ANALYTIC APPROACHES TO MANAGING RISK: The I-D-Q Approach Presented by: Steven D. Staugaitis, Ph.D. Connecticut DMR - and the UMASS Medical School, Family.

ANALYTIC APPROACHES TO MANAGING RISK:The I-D-Q Approach

Presented by:Steven D. Staugaitis, Ph.D.Connecticut DMR - and theUMASS Medical School, Family Medicine & Community HealthCenter for Developmental Disabilities Evaluation and Research

AN INTRODUCTION: Two Analytic Risk Management Tools:

• Root Cause Analysis

• Failure Mode & Effects Analysis

DECEMBER 2003

It has been reported in the medical literature that as many as 180,000 deaths occur in the United States each year due to errors in medical care, many of which are preventable. In order to take actions that will improve this situation, it is necessary to have a clear picture as to what is actually happening so that appropriate steps can be taken that will prevent such occurrences.

Only by viewing the health care continuum as a system can truly meaningful improvements be made. A systems approach that emphasizes prevention, not punishment is the best method to accomplish this goal. Other high-risk industries/companies such as airlines and nuclear power have used this approach to accomplish safety. To make the prevention effort effective, we use methods of gathering and analyzing data from the field that allow the formation of the most accurate picture possible.

Because people on the frontline are usually in the best position to identify issues and solutions, Root Cause Analysis teams at each of the VA Healthcare facilities formulate solutions, test, implement, and measure outcomes in order to improve patient safety. Findings from the teams are shared nation-wide.

This is really at the core of what we mean by building a culture of safety. This kind of cultural change does not happen over night. It can only happen as a result of effort on everyone’s part to take a different approach to the way we look at things. We must constantly question if we can do things in a better, more efficient, and safer manner. We must never let ‘good enough’ be good enough. We must be relentless in our pursuit of finding ways to improve our systems. We don’t believe people come to work to do a bad job or make an error, but given the right set of circumstances any of us can make a mistake. We must force ourselves to look past the easy answer that it was someone’s fault - to answer the tougher question as to why the error occurred. It is seldom a single reason.

Through understanding the real underlying causes we can better position ourselves to prevent future occurrences. As has been said, “Experience is the best teacher” but is also one of the most expensive teachers as well. One of the best ways to reduce the expense, is to take advantage of lessons present in close calls where things almost go awry, but no harm is done. Establishing a culture of safety where people are able to report both adverse events and close calls without fear of punishment is the key to creating patient safety.

Creating a Culture of Safety

Would you stay aboard? Luckily, the safety statistics in airline travel are far, far better than these figures. Between 1990 and 1994 United States airline fatalities were 0.27 per 1000000 aircraft departures, less than one third the rate in mid-century, despite vast increases in the complexity and volume of our aviation systems. One estimate is that a modern passenger would have to fly continuously for 20 000 years in order to reach a 50% chance of

injury in an airplane accident. In health care it is a totally different story. With the rising complexity and reach of modern medicine have come startling levels of risk and harm to patients. One recent study in two of the most highly regarded hospitals in the world

discovered serious or potentially serious medication errors in the care of 6.7 out of every 100 patients, and the Harvard Medical Practice Study, which reviewed over 30 000 hospital records in New York state, found injuries from care itself (“adverse events”) to

occur in 3.7% of hospital admissions, over half of which were preventable and 13.6% of which led to death. If

these figures can be extrapolated to American health care in general then over 120 000 Americans die each year as a result of preventable errors in their hospital care.

Reducing errors in HealthcareIt's time to take this more seriously(Berwick and Leape, BMJ, 1999)

“Ladies and gentlemen, welcome aboard Sterling Airline's Flight Number 743, bound for Edinburgh. This is your captain speaking. Our flight time will be two hours, and I am pleased to report both that you have a 97% chance of reaching your destination without being significantly injured during the flight and that our chances of making a serious error during the flight, whether you are injured or not, is only 6.7%. Please fasten your seatbelts, and enjoy the flight. The weather in Edinburgh is sunny.”

RISK is all around usIn one NE state, if you receive residential services from the MR/DD agency you have a:

1 in 40 chance of suffering a severe injury1 in 50 chance of experiencing abuse or neglectAnd, you will probably experience 1½ unusual incidents each year.

Or, if you have mental retardation and are served by the state agency, and you are:

Mobility dependent - you are 6X as likely to die as a person who is mobility independent.

On the other hand, if you are:Female, you are 1.06X as likely to die as a male served by the state agency.

And, if you are: Over 85-yrs of age, you have a 50:50 chance of dying, but less than a 5% chance if you are younger than 65.

Are these risks “reasonable?”

Problem: We don’t know unless we use data & develop benchmarks!

Today’s Presentation will look at:

Analytic Approaches to Quality General Principles and Cautions Managing Risk and Improving Safety Retrospective Analysis: ROOT CAUSE

ANALYSIS Prospective Analysis: FAILURE MODE &

EFFECTS ANALYSIS For More Information: Resources &

References

CONTENTS Analytic Approaches to Quality General Principles and Cautions Managing Risk and Improving Safety

Retrospective Analysis: ROOT CAUSE ANALYSIS

Prospective Analysis: FAILURE MODE & EFFECTS ANALYSIS

For More Information: Resources & References

FOCUS Today

QUESTIONSWELCOMED

GENERAL OVERVIEW

WHY IS ANALYTIC RISK MANAGEMENT USED IN HEALTH CARE SETTINGS?

• 40 potentially harmful drug errors EVERY DAY in an average hospital.

• 4% of all hospitalizations have an adverse event. Of these, 16% lead to death or serious disabling injuries.

• 1 in 2.5 chance of experiencing a complication on a general surgical service following surgery.

• One million people are injured in the hospital every year.

• Each year 44,000 to 120,000 die as a result of medical error.

Root Cause Analysis, Incident Reporting and Failure Mode and Effects Analysis have been incorporated into health care due to the growing complexity of the delivery system and the high rate of adverse events that have resulted in patient harm and even death.

Accidental Deaths in U.S.

0

20000

40000

60000

80000

100000

120000

Med Error Motor VehAcc

Falls CommAviation

No

. D

ea

ths

Medical Errors are the8th Leading Cause of Deathin the United States

Source: National Safety Council

•Learn from previous mistakes

•Change organizational processes – don’t just “blame” people

•RCA, IR & FMEA are good tools for doing this

IOM: must focus on reducing faulty SYSTEMS to improve patient safety.

A Position Statement of the Joint Commission on Accreditation of Healthcare Organizations

JCAHORequires analytic methods to reduce risk

The Joint Commission on Accreditation of Healthcare Organizations is committed to improving patient safety through its accreditation process. Meaningful improvement in patient safety will eventually be reflected by a significant reduction in the number of medical/health care errors that result in harm to patients. Achieving this reduction is dependent upon:• Identification of the errors that occur• Analysis of each error to determine the underlying

factors – the “root causes”• Compilation of data about errors (frequency, type, cause)

• Dissemination of information about errors and causes

• Assessment of effectiveness of risk reduction efforts.

CMS requires that states have a comprehensive

QUALITY MANAGEMENT SYSTEM Planned, systemic, organization-wide approach to design,

performance measurement, analysis and improvement Assures compliance with standards Reduces adverse events Leads to ongoing improvement Crosses all waiver programs

States must therefore have policies and practices that include:• Assessment and protection of participant health and safety on an individual and aggregate basis

• Monitoring compliance with assurances on individual & aggregate basis

• Assessment of quality of services on individual & aggregate basis

USING DATA & ANALYTIC METHODS CAN HELP MEET THESE DEMANDS!

J. Tuller, Oregon Technical Assistance Corporation, 2003

HASHAPPENED

IS HAPPENING

WILL (MOST LIKELY) HAPPEN

I-D-QINFORMATION DRIVEN QUALITY

requires procedures that analyze

INFORMATION ABOUT

WHAT:

PAST

PRESENT

FUTURE

THAT CAN HELP US

LEARN based on the:

One way of lo

oking

at solutio

ns.

PASTInvestigation

Mortality ReviewRoot Cause Analysis

Trends Analysis Comparative Analysis

Audits

PRESENTIncident ManagementPersonal Outcomes

Monitoring/Site Reviews

FUTUREFMEA

(Failure Mode & Effects Analysis)

TCA (Task Criticality Analysis)

Predictive Statistics

I-D-Qis an ANALYTIC Process

and can help create a HOLISTIC view by using

and can help meet the expectations within the

CMS QUALITY FRAMEWORK

RETROSPECTIVE analyses that look at the

CONTEMPORARY analyses that look at the

PROSPECTIVE analyses that look toward the

IRIncident Management

LICLicensing

INVInvestigations

IRIncident Data

MRMortality Review

RCARoot Cause Analysis

FMEAFailure Mode & Effects Analysis

TCATask Criticality Analysis

PSPredictive Statistics

I-D-Q is an INTERACTIVE PROCESSUse retrospective and contemporary analyses to initiate prospective analyses.

RETROSPECTIVE PROSPECTIVECONTEMPORARY

I-D-Q is an INTERACTIVE PROCESSThen use results of prospective analyses to enhanceSystems and methods for Retrospective and Contemporary Analyses

Policies Practices StandardsServices

SYSTEM


LICLicensing

INVInvestigations

IRIncident Data

MRMortality Review





ANALYTIC METHODS

The Purpose of I-D-Q: CONTINUOUS IMPROVEMENT


LICLicensing

INVInvestigations

IRIncident Data

MRMortality Review





RETROSPECTIVE PROSPECTIVECONTEMPORARY

SYSTEMIMPROVEMENT• Changes in Policy• Protocol Development• Practice Standards• Focused Training• Better Design

Participant Safeguards

Participant-CenteredService Planning andDelivery

Provider Capacity andCapabilities

Focus

Quality Management Functions

QUALITYFRAMEWORK

ANALYTICapproaches to

performing

QUALITYFUNCTIONS can apply to most

FOCUS AREAS

Participant Access

Participant Rights andResponsibilities

Participant Outcomesand Satisfaction

System Performance

ImprovementDesign Discovery Remediation

PASTRetrospective

PRESENTContemporary

FUTUREProspective

DESIGN

DISCOVER

REMEDY

IMPROVE

TYPE OF ACTIVITY

Build into your QI system the use of data and analytic proceduresthat incorporate retrospective, contemporary and prospective approaches

Methods to Analyze Data and Events using structured and formal processes: • Root Cause Analysis• Investigations• Mortality Review• Audits• Comparative Analyses• Trends Analyses• Pattern Analyses

Methods to Collect & Use Information and Data to:• Take Corrective Actions• Track Follow-up and Outcomes• Assess Compliance • Manage Contracts and Staff• Assess Effectiveness of Changes

Methods to Use Results of Data Analyses to: • Establish Improvement Objectives• Design/change Policy• Modify Practices• Inform Consumers/Families• Adjust Consumer Plans/Services

Methods to Collect & Analyze Data re: • Abuse and Neglect Allegations • Unusual Incidents• Restraint Utilization• Medication Errors• Licensing Citations• Complaints• Personal Outcomes • Deaths

Methods to Analyze Data and Events using structured and formal processes: • Predictive Statistics• Failure Mode & Effects Analysis• Task Criticality Analysis

Examples of HOW Data and Analytic Methods can address Framework Functions

Specific analytic approaches (e.g., RCA & IR) have NOT yet been established as an evidence-based practice in health care – or DD.

FACE VALIDITY ONLYFor Analytic Processes – and most other “DATA”

Data is a tool for inclusion in a more comprehensive system. It is NOT a “be all, end all” solution to risk management, quality assurance or quality improvement.

According to Wald & Shojania (AHRQ, 2001),

Flawed analysis can result in:• blind pursuit of ABSOLUTE SAFETY• increasingly COMPLEX safeguards – prone to system failures• DECREASED INTEREST in quality improvement (avoidance)• EXPENSIVE “solutions” to the wrong problem

However, NO ANALYSIS can result in even GREATER RISK of harm to people - and - failure of our systems!

Much of the “data” – information - we use in DD has NOT been tested for its reliability or validity. And, a lot of it has NOT been properly analyzed, if analyzed at all. Be CAREFUL!

AHRQ = Agency for Healthcare Research & Qualitywww.ahrq.gov

BALANCE IS ESSENTIALDo not over-emphasize data!

We work in a REAL world with:• Limited Staff Resources• Competing Demands• Budgets

Must approach use of DATA & ANALYSIS

from a PRACTICAL perspective!

DATA can be powerful – but it is only one of many different tools we should use to build a strong Quality Management System.

IDEAL

PRACTICAL

If we try to buildthe IDEAL Analytic System we will need to pull resources from:

Good Consumer Planning, Direct Service & Support, Solid Clinical Services, Attentive Service Coordination, Timely IDT Review of Progress, Evaluation of Consumer Outcomes, Licensing and Certification, Risk Assessments, Decent Investigation Systems, Family Surveys, Contract Monitoring, Family Feedback, etc.

NECESSARY – BUT NOT SUFFICIENTData and analysis are only a small part of an overall quality management system.

EXAMPLE: Incident ManagementA truly comprehensive Incident Management System contains 8 Essential Elements:

1. Policies and Procedures2. Identification3. Notification4. Triggers Response

5. Documentation & Follow-up6. Analysis7. Reporting8. Quality Improvement

Data is a NECESSARY BUT NOT SUFFICIENT CONDITION. You need it. BUT, it must be logically integrated into the overall MANAGEMENT PROCESS.

ROOT CAUSE ANALYSIS

A Tool for Retrospective AnalysisDesigned to Reduce Risk of Harm

WHAT IS ROOT CAUSE ANALYSIS?

• Analytic technique designed to identify a broad range of factors that have contributed to or directly caused an adverse event

• Origins in engineering, transportation safety, information technology

• Expanded to health care in late 1990’s– JCHAO requirement for hospitals – Starting to be used in home health care services

• Focus is on:– Understanding WHY human errors occur– Developing prevention strategies

• Structured and formal process using a team approach

Root Cause Analysis is a powerful tool for evaluating adverse events in order to identify why they happened and what can be done to prevent them from happening again.

Common Response to

Adverse Events

“Would you pleeeease elaborate on ‘then something really bad happened’?”

WHY USE RCA IN MR/DD SYSTEMS?

MR/DD Director

Why use it in DD/MR?For the very same reasons it has become standard practice in Healthcare!

ARE WE DOING ENOUGH TO REDUCE

REASONABLE RISK TO THE PEOPLE WE SERVE? (And, how do we know?)

HOW MUCH & WHAT KIND OF RISK ARE

WE WILLING TO TAKE? (For acting – and for not acting.)

We serve thousands of people with a variety of different needs. Our service system is growing more complex every day. There are a lot of adverse events, most of which we don’t even know about.

We Must Ask 2 BIG Questions:

Omaha World Herald

July 11, 1998Deaths Spur An Inquiry at Facility

Iowa officials are investigating procedures at Woodward State Hospital- School….

Peoria Journal StarFebruary 27, 1999Six Doctors Charged with Mistreating Patients at Hospital for Retarded……

The Daily Record

November 1, 1997

AG Prosecuting Former Caregivers for

Neglect of Mentally Retarded Adults

The Milwaukee Journal Sentinel

April 3, 2002

Charges Filed in Group Home Neglect

CaseThe Cincinnati Enquirer 2001

Ohio’s Secret ShameIn Dayton, a mentally retarded woman is so severely dehydrated she curls up on an old

couch and dies of thirst.

In Cincinnati, a mentally retarded man having a seizure dies after a caregiver hangs up on

911 operator offering help.

In Columbus, a mentally retarded man on a dairy-free diet chokes to death of a “golf ball-

sized wad” of bread and cheese.

Whether by neglect, incompetence or accident, these are three of at least 12

people who died in questionable circumstances inside a state system that’s

supposed to protect them.

The Hartf

ord C

ourant

October

11, 1998

Deadly Rest

rain

t: A

Natio

nwide

Pattern

of D

eath

Managing

Risk.Or, failing to!

The DayNovember 3, 2003

Norwich Nursing Home Cited in Man’s DeathState Inquiry Finds Hamilton Neglected Retarded Patient, 36

RCA – ONE TOOL FOR HELPING MANAGE RISKRoot Cause Analysis is NOT a panacea. It is simply a tool that should be part of a broader system of incident management that focuses on the:

Discovery & Prevention of ERROR

Health Care – historically relied on retrospective identification:• Mortality and Morbidity Committees• Chart Reviews/Audits• Computerized Surveillance (e.g., pharmacy ordering)• Incident Reporting

Very time consuming and expensive options.Limited scope.Not usually used in DD/MR systems.

RCA & INCIDENT REPORTINGHistorically Closely Linked

1954 Flanagan began Critical Incident Reporting – military aircraft training accidents

Critical Incident Reporting began to be more widely used in the Military, Civilian Aviation, the Nuclear Power, Petrochemical Processing, and Steel Production industries. Engineering introduced RCA as a way of further analyzing critical incidents.

IR = Identification of preventable incidents; events that have or could have led to an undesirable outcome and that are directly reported by personnel involved in the event or its discover.

1975USFDA mandated IR for blood transfusion reactionsU.S. HEALTH CARE INDUSTRY:

1978Cooper introduced IR to anesthesiology

1990sCDC included IR in National Nosocomial Infection Surveillance SystemNew York State introduced IR into NY Statewide Transfusion System

1995 JCAHO mandated hospital-based surveillance system for reporting sentinel events (voluntary reporting to JCAHO)1997 JCAHO mandates Root Cause Analysis for sentinel events

RESEARCHOn IR/RCA Effectiveness in Health Care is Lacking

Incident ReportingAviation Safety Reporting System (ASRS): 30,000 reports each year

IF same level of reporting in Health Care = Millions of reports annually

JCAHO: Only 1,152 events in 6 years!And, 76% of these were untimely deaths.

Data represents significant UNDER Reportingin Health Care

Existing U.S. research studies suggest:• only 1.5% of all adverse events are reported• 6% of ADE are identified through IR systems• only 5-30% of surgical events are captured through IR (ACS)

Australian research (AIMS):• 2% of reported incidents resulted in patient death• 44% had negligible effect on patient outcomes• 90% had identified SYSTEMS failures• 79% had human failures

NO research studies to date - to establish BENEFIT to patient outcomes from implementing IR system in health care settings!

RESEARCHOn IR/RCA Effectiveness in Health Care is Lacking

Root Cause Analysis

Texas Hospital (Rex, et al., 2000):• applied RCA to all serious adverse drug events (ADE) for 12 mo.• 45% decrease in rate of reported ADEs (17 mo. Follow-up)• No fatal ADE on follow-up (however, low baseline)• Attribute change to “blame-free” RCA – led to:

• leadership focus on safety• numerous improvements to med ordering/distribution• targeted changes in staffing levels

Little published literature that evaluates the impact of RCA on sentinel events or error rates.

Blood Transfusion System (Kaplan, et al., 1998):• applied RCA to “unique events”• used classification system from petrochemical industry• 503 events reported – 1238 causes identified• Distribution of causes similar to research in petrochemicals:

• 46% - human failure• 27% - technical failure• 27% - organizational failure

SPECIAL NOTE: Be prepared for debate!

Adapted

Structured analytic processes can expose weaknesses in your systems that may pose moral – clinical - fiscal – programmatic – political – and especially legal DEBATE!

FIS

CALLY

RIG

HT

POLITICALLY RIGHT

Inevitable intersection

on theroad of IDQ

Introduction to

ROOT CAUSE ANALYSIS

TRADITIONAL APPROACHto significant adverse events

PROCESS =

INVESTIGATION

PURPOSE =To Identify

WHAT happened &

WHO was responsible

TYPICAL QUESTIONS:

What happened?

Did anybody do something wrong?

If so, who messed up?

What disciplinary or enforcement action is needed?

FOCUS = is often on

INDIVIDUAL fault

PROCESS =

ANALYSISPURPOSE =To Identify

WHY it happened &HOW to prevent its reoccurrence

FOCUS =

SYSTEMS change

TYPICAL QUESTIONS:

What exactly happened and in what sequence?

What factors may have contributed to human error?

What “barriers” could have prevented it?

What changes to our systems and processes should we make?

RCA APPROACHto significant adverse events

SOME BASIC TENETS OF RCA

• The goal of RCA is PREVENTION• Belief = errors and failures result from

flaws in the system – not just people’s actions or inactions.

• Try to find out WHY someone made a mistake, not just who made it.

• RCA requires thorough analysis of• Human factors• Organizational support systems• Formal and informal processes

• Based on a series of “WHY?” questions to identify actual and potential Contributory Factors that led to the event

Root Cause Analysis is a formal process of discovery that works to identify all of the human and system factors that contributed to – or allowed – an adverse event to happen.

ROOT CAUSE ANALYSIS

Process to Discover • WHAT happened• WHY it happened• HOW it can be prevented

Focus is on UNDERSTANDING, not BLAMING

Analyses CAUSE & EFFECT Relationships

Emphasis is on SOLUTIONS and System IMPROVEMENT

Analog: Something we are familiar with in MR/DD

If a consumer acted out and seriously injured other people.

WOULD YOU ACCEPT:• Staff blaming the consumer.• Kicking the person out of the program.• One person deciding what to do. • Use of a “trial and error” approach.• Using a “quick fix” based on assumptions.• Ignoring whether or not the behavior changes.

WOULDN’T YOU EXPECT

A systematic process to change the conditions that promote and maintain the problem behavior?

1. A team of people who know the consumer and have expertise in modifying behavior will meet to develop an intervention plan.

2. They will analyze the situation to gain an understanding of what antecedents and consequences are maintaining the target behavior(s).

3. Then they will implement strategies to modify those antecedents and consequences.

4. They will evaluate whether or not their intervention is working.

5. If it isn’t, they will make revisions – until change takes place.

GOOD program development for treating challenging consumer behavior is an analog for GOOD management of adverse events.

INTERVENTION Behavior Support PlanDevelop a plan to modify environmental

antecedents and consequences that promote and maintain the behavior/incident.

Prevention Strategies

TEAM APPROACH IDTPeople with knowledge of the person/incident and content expertise

RCA Team

SYSTEMATIC ANALYSIS Functional AnalysisSystematically evaluate environmental and

personal characteristics related to the target behavior/incident.

Root Cause Analysis

EVALUATION Track Target Behavior(s)Establish method to collect data and track

impact or effectiveness of intervention.

Track Incidents

REVISION Modify the Support PlanPlan for changing the intervention if

evaluation indicates it isn’t having the desired effect.

Modify Prevention Strategies

CONSUMER INCIDENT

TEAM APPROACH

Consumer, and people who: • Know the consumer• Special knowledge of behavior analysis• Will implement the intervention

People who: • Know about the specific incident• Special knowledge of incident analysis (RCA)• Special content knowledge (based on the type

of incident)

• Manage or supervise staff (based on the type of incident)

RCA

IDT

When developing any effective intervention you need to involve people who have specific knowledge of the person/incident under study and specialized content expertise.

RCA is just like FA

A – B - CFUNCTIONAL ANALYSIS includes an examination of the behavioral sequence associated with the target behavior.

ANTECEDENTS BEHAVIOR CONSEQUENCE

SETTING EVENTS

ANTECEDENT STIMULI

WHY? So we can understand under WHAT CONDITIONS the behavior occurs and does not occur. This provides information about the environmental cues and “chain” of events that can trigger or set off the target behavior, and which can be modified to prevent or reduce the probability the target behavior will occur in the future.

Both use SYSTEMATIC ANALYSIS

A – B - C

Like FA, ROOT CAUSE ANALYSIS also includes an

examination of the sequence of events that were associated with the adverse incident.

SETTING EVENTS

ANTECEDENT STIMULI

WHY? So we can understand WHAT were the CONDITIONS that contributed to the incident. This provides information about the environmental factors that increase the probability the incident will occur in the future – so we can change or modify them.

TEAM APPROACHPeople with knowledge of the person/incident and content expertise

SYSTEMATIC ANALYSISSystematically evaluate environmental and personal characteristics related to the target behavior/incident.

The same logic applies to the development of effective interventions for consumers as well as for staff. An effective behavior support plan AND an effective risk management plan

strive to REDUCE THE FUTURE PROBABILITY of certain behaviors. The foundation of both rests on:

USE THE SAME APPROACH

WHEN TO USE ROOT CAUSE ANALYSISRoot Cause Analysis can be used to study a wide variety of incidents. It is NOT a substitute for investigations. And it is a demanding process that should be used selectively.

To review EVERY INCIDENT or potential system failure.

As a SUBSTITUTE for Investigation.

If the reason is OBVIOUS.

When there is NOT TIMEto go through a FormalProcess of discovery.

The cause of a MAJOR system Failure is NOT CLEAR.

There are REPEATED PROBLEMS that are ascribed to human error.

A sentinel event results in DEATH or SERIOUS INJURY.

There is TIME to gatherInformation, use a Team Process and “Think” beforeActing.

USE ROOT CAUSE ANALYSIS WHEN: DO NOT USE IT:

PROCESS OF

ROOT CAUSE ANALYSIS

ADVERSE EVENT. Undesirable incident that causes harm or calls into question the adequacy of care.

BARRIER. Represents a process, environmental change, or system that can prevent or reduce the probability of an adverse event.

CONTRIBUTORY FACTOR. Additional reasons, beyond the root cause, that an adverse event has occurred. They increase risk by setting the stage for or contributing to the probability of an adverse event.

INCIDENTIAL FINDING. Problems, inefficiencies or failures in a process or system that did not contribute to the incident under consideration but which require attention by the organization in order to prevent other types of adverse events.

PROXIMATE CAUSE. Most obvious reason an adverse event occurred. Immediately precedes the incident. Usually identified in investigations.

ROOT CAUSE. Most basic and fundamental or underlying cause of an adverse event. It is the “root” of the problem and must be addressed if prevention strategies are to be effective. In most instances the root cause is a systems failure, not human error.

SENTINEL EVENT. Adverse event that is unexpected and that

leads directly to or places someone at risk of death or serious harm.

SOME SPECIAL TERMSAND CONCEPTS IN RCA

HOW DOES IT WORK IN CONNECTICUT?

• DMR Policy and Procedure guide utilization of Root Cause Analysis to evaluate selected SENTINEL EVENTS.

• The Commissioner, Deputy Commissioner, Director of Quality Assurance or the Chairperson of the IMRB can request a RCA for any significant incident.

• DMR Strategic Leadership Center is responsible for managing and coordinating the RCA process. Team membership includes:

• Team Leader and Meeting Facilitator• Personnel familiar with the incident under review• Representatives from QA, Investigations, Management• Additional members – based on issue under study

• Each RCA should review and analyze:• The sentinel event• Background, context, and potential contributory factors• Risk reduction actions already taken• Recommended prevention strategies• Incidental findings, when present

CT DMR has established new procedures that activate a formal RCA for special incidents that may provide valuable information for prevention and systems improvement.

ADVERSE EVENT

Reporting, Protection,Investigation,

Management Review& Action

Is it a SERIOUS Event

?

Does Comm,

Dep Comm, Dir QAor IRMB Chair request

a review?

Is a SystemFailure Possible

Cause/Factor?

SLC assigns aRCA TEAM

COLLECTINFORMATION -

SEQUENCEANALYSIS

Team IdentifiesCONTRIBUTORY

FACTORS &Incidental Findings

Group Factors –Identify

ROOT CAUSE

Team IdentifiesPREVENTIONSTRATEGIES

Team Prioritizes Strategies & MakesRecommendations

Report reviewedby DMR Leadership

LeadershipActs

1 2 3 4

5 6 7 8

NO

YES

NO

YES

NO

YES

Process in Connecticut

8 STEP APPROACH FOR

ROOT CAUSE ANALYSIS

1. ASSIGN TEAM

• Team membership will usually include:• Team Leader and Meeting Facilitator (with knowledge of RCA process)• Personnel familiar with and/or involved in the incident under review• Representatives from QA/QI, Risk Management, Special Protections,

Program Management• Additional members with special content knowledge – based on issue

under study

• Members should be formally appointed an agency executive

• In most instances the team will include 5-7 individuals

RCA is best conducted using a team of knowledgeable staff that includes representation from management, quality improvement, and the program unit where the incident under study took place.

FORMAL ROOT CAUSE ANALYSIS WILL USUALLY REQUIRE at least 2 TEAM MEETINGS - PLUS ANALYSIS AND RESEARCH WORK IN-BETWEEN MEETINGS.

2. COLLECT INFORMATION

• Team leadership will usually assign tasks to members prior to the first meeting.

• Documentation may include the following type of information:

• Investigation report(s) – (VERY helpful when available)

• Autopsy or Police reports – available photographic evidence (if available and relevant)

• Interviews and/or staff notes/log

• Consumer file

• Related Incident reports

• Related Polices and agency guidelines

• Training records

• Program and/or site descriptions

• Sequence Analysis

The RCA team will need detailed information about the incident under review, organizational practices – policies - procedures, and accepted practice standards.

SEQUENCE ANALYSIS

• Sequence Analysis is a simple but essential process in RCA that helps avoid “assumptions” by identifying what really took place and in what order.

• It should be completed by personnel very familiar with the incident • Source data can come from:

• Investigation reports• Program notes/logs• Statements/recollections

Sequence Analysis provides the “raw data” for Root Cause Analysis by objectively summarizing exactly WHAT HAPPENED and in WHAT ORDER.

DATE TIME WHO EVENT or ACTIVITY

“First – This Happened.Then – That Happened.Then – This and This.Followed by……”

FORMAT FOR SEQUENCE ANALYSIS

IT IS OFTEN HELPFUL TO FLOWCHART THESEQUENCE ANALYSIS TO VISUALLY ILLUSTRATEWHAT TRANSPIRED BEFORE, DURING AND AFTERTHE INCIDENT UNDER STUDY.

START EVENTInclude WHO & WHEN

DECISIONPOINTS

END

EVENTInclude WHO & WHEN

FLOWCHARTS

Excellent analytic tool:• If you can’t easily show a process, it’s probably not functional• Help visualize what happened – or what should happen• Can compare actual event sequence with procedural requirement• Help locate critical points to insert “barriers”• Instructive aid when training in new procedures

Yes

No

3. IDENTIFY FACTORS THAT MAY HAVE CAUSED THE INCIDENT

• Review the INCIDENT – provide background information• Review SEQUENCE ANALYSIS and flowchart for the actual incident• Compare to WHAT SHOULD HAVE HAPPENED

• Develop flowchart based on agency policy/practice and/or national practice standards

• Identify deviations and system errors/failures• Identify the PROXIMATE CAUSE – what was or was not done

immediately before the incident that resulted in the event.

The Team will carefully review the incident, all relevant documentation and begin the process of analysis to discover not only what happened, but WHY it happened. This leads to a better understanding of what factors allowed or contributed to any error.

Focus is NOT on BLAMING A PERSON or group!

It IS on FACTORS that did/could have led to or contributed to any error.

REVIEW COMMON FACTORS

COMMON FACTORS CHECKLISTFor Identifying Causes and Contributory Factors

When attempting to identify potential causes and contributory factors related to an incident or system failure review the following common factors. Place a checkmark next to each factor that was or may have been present and therefore influenced or caused the adverse incident or failure. Add additional factors discussed by the team.

STAFFING FACTORS Workload There were not an adequate no. of staff present to carry

out the assigned duties in the expected timelines. The amount of work was not consistent with the duties and assigned tasks.

Awareness Staff were not aware of what tasks they needed to do and when those tasks were to be performed

Staff Skills Staff did not have the necessary skills and knowledge to perform the assigned duties – staff didn’t know how to perform required tasks.

Prioritization Staff or supervisors did not set priorities for multiple tasks and responsibilities. Assumed priorities were not consistent with organizational expectations.

Time Allocation

Staff did not set aside sufficient time to perform priority tasks within expected timelines.

Working with Others

Staff were not aware of whom they needed to work with to perform tasks and work together as a team when required.

Experience Staff did not have necessary experience performing the task. If not experienced, they didn’t know whom to access for direction and support.

Supervision Staff were not provided with consistent supervision from an assigned supervisor or peer network.

Relationships with Other Staff

Staff did not have positive and mutually supportive relationships with other personnel. Conflicts or poor relationships interfered with the ability of staff to properly complete assigned tasks in a safe and/or efficient manner.

Relationships with Consumers

Staff did not have a positive and supportive relationship with the people they were assisting.

Motivation Staff did not have incentives for performing the required activity, i.e., it was seen as unimportant or not recognized by the organization. There were no clear consequences for completing or not completing the activity.

The team should work to identify what conditions or factors could have CONTRIBUTED to any perceived errors by referencing theCOMMON FACTORS CHECKLIST.

• The Checklist provides a series of potential causes or contributory factors common to DD service systems that can serve as a prompt or cue.

• Factors are categorized into clusters:• Staffing Factors• Consumer Factors• Communication• Equipment• Assessment/Planning• Policies• Environment• Organization

• Select those that are relevant • Modify as necessary• Add additional factors

DESCRIBE FACTORS IN OBJECTIVE TERMS & FOLLOW THE “FIVE RULES OF CAUSATION!”

5 RULES OF CAUSATIONThe following rules are adapted from the National Council on Patient Safety

1. Describe the “cause and effect” relationship.• Describe the linkage between the stated cause and the adverse event or

outcome. • Do not assume others will be able to automatically understand the

relationship. • Describe how and why something led to a mistake

2. Do NOT use negative descriptors. • Avoid shorthand statements like “inadequate” or “poorly written.” • Include detailed explanations.

3. Each human error should have a preceding cause. • Most adverse events involve human error - the goal of RCA is to understand

WHY the error occurred. • Do not simply list the error. Remember, it is the cause of the error, not the

error itself, that will generate productive solutions.

4. Each procedural deviation should have a preceding cause.

• Failure to follow a prescribed procedure is not directly manageable. • Need to understand what caused or contributed to the deviation (didn’t

understand it, forgot, not enough time, distraction, etc.)

5. Failure to act is not causal unless there is a clear “duty to act.”

• Formal requirements, practice standards, or other established guidelines must have been present.

• Must be a clear recognition in the system that there was a duty to perform.

NOTE INCIDENTAL FINDINGSSpecial concerns and practices that could lead to errors or contribute to other types of system failures should be documented.

• The process of review will usually identify a variety of concerns, issues, and factors that could lead to other types of errors (not directly related to the adverse event under study). These are called INCIDENTIAL FINDINGS – and should be:• Noted on a separate wall chart• Included in the final report for

consideration by agency leadership• Only BRIEFLY discussed – they are

“incidental” to the issue under review

Incidental Findings

Note and include in final report, but do NOT spend time analyzing or discussing possible solutions.

Avoid becoming side-tracked on issues thatmay be important, but are not a cause of the incident under study.

4. GROUP THE FACTORS

• Continue the process of asking “WHY” to identify additional factors

• Group factors (use headers in CF Checklist or develop new headers)

• Develop a CONTRIBUTORY FACTORS DIAGRAM to illustrate the relationships between factors• Hierarchal illustration• Starts with the error or failure• Next is proximate cause• Arrange Groupings according to priority – list individual factors

under headings• Go back and elicit additional reasons or factors that may have led to the

error – modify the diagram as needed

Once a number of contributory factors have been identified it is helpful to group them into CLUSTERS and organize them according to priority or importance.

Not tested on procedure

Policy not clear on requirementsInexperienced staff on

duty

Staff hesitant to call nurse on weekends

Classroom Training only

New admission – staff not expecting noncompliance

Staff felt pressured to finish –Busy evening recreation schedule

GROUP FACTORS INTO CLUSTERS

Training Factors

Staff Factors

Client Factors

Organizational Factors

FailureNo review or protective servicesplan following a major incident.

Proximate CauseCase Manager did not review the incident report.

Contributory FactorsSystems deficiencies that led to failure and proximate cause

Staffing IssuesNo Training in IR requirementsLimited SupervisionLack of Experience

EnvironmentLimited computer accessNoisy workplace/distraction

Policy/ProcedureOld Policy- not updatedVerbal instruction contradicts policyLack of routing/triage protocol for IRs

Root CauseSystems to assure prompt follow-up to IRs not available

FailureNo review or protective servicesplan following a major incident.

Proximate CauseCase Manager did not review the incident report.



Staffing IssuesNo Training in IR requirementsLimited SupervisionLack of Experience

EnvironmentLimited computer accessNoisy workplace/distraction

Policy/ProcedureOld Policy- not updatedVerbal instruction contradicts policyLack of routing/triage protocol for IRs

Root CauseSystems to assure prompt follow-up to IRs not available

EXAMPLE OF CF DIAGRAM

IDENTIFY ROOT CAUSEOnce all of the potential contributory factors have been identified and organized the team should zero-in on the most essential or root cause of the error/failure.

• The ROOT CAUSE represents the most fundamental or essential reason that the failure took place.

• May be a combination of factors (e.g., lack of clear policy and practice guidance plus inadequate training and supervision led to staff not knowing what to do or how to do it)

• Often related to a system or process problem• Should not be focused on an act of omission or commission by a person – but

rather what was present or absent that allowed the human error or equipment failure to occur

FAILURE orERROR

PROXIMATE CAUSEPROXIMATE CAUSE

MULTIPLECONTRIBUTORYFACTORS

MULTIPLECONTRIBUTORYFACTORS

ROOTCAUSEROOTCAUSE

Think About the 7 COMMON REASONS FOR STAFF “NONCOMPLIANCE”

Don’t Know

WHAT to do

• Incomplete instructions• Confusing or contradictory instructions• Instructions too general• Reading levels too advanced• Afraid to ask questions• Tasks are too complex• Information overload

Don’t Know

HOW to do it

ALL of the above, plus: • Using new or novel equipment• No demonstration of how something is done• No guidance in the program setting

Don’t Understand

WHY

• Lack “big picture”• Risk seems too general, not personalized• See their situation as different from others• Don’t believe it really matters• Co-workers and supervisors do not stress importance

They

FORGET• Environment is distracting• Numerous competing demands and activities• Become sidetracked with other consumers• No visible cues or prompts present

1

2

3

4

7 COMMON REASONS FOR STAFF “NONCOMPLIANCE”

Don’t Manage

TIME well

• Poor time management skills• Hectic work schedule• Lack structured routines• Schedule too many tasks in limited time period

5

Lack Necessary

RESOURCES• Not enough staff on duty for type of activity• Inadequate or antiquated equipment for tasks• Limited supplies• Limited supervision or technical assistance

6

Don’t WANT to do it

• Receive little or not recognition for the activity• Task is unpleasant or boring• Task requires more effort than other tasks• Task can lead to embarrassment or injury• History of being criticized or “punished”• Interferes with more pleasant activities• Don’t like the consumer• Want to avoid interactions with others involved in task• Afraid they will make a mistake

7

Take the time to better understand WHY an error has occurred. Don’t be satisfied with “Joe failed to……” as an adequate explanation for an adverse event.

• BRAINSTORM solutions that focus on preventing or modifying the most important contributory factors

• Think about WHY someone did or didn’t do something - the 7 Common Reasons (solutions usually automatically flow)

• Use Flowcharting to revise a process – make it “make sense”

• Identify BARRIERS• Equipment, technology, processes• Automatically blocks unwanted action

• Look outside to learn about alternatives - Perform literature search and contact other industries, agencies or state DD systems to identify potential strategies and solutions: Find out what has worked elsewhere!

• PRIORITIZE SOLUTIONS according to their potential effectiveness, feasibility and resource requirements.

• Use Efficiency Diagram to illustrate• Assess agency “readiness” for change

5. IDENTIFY PREVENTION STRATEGIES

The team should brainstorm potential solutions for each of the identified contributory factors.

EFFORTCost/Resources

LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

ven

ess


LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

ven

ess

LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

ven

ess

GOOD

VERY

GOOD

POOR

EFFICIENCY DIAGRAMS


LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

veness

Develop Checklist Reduce Caseload

Provide Training

Revise Protocol toclarify responsibilities

Develop new InformationManagement System


LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

veness


Provide Training



LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

veness


Provide Training




LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

veness


Provide Training




LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

veness


Provide Training



LOW HIGH

LOW

HIGH

IMP

OR

TA

NC

EIm

pact/

Eff

ecti

veness


Provide Training



EXAMPLE

6-7. COMMUNICATE FINDINGS

• FINDINGS should focus on objective information and include information about:

• Root cause and major contributory factors that resulted in the error or failure• Incidental findings that require follow-up• Recommended actions to prevent future occurrence(s)

• REPORT can utilize different formats – it should always• Be reviewed by all team members before being issued• Incorporate changes recommended by team members (unless they conflict with

majority opinion)• Be signed by the team leader

• FORMAL PRESENTATION of the findings and recommendations to the agency leadership should be scheduled by the Team Leader and one or two representatives from the team - after sharing draft report with the person who requested the RCA in the first place.

• Focus on facts and rationale for findings • Include pros and cons of various recommendations• Be prepared to discuss alternative strategies

After all potential contributory factors and reasonable solutions have been identified the team needs to develop a set of recommendations and communicate its findings to leadership.

REPORTS CAN BE WRITTEN IN:• Traditional Narrative Format• RCA Form

Whatever the format, reports should include:• Description of Adverse Event• Purpose of the RCA• Who requested it and Team membership• Proximate cause of the event• Contributory factors (grouped with explanation of how)• Root Cause – if identified• Corrective Actions already taken• Potential barriers and “solutions” – prioritized• Incidental findings• Recommendations• Signature of Team Leader

REPORTS

Possible Format for Documenting Results of RCAAdapted from a template utilized by the Australian Department of Human Services for use by Health Care Organizations and Hospitals

AGENCY: Sentinel Event Reference No.:

City/Town: Date of Event: Date RCA Completed:

1. THE EVENT – Describe what happened and any harm that resulted. Identify the proximate cause, if known.

RCA Team Members:Team Leader:

2. BACKGROUND & FACTORS SUMMARY– Answer the following questions (brief summary only- attach supporting documents).

2.1 What was the sequence of events that was expected to take place? Attach flowchart if available.

Description:

2.2 Was there a deviation from the expected sequence?

Yes No

If YES, describe the deviation. Attach flowchart if available.

2.3 Was any deviation from the expected sequence likely to have led to or contributed to the adverse event?

Yes No NK

If YES, describe with causal statement.

2.4 Was the expected sequence described in policy, procedure, written guidelines, or included in staff training?

Yes No NK

If YES, cite source.

RCA FORM

The team is responsible for analyzing and recommending. It is the responsibility of Agency Leadership - NOT the RCA Team - to make final decisions about what actions – if any – the agency should and will take.

8. AGENCY ACTION

RCA TEAM AGENCY LEADERSHIP

• Gather information

• Review the incident

• Analyze and identify what contributed to the problem

• Study possible solutions

• Make recommendations

• Present findings

• Review findings

• Decide what action to take

• Assign responsibilities

• Manage any change

• Follow-up and track

RESPONSIBILITIES

FAILURE MODE & EFFECTS ANALYSIS

A Tool for Anticipating Problems and Correcting them Before they Happen

Failure Mode and Effect Analysis

(FMEA) is a systematic method of

identifying and preventing failures

BEFORE they happen.

FMEA is a prospective analytic tool. It is not based on an actualadverse event - but rather anticipates and corrects system failures before they actually occur.

WHAT IS FMEA?

Since FMEA is not dependent upon or require the examination of something “bad” that has already happened, it is less “threatening” and controversial than Root Cause Analysis.

“Borrowed” from Engineering

Like RCA, Failure Mode and Effects Analysis was used for decades in engineering before being introduced to healthcare:

• Military and civil aviation• NASA• Nuclear power generation• Chemical processing• Automotive design and engineering

In these industries a systems failure can have catastrophic results that include significant loss of life, serious injury, and major economic loss.

Tools have therefore been developed to ANTICIPATE failures and introduce product or process design changes before a problem becomes manifest.

Historically in healthcare AND in MR/DD:

We have NOT focused much energy on preventing human (staff) error.

We HAVE “relied upon” errorless performance of staff – professional and paraprofessional alike.

We tend to REACT to incidents, introducing changes to practice only after major problems have taken place.

REACTIVE SYSTEMS

JCAHO now requires proactive approaches to risk management

JCAHO Standard LD.5.2

Identify and prioritize HIGH RISK PROCESSES Annually Select ONE Analyze potential “FAILURE MODES” Identify the possible EFFECTS of the failure Conduct a ROOT CAUSE ANALYSIS for the most

critical effects REDESIGN the process to minimize risk of failure TEST the new process Implement MEASURES of effectiveness Implement strategies to MAINTAIN the effectiveness

Failure Mode and Effects Analysis is becoming the “preferred” method to meet this new standard.

FMEA v RCAHOW ARE THEY

ALIKE?HOW ARE THEY

DIFFERENT?• Teams• Use Flow Charts• Focus on Systems• Avoid Blame• Cause & Effect Analyses• Use Brainstorming• Focus on Prevention• Result in Change

• FMEA studies Process – RCA looks at Chronological sequence• FMEA is prospective – RCA is retrospective• FMEA “chooses” a topic – RCA looks at an actual event• FMEA asks “What If” – RCA asks “Why”• FMEA emphasizes Testing the change

FMEA:10 STEP PROCESS

1. Identify High Risk processes a. Use Incident data, results of Investigations, RCA and/or Mortality Reviewb. Select a process for FMEA

2. Assemble Team

3. Complete Process Analysis & Flow-chart 4. Identify all possible Failure Modes for each step

5. Develop RPN for each failure mode and Prioritize 6. Identify Improvements and prioritize them

7. Develop Change Plan

8. Identify Outcome Measures9. Implement Changes

10. Measure and Evaluate Effects

STEP 1: IDENTIFY HIGH RISK PROCESSES Leadership, Quality Staff, Risk Managers,

Direct Support Professionals, and others can RECOMMEND activities they believe may be or could pose high risk of failure.

Use existing DATA to target activities Incident reports and analysis Mortality review results Incidental findings from RCA A/N Investigations

Keep the scope limited – focus on SPECIFIC processes or activities

STEP 2: ASSEMBLE TEAM Like RCA, FMEA benefits from analysis by an

“Interdisciplinary Team” Different perspectives and experiences Broadens knowledge base

MINIMUM requirements for Team: Content experts (knowledge of best practices in area under study) Persons who actually use the process Management QI and/or risk management staff Person familiar with FMEA and/or RCA

Define expectations and ROLES Team Leader Meeting Facilitator Note Taker

Failure Mode and Effects AnalysisSAMPLE WORKSHEET FOR CALCULATING THE RPN

Risk Prioritization Number

Process Step Failure Mode A B C AXBXC Priority

Impact of Failure

1 - 10

Probability1 - 10

Discoverability1 - 10

Risk Prioritization No.

RPN1 - 1000

FMEA is a STRUCTURED PROCESS: WORKSHEET

STEP 3: PROCESS ANALYSIS Process analysis is the same thing as Task Analysis in special education. It requires you to list the steps that take place in any given activity or process.

1. List the steps (task analysis)2. Identify “sub-processes” for higher risk or more

complex steps3. Flow Chart (process)4. Label each step (1, 2, 3….)

Bring to Bathing Area

Prepare Water in Tub

Undress Place in Seat Bathe

• Check temp• Fill halfway

• Get assistance• Properly position• Prep for transfer

• Get supplies• Notify staff

• Check temp • Use lift• Position seat• Use straps• Lower seat

etc.

1 2 3 4 5


Impact of Failure

1 - 10

Probability1 - 10



RPN1 - 1000

Failure Mode and Effects AnalysisWORKSHEET FOR CALCULATING THE RPN

Risk Prioritization NumberSTEP 3

STEP 4: HAZARD ANALYSISFor each step or process, identify all of the potential Failure Modes that could result in error. Use literature and group experience as a guide.

1. Use Brainstorming2. Identify ALL potential failures – mechanical and

human3. List each failure in 2nd column on worksheet

QualityHealthCare.org has an interactive FMEA Tool on-line that also includes the results of FMEAs performed by other users. These examples can be informative if the process you are studying is similar.

STEP 4


Impact of Failure

1 - 10

Probability1 - 10



RPN1 - 1000

STEP 5: CALCULATE THE RPNThe Risk Prioritization Number is a simple quantitative method for rank ordering – prioritizing – each of the failure modes, to allow identification of the most to least critical.

RPN = Impact X Probability X Discoverability

How bad is it if a failure occurs?

How often will this type of failure take place?

What is the likelihood we would discover the failure and be able to stop it?

Many different scoring methods can be used as long as they allow differentiation of the various failures and process steps.

STEP 5A: CALCULATE THE

IMPACT SCORE

CATASTROPHIC Failure would result in the entire process failing and would have a very

negative impact on individuals or result in major economic loss. Potential for loss of life and/or severe injury.

SEVERE Failure would result in high probability that the entire process might fail. Potential for severe injury and/or economic loss if not detected.

MODERATE Failure would result in negative impact on the entire process although would most likely not result in complete failure. Potential for moderate injury or compromise of a clinical or personal outcome, major disruption and inconvenience, and/or possible economic loss.

MILD Failure would result in mild negative impact on the process. Potential for minor injury and only a minimal impact on a clinical or personal outcome. Consumers and/or staff would be somewhat inconvenienced.

Negligible economic loss.

MINOR/NONE Failure does not impact the entire process. No potential for injury or negative impact on a clinical or personal outcome. May involve minor inconvenience. No noticeable economic loss.

10

75

3

1

POINTS CATEGORY DESCRIPTION

STEP 5a


Impact of Failure

1 - 10

Probability1 - 10



RPN1 - 1000

STEP 5B: CALCULATE THE

PROBABILITY SCORE

VERY HIGH Very likely to occur. High frequency (multiple times each day) is probable and/or has been observed.

HIGH Likely to occur. Relatively high frequency (weekly) is probable. Has been observed to take place.

MODERATE Somewhat likely to occur. Estimate monthly or quarterly occurrence rate. Sometimes observed.

LOW Rather unlikely to occur. Estimate only yearly or every few years. Very rarely observed in the past.

VERY LOW Very unlikely to occur. Estimate occurrence only every five or more years. Has not been observed in the past.

10

7531

POINTS PROBABILITY DESCRIPTION

STEP 5b


Impact of Failure

1 - 10

Probability1 - 10



RPN1 - 1000

STEP 5C: CALCULATE THE

DISCOVERABILITY SCORE

VERY DIFFICULT Very hard to detect a failure. Activity or step not seen by others. Not followed by obvious step in process. Signs of failure or problem take place too late to remedy or correct.

DIFFICULT Hard to observe or detect when a failure takes place. Activity usually not seen by others. Signs of failure probably won’t be detected until after the activity is completed.

MODERATE Somewhat difficult to detect. May be observed by others who could intervene if vigilant. Signs of failure usually present, but only after

failure has taken place but activity is not completed.

EASY Fairly easy to detect errors. Usually observed by others. Mistake is rather noticeable and evident. Signs of failure are present immediately after step is completed.

VERY EASY Very easy to detect errors. Always done in the presence of others. Mistakes are very noticeable. Next steps cannot take place until error is corrected.

10

7

5

3

1

POINTS DISCOVERABILITY DESCRIPTION

STEP 5c


Impact of Failure

1 - 10

Probability1 - 10



RPN1 - 1000

CALCULATE RPN

RPN = Impact X Probability X Discoverability

RPN will range from a

LOW of 1 to a

HIGH of 1000

How Severe is the result of Failure?

How Likely is Failure? What is the Likelihood Failure will be observed and stopped?

EXAMPLE 1:Failure to grind food properly for a person on a special diet who has a significant swallowing disorder and who lives alone with PAs sent by a home health agency.

IMPACT ?PROBABILITY ?DISCOVERABILITY ?

Total RPN =

EXAMPLE 2:Failure to serve two vegetables to residents of a small CLA that has at least two staff on duty at meal time.

IMPACT ?PROBABILITY ?DISCOVERABILITY ?

Total RPN =

EXAMPLES

757

245 3

131

STEP 5d


Impact of Failure

1 - 10

Probability1 - 10



RPN1 - 1000

COMPLETE STEP 5: SET PRIORITIES

RANK ORDER FAILURE MODES (Highest RPN is the most important or critical – Lowest RPN is the least critical)

Evaluate STEPS or Processes for the number of highly critical failures

Focus your prevention strategies on the failures and/or steps that have the highest RPN and therefore pose the greatest risk of error and harm.

STEP 6: IDENTIFY IMPROVEMENTS

Brainstorm POTENTIAL CHANGES & ENHANCEMENTS that could CONTROL or prevent the failure(s)

Evaluate options for practicality and efficiency

Common Strategies

• Build in Redundancy• Simplify• Reduce Reliance on Memory• Standardize • Use Checklists• Install Visual Prompts/Cues• Eliminate “Look and Sound Alikes”• Institute “Fail-Safe” Designs

Process Redesign• Staff “Buddy” System• Mentoring• Teamwork• Increase Direct Communication• Operational Procedures• Reduce “Fear”• Reward Reporting/Debriefing• Culture of Safety

Organizational Change

STEP 7: CHANGE PLAN

Identify the “best” solutions Develop and write-up a simple CHANGE

PLAN or proposal REVIEW with agency leadership

Rationale Options Pros and Cons Obstacles and Costs

Take the time to carefully “spell-out” the results of the analysis and the reasons for proposing change. Honestly identify obstacles and costs. Show the risk if nothing is done.

STEP 8: OUTCOME MEASURES

Identify series of MEASURES to assess if proposed changes actually make a difference

Keep it SIMPLE and “Doable” Include staff feedback (after all they are usually the

target for intervention)

Goal = implement change and see if it has the intended impact – Increased Safety and Reduced Risk of Harm

Do the same thing you would for a new teaching plan or behavior management program for people you serve. Plan, Do, Measure, Change.

STEP 9: IMPLEMENT CHANGE

Consider using a Project Plan to map out steps for implementation

Consider simulation or setting up a PILOT before expanding to entire system Work out the “kinks” first Test your ideas before you roll-out the change

Make sure you follow through – DO IT!

STEP 10: EVALUATE CHANGE

Use identified Outcome Measures to assess whether or not the change: Is working Is being accepted Is actually being implemented Has any unanticipated negative consequences

MODIFY as needed REPORT BACK to leadership Include outcome in final report or

addendum to original FMEA report

RESOURCESfor learning more about

Root Cause Analysis and

Failure Mode & Effects Analysis

Some Helpful ResourcesBongiorno, J. (2000) Improving FMEAs. Quality Digest, 20 (10), October 2000, p.37.

Cohen, M.R., et al. Failure Mode and Effects Analysis: A Novel Approach to Avoiding Dangerous Medication Errors and Accidents. Hospital Pharmacy, 29 (4), April, 1994, p. 319. (2000)

DeRosier, P.E., Stalhandske, E., Bagian, J.P. and Nudell, T. (2002) Using Health Care Failure Mode and Effects Analysis: The VA National Center for Patient Safety’s Prospective Risk Analysis System. The Joint Commission Journal, May, 2002, p. 248.

Doing What Counts (2000). Doing What Counts for Patient Safety. Federal Actions to Reduce Medical Errors and their Impact. Washington D.C. Report of the Quality Interagency Coordination Task Force, February, 2000. To access the report go to: http://www.quic.gov/Report/

Failure Modes and Effects Analysis Tool. QualityHealthCare.org. www.qualityhealthcare.org/qhc/workspace/tools/fmea/

Hirsch, K and Wallace, D. (2001) Step by Step Guide to Effective Root-Cause Analysis. Opus Communications, Marblehead, MA

Hohmann, S.F. and Hoppes, M.M. (2003) Improving Patient Safety with Statistical Correlation. HCPro, Inc., Marblehead, MA.

Institute of Medicine (1999). To Err is Human: Building a Safer Health System. Washington D.C. National Academy Press.To access the report go to: htttp:// www. nap.edu/books/0309068371/html

Marder, R. and Sheff, R.A. (2002) The Step by Step Guide to Failure Modes and Effects Analysis. Opus Communications, Marblehead, MA.

Morey (1994). Error Reduction as a Systems Problem. In Human Error in Medicine, Marilyn S. Bogner (Ed)., Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Reason, J. (1990) Human Error. Cambridge Univ Press, Cambridge, U.K.

Spath, P.L. (2000) Error Reduction in Health Care: A Systems Approach to Improving Patient Safety. Jossey-Bass, San Francisco.

Staugaitis, S.D. (2002) Root Cause Analysis: A Summary of Root Cause Analysis and its Use in State Developmental Disabilities Agencies. Developed by Human Services Research Institute, under subcontract with the MedStat Group, Inc., for the Centers for Medicare and Medicaid Services. HSRI, Cambridge, MA 2002. www.hcbs.org/quality/hsri/RCAManual.doc

Wachter, R. (2001) Making Health Care Safer: A Critical Analysis of Patient Safety Practices. Evidence Report/Technology Assessment No. 43, AHRQ Publication 01-E058, July 20, 2001.

ADDITIONAL WEB RESOURCESThe following resources may provide additional assistance to readers wishing to learn more:

www.rootcauseanalyst.comwww.patientsafety.govwww.outsights.com/systems/rca/rootca.htmwww.clinicalrisk.health.vic.gov.au/rca.htmwww.accreditinfo.com/howto/HHH/howtoex.cfmwww.va.gov/ncps/causation.htmlwww.va.gov/ncps/tools.html

For more information or assistance please feel free to contact:

Steve Staugaitis at:

[email protected] or

[email protected]

Case Example 48 y/o woman – Midwest State Severe MR – bi-polar disorder Living in private CLA – attending full time day program Lithium levels WNL (.9 mmol/L) Medication packaged in blisterpak 1 week after starting new blisterpak signs of distress begin – last one month

Crying and whining Unsteady gait progressing to inability to walk or sit Diarrhea and refusal to eat Bruise on lower back

Repeatedly missed day program due to illness Ongoing communication between residential and day providers Seen by PCP multiple times

Annual exam Evaluate for possible abuse (bruise)

Parents demand she be sent to hospital Hospital finds severe dehydration, renal failure, in “grave condition,” Lithium levels in toxic range

(6.8 mmol/L) Dies within 48 hrs of hospital admission 3 weeks later Public Health inspector discovers 300 mg Lithium packaged in 150 mg blisterpak

(double prescribed dose)

Psych evaluates & inc Lithium Rx on March 2nd -

Blood Levels orderedfor 8 weeks later –

WNL (.9)

New Lithium Rx Picked up by Res

(30 day supply in blisterpak)

5/13/03

JS has diarrhea atDay Program –

Needs RTW order.Complaint of back pain –

5/23/03

Absent from DayProgram

May have had diarrhea atHome

(No Documentation)

5/24-5/25/03

Seen by PCPOrders inc in fluids – call if

dec in oral intake or urination orchange in activity level or if

diarrhea –RTW clearance issued

5/25/03

JS returns to Day Program - has continued diarrheaRes staff notified

JS continues to beSick at Day Program

Whining, unsteady, difficultySitting – small bruise noted

on lower back

5/26/03 5/29/03

Res staff notify PCPAppt already scheduled for

next dayfor Annual Physical Exam

5/29/03

JS seen by PCPReferral problem = bruising –Orders x-ray and bloodwork

(No lithium levels) – JS uncooperative for x-ray

5/30/03

Res staff and Day staffnote continued

problemsUnsteady, crying, diarrhea

6/1/03

Psychiatrist Orders dec in Zyprexia – did not see JS -

Told JS irritable, restless andfalling

Day Program notescontinued problems

and won’t accept back -Calls SC who calls home –

no answer at home- original RTW order provided

6/1-6/2/03 6/2-6/3/03

Not at Day ProgramRes staff notes JS very

unstable –Res Prog supervisors notified of

need for new RTW order

6/3-6/7/03

Res Prog Clin Dir callsPCP Office in resp to RTW -

JS shows continued decline –PCP orders inc in Levoxyl

(Diarrhea STOPS)

6/8/03

SUMMARY OF INVESTIGATION REVIEWSequence Analysis: Joanne Smith

PCP Office refuses toissue new RTW order

may believe no continued Diarrhea

Res Prog staff and supervisors aware of

continued declinePCP called and asked for

X-ray – appt set for next day

6/9/03 6/9/03

SC visits – looks very bad –Told JS going to PCP –

JS gets x-ray Ativan adm to presedate –

Not seen by PCP – only issueof x-ray is reviewed

6/10/03

JS very weakstaff note she needs extra

assistance – did not sleep well,eyes moving side/side – seen

as “much worse”

6/11/03

Parents visit homeparents alarmed at JS’s

condition – insist on hospital visit

6/11/03

EMTs transport toHospital ER

Evaluated at ER – in gravecondition – severe dehydration –

attempt to hydrate

Hospital MD reviews options –requests nephrology consult

6/11/03 6/11/03

Lithium levels showextreme elevation

bloodwork shows lithium levelat 6.8

6/11/03

Comfort measuresonly provided

6/11-6/13/03

JS diesNephrology consult same day

as death – Cause of deathnoted as renal failure and

hyperkalemia

6/13/03

3/2/03 – 5/9/03

Day and Res Clinical staff did NOT directly speak

Home for weekend – ResNote re: JS not eating

5/27-5/28/03

6/6/03: Res Program Director resigns without notice

Conclusions

An examination of the contributory factors reflected a pattern of misinformation across all systems. A key finding revealed: that no one individual was designated with the ultimate responsibility for [JS’s] health care case management. A series of erroneous assumptions by all relevant participants pointed to inadequate communication regarding the critically changing health status of [JS].

The root cause of the event was acknowledged by the team to be a lack of adequate systems to identify high-risk health situations, document target symptoms and assure accurate communication and timely response. The RCA team’s recommendations are intended to reduce the probability of future communication errors in a multi-service delivery system.

From final report by RCA team to Commissioner/Director of State MR/DD agency.

ANALYTIC APPROACHES TO MANAGING RISK: The I-D-Q Approach Presented by: Steven D. Staugaitis, Ph.D. Connecticut DMR - and the UMASS Medical School, Family.

Documents

patient safety

culture of safety slide

safety statistics

best ways

systems approach

best position

medical care

q approach