Total Quality Managment - USPS (BUS445-Final Paper)

mpressive examples over the past several years illustrate the value of uti- lizing Six Sigma and related best practices for healthcare quality and process improvement.

Providers, however, continue to face a daunting and escalating array of challenges. Regulatory pressures, increased competition, cost management issues, workforce shortages and rising consumerism all vie for attention and remediation.

Occupying an increasingly prominent place on the healthcare execu- tive’s radar screen are issues involving clinical quality and patient safety. Instances of overuse, underuse and misuse of healthcare services have been costly to patients, providers and payers.

Prompted by illuminating reports from the Institute of Medicine1, 2 and scrutiny from groups such as Leapfrog, providers are seeking effective methods for both optimizing the care they deliver and documenting the improvements.

It is a pivotal moment in the history of medicine—one offering great promise through rapidly advancing technology and tremendous pressure to deliver better care to more people for less cost.

At this juncture, then, it seems an appropriate time for reflection—both on the progress made through Six Sigma applications and the realm of opportunities for the future. Drawing from research and organizational experience, we can evaluate achievements and explore the next phase in reshaping the industry.

Applications to Healthcare

The DMAIC (define, measure, analyze, improve and control) approach works quite well for any service line or process that can furnish measurable response variables.

Generally, four groups of metrics or response variables in healthcare may define a delivery system’s performance:

• Service level.

• Service cost.

• Customer satisfaction.

• Clinical excellence. Service level metrics indicate the ability of the system’s performance to

meet the expectations of patients, referring physicians and other stake- holders—critical to quality parameters (CTQs).

Each set of metrics has specific parameters. Service level indicators may be generalized as access to care, wait time, service time and information conveyance time. Service cost indicators include cost per unit of service, labor productivity and other factors associated with the cost of providing service. Customer satisfaction indicators may be segmented into specific

Healthcare’s Horizon

FROM INCREMENTAL

IMPROVEMENT

TO DESIGNING

THE FUTURE.

I

By Richard Stahl,

MD, Yale-New

Haven Hospital;

Bradley Schultz

and Carolyn

Pexton, GE

Medical Systems

S I X S I G M A F O R U M M A G A Z I N E I F E B R U A R Y 2 0 0 3 I 17

H E A L T H C A R E

groups such as patient and family, referring physician, staff and payer.

Clinical excellence indicators may relate to a partic- ular treatment pathway or department, such as com- pliance with guidelines for prescription of aspirin to myocardial infarction patients or reduction of rates of infection contracted in a hospital or other healthcare facility. Figure 1 illustrates sample metrics from an emergency department.

Most healthcare organizations measure perform- ance using some combination from these four groups, but such analysis can be misleading since the metrics often represent an average. Customers rarely experi- ence the average performance of a system—instead, they tend to experience the variability.

From Manufacturing to Medicine

Six Sigma came slowly to healthcare and initially was met with some skepticism. This hesitancy stemmed in part from disparities between processes driven by humans vs. automated or engineered processes.

In manufacturing, it’s quite possible to eliminate most—if not all—human variability through automa- tion, creating precise measurement of assignable caus- es of variation. In healthcare, however, the delivery of patient care is largely a human process, and the caus- es of variability are often more subtle and difficult to quantify.

The challenge for healthcare institutions and staff as they begin to embrace Six Sigma is to find a way to leverage the data to drive human behavior. Where the approach seems to have had greatest success, providers combined a strong technical strategy (Six Sigma) with a strong cultural strategy, such as change acceleration process, and a sound operationalizing mechanism, such as GE Medical Systems’ Work-Out, Motorola’s Leadership Jump Start, lean, Pareto analy- sis or decision trees. This is illustrated in Figure 2.

Leveraging all three aspects has led to notable results. Most projects, however, involved optimizing existing processes and retaining systems and struc- tures bound by capital investment and traditional grouping by function. A hospital’s IT system, for example, may not fully support changing a given process, but the facility might decide to simply opti- mize around it until the investment is retired.

H e a l t h c a r e ’ s H o r i z o n

18 I S I X S I G M A F O R U M M A G A Z I N E I W W W . A S Q . O R G

Figure 3. Capability Analysis: Report Turnaround Time

Customer (physician) defined specifications: < 24 hours

Excellent mean performance, but 100,000 patient visits per year equates to 22,000 physician disappointments.

N = 300 Mean = 14 hours Standard deviation = 12 hours Report turnaround time > 24 hours = 22% Defect per million opportunities = 220,000

-40 -20 0 20 40

Defects

60 80

Source: GE Medical Systems Figure 1. Response Variables

Service cost

Service level

Clinical excellence

Customer satisfaction

Financial analysis

Customer defined/ capability analysis

Outcomes research

Survey/focus groups

Translate customer critical to quality into process specifications measurable specific response variables

Y11—Cost per procedure Y12—Labor productivity

Y21—Triage Y22—Assessment Y23—Treatment Y24—Disposition

Y31—Return rate Y32—Cardiac patient time to aspirin Y33—Cardiac patient discharged

with beta-blocker Y41—Patient Y42—Referring MD Y43—Employee

Y1

Y2

Y3

Y4

Source: GE Medical Systems

Figure 2. Formula for Effective Results

The effectiveness (E) of the result is equal to the quality (Q) of the solution times the acceptance (A) of the idea.

Six Sigma methodology

Change acceleration

process

Work-Out or other operationalizing mechanism

Effective results

Q x A = E

Source: GE Medical Systems

H e a l t h c a r e ’ s H o r i z o n

Service delivery methods in healthcare have also become entrenched and often run counter to the notion of customer centricity. It’s common in many facilities, for instance, to take the patient to the care rather than bring the care to the patient. Clearly, we need new models to create a system that genuinely meets patient needs.

A Brief Overview of DMAIC

To implement the right solution to a problem, you need to understand the degree to which different fac- tors may impact the variability of the project’s response variable (Y) before specific solutions are designed. Projects tend to focus on response variables from the four groups mentioned earlier.

The initial define and measure phases of a project essentially involve translating the voice of the customer, or CTQs, into measurable re- sponse variables. Customer ex- pectations—whether patients, referring physicians, staff or payers—are then used to estab- lish process specifications for those response variables. A measurement of the process capability to meet CTQs is per- formed, and the end result is expressed as a sigma level or defects per million opportuni- ties (DPMO). This concept is shown in Figure 3, using cycle time for reporting radiology results.

In the analyze phase, the team identifies the causal fac- tors (X’s) likely to have the greatest impact on the response variable (Y). These factors are classified as either controllable or uncontrollable. If a factor (X) is controllable and con- tributes significantly to variabil- ity in the response variable (Y),

then an opportunity to achieve a better result presents itself by controlling the causal factor.

On the other hand, if the primary causal factors are uncontrollable, a new process must be built to with- stand that variability to the degree possible. Many fac- tors in healthcare are quite predictable, though uncontrollable—such as arrival rate at the emergency room. See “Common Emergency Department Critical to Quality Factors.”

In healthcare, the improve and control phases can be most challenging since they often involve changing human behavior. It probably comes as no surprise to healthcare professionals that organizational structure can actually inhibit process thinking. Inherently, there are multiple silos across a typical facility and few exam- ples of big picture oversight to unify conflicting agen- das and constituencies.

S I X S I G M A F O R U M M A G A Z I N E I F E B R U A R Y 2 0 0 3 I 19

Common Emergency Department Critical to Quality Factors

Quality. • Accuracy of diagnosis.

• Appropriateness of treatment.

• Timeliness of ser vice

• Wait times.

• Exam and treatment.

• Testing and report turnaround.

• Staff availability.

• Bed availability in emergency department and hospital.

• Responsiveness to squads.

Satisfaction of patient and referring doctor.

Cost of operations.

Productivity and workflow.

CUSTOMERS RARELY FEEL THE AVERAGE PERFORMANCE OF A

SYSTEM––INSTEAD, THEY TEND TO EXPERIENCE THE VARIABILITY.

H e a l t h c a r e ’ s H o r i z o n

The control phase, therefore, may require dis- mantling root-bound bureaucracies growing around ancient processes. To achieve long-term success, this must be accompanied by new control measures and process metrics to drive behavior changes.

Another challenge for healthcare is to institution- alize the wins—in other words, to translate the results from one area to another. For example:

• Adopt best practices to improve bed turnover time from a given inpatient unit to all hospital units.

• Translate ventilator weaning protocols from one intensive care unit to another.

From Here to Futurity

Mistakes can be costly in any industry, and there are essentially three ways to approach them. Ignore them and hope for the best (not advisable in most cases); find and fix them within existing processes; or prevent them from occurring in the first place by designing processes correctly from the ground up.

Using the DMAIC approach (the find and fix method), many institutions have seen significant improvement in various clinical and operational processes. When coupled with proven change man- agement and decision making techniques, some have even been able to induce a beneficial transfor- mation in the organizational culture.

But quantum leap changes in the delivery of healthcare (and the prevention of errors through ground floor development) will not come about until providers begin the process of actually design- ing for Six Sigma. In Six Sigma: The Breakthrough Management Strategy Revolutionizing the World’s Top Corporations, author Mikel Harry discusses the limits of traditional Six Sigma initiatives:

The closer companies come to achieving Six Sigma, the more demanding the improve- ments become. At 4.8 sigma, companies hit a wall that requires a redesigning of processes, known as design for Six Sigma.3

This wall is often felt at significantly lower sigma levels in healthcare and consists of bricks retained from old systems and structures. To get through this wall and create quantum leap change, healthcare will have to adopt breakthrough or revolutionary thinking in how systems are designed and built to optimize the interaction of people, processes and technology.

20 I S I X S I G M A F O R U M M A G A Z I N E I W W W . A S Q . O R G

Figure 4. Design Process Map

Voice of the customer

Critical to quality (CTQ) parameters

Service delivery system design requirements

Subsystem/process design requirements

Alignment of supporting systems and structures

• Listening to the customer. • Understanding what the customer wants.

• Segmenting customer needs or wants. • Identifying the must haves and delighters.

• Translating customer CTQs into system performance specifications.

• Translating performance specifications into design considerations.

• Flowing system design requirements down to each process step.

• Translating subsystem design requirements and capability assessment.

• Organizational design. • Staffing. • Development. • Measurement systems. • Rewards and recognition. • Communication. • Information technology.

Source: GE Medical Systems

Figure 5. The Wall of Change Efforts

In te

ns ity

o f c

ha ng

e ef

fo rt

The wall

Stabilization Optimization

Time

Transformation

Source: GE Medical Systems

H e a l t h c a r e ’ s H o r i z o n

A Brief Overview of DFSS

The primary difference between DMAIC and design for Six Sigma (DFSS) is that statistical tools are used to design a new service delivery system, process or tool rather than to improve the existing system. Customer expectations are translated into process specifications and then into system design requirements. These, in turn, flow down into subsystem and process design requirements.

Elements such as service, the care delivery model, supporting systems and structures and facilities are aligned with the resulting design specifications. Similar to DMAIC, DFSS is a five-step process repre- sented by the acronym DMADV (define, measure, analyze, design and validate). Figure 4 is a design process map, basically a criteria-rating matrix that translates iteratively into system design requirements and then into subsystem requirements—drilling down into each level in order to design the process correct- ly the first time.

Organizational Readiness for DFSS

It’s important to note not all organizations are ready for DFSS. Healthcare institutions can be assessed for

readiness along a change continuum, illustrated in Figure 5. Those at the far left have fundamentally unstable operations and service delivery processes. The environment is typically chaotic and repeatability is often dependent on the performance of a few who seem to understand the “magic” involved.

In these institutions, substantial improvement may be achieved through developing and operationalizing procedures that document the magic and begin mov- ing it into the world of science.

This approach is often referred to in DMAIC as a PM/CE/CNX/SOP approach—simply a shorthand method of communicating the following:

• PM = process map.

• CE = cause-effect.

• CNX = controllable, not controllable, experimen- tal variables.

• SOP = standard operating procedure. The team first gains a common understanding

through process mapping (PM). Brainstorming then follows to discover causes of process variability and assess the effect (CE). Drivers of variability are classi- fied as controllable, not controllable or experimental (CNX).

In the analyze phase of a project, the contribution

S I X S I G M A F O R U M M A G A Z I N E I F E B R U A R Y 2 0 0 3 I 21

Table 1. Process Improvement/Solution Design Continuum

Improvement objectives >>> Stabilization Optimization Transformation

Methods Six Sigma—DFSS

When to use

Issues or drivers of variability are well understood. Primary concern is building consensus on solutions.

Causal factors or drivers of variability not well understood. Not in a position to build consensus on solutions.

White paper improvement initiative or development of new and future services designed to exceed customer expectations.

Examples

Stabilization of service level metrics such as wait time through role clarity and standard operating procedures.

• Optimization of operating room capacity utilization.

• Optimization of emergency department or radiology throughput.

• New operating room pick sheet.

• New service line.

• Renovated facility.

• New hospital.

Work-Out

Kaizen

Quality circles

Six Sigma—DMAIC

Lean thinking

Total quality managment

to variation of the experimental variables is quantified, but the institution may not realize imme- diate gains by developing SOPs targeted at controllable variables.

In the second stage, processes are stabilized but not yet opti- mized. The performance service delivery may be stable and repeatable, but still fail to meet customer expectations or oper- ate at lower efficiency and high- er cost. In these cases the appli- cation of DMAIC will provide the mechanism for process opti- mization. This occurs by devel- oping a sound understanding of the mathematical relationship between specific response vari- ables (Y’s) and their causal fac- tors (X’s).

Organizations eventually reach the previously mentioned wall where further optimization of existing systems and structures is no longer feasible. The wall is unavoidable as customer expecta- tions increase and the retention of legacy systems restricts improvement. Design then becomes an important component of the strategy for transformation.

When considering the application of DMAIC or DFSS to a process, the following considerations become relevant:

• To what extent does the current process meet cus- tomer expectations?

• Does it require decreased variability alone or a radical shift in mean?

• How committed are you to current legacy systems that support this process?

• What new developments are on the horizon? For example, new pick sheet of materials needed for operating room cases; new service line or center of excellence; renovation of facility or new facility.

An example of a process improvement or solution design continuum is shown in Table 1 (p. 21).

DFSS may be the better approach in cases where in which the process is simply too broken to satisfy cus- tomer expectations or further optimization is con- strained by legacy systems and structures. The devel- opment of new opportunities also invites DFSS as a

mechanism to design specifically for customer CTQs as opposed to cloning old processes that may fall short.

The DMADV Process

The define and measure phases of a DMADV proj- ect are similar to those of DMAIC in collecting and using voice of the customer data to develop process performance specifications.

The difference with DMADV is that we’re often dealing with new products or services, so measuring existing performance against specifications is not pos- sible. With DMADV, the goal is to predict the per- formance of the new product or service and facilitate evaluation and selection of the best design alternative.

To accomplish this ambitious task of translating voice of the customer data into actionable design cri- teria, there is a commonly used tool known as quality function deployment (QFD). QFD is an advanced cri- teria rating matrix, used in DMADV to:

• Identify customer needs or CTQs (the whats).

H e a l t h c a r e ’ s H o r i z o n

22 I S I X S I G M A F O R U M M A G A Z I N E I W W W . A S Q . O R G

Table 2. Quality Function Deployment for New Emergency And Trauma Center — Requirements

Y's X's

Customer expectations Importance

Parking and location signage well identified Dedicated emergency room parking within 100 yards

of emergency department Internal building signage clearly identifies location Standardized triage process No triage delays Door to doctor time under 30 minutes User friendly concentric wait area Metal detectors at entrance Staffing by arrival pattern demand Patient communication model designed and

caregivers trained Mechanisms to ensure staff accountability Accurate diagnostics Diagnostic cycle time specifications Quality therapeutics Decision to disposition cycle time less than 15 minutes Coordination of aftercare by emergency department Frequent patient contact and information exchange Privacy of triage spaces Privacy of treatment spaces Availability of treatment supplies at point of care Point of care testing

No delays to service due to parking 1 Ease of entrance identification and location 1 Ease of triage identification and location 2 Comfort and safety of wait area 3 Minimal wait time to see doctor 4 Minimal wait time to receive treatment 4 Privacy 3 To be kept informed of process status 4 Quality care 5 Positive caregiver interaction 5 Understand aftercare requirements 4

System requirements

H e a l t h c a r e ’ s H o r i z o n

• Weight customer needs by order of importance.

• Identify product and service features that meet CTQs (the hows).

• Evaluate the ability of each feature to satisfy each need.

For a new hospital service line, this process would be repeated three times. The first iteration would match customer needs with specific service line features. In successive efforts, the hows become the whats.

The second iteration matches the features against system level requirements. In the last iteration, system level requirements are translated into subsystem level

requirements. A QFD for a new emergency and trau- ma center illustrates the concept in Table 2.

Customer expectations are brainstormed and weighted in the left-hand column. Potential system requirements to satisfy these expectations are brain- stormed in the right-hand column.

Each system level requirement is evaluated for its ability to satisfy each customer requirement using a high, medium and low scoring system, as in Figure 6. The system requirement score in the bottom row of the table indicates its relative importance. The custo- mer expectation score in the table’s right-hand col- umn indicates the extent to which this expectation is

S I X S I G M A F O R U M M A G A Z I N E I F E B R U A R Y 2 0 0 3 I 23

Figure 6. Quality Function Deployment for New Emergency and Trauma Center — Results

Source: GE Medical Systems

H e a l t h c a r e ’ s H o r i z o n

covered by the listed systems requirements and can be compared to the weight of the associated customer expectation.

System design requirements can then be sorted in order of importance, as in the Pareto chart in Figure 7. The system design requirements become the expec- tations (Y’s) for the next flow down, and the process is repeated two more times.

On the surface, this process may seem arbitrary and subjective. If executed correctly, however, using voice of the customer data to drive the importance of the whats and sound capability data to impact the hows, a very clear picture of overall design requirements and the trade-off between competing interests will emerge with clarity.

It is important to note all customer needs are not created equal in this process. Features that currently exceed customer expectations and are considered delighters will quickly become expected must haves tomorrow.

This concept is illustrated using Kano’s model in Figure 8 and must be considered carefully in assigning importance to customer needs. The model is followed by an illustration of the first iteration of the QFD process that matches features with customer needs.

The define and measure phase of a DMADV project may be summarized as a process of CTQ flow-down. The analyze phase can be summarized as a process of capability flow-up. This is where DMAIC and DMADV differ significantly.

24 I S I X S I G M A F O R U M M A G A Z I N E I W W W . A S Q . O R G

Figure 7. Emergency and Trauma Center Pareto Chart

Staffing by arrival pattern demand Accurate diagnostics

Mechanisms to ensure staff accountability Quality therapeutics

Door to doctor time under 30 minutes Coordination of aftercare by emergency department Frequent patient contact and information exchange

No triage delays Patient communication model designed and caregivers trained

Diagnostic cycle time specifications Decision to disposition cycle time less than 15 minutes

Standardized triage process Availability of treatment supplies at point of care

User friendly concentric wait area Metal detectors at entrance

Privacy of triage spaces Privacy of treatment spaces

Internal building signage clearly identifies location Point of care testing

Parking and location signage well identified Dedicated emergency room parking within 100 yards of emergency department

0 50 100 150 200

Source: GE Medical Systems

WHEN APPROPRIATELY IMPLEMENTED WITH LEADERSHIP SUPPORT AND

THE UTILIZATION OF CHANGE MANAGEMENT TECHNIQUES TO ADDRESS

CULTURAL BARRIERS AND BUILD ACCEPTANCE, SIX SIGMA HAS

ACHIEVED MEASURABLE SUCCESS.

H e a l t h c a r e ’ s H o r i z o n

In DMAIC, an understanding of causal factors on a specific process outcome is quantified mathematically. In DMADV, a specific process supporting a service line feature and customer need may not exist. Where it does, capability can be measured directly as illustrated in the DMAIC section.

In cases involving new processes, systems and struc- tures, the capability may be projected or forecast using modeling. In healthcare, the models most relevant to a new service line are those targeted at understanding capacity, patient queuing, provider resource alloca- tion and patient routing. This concept is illustrated in Figure 9.

For a healthcare service line, the end result of the analyze phase of a DMADV project is twofold:

1. Develop a mathematical expression of customer needs translated to specific service line features, service delivery system, and service subsystem and process design specification.

2. Match needs and requirements against a mathe- matical expression of existing or forecast pro- posed process capabilities.

During the design phase, an optimal design is select- ed and implemented based upon the merging of the CTQ flow-down and the capability flow-up into an integrated design scorecard. Capability forecasting and analysis provides insight into how well design requirements will be met, and the QFD translates this into customer satisfaction. The result is a formalized, mathematical model for understanding customer impact associated with specific design alternatives and trade-offs.

Finally, in the validate phase, the actual perform- ance from a subsystem is measured against predicted performance through the confirmation of customer satisfaction. In a manufacturing environment this is achieved through component, subsystem and system level testing.

In healthcare, however, the opportunity to test seg- ments of the service line may or may not exist. What becomes more important in a healthcare application of DMADV is the degree to which appropriate con- trols are operationalized to consistently yield pre- dictable results.

Full realization of designed service line entitle- ments depends on translating the vision of these entitlements to specific behaviors. This requires appropriately targeted changes in recruitment, staff development, measurement systems, performance evaluation, incentives, communication and informa- tion technology.

The validation phase of a DMADV project also

affords the opportunity to rethink many institutional processes, systems and structures. For example, evalu- ating design alternatives for a new imaging depart- ment may indicate the existing patient registration process will not meet customer expectations. Redesign of this process should trigger rethinking of patient registration across the institution and, at the very least, provide a structured approach to institu- tional transformation of service.

S I X S I G M A F O R U M M A G A Z I N E I F E B R U A R Y 2 0 0 3 I 25

Figure 8. Kano’s Model

Customer satisfaction +

+

-

-Dysfunctional

Delighters

Must be

Features

Fully functional

Needs not created equal

Differing impact on customer satisfaction

Satisfaction proportional to service and process functionality

Types of needs: 1. Basic expectations (must be) 2. Features 3. Delighters

3

2

1

Figure 9. DFSS Process

Customer needs

Service line features

Service system design requirements

Subsystem and process design requirments

Cr iti

ca l t

o Qu

al ity

fl ow

-d ow

n

Ca pa

bi lit

y flo

w -u

p

Source: GE Medical Systems

H e a l t h c a r e ’ s H o r i z o n

Measurable DMAIC Successes

As a methodology for process and quality improve- ment, Six Sigma has demonstrated its ability to adapt to virtually any process—including patient care. Recorded achievements do not seem to be based on the type or demographics of the organization. Six Sigma has taken root in a wide variety of settings: with- in individual departments, throughout small, rural hospitals, within large teaching facilities and across multihospital systems.

When appropriately implemented with leadership support and the utilization of change management techniques to address cultural barriers and build accept- ance, Six Sigma has achieved measurable success.

The DMAIC approach has been deployed in hospi- tals and health systems to improve service levels, cost productivity and customer satisfaction. Conceding the inherent distinctions between manufacturing and medicine, however, it’s important to acknowledge the impact of human variability on statistical process con- trol and the importance of cultivating acceptance for service based change initiatives.

What’s Next?

Building on the success of the DMAIC model, the next platform for healthcare will likely follow the DFSS approach with continued emphasis on accept- ance. DMAIC optimizes existing processes, while DFSS can be used to create and institute an entirely new model for healthcare. Both promises and pitfalls accompany current applications of Six Sigma within healthcare, and organizations will need to carefully assess their own unique needs and preparedness for either targeted or systemic change.

The 21st century healthcare organization faces mul- tiple challenges. Some are complex, longstanding and unresolved issues, and others are emerging trends:

• Workforce shortages.

• Rising consumerism and patient expectations.

• The Health Insurance Portability and Accountability Act of 1996 and other compliance issues.

• Quality and patient safety.

• Reimbursement issues.

• Aging of the population.

• Regulatory constraints.

• Increasing acuity of illness.

• Disaster preparedness. Driven by a confluence of such significant factors,

the healthcare industry may soon gravitate toward an evidence based design of new systems and structures as a more verifiable and sustainable way to deliver optimal patient care.

There are no easy answers and no overnight solu- tions. It will take a considerable commitment and a concerted effort on the part of all stakeholders to embrace a new paradigm and build a better health- care system by design.

REFERENCES

1. Institute of Medicine, Crossing the Quality Chasm: A New Health System for the 21st Century, National Academy Press, 2001.

2. Institute of Medicine, To Err is Human: Building a Safer Health System, National Academy Press, 1999.

3. Mikel Harry and Richard Schroeder. Six Sigma, The Breakthrough Management Strategy Revolutionizing the World’s Top Corporations, Currency, 2000.

BIBLIOGRAPHY

A New Vision for Healthcare, Committee for Economic Development, 2002.

Burda, David, ed., “By the Numbers,” Modern Healthcare, Dec. 24, 2001.

Chassin, M.R., “Is Health Care Ready for Six Sigma Quality?” Milbank Quarterly, Nov. 4, 1998.

Chowdhury, Subir, Design for Six Sigma: The Revolutionary Process for Achieving Extraordinary Profits, Dearborn Trade Publishing, 2002.

Pande, Peter S., Robert P. Neuman and Roland R. Cavanagh. The Six Sigma Way: How GE, Motorola and Other Top Companies are Honing Their Performance, McGraw-Hill, 2000.

26 I S I X S I G M A F O R U M M A G A Z I N E I W W W . A S Q . O R G

WHAT DO YOU THINK OF THIS ARTICLE? Please share

your comments and thoughts with the editor by e-mailing

[email protected].

THE CONTROL PHASE MAY REQUIRE DISMANTLING ROOT-BOUND

BUREAUCRACIES GROWING AROUND ANCIENT PROCESSES.