Display Settings:

Format

Send to:

Choose Destination
See comment in PubMed Commons below
Radiol Manage. 2003 Mar-Apr;25(2):40-53.

Six Sigma: not for the faint of heart.

Author information

  • University of Texas MD Anderson Cancer Center, GE Medical Systems-Healthcare Services, USA. arbenedetto@pdq.net

Abstract

Six Sigma is an excellent quality and performance improvement tool. Like any tool, the results of using it are highly dependent on whether you use it with competence and on the right problem. This article will help you decide if your problem is well-suited for a Six Sigma approach and will suggest the optimum approach for planning and implementing Six Sigma methodology. Performance improvement methods can be grouped into two broad categories, based on the problem to be addressed. When the problem is relatively minor and localized, "evolutionary" methods may be suitable (e.g., quality circles, problem-solving staff meetings, continuous quality improvement [CQI], total quality management [TQM]). These tools work best when modest incremental improvements are sought, when major process redesign is not thought to be necessary, and when the avoidance of workplace disruption is desired. Reengineering and Six Sigma are the best-known examples of the "revolutionary" performance improvement methods. These methods should be used when major (drastic, do or die, etc.) improvements are needed. Problems that cross departmental boundaries need these methods. When a process is so dysfunctional that you feel like you need to tear up the standard operating procedure (SOP) and start all over again, you need a revolutionary method. A Six Sigma project requires a major expenditure of money and employee time, and a willingness to make some hard decisions about jobs, employee retention and relationships among stakeholders. An institution's culture should be considered as part of the decision about using Six Sigma. If the institution has a history of making data-driven decisions, or at least has displayed openness to operating in that manner, Six Sigma has a good chance of success. A radiology-driven Six Sigma project should not be undertaken until a comprehensive written description of the scope of the project is approved by the radiology department leadership team and by the appropriate higher-level institutional leaders. This document should address the specific outcomes desired, the resources available, a rough tentative timeline, and any political constraints imposed on the project (e.g., inviolate HR policies, compatibility with enterprise strategic goals). The document should be comprehensive enough and explicit enough to be useful as a major component of the bid package for hiring a consultant or for writing the job description for the hiring of an in-house expert. A full-time project manager should be designated if an in-house expert is not hired. This person should be a senior leader in the department, but not the department director. Leading a Six Sigma project is a full-time job in itself and cannot be performed as an additional duty. Although it may be tempting to appoint a senior staff technologist or nurse, keep in mind that the project manager must have sufficient authority to expect cooperation from departmental supervisors without resorting to frequent appeals to the department director. Contact the institution's CIO and ask that a knowledgeable person on the IT staff be appointed to serve as the IT liaison for the project. This person should have in-depth familiarity with the HIS and the ways that it interfaces with the RIS (if the RIS is not a module within the HIS). Most importantly, this liaison must understand the exact data definitions and the origins of the data that are passed between the HIS and the RIS. The steering committee should consist of at least one physician and one department-level administrator from outside of the radiology department. From within the radiology department, there should be at least one radiologist and one senior modality manager (whose modality is not the primary focus), the project manager, and the manager of a radiology component that is a focus of the project (e.g., the film library manager). The consultant should be an ex officio member without vote. The steering committee should be small enough to be manageable, yet large enough to provide insight from both the radiology department and the rest of the institution. Because of the size of the steering committee and the difficulty of bringing so many people together for meetings, the day-to-day governance of the project will be provided by an informal "operations committee." When we consider "change" in the context of a Six Sigma project, we talk about a wide variety of topics, but they can be summarized usefully as dealing with processes, policies, physical plant and equipment, personnel, and politics (or culture). The first three of these lend themselves to quantitative analysis, or at least rigorously qualitative analysis. The final two, however, are subjective, ill-defined or not readily acknowledged, and fraught with potentially major challenges when it becomes necessary to implement changes in the first three. The Six Sigma process, as taught by GE, consists of five phases summarized by the acronym DMAIC: Define, Measure, Analyze, Improve and Control. This article deals mostly with the time period from first consideration of a possible Six Sigma project through the early part of the Define phase. It also discusses pitfalls that must be considered anytime throughout a project, from first thoughts of conducting a project until recommended changes become ingrained in the department's operations. Six Sigma compares baseline or historical data to data obtained after implementation of Six Sigma-driven changes in order to determine if desired changes in performance have been achieved. When historical data are not available, the Six Sigma team must collect baseline data as one of their earliest tasks. A Six Sigma project can easily last 18 to 24 months or longer. We must be sure that the data we collect during Month 18 are collected identically to the data we collected at baseline. A major performance improvement project is likely to require 12 to 24 months or longer. Upper management initially may be reluctant to appoint the "cream of the crop" to the project teams. Success is predicated on having the most knowledgeable personnel involved in the project. Without them, the chances of success are reduced. A Six Sigma project's length always exceeds the attention span of the vast majority of its participants. The department director and project manager must anticipate this and devote special efforts to maintaining motivation and momentum after the initial flurry of activity. The complexity of a Six Sigma project will be greatly increased, and all of the pitfalls discussed here will be exacerbated, if your facility has multiple sites. At the simplest, the multiple sites will introduce complications in getting personnel to come to project meetings. The complications will escalate if the sites are under different management, such as a confederated health system. The project manager and the consultant will expend additional time and effort dealing with these issues, which likely will lengthen the project unavoidably. The project manager must spend time with the department's external customers who have significant stakes in the project. At a minimum, this should include relatively formal meetings with other department directors or subordinate managers and key physician and nurse leaders, and attendance at their managerial or staff meetings (you may need to ask to be invited). Although paper or e-mail surveys can be helpful, only sustained personal contact with a stakeholder will truly allow you to understand how they interact with radiology and what their concerns are. As with daily operational management, a performance improvement project requires attention to policies, procedures, processes, physical plant and infrastructure, personnel, and perhaps most importantly, to politics.

PMID:
12800564
[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Loading ...
    Write to the Help Desk