to registry data (e.g. Myocardial Ischaemia National Audit Project), providing new means of ascertaining, validating and phenotyping disease. (iii) A common data model in which routine electronic health record data are made research ready, and sharable, by defining and curating with meta-data >300 variables (categorical, continuous, event) on risk factors, CVDs and non-cardiovascular comorbidities. (iv) Transparency: all CALIBER studies have an analytic protocol registered in the public domain, and data are available (safe haven model) for use subject to approvals. For more information, e-mail s.denaxas@ucl.ac.uk
doi:10.1093/ije/dys188
PMCID: PMC3535749  PMID: 23220717
electronic heath records; linkages; cardiovascular
19.  Influenza Infection and Risk of Acute Myocardial Infarction in England and Wales: A CALIBER Self-Controlled Case Series Study 
The Journal of Infectious Diseases  2012;206(11):1652-1659.
Background. An association between infections and vascular events has been observed, but the specific effect of influenza and influenza-like illnesses on triggering acute myocardial infarction (AMI) is unclear.
Methods. Episodes of first AMI from 1 January 2003 through 31 July 2009 were identified using linked anonymized electronic medical records from the Myocardial Ischaemia National Audit Project and the General Practice Research Database. Self-controlled case series analysis was used to investigate AMI risks after consultation for acute respiratory infection. Infections were stratified by influenza virus circulation, diagnostic code, and vaccination status to assess whether influenza was more likely than other infections to trigger AMI.
Results. Of 22 024 patients with acute coronary syndrome, 11 208 met the criterion of having had their first AMI at the age of ≥40 years, and 3927 had also consulted for acute respiratory infection. AMI risks were significantly raised during days 1–3 after acute respiratory infection (incidence ratio, 4.19 [95% confidence interval, 3.18–5.53], with the effect tapering over time. The effect was greatest in those aged ≥80 years (P = .023). Infections occurring when influenza was circulating and those coded as influenza-like illness were associated with consistently higher incidence ratios for AMI (P = .012).
Conclusions. Influenza and other acute respiratory infections can act as a trigger for AMI. This effect may be stronger for influenza than for other infections.
Clinical trials registration. NCT01106196.
doi:10.1093/infdis/jis597
PMCID: PMC3488196  PMID: 23048170
20.  The freetext matching algorithm: a computer program to extract diagnoses and causes of death from unstructured text in electronic health records 
Background
Electronic health records are invaluable for medical research, but much information is stored as free text rather than in a coded form. For example, in the UK General Practice Research Database (GPRD), causes of death and test results are sometimes recorded only in free text. Free text can be difficult to use for research if it requires time-consuming manual review. Our aim was to develop an automated method for extracting coded information from free text in electronic patient records.
Methods
We reviewed the electronic patient records in GPRD of a random sample of 3310 patients who died in 2001, to identify the cause of death. We developed a computer program called the Freetext Matching Algorithm (FMA) to map diagnoses in text to the Read Clinical Terminology. The program uses lookup tables of synonyms and phrase patterns to identify diagnoses, dates and selected test results. We tested it on two random samples of free text from GPRD (1000 texts associated with death in 2001, and 1000 general texts from cases and controls in a coronary artery disease study), comparing the output to the U.S. National Library of Medicine’s MetaMap program and the gold standard of manual review.
Results
Among 3310 patients registered in the GPRD who died in 2001, the cause of death was recorded in coded form in 38.1% of patients, and in the free text alone in 19.4%. On the 1000 texts associated with death, FMA coded 683 of the 735 positive diagnoses, with precision (positive predictive value) 98.4% (95% confidence interval (CI) 97.2, 99.2) and recall (sensitivity) 92.9% (95% CI 90.8, 94.7). On the general sample, FMA detected 346 of the 447 positive diagnoses, with precision 91.5% (95% CI 88.3, 94.1) and recall 77.4% (95% CI 73.2, 81.2), which was similar to MetaMap.
Conclusions
We have developed an algorithm to extract coded information from free text in GP records with good precision. It may facilitate research using free text in electronic patient records, particularly for extracting the cause of death.
doi:10.1186/1472-6947-12-88
PMCID: PMC3483188  PMID: 22870911
21.  Is access to specialist assessment of chest pain equitable by age, gender, ethnicity and socioeconomic status? An enhanced ecological analysis 
BMJ Open  2012;2(3):e001025.
Objectives
To determine whether access to rapid access chest pain clinics of people with recent onset symptoms is equitable by age, socioeconomic status, ethnicity and gender, according to need.
Design
Retrospective cohort study with ecological analysis.
Setting
Patients referred from primary care to five rapid access chest pain clinics in secondary care, across England.
Participants
Of 8647 patients aged ≥35 years referred to chest pain clinics with new-onset stable chest pain but no known cardiac history, 7570 with documented census ward codes, age, gender and ethnicity comprised the study group. Patients excluded were those with missing date of birth, gender or ethnicity (n=782) and those with missing census ward codes (n=295).
Outcome measures
Effects of age, gender, ethnicity and socioeconomic status on clinic attendance were calculated as attendance rate ratios, with number of attendances as the outcome and resident population-years as the exposure in each stratum, using Poisson regression. Attendance rate ratios were then compared with coronary heart disease (CHD) mortality ratios to determine whether attendance was equitable according to need.
Results
Adjusted attendance rate ratios for patients aged >65 years were similar to younger patients (1.1, 95% CI 1.05 to 1.16), despite population CHD mortality rate ratios nearly 15 times higher in the older age group. Women had lower attendance rate ratios (0.81, 95% CI 0.77 to 0.84) and also lower population CHD mortality rate ratios compared with men. South Asians had higher attendance rates (1.67, 95% CI 1.57 to 1.77) compared with whites and had a higher standardised CHD mortality ratio of 1.46 (95% CI 1.41 to 1.51). Although univariable analysis showed that the most deprived patients (quintile 5) had an attendance rate twice that of less deprived quintiles, the adjusted analysis showed their attendance to be 13% lower (0.87, 95% CI 0.81 to 0.94) despite a higher population CHD mortality rate.
Conclusion
There is evidence of underutilisation of chest pain clinics by older people and those from lower socioeconomic status. More robust and patient focused administrative pathways need to be developed to detect inequity, correction of which has the potential to substantially reduce coronary mortality.
Article summary
Article focus
Is access to chest pain clinics of people with recent onset symptoms equitable according to local need and consistent with national policy.
Key messages
Need for evaluation in chest pain clinics will vary according to the variable incidence of heart disease in different age, gender, socioeconomic and ethnic groups.
There is evidence of underutilisation of chest pain clinics by older people and those from lower socioeconomic status.
Strengths and limitations of this study
Large, diverse and unselected patient population with uniformly collected patient-level data, allowing robust comparisons between demographic and clinical groups.
Ecological fallacy with respect to age and sex has been avoided by applying an enhanced ecological analysis.
Need to use census wards, not postcodes, as the smallest geographical areas for which mortality and demographic data were available.
Ethnicity was not based on self-ascription.
doi:10.1136/bmjopen-2012-001025
PMCID: PMC3378943  PMID: 22700834
22.  Treading carefully: a qualitative ethnographic study of the clinical, social and educational uses of exercise ECG in evaluating stable chest pain 
BMJ Open  2012;2(1):e000508.
Objective
To examine functions of the exercise ECG in the light of the recent National Institute for Health and Clinical Excellence guidelines recommending that it should not be used for the diagnosis or exclusion of stable angina.
Design
Qualitative ethnographic study based on interviews and observations of clinical practice.
Setting
3 rapid access chest pain clinics in England.
Participants
Observation of 89 consultations in chest pain clinics, 18 patient interviews and 12 clinician interviews.
Main outcome measure
Accounts and observations of consultations in chest pain clinics.
Results
The exercise ECG was observed to have functions that extended beyond diagnosis. It was used to clarify a patient's story and revise the initial account. The act of walking on the treadmill created an additional opportunity for dialogue between clinician and patient and engagement of the patient in the diagnostic process through precipitation of symptoms and further elaboration of symptoms. The exercise ECG facilitated reassurance in relation to exercise capacity and tolerance, providing a platform for behavioural advice particularly when exercise was promoted by the clinician.
Conclusions
Many of the practices that have been built up around the use of the exercise ECG are potentially beneficial to patients and need to be considered in the re-design of services without that test. Through its contribution to the patient's history and to subsequent advice to the patient, the exercise ECG continues to inform the specialist assessment and management of patients with new onset stable chest pain, beyond its now marginalised role in diagnosis.
Article summary
Article focus
Given the widespread use of the exercise ECG in assessments of patients with stable chest pain, this paper seeks to understand its role in the light of emerging evidence about its poor performance as a diagnostic test.
This paper reports on the functions of the exercise ECG in UK chest pain clinics, highlighting those uses that go beyond its diagnostic function.
This paper is part of an international debate about the appropriate initial tests for patients with new onset stable chest pain.
Key messages
The exercise ECG has additional functions that transcend its technical contribution to diagnosis: it can help clarify symptoms and other aspects of the clinical history, engage the patient in the diagnostic process, provide a context for guidance on reversible cardiovascular risk factors, be used to better involve and reassure patients and has the potential use for tailored lifestyle advice.
Through its contribution to the patient's history and to subsequent advice to the patient, the exercise ECG continues to inform the specialist assessment and management of patients with new onset stable chest pain, beyond its now marginalised role in diagnosis.
Many of the practices that have been built up around the use of the exercise ECG are potentially beneficial to patients. As chest pain clinic services are re-configured without the test, in line with UK national (National Institute for Health and Clinical Excellence (NICE)) guidance, these practices need to be integrated into new diagnostic pathways.
Strengths and limitations of this study
A strength of our study is its ethnographic design incorporating the observation of patient–clinician consultations and combining these data with interviews: we knew what participants did as well as said.
The fieldwork was undertaken at a key time just before the introduction of the UK's 2010 NICE guidelines and therefore provides an understanding of current practice that can inform their implementation.
A limitation of our study is that data were collected largely from two chest pain clinics, potentially limiting the transferability of the findings, although the clinicians in the research team thought that the clinics were not atypical compared to others they had experienced.
doi:10.1136/bmjopen-2011-000508
PMCID: PMC3277903  PMID: 22318662
23.  Extracting Diagnoses and Investigation Results from Unstructured Text in Electronic Health Records by Semi-Supervised Machine Learning 
PLoS ONE  2012;7(1):e30412.
Background
Electronic health records are invaluable for medical research, but much of the information is recorded as unstructured free text which is time-consuming to review manually.
Aim
To develop an algorithm to identify relevant free texts automatically based on labelled examples.
Methods
We developed a novel machine learning algorithm, the ‘Semi-supervised Set Covering Machine’ (S3CM), and tested its ability to detect the presence of coronary angiogram results and ovarian cancer diagnoses in free text in the General Practice Research Database. For training the algorithm, we used texts classified as positive and negative according to their associated Read diagnostic codes, rather than by manual annotation. We evaluated the precision (positive predictive value) and recall (sensitivity) of S3CM in classifying unlabelled texts against the gold standard of manual review. We compared the performance of S3CM with the Transductive Vector Support Machine (TVSM), the original fully-supervised Set Covering Machine (SCM) and our ‘Freetext Matching Algorithm’ natural language processor.
Results
Only 60% of texts with Read codes for angiogram actually contained angiogram results. However, the S3CM algorithm achieved 87% recall with 64% precision on detecting coronary angiogram results, outperforming the fully-supervised SCM (recall 78%, precision 60%) and TSVM (recall 2%, precision 3%). For ovarian cancer diagnoses, S3CM had higher recall than the other algorithms tested (86%). The Freetext Matching Algorithm had better precision than S3CM (85% versus 74%) but lower recall (62%).
Conclusions
Our novel S3CM machine learning algorithm effectively detected free texts in primary care records associated with angiogram results and ovarian cancer diagnoses, after training on pre-classified test sets. It should be easy to adapt to other disease areas as it does not rely on linguistic rules, but needs further testing in other electronic health record datasets.
doi:10.1371/journal.pone.0030412
PMCID: PMC3261909  PMID: 22276193
24.  Clopidogrel discontinuation after acute coronary syndromes: frequency, predictors and associations with death and myocardial infarction—a hospital registry-primary care linked cohort (MINAP–GPRD) 
European Heart Journal  2011;32(19):2376-2386.
Aims
Adherence to evidence-based treatments and its consequences after acute myocardial infarction (MI) are poorly defined. We examined the extent to which clopidogrel treatment initiated in hospital is continued in primary care; the factors predictive of clopidogrel discontinuation and the hazard of death or recurrent MI.
Methods and results
We linked the Myocardial Ischaemia National Audit Project registry and the General Practice Research Database to examine adherence to clopidogrel in primary care among patients discharged from hospital after MI (2003–2009). Hospital Episode Statistics and national mortality data were linked, documenting all-cause mortality and non-fatal MI. Of the 7543 linked patients, 4650 were prescribed clopidogrel in primary care within 3 months of discharge. The adjusted odds of still being prescribed clopidogrel at 12 months were similar following non-ST-elevation myocardial infarction (NSTEMI) 53% (95% CI, 51–55) and ST-elevation myocardial infarction (STEMI) 54% (95% CI, 52–56), but contrast with statins: NSTEMI 84% (95% CI, 82–85) and STEMI 89% (95% CI, 87–90). Discontinuation within 12 months was more frequent in older patients [>80 vs. 40–49 years, adjusted hazard ratio (HR) 1.50 (95% CI, 1.15–1.94)] and with bleeding events [HR 1.34 (95% CI, 1.03–1.73)]. 18.15 patients per 100 person-years (95% CI, 16.83–19.58) died or experienced non-fatal MI in the first year following discharge. In patients who discontinued clopidogrel within 12 months, the adjusted HR for death or non-fatal MI was 1.45 (95% CI, 1.22–1.73) compared with untreated patients, and 2.62 (95% CI, 2.17–3.17) compared with patients persisting with clopidogrel treatment.
Conclusion
This is the first study to use linked registries to determine persistence of clopidogrel treatment after MI in primary care. It demonstrates that discontinuation is common and associated with adverse outcomes.
doi:10.1093/eurheartj/ehr340
PMCID: PMC3184230  PMID: 21875855
ACS; Myocardial infarction; Clopidogrel; Anti-platelet; Observational
25.  NICE clinical guideline: chest pain of recent onset 
doi:10.3399/bjgp10X515124
PMCID: PMC2913741  PMID: 20822694

Results 1-25 (55)