Monitoring technology and firm boundaries: Physician-hospital integration and technology utilization

Jeffrey S. McCullough
Assistant Professor
Division of Health Policy and Management
University of Minnesota

Eli M. Snir
Assistant Professor
Information Technology and Operations Management Department
Southern Methodist University

Recent years have witnessed fundamental changes to medical care management in efforts to contain cost. We study two noteworthy trends. First, physicians and hospitals have increasingly engaged in exclusive long-term contracts, a practice also known as "vertical integration." In theory, vertical integration may improve coordination between hospitals and physicians thus reducing inefficiency. Second, hospitals have invested in information technology (IT) specifically designed to monitor physicians' hospital-based activities. Potentially, hospital administrators could use these technologies to coordinate and influence the behavior of either integrated or non-integrated physicians. In particular, we study whether vertical integration and physician monitoring technologies are complementary practices in U.S. hospitals.

Background

Research Brief is a monthly update of the latest research from School of Public Health faculty. Subscribe Unsubscribe

Early hospital IT played little role in monitoring or coordinating physician activities. Hospitals focused IT resources on services for which hospitals were reimbursed, rather than those for which physicians were reimbursed; thus, laboratory, pharmacy, and radiology systems were widely diffused by the early 1990s while physician-related applications were relatively rare. Physician monitoring IT gained prevalence during the late 1990s. These applications were specifically designed by hospitals to monitor physicians' clinical behavior within hospitals. Hospitals could, for example, monitor physician adherence to treatment guidelines intended to reduce errors and their resultant costs. Monitoring technology diffused rapidly in the late 1990s, increasing from 40 percent in 1998 to 69 percent in 2001.

The interaction between IT and organizational structure is at the heart of IT adoption and may be responsible for much the recent IT-driven productivity gains. Consequently, there is a growing literature describing IT and firm transformation. This literature seeks to explain how IT is transforming society, whether IT leads to large integrated organizations by reducing the cost of coordinating resources or whether IT leads to smaller and leaner organizations that use IT for outsourcing and contracts. Overall, researchers have found that aggregate IT investment is associated with smaller firms and less vertical integration.

We build upon the literature in three ways. First, we develop a theoretical model that relates the potential value of monitoring technology to the vertical relationship between physicians and hospitals. Second, we utilize data that describe more than 1,200 hospitals' specific IT applications as well as their contractual relationships with physicians. The use of IT application data rather than aggregate IT investment is important as IT is general purpose technology that may serve different, even contradictory objectives.

Finally, we utilize an empirical strategy that addresses the interdependence of hospitals organizational and technology adoption decisions. For example, hospitals may have integrated with physicians in response to financial pressure from insurers or their local competitors. This same pressure may reduce hospitals' ability to invest in monitoring IT. This type of interdependence has largely been ignored in the IT and organization literature and has never been addressed in the physician-hospital integration literature. Failure to address this interdependence would result in an underestimate of the affect of organization on monitoring IT adoption. We employ a statistical strategy, instrumental variables, that IT uses physician labor market variation to purge this bias from our estimates.

Findings

We find that, ceteris paribus, vertical integration increases monitoring technology demand by 27 percent. Our results suggest that monitoring IT provides hospitals with valuable information regarding physicians' clinical behavior. This monitoring information is more valuable for vertically integrated health systems. One possible interpretation is that the complexity of clinical practice makes it difficult to write detailed contracts over physicians' clinical decisions; thus, a health system that already enjoys relatively strong managerial authority could make subjective use of this monitoring while a non-integrated health system could not write a contract based on objective monitoring results.
We find that there is a substantial difference in the behavior of for-profit and not-for-profit health systems. For-profit hospitals are about 13 percent less likely to vertically integrate with physicians, this lack of integration decreases their propensity to adopt monitoring IT. They are, however, about 14 percent more likely to adopt monitoring IT conditional upon their integration decisions. In aggregate, these apparently contradictory effects roughly cancel each other out.
Overall, our results contradict those of most studies of IT and firm organization ' we find that vertical integration is complementary to monitoring IT investments. These results contradict those of most previous empirical studies, which found that aggregate IT investment was a substitute for, rather than a complement to vertical integration. We believe that there are two explanations for this contradiction. First, we utilize IT application data rather than aggregate IT data, some forms of IT are probably substitutes for integration while others are complementary to integration. Second, we utilize an empirical strategy that addresses the interdependence of hospitals' IT and vertical integration decisions.

From a policy perspective, these results suggest organizational change is likely to play an important role in realizing the potential of health care IT. In particular, physician monitoring applications may have little value without complementary organizational investments that can take advantage of the new information.