Content area
Full Text
"Mathematics can give your ideas tremendous clarity. Used correctly, the equations can tell you which parts of your theory work and which don't. When you mathematize something, you distill its essence."(1)
THE HEALTH CARE SETTING HAS MANY generic functional areas, such as accounting, marketing, purchasing, dietary, central supply, facilities maintenance, and hospital management, that share elements with similar operations in any other business setting. In this way, they are amenable to process improvement through standard total quality management (TQM) techniques, many of which are based on statistical process control (SPC The health care institution also has other areas in which the error rate must be so low, due to the seriousness of adverse outcome, that systems must be developed to keep error to the theoretical minimum. From this perspective, the health care delivery system can be divided into two main systems: one dealing with critical processes in which error must be reduced to the theoretical minimum and the other with noncritical generic service delivery. Process design and management of the two types of systems are shown in Figure 1.(Figure 1 omitted)
When SPC is inadequate
SPC is inadequate in some health care settings. For example, the clinical laboratory North American Biologicals, Inc. maintains various federal contracts to test for HIV infection. U.S. Army Col. James Damato summarized the testing steps for the Department of Defense in 1991 and concluded that the system resulted in a false-positive output of less than one in 1 million.(2) In 1993, the lab documented testing for the U.S. Navy in which more than 1.2 million individuals were tested without a single misclassification. Both of these results surpassed Motorola's six-sigma process capability and were not achieved through SPC.
Settings other than health care also require a low error rate. For instance, Florida Power and Light (FPL) didn't wait for a number of people to climb up pylons and get hit by high-voltage electricity, generate a control chart, establish process capability, and then redesign the process for improvement. FPL designed barriers to this up front, and when a rare event did occur, it was analyzed not by statistical methods but by single-case boring, a technique by which a few isolated cases are studied in great detail. NASA also didn't wait for a...