Owen Nicholas

Comparison of random forest and parametric imputation models for imputing missing data using MICE: a CALIBER study.

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.

Prognostic models for stable coronary artery disease based on electronic health record cohort of 102 023 patients

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.

Threshold haemoglobin levels and the prognosis of stable coronary disease: two new cohorts and a systematic review and meta-analysis

Anoop Shah and colleagues performed a retrospective cohort study and a systematic review, and show evidence that in people with stable coronary disease there were threshold hemoglobin values below which mortality increased in a graded, continuous fashion.

Neutrophil Counts and Initial Presentation of 12 Cardiovascular Diseases: A CALIBER Cohort Study

Background Neutrophil counts are a ubiquitous measure of inflammation, but previous studies on their association with cardiovascular disease (CVD) were limited by small numbers of patients or a narrow range of endpoints. Objectives This study investigated associations of clinically recorded neutrophil counts with initial presentation for a range of CVDs. Methods We used linked primary care, hospitalization, disease registry, and mortality data in England. We included people 30 years or older with complete blood counts performed in usual clinical care and no history of CVD. We used Cox models to estimate cause-specific hazard ratios (HRs) for 12 CVDs, adjusted for cardiovascular risk factors and acute conditions affecting neutrophil counts (such as infections and cancer). Results Among 775,231 individuals in the cohort, 154,179 had complete blood counts performed under acute conditions and 621,052 when they were stable. Over a median 3.8 years of follow-up, 55,004 individuals developed CVD. Adjusted HRs comparing neutrophil counts 6 to 7 versus 2 to 3 × 109/l (both within the ‘normal’ range) showed strong associations with heart failure (HR: 2.04; 95% confidence interval [CI]: 1.82 to 2.29), peripheral arterial disease (HR: 1.95; 95% CI: 1.72 to 2.21), unheralded coronary death (HR: 1.78; 95% CI: 1.51 to 2.10), abdominal aortic aneurysm (HR: 1.72; 95% CI: 1.34 to 2.21), and nonfatal myocardial infarction (HR: 1.58; 95% CI: 1.42 to 1.76). These associations were linear, with greater risk even among individuals with neutrophil counts of 3 to 4 versus 2 to 3 × 109/l. There was a weak association with ischemic stroke (HR: 1.36; 95% CI: 1.17 to 1.57), but no association with stable angina or intracerebral hemorrhage. Conclusions Neutrophil counts were strongly associated with the incidence of some CVDs, but not others, even within the normal range, consistent with underlying disease mechanisms differing across CVDs. (White Blood Cell Counts and Onset of Cardiovascular Diseases: a CALIBER Study [CALIBER]; NCT02014610)

Total Center Percutaneous Coronary Intervention Volume and 30-Day Mortality: A Contemporary National Cohort Study of 427 467 Elective, Urgent, and Emergency Cases.

Background— The relationship between procedural volume and prognosis after percutaneous coronary intervention (PCI) remains uncertain, with some studies finding in favor of an inverse association and some against. This UK study provides a contemporary reassessment in one of the few countries in the world with a nationally representative PCI registry. Methods and Results— A nationwide cohort study was performed using the national British Cardiovascular Intervention Society registry. All adult patients undergoing PCI in 93 English and Welsh NHS hospitals between 2007 and 2013 were analyzed using hierarchical modeling with adjustment for patient risk. Of 427 467 procedures (22.0% primary PCI) in 93 hospitals, 30-day mortality was 1.9% (4.8% primary PCI). 87.1% of centers undertook between 200 and 2000 procedures annually. Case mix varied with center volume. In centers with 200 to 399 PCI cases per year, a smaller proportion were PCI for ST-segment–elevation myocardial infarction (8.4%) than in centers with 1500 to 1999 PCI cases per year (24.2%), but proportionally more were for ST-segment–elevation myocardial infarction with cardiogenic shock (8.4% versus 4.3%). For the overall PCI cohort, after risk adjustment, there was no significant evidence of worse, or better, outcomes in lower volume centers from our own study, or in combination with results from other studies. For primary PCI, there was also no evidence for increased or decreased mortality in lower volume centers. Conclusions— After adjustment for differences in case mix and clinical presentation, this study supports the conclusion of no trend for increased mortality in lower volume centers for PCI in the UK healthcare system. Clinical Trial Registration— https://www.clinicaltrials.gov. Unique identifier: NCT02184949.

Low eosinophil and low lymphocyte counts and the incidence of 12 cardiovascular diseases: a CALIBER cohort study

Background Eosinophil and lymphocyte counts are commonly performed in clinical practice. Previous studies provide conflicting evidence of association with cardiovascular diseases. Methods We used linked primary care, hospitalisation, disease registry and mortality data in England (the CALIBER (CArdiovascular disease research using LInked Bespoke studies and Electronic health Records) programme). We included people aged 30 or older without cardiovascular disease at baseline, and used Cox models to estimate cause-specific HRs for the association of eosinophil or lymphocyte counts with the first occurrence of cardiovascular disease. Results The cohort comprised 775 231 individuals, of whom 55 004 presented with cardiovascular disease over median follow-up 3.8 years. Over the first 6 months, there was a strong association of low eosinophil counts (<0.05 compared with 0.15–0.25×109/L) with heart failure (adjusted HR 2.05; 95% CI 1.72 to 2.43), unheralded coronary death (HR 1.94, 95% CI 1.40 to 2.69), ventricular arrhythmia/sudden cardiac death and subarachnoid haemorrhage, but not angina, non-fatal myocardial infarction, transient ischaemic attack, ischaemic stroke, haemorrhagic stroke, subarachnoid haemorrhage or abdominal aortic aneurysm. Low eosinophil count was inversely associated with peripheral arterial disease (HR 0.63, 95% CI 0.44 to 0.89). There were similar associations with low lymphocyte counts (<1.45 vs 1.85–2.15×109/L); adjusted HR over the first 6 months for heart failure was 2.25 (95% CI 1.90 to 2.67). Associations beyond the first 6 months were weaker. Conclusions Low eosinophil counts and low lymphocyte counts in the general population are associated with increased short-term incidence of heart failure and coronary death. Trial registration number NCT02014610; results.

Impact of operator volume for percutaneous coronary intervention on clinical outcomes: what do the numbers say?

The impact of operator and centre volume on clinical outcomes and quality of care has been of considerable debate in recent years in a number of surgical- and procedural-based specialities. A relationship between higher volumes at both the institutional and operator levels and better clinical outcomes would at first appear intuitive, based on the premise that performing a procedure very infrequently would be likely to lead to unfamiliarity, complications, and poorer outcomes. In the current review, we study the relationship between operator volume and outcomes in the setting of percutaneous coronary intervention (PCI), and examine the evidence for current clinical competency guidelines that advocate that a minimum number of PCI procedures be undertaken annually. Whilst both high institutional and operator volumes have been shown to be associated with better outcomes by reducing death and in-hospital mortality, these data are often derived from the pre-stent era, or when high-volume operators undertook far smaller numbers of procedures than is currently recommended to maintain clinical competency. The emphasis of specific volume requirements for optimal outcomes needs to be interpreted with caution, as volume is not a surrogate for quality and merely one of the variables associated with outcome. Healthcare providers should focus on other measures of quality such as robust clinical care pathways, evidence-based treatments, periodic case review, using validated risk assessment scores, and ascertainment of outcome to improve care and reduce adverse events.

Reply: Neutrophil Count Is Associated With Risks of Cardiovascular Diseases

We thank Li and colleagues for commenting on our study, which showed that neutrophil counts are strongly associated with incidence of some cardiovascular diseases [(1)][1]. First, regarding adjustment for potential confounders, we adjusted for a wide range of factors associated with cardiovascular