| Over the past decade, the federal government has launched several initiatives aimed at increasing the rate of adoption of Health Information Technology with the goal of improving the quality of care while reducing the cost of care. One of the primary mechanisms suggested for reducing the cost of care is to decrease the ordering of unnecessary and redundant diagnostic tests. This thesis examines the effects of HIT adoption on trends in test ordering to estimate the real effect of technology use on ordering practices. Analyzing over 100,000 patient visits to providers and using an instrumental variable to eliminate endogenous variation in provider characteristics that might bias the results, I find that the use of HIT actually increases the ordering of diagnostic tests, contrary to the claims of policy makers. This increase is consistent with the theory that HIT adoption increases the efficiency with which tests are ordered, performed and resulted, and that this efficiency increase dominates any decrease in the medical necessity of the tests and procedures. Further investigation shows that the effect of HIT use is not homogenous for all patients and physicians. HIT use is associated with less testing for certain patient groups, such as new patients and patients who visit the same physicians several times in a year, and it is associated with more testing for certain provider groups, such as those who work in clinics with laboratory equipment.;While this work suggests that the promise of HIT may not be met in its current implementation across the country, more work needs to be done to understand the exact mechanisms through which this technology increases testing, whether the additional tests provided benefit to patients, and how future policies might be crafted to achieve the basic goal of improving care while reducing costs over the long run. |
