Process Variation

Reducing unwanted variation in a process and eliminating defects represent the central themes of Six Sigma methodology. While certain level of process variation may be acceptable, variation is generally deemed undesirable because it brings uncertainty to the organization’s ability to deliver high quality of patient care on a regular basis. Process variation may occur due to a wide variety of factors and influences within or outside a process including individual employees, equipment, work methods, materials, work environment, and measurement methods. Predictable causes of variation that are inherent in any given process are called common causes. On the other hand, special or assignable causes of variation occur due to unpredictable special events or unusual circumstances and they are not normally part of a process. It should be mentioned that outputs of a stable and predictable process are not automatically desirable or acceptable. The aim of Six Sigma is to quickly eliminate any source of unexpected variation and also systematically reduce expected variation over longer period of time. It is important to avoid taking unnecessary and hasty actions to compensate for variation within a stable process. This kind of tampering with stable processes can only increase variation. Technically, Six Sigma can be viewed as a statistical measure of variability present in any given business process in relation to customer requirements. The prime concentration of values in a normally distributed process is around the mean and any value that falls outside the customer specifications is categorized as a defect or error. In the context of healthcare, defects may include wrong diagnosis, untimely diagnosis, medication error, delay in treatment, wrong treatment, missing patient records, wrong side surgery, patient misidentification, and many others. The voice of the customer (VOC) is the term used to describe the process of capturing customer needs, expectations, preferences, and requirements. Some of the typical data collection methods include written surveys, telephone surveys, in-person interviews, technical specifications, direct observations, complaint logs, market research, data mining, Kano analysis, quality audits, and comment cards. To minimize bias, inconsistency and ambiguity in the customer responses, it is important to use multiple data collection methods and test instruments for validity and reliability. Quality function deployment (QFD), also known as the House of Quality, is a commonly used method to translate customer needs, priorities and requirements into product or service design features and technical characteristics. The voice of the process (VOP) communicates information about the actual process performance against customer requirements.