Six Sigma Flashcards Preview

HMGT 6332 > Six Sigma > Flashcards

Flashcards in Six Sigma Deck (44):

What will we cover in define

Identify customer
Collecting VOC
Deriving CTQs
Developing a map of the process
Deciding what the project should focus on
Assembling a team to execute the project
Establishing a project charter


Define of customer

Who are downstream in the supply chain, no matter they are inside or outside the same organization


What is VOC voice of customer

Refers to customer’s wants and desires, customer can be any stakeholder.
Goal 1: understand customer wants
Goal 2: engage the customer in the improvement process


Why we need to listen to VOC

-Changing healthcare environment (Customer’s voice is instrumental in shaping the nature and characteristics of the service. Transfer customers from passive consumers to advocates of the service.)
-Patient as partner in healthcare delivery
-therapeutic value(心理治疗)
-avoid malpractice lawsuits
-Understand patient expectations(reduce the gap between expectation and reality)


Collecting VOC

-direct observation
-surveys(open ended & close ended)
-focus group
-patient complain



Represents a specific and measurable aspect of product or service that reflects quality for the customer. CTQs are derived from the VOC and are the specific drivers for improvement.


The relationship between VOC and CTQ

VOC and CTQ are mapped to each other. Each CTQ is mapped to one VOC, however each VOC may be linked to one or more CTQs.


CTQs can be defined at the level of an individual patient’s needs, at the level of a project that impacts a community, or from an even broader public policy perspective.

6 aims for improvement in healthcare.



Defects per million opportunities.
In six sigma performance, no more than 3.4 defects per million opportunities


QFD( quality function deployment)

Is a tool help us to translate customer. Wants to the design of a product or service that will satisfy customer requirements. It is a systematic way.
-customer needs expressed through the VOC
-Critical to Quality criteria derived from the VOC
-the translation of these CTQs into potential service attributes
-prioritizing these attributes
-determining product or service alternatives to meet these service requirement


Project scope

-Affinity diagram
-prioritizing CTQs— Kano model (1980s)
-cost benefit matrix


Kano model

Delighters—feathers that customers do not expect.
Satisfiers— needs are stated by the customers
Dissatisfiers—not stated by the customers but are assumed as given.


Cost benefit model

Project cost
Low. High
Project. Low. Study further. Do not execute
Importance. High. Execute. Study further



Customer, output, process, input, supplier.
It helps facilitate the development of a process map
It summarizes the interactions between different parts of a process.


Process mapping

Is a REPETITIVE set of activities executed in a structured and sequential manner for the purpose of accomplishing a given object.


Process map - objectives

-generates insights into need for each step
-helps eliminate non-value added activities
-provides a common reference point
-ensures everyone has same understanding
-indicates how roles interact with each other
-helps eliminate duplication of tasks
-documents and communication business processes.


In developing or re-engineering a process

The current process is typically designated the “as is” process, the ideal state of the process that we are striving for is designated the “to be” process. We shall revisit the process map and create the “to be”process in the IMPROVE phase of the damic process


Project charter elements

Business case
Problem statement
Goal statement
Project timeline
Team composition and roles


SMART goals



Prospective tools
Retrospective analysis

Fishbone chart and failure modes effects analysis or FMEA
Root cause analysis and includes the five whys.


Fishbone diagram

Also known as Ishikawa diagram & cause and effect diagram
Graphically presents possible causes and effect


Possible causes in five groups, five Ms



Failure modes effects analysis (FMEA)

NASA defines FMEA as
A methodology to analyze and discover
1. all potential failure modes of a system
2. The effects these failures have on the system
3. How to correct and or mitigate (decrease impact) the failures or effects on the system.

It is a prospective tool that can be used to identify potential risks to patient safety and that can help in reducing medical /healthcare errors.
Unlike root cause analysis, is a retrospective tool which is done afterwards, creates the possibility that someone is blamed for contributing to the event.


Acute errors
Latent errors

Acute errors—incidence of errors(护士没有足够消毒)
Latent errors—potential errors (病人房间内没有洗手池)

FMEA is based on the premise that the incidence of errors can be reduced by decreasing the potential for errors.


Steps in conducting FMEA

1.Identifying the various steps in a process(each step in COPIS and the process map should have a place in the FMEA)
2.Anticipating what can fail in a process step
3. Estimating the severity of the effect if such a failure were to happen(what is the impact of the failure. The severity is classified as catastrophic,major,moderate or minor)
4. Estimating the frequency of occurrence.
5. Estimating the detectability of the failure.
6. Computing the combined impact of severity, occurrence and detectability.( risk priority number RPN from 1 to 1000)
7. Prioritizing the steps in order of risk magnitude.
8. Taking corrective action
9. Re-estimating severity, occurrence and detectability(FMEA can be applied at any time, the RPN should be lower


Limitations of FMEA

1. Labor intensive, and typically consumes many hours.
2. The prospective nature of FMEA suggests that we are evaluating potential failures and there is no certainty that - in the absence of an FMEA - a specific adverse event would have resulted.
3. By considering one failure at a time the FMEA tool does not take into account the possibility of typically complex with multiple factors in play.


Root cause analysis

Most basic reason a problem occurs or could occur
Is a retrospective analysis.
It’s a 3-step process: execution, interpretation and follow-up action
RCA needs to be done carefully. Sufficient Time and the resources need to be allocated. It might be helpful to get a variety of opinions on the analysis, we need to keep in mind that rather than a single root cause there might be a complex of factors contributing to an adverse event . Many RCA’s are performed incorrectly or in completely and do not produce useable result


Five whys and its limitation

Ask why as many time as required.
1. It encourages focus on single root cause. Better for simple adverse events not for complex events.
2. Tends to be liner. Step A to step B to step C.
3. Subject to bias in decision-making
- heuristics(直觉)
- anchoring bias(this bias influences the decision maker to base judgment on a single piece of information, or to stick with initial impressions.
- framing bias(固定思维)
- blind obedience( an expert might influence a decision in the wrong direction)


Sentinel events

Events that should never happen.

66% root causes of sentinel events comes from COMMUNICATION, which is the most often root cause.
Then orientation and training 57%
Then patient assessment processes 42%


The structure-procrastinate-outcome model`

Introduced in 1966 by Avedis Donabedian.
It is a way of describing and evaluating methods for assessing the quality of medical care.


Difference between quality assessment and quality measurement

Assessment refers to a more judgmental or subjective approach to the determination of quality while measurement represents greater objectivity.
Assessment主观 measurement客观


Difficulties in evaluating healthcare quality

1. A lot of variations.
2. Healthcare providers view measuring quality as a slight to their professionalism, hence there is a natural resistance to measuring quality.
3. In service industry, can’ be standardized. The severity of disease.
4. Co-morbidity, which means the presence of more than one disease or condition.
5. Lack of relevant data
6. Inadequacy of definitions.
7. Healthcare purchasers, physicians and patients have different perspective on quality.


What to measure in healthcare

The selection and use of performance measures is an important consideration in improving performance and managing change



Framework, physical infrastructure and organizational resources for providing healthcare.
They are essentially measures of capacity to deliver quality health care not measure of the care itself.
It is necessary not sufficient, not quantifiable like process and outcome.
Largely domain of individual physician, the institution or public policy, patients only influence indirectly.



1. Refers to the application of our current knowledge of medical science.
2. Refers to measurements of the things done to and for the patient by practitioner in the course of treatment.
3. Pertains to both technical skills and interpersonal skills人文关怀)
4. Processes an outcomes influenced directly by the patient


Drawback of process

1. Process measures are based on the assumption that a difference in the process is indicative of a difference in health outcomes.
2. It is easier to follow changes in health outcomes than changes in process measure



A change in the patient’s current and future health statu that can be attributed to antecedent health care
Five Ds:


Drawbacks of outcome

1. Outcomes can be influence by processes and the structure, but are not fully within our control. So relevance of the outcome.
2.long gestation periods


AHRQ quality indicators

Prevention quality indicators(simple procedure shouldn’t have outcome of death)
Inpatient quality indicators(quality of care inside hospital)
Patient quality indicators(防止病人自杀)
Pediatric quality indicators


Measurement approaches

Item by item
All or none(if you don’t remove all cancer cell, still gonna die)





Types of data

Ordinal variable (one to five stars)
Nominal variable(male&female, eye color from black, brown, blonde,gray, other)

Interval (like temperature, but they don’t have “true zero”as no temperature. No matter how hot or how cold, there will be a temperature)
Ratio (like height, weight, can have 0 gram means no weight)


Primary data
Secondary data

Focus group

Clinical data(from medical records)
Administrative data(more uninformed, billing and claims)


True variation
Observed variation

True variation:
1. Common cause(due to random variation. Does not affect the mean)
2. Special cause(also called assignable cause variation, affect the mean)

Observed variation
1. Variation due to operator
2. Variation due to gage