Today, about 65% of emergency rides among Medicare beneficiaries are on Advanced Life Support (ALS) ambulances, and most of the remaining are on Basic Life Support (BLS) trucks. The key difference is that while ALS provides more onsite care, BLS prioritizes time to the hospital. However, there is no strong support for the use of ALS over BLS, which is also less expensive. On the contrary, most research has shown that ALS provides no additional benefits and may even increase rates of morbidity and mortality in some patients. Almost all of this work, though, has been done outside of the United States, and has serious weaknesses related with data quality, study design, and methodology. This uncertainty about optimal pre-hospital care is also apparent in the tremendous variation of ALS use across geographic areas in the United States. The current proposal aims to move the scientific understanding of pre-hospital care systems significantly forward by rigorously evaluating the use of ALS in ambulances. The first aim is to describe the differential uptake in ALS use across the country, and the hypothesis is that much of the variation is due to differences in factors exogenous to health and healthcare in an area. The second aim is to compare the effectiveness of BLS and ALS on a range of health outcomes for different subgroups of patients. The third aim is to analyze inconsistencies in Medicare ambulance claims that arise during data preparation for the first two aims. This project will use a large sample of Medicare claims data from ambulance and hospital providers between 1992 and 2010, which includes individuals living in inner cities and rural areas, women, and the elderly. In Aim 1, the county-level variation in ALS use across the country and over time will be modeled using individual and county-level predictors in a multilevel framework. This model will be used to predict the county-level propensities to provide ALS level of care. In the second aim, these probabilities will be used to compare BLS and ALS on outcomes, including mortality and neurological functioning. A few different study designs will be used here. For example, areas that experience sudden temporal shifts in ALS use will be analyzed, as well as areas that have different rates of ALS use but are otherwise comparable. Finally, the inconsistencies in the Medicare claims will be systematically analyzed by testing for valid explanations. This will be the first American study on the comparative effectiveness of BLS and ALS that is based on a large, population-level sample and rigorous quantitative analysis. The first aim will be useful for developing pre-hospital care policy solutions that are effective in diverse local settings. The second aim will help identify subgroups of patients that benefit from each ambulance type, and will help standardize pre-hospital care policy. The last aim has a direct connection with reducing health care costs of fraud and abuse, and the outcomes may be used to develop data-driven solutions. Thus, this project will have important policy implications for reducing morbidity and mortality in emergency health events, and improving spending efficiency for ambulance services.

R36HS022798

2013-09-01

2015-04-30