Enhancing the cost-effective practice of medicine: the physician's perspective

Data-driven process improvement, such as that at Harper Hospital described earlier, requires several critical elements. First, data must be readily available, accurate, timely, and easy to analyze. Over 200 charts and graphs were produced at Harper Hospital to support the work of the five teams. Often, teams met weekly and data had to be accessed, analyzed, and turned around in five working days. Without the right systems, this is impossible. Second, accuracy of data and accessibility are issues. Sometimes, data are not accessible because of disparate systems or databases, or because they are not collected in a format to support analysis. An example of the latter occurred at Harper Hospital when it was discovered that surgical OR supplies were billed at one inclusive rate and that the detail data were not collected in an automated system. Manual analysis could not be done in the time frame required and momentum was lost. Next, timeliness is an issue when physicians are asked to make decisions using dated data. For most purposes, data should be no more than three months old. Ideally, two weeks should be the limit. Some hospitals are making new data available hourly. Trends change too rapidly, lengths of stay shorten, new antibiotics are introduced, new treatments are available, and reimbursement changes too often to use old data for decision making. Lastly, tools should be readily available to bring the story from the numbers. Graphs, charts, and statistics need to be used much more often and in more imaginative ways. As the health care information industry makes databases more user-friendly, physicians will understand and regain their potential to use medical science to improve the lives of their patients.