Providers' improved access to electronic patient information through health information exchange is a national policy priority and a key infrastructure requirement for the US health care system. Thanks to significant public investments, providers have access to different health information exchange approaches to meet their information needs. One approach is referred to as ?pull,? which allows providers to query community-wide, longitudinal patient records. A second approach is ?push,? where key information, such as test results, is automatically delivered to providers. Unfortunately, there is a lack of evidence of the effectiveness of either the ?push? and/or ?pull? approach in primary care. Furthermore, which approach to sharing information best fits into primary care is also unknown. Despite the lack of evidence, federal policy encourages the exchange of patient information based on the assumption that the two approaches are equivalent. The objective of this proposal is to address these shortcomings in the evidence-based by clarifying the relationship between ?pull? and ?push? health information exchange usage in primary care settings and by determining the impact of each approach on potentially avoidable and costly health care utilization.
Aim 1, Determine whether primary care providers use ?push? and ?pull? as complementary or alternative approaches to health information exchange, leverages a novel dataset of individual provider and staff behavior tracked within an electronic health record system combined with detailed measures of ?push? and ?pull? health information exchange usage. These data furnish a complete, detailed temporal sequence of providers' behavior revealing how each approach to health information exchange is used during a patient visit.
Aim 2, Quantify the effect of ?push? and ?pull? health information exchange on potentially avoidable health care utilization, tests the hypothesis that health information exchange usage will be associated with reductions in readmissions and hospitalizations for ambulatory care sensitive conditions. This hypothesis will be tested among patients attributed to a 7-year, retrospective panel of more than 240 ambulatory physicians. The differential timing of implementation will be used to identify the effect of ?push? and/or ?pull? with a regression model that includes physician and year fixed-effects. This approach allows physicians to serve as their own controls eliminating unmeasured time-invariant confounding. This proposed analysis of secondary data is significant because the US has invested billions on interoperable health information technologies, but there is very little evidence of health information exchange's effects on utilization and on how it is used in primary care practice. This proposal is innovative because it will be the first to examine the impact of ?push? and ?pull? health information independently, and jointly, in primary care. This combination goes beyond existing research to accurately reflect current approaches to health information exchange available to physicians and supported by health information technology policy.
The proposed research is relevant to public health because health information exchange has the potential to improve the quality, safety and efficiency of health care. Identifying the relationship of ?push? and ?pull? information exchange in primary care will provide evidence to guide policy, the structuring of clinical workflows, and the design of information technology. Identification of health information exchange as a valuable, evidence-based intervention would support AHRQ's goal of quality care.
