Modeling Second Order Impacts of Healthcare Innovation

Any single health service organization today is likely engaged in dozens of concurrent, often times unrelated change initiatives. Each of these change initiatives is likely supported by evidence that demonstrates the innovation?s intended, first order impact. However, very little attention has been paid to the unintended, second order impacts of innovation. In this dissertation we introduce a model to provide a framework for inquiring about this very type of non-immediate impact. Next, using three innovations currently being implemented in the healthcare industry?training primary care residents to perform in-office colonoscopies, Studer Group?s ?Evidence Based Leadership,? and implementation of electronic health records in a hospital-integrated pediatric network?we model the innovations? second order impacts within the context of our second order impact conceptual model. Cost effectiveness analysis, multiple analysis of variance (MANOVA), and two-level fixed effects modeling are used to across the three interventions. Results from the primary care residency intervention support further investment in colorectal cancer screening training for primary care residents. Results from the Studer Group?s ?Evidence Based Leadership? intervention demonstrate mixed results across change interventions and across categories of tenure, suggesting receptivity towards change and organization tenure is highly dependent upon the nuances of a specific change intervention. Finally, results from the implementation of the electronic health record demonstrate improved charge capture.

We conclude that this further probing of popular innovations in the industry is warranted for multiple reasons. For one, it is entirely possible that social scientists and economists are prematurely ?moving on? to other innovations as soon they have published results from an initial round of inquiry. However, as we will demonstrate in our model, it is conceivable that after the ?lights have dimmed? on an innovation?s initial glow, the artifacts of the innovation could very well continue to disrupt structures and processes long after its implementation. If these latent disruptions adversely affect the organization, one could argue that any initial positive impacts were likely overstated. Conversely, if these latent disruptions go on to produce additional benefit to the organization one could argue that any initial positive results were actually understated.