2.3. Independent variables

Laboratory size was calculated by dividing the total volume of test results into quartiles. The total volume of test results sent was considerably higher for hospital laboratories compared to independent laboratories, so lab size quartiles were computed separately for hospital and independent laboratories. We also examined laboratory type and affiliation. Setting (e.g., rural, urban) was determined by whether the clinical laboratory was located in a metropolitan statistical area.

In order to assess the potential availability of exchange partners for clinical laboratories to send test results to an ordering practitioners’ EHR, using data from the Medicare and Medicaid EHR Incentive Programs, we calculated the share of eligible professionals paid by county for completing the first stage of the program. Providers who completed the first stage of the program had EHRs with the capability to incorporate laboratory test results. For hospital laboratories, using data from American Hospital Association (AHA) Health IT Supplement survey conducted in 2013 (but reflecting the calendar year 2012), we merged data regarding whether the hospital had a basic EHR and/or participated in HIO. Using data from the main AHA survey from that same year, we examined whether the hospital (or system) offered a health maintenance organization (HMO) product and constructed a measure of market concentration, known as the Herfindahl-Hirschman Index [22]. Greater competition as well as market consolidation has been shown to be associated with lower rates of electronic exchange of health information among hospitals [23,24]. Hospitals offering HMO products possess financial reasons for sharing information to better manage and coordinate patient care, and have been found to be more likely to electronically exchange health information [25].