When to initiate beta-blockers in heart failure: is it ever too early?

Analytical Approaches to Reduce Selection Bias in As-Treated Analyses with Missing In-Hospital Drug Information

INTRODUCTION

While much attention has focused on immeasurable time bias as a potential exposure misclassification bias, it may also result in potential selection bias in cohort studies using an as-treated (or per protocol) exposure definition in which patients are censored upon treatment discontinuation.

METHODS

We examined analytical approaches to minimise informative censoring due to the absence of in-hospital drug data using a case study of β-blocker use and mortality in heart failure. We conducted a cohort study using Korea's healthcare database, including inpatient and outpatient drug data. Using an as-treated exposure definition, patients were followed up until death, β-blocker discontinuation (in the exposed), β-blocker initiation (in the unexposed), or end of study period. In 'complete prescription' analysis using inpatient and outpatient drug data, we estimated hazard ratios (HR) and 95% confidence intervals (CI) using a Cox proportional hazard model. In outpatient drug-based analyses, we attempted to reduce the bias using stabilised inverse probability weighting (IPW) for treatment crossovers, hospitalisation, and all artificial censorings.

RESULTS

An HR of 0.89 (95% CI 0.74-1.07) for β-blocker use versus non-use for all-cause mortality was found in 'complete prescription' analysis. Benefits were exaggerated when follow-up was assessed using outpatient drug data only (HR 0.71; 95% CI 0.57-0.89). Weighting by stabilised IPW for treatment crossovers and hospitalisation reduced the bias.

CONCLUSIONS

When using an as-treated exposure definition, missing in-hospital drug data induced selection bias in our case study. Using IPW for censoring mitigated bias from the hospitalisation-induced censorings.

Importance of measuring glycosylated hemoglobin in patients with myocardial infarction and known diabetes mellitus

Although medical co-morbidities commonly affect clinical outcomes after acute myocardial infarction (AMI), current performance measures of AMI quality focus exclusively on the management of the AMI itself. However, patients with AMIs frequently present with other co-morbidities, such as diabetes mellitus (DM), that also warrant assessment and management. To date, the quality of DM evaluation in patients presenting with AMIs has not been described. From January 2003 to June 2004, the Prospective Registry Evaluating Myocardial Infarction Patients: Events and Recovery-Quality Improvement (PREMIER-QI) enrolled 3,953 patients with AMIs at 19 centers in the United States. The frequency of glycosylated hemoglobin (HbA(1c)) assessment, either during the hospitalization or documented in the chart from the preceding 3 months, was prospectively evaluated. Among 1,168 patients with AMIs with preexisting DM, only 47% had recent HbA(1c) levels available, with marked variability in HbA(1c) assessment among hospitals (range 7% to 81%). Among those with available HbA(1c) levels, 39% had good control (HbA(1c) <7%), 36% had suboptimal control (HbA(1c) 7% to 9%), and 25% had poor control (HbA(1c) >9%). Patients with suboptimal and poor control were more likely to have their DM treatment intensified than those without HbA(1c) assessment (for HbA(1c) 7% to 9%, rate ratio 1.38, 95% confidence interval 1.03 to 1.85; for HbA(1c) >9%, rate ratio 2.20, 95% confidence interval 1.68 to 2.88). Similarly, patients with DM who had HbA(1c) measured were more likely to receive instructions on DM disease management before discharge. In conclusion, the assessment of chronic glycemic control is highly variable among patients with AMIs and DM. Because much of this variability occurs at the hospital level, the evaluation of DM control could represent an additional quality indicator and an opportunity to advance patient-centered AMI care.