Drug-induced sudden cardiac death (also called sudden death, SD) and ventricular arrhythmia (VA) have arisen as major public health concerns in the last decade. SD/VA has resulted in the withdrawal of more drugs in recent years than any other adverse drug reaction, and the identification of over 100 non-cardiac drugs as suspected of being high-risk. Unfortunately, controlled studies measuring the risks associated with specific drugs are very few in number, presumably because of the complexity of such studies and the massive sample sizes needed to study this outcome. Even studies in """"""""large"""""""" databases have lacked adequate statistical power for crucial subgroup analyses. This lack of controlled data on clinical SD/VA has necessitated reliance on uncontrolled observations and on studies of putative markers of risk such as QTc prolongation. However, the utility of uncontrolled observations is always subject to question, and the validity of these putative markers remains unknown. As a result, clinicians, patients, regulators, and drug manufacturers are ill-equipped to address the critical clinical and public health decisions concerning drug-induced SD/VA. This study will compile a massive new pharmacoepidemiologic database of Medicaid data (linked with Medicare data for those enrolled in both programs, and with the Social Security Administration Death Masterfile) from five large Medicaid programs. This will be combined with the UK General Practice Research Database. This combined database will be used to conduct a series of nested case-control and case-crossover studies to measure the absolute and relative rate of all-cause death and SD/VA associated with five of the most commonly used drug classes of greatest concern: antipsychotics, antidepressants, opioid analgesics, quinolone antibiotics, and macrolide antibiotics. A multi-stage investigative strategy will be used: Stage 1 will compile the database, assure its quality, and reproduce known associations. Stage 2a will compare drugs among the classes of interest. Stage 2b will use the case-crossover design to look for associations controlling for stable patient factors. Stage 2c will examine the effect of dose and inhibitors of pharmacokinetic clearance, the functional equivalent of high-dose use. Stage 3 will develop predictive indices to stratify patient subgroups receiving high-risk drugs. This study will address an extremely important public health concern, aiding clinical and regulatory decision making by providing critically needed comparative data on the risk of death and SD/VA. In addition, this study will provide clinical outcome data that will help to elucidate the relationship between widely used markers and clinical risk. Finally, this project will create a valuable and lasting pharmacoepidemiologic data resource that can be brought to bear in future drug safety studies. ? ?
