The tools of quality improvement: CQI versus epidemiology.

When Your Only Tool Is a Hammer, Everything Looks Like a Nail In this issue, Burnett and Chesher describe how they applied tools developed for continuous quality improvement (CQI) to reduce the risk of sharps injuries in their hospital.1 They charted the number of syringes received in the lab with the needle still attached, then contacted the medical officers involved in these instances to determine the reason why. In most cases, the needle was left on because the medical officers could not find the syringe cap. Using this information, the hospital changed to preheparinized syringes that were prepackaged with their caps. Subsequently, the number of syringes returned with attached needles fell significantly. Burnett and Chesher found the solution to their quality problem by using the tools of CQI. Why didn’t they use the tools of epidemiology? Although epidemiology long has been used to solve quality problems in the hospital, especially in infection control, it probably wouldn’t have been efficient for solving Burnett and Chesher’s problem. On the other hand, CQI cannot be used to solve many hospital quality problems, such as finding a staphylococcal disseminator or a healthcare worker purposely killing patients.2 Given the variety of quality problems encountered in healthcare, it is important to ask: When is it more advantageous to use the tools of CQI versus those of epidemiology to improve quality? There are clear instances when epidemiology is superior to CQI and vice versa. To understand the strengths of each set of tools, one first should understand their historical development. CQI developed out of industrial management science to improve the quality and efficiency of manufacturing processes.3 Manufacturing processes are explicitly designed. Every step of the process is known, and the desired result of every process step is specified. In this situation, one can examine the process to see if each process step is capable of producing the desired output. If it is not, the process can be redesigned until it is capable of performing to expectations. The tools of CQI are geared to assist in this endeavor. What aspect of the process should receive attention? That depends on the kinds of defects the overall process is producing. A Pareto chart displays the relative frequency with which certain kinds of defects occur. Burnett and Chesher’s Pareto chart shows the relative frequency of various reasons why the syringes were not capped properly. Knowing how the process was failing allowed the hospital to select a process redesign (in this case changing to syringes prepackaged with their caps) that reduced the opportunity for the process failures to occur. The tools of epidemiology have a very different lineage from those of CQI. The science of epidemiology developed because there were outcomes—infectious illnesses—with unknown processes generating them. Epidemiology seeks to deduce the processes generating outcomes. It does this by examining patterns of occurrence of putative causes in association with the events of interest. When applied to the case-control study, this logic leads to the systematic comparison of the frequency of putative causes between two groups: those with the outcome and those without the outcome. When applied to the