ROOT CAUSE ANALYSIS PROTOCOL
ASSURANCE AND PERFORMANCE COMMITTEE
Approved by:
November 2007
On:
June 2009
Review Date:
Directorate responsible for Review
Nursing and Quality
Policy Number:
NQ014
Signed by:
Mrs Anna Barrett Director of Nursing and Quality
Page 1 of 20 NQ14 - Root Cause Analysis Protocol/Version 1/Authors, Heather Darlow & Helen Wallace, Corporate Governance Managers/Approved A&P Nov 2007/Review date June 2009
Equality Impact Assessment - ROOT CAUSE ANALYSIS PROTOCOL
Yes/No 1.
Comments
Does the policy/guidance affect one group less or more favourably than another on the basis of: No
•
Race, Ethnic origins (including gypsies and travellers) and Nationality
•
Gender
•
Age
•
Religion, Belief or Culture
•
Disability – mental and physical disabilities
•
Sexual orientation including lesbian, gay and bisexual people
No No No No
No
2.
Is there any evidence that some groups are affected differently?
No
3.
Is there a need for external or user consultation?
No
4.
If you have identified potential discrimination, are any exceptions valid, legal and/or justifiable?
No
5.
Is the impact of the policy/guidance likely to be negative?
No
6.
If so can the impact be avoided?
NA
7.
What alternatives are there to achieving the policy/guidance without the impact?
NA
8.
Can we reduce the impact by taking different action?
NA
Page 2 of 20 NQ14 - Root Cause Analysis Protocol/Version 1/Authors, Heather Darlow & Helen Wallace, Corporate Governance Managers/Approved A&P Nov 2007/Review date June 2009
TITLE: ROOT CAUSE ANALYSIS PROTOCOL
VERSION CONTROL
First Version NQ14
Date June 2007
Amendment
Page 3 of 20 NQ14 - Root Cause Analysis Protocol/Version 1/Authors, Heather Darlow & Helen Wallace, Corporate Governance Managers/Approved A&P Nov 2007/Review date June 2009
ROOT CAUSE ANALYSIS PROTOCOL Supporting the investigation of all types of incidents
Heather Darlow/Helen Wallace Corporate Governance Manager June 2007
Page 4 of 20 NQ14 - Root Cause Analysis Protocol/Version 1/Authors, Heather Darlow & Helen Wallace, Corporate Governance Managers/Approved A&P Nov 2007/Review date June 2009
Contents Paragraph
Page
1
Introduction
6
2
PCT’s Investigation Procedure
6
3
Undertaking an Investigation
7
3.1
Fact Finding Stage
8
3.2
Developing a Chronology of the Incident
8
3.3
Identifying Care/Service Delivery Problems
8
3.4
Analysis Stage
8
3.5
Solution Stage
8
3.6
Investigation Report
8
3.6.1
Minimum Standards for Investigation Reports
9
3.6.2
Minimum Standards for Action Plans
9
4
Sharing Safety Lessons
10
5
Staff training
10
6
Support mechanisms for patients, carers and staff
11
7
Accountability and Performance Management
11
Appendix 1
Five Whys
12
Appendix 2
Representing Contributory Factors
15
Appendix 3
Action Plan Example
20
Introduction
1.
Reporting when things go wrong is essential in healthcare. But it is only part of the process of improving patient safety. It is equally important that organisations look at the underlying causes of patient safety incidents and learn how to prevent them from happening again. Often the underlying causes are many and in the majority of cases extend beyond the individual staff member or team involved. Step 6 of the NPSA’s Seven Steps to Patient Safety details the following key principles: •
When a patient safety incident occurs the important issue is not ‘who is to blame for the incident?’ but ‘how and why did it occur?
•
To promote learning NHS organisations need a systematic approach in which staff know what type of incidents should be reported, what information is needed and when, and how to analyse and act on this information.
2.
PCT’s Investigation Procedure
A vital part of the Trusts incident procedure (NQ007 Policy for managing, reporting and investigating incidents and SUIs) is the role of the investigation and this needs to be undertaken as soon as possible after the incident occurs. The aims of investigation as detailed in the Investigation Protocol (part of NQ007 Policy for managing, reporting and investigating incidents and SUIs) are: • •
To determine the root cause of an incident/accident and any contributory causes to enable preventative actions to be implemented to eliminate or reduce the risk of recurrence To provide documented evidence if required for litigation purposes
It should be noted that the principles identified within the protocol apply to claims or complaints resulting in the need for an investigation. A root cause is the cause or causes that if addressed will prevent or minimise the chances of an incident recurring. Root cause analysis (RCA) is a technique for undertaking a systematic investigation that looks beyond the individuals concerned and seeks to understand the underlying causes and environmental context in which the incident happened. A full RCA should be undertaken for all SUIs including: • • • • •
All incidents which are graded as major or catastrophic, or comply with the PCTs SUI definition. This can include: All unexpected deaths that were directly related to an incident; All incidents that resulted in suspected permanent injury, loss of function or loss of a body part. All accidents/incidents that have been reported to the Health & Safety Executive (HSE). Incidents to staff and patients that result from equipment failure.
• • •
Incidents to patients and others that result in the injured needing medical attention through a direct visit to the Walk In Centre or Accident & Emergency Department dependent on severity. Incidents where there is extensive property damage to Trust and/or Personnel and potentially any criminal investigation that the police are involved in Incidents where the injured party actually states they will take legal action against the Trust.
RCA should be considered for: • • •
Incidents where the patient needed further surgical intervention or transfer to intensive care; Incidents that were prevented but considered by the investigator to be worth an in-depth review of not only what, why and how it happened but also what were the actions that prevented the incident from affecting the patient(s); All incidents that trigger external investigations, such as a coroner’s inquest, complaints, legal claims or criminal investigations.
The principle of establishing the root cause of an incident should be adopted for all incidents irrespective of grade, however the degree of investigation and application of the tools and techniques available will be wholly dependent on the severity of the incident in question as detailed above. Some degree of judgment is necessary and if in doubt contact your senior manager or the Managers responsible for clinical governance, Health and Safety Advisors etc for advice. It is imperative that where there are legal implications to an investigation that this is discussed with the senior management team and contact made with the Trusts Solicitors if appropriate. The investigation should be carried out by a review team or designated executive selected by the Director of Nursing and Quality in consultation with the Chief Executive. Depending on the level of complexity needed the investigation could take between 20-60 days. Where necessary, specialist input should be sought from the relevant individuals (e.g. Infection Control Nurses, Fire Safety Advisors, Health and Safety Advisors, Moving and Handling Team, etc). 3.
Undertaking an Investigation
The Investigation Protocol (part of NQ007 Policy for managing, reporting and investigating incidents and SUIs) details the key steps to be undertaken once an incident has occurred. There are 6 key stages to the investigation, which are supported by RCA tools and techniques.
3.1
Fact Finding Stage
This is about framing the incident through data collection e.g. review of records and interviews. Depending on the complexity of the incident the timeline may suffice in terms of determining what happened or further data collection may be required especially in the most serious of cases. 3.2
Developing a chronology of the incident
This is about establishing the key facts of the incident and is helped by establishing the timeline of events. This is a method for mapping and tracking the chronological chain of events in an incident. The chronology will be formed by pulling together information about the incident derived from case notes, statements interviews etc. It allows the investigating team to view the episode in a clear and concise format as well as enabling them to identify information gaps. This can be very simple or very complex depending on the incident. 3.3
Identifying Care/Service Delivery Problems
During and after the information gathering stage the precise points at which things went wrong need to be identified, these are referred to as Care Delivery Problems (CDPs) or Service Delivery Problems (SDPs). Some of these will be clear at an early stage and others will emerge through discussion and gathering of more information. 3.4
Analysis stage
Once information about an incident has been gathered mapped and the problems identified, each problem needs to be analysed to identify underlying causes. This is the stage where contributory factors and any root causes will be identified. There are many tools to aid this part of the process including: • • 3.5
5 whys (see Appendix 1) Contributing factors (see Appendix 2) Solution Stage
This stage is key in ensuring that real change happens. It ensures that solutions are identified and that actions are put in place to address areas of concern (see Appendix 3). 3.6
Investigation Report
This is an essential last step to ensure that all the previous stages are captured and can be used for learning and sharing. To ensure that the report is comprehensive and meaningful it is important that the minimum standards set out below are followed and are submitted to the Trusts’ Patient Safety Officer within the 60-day timeframe.
3.6.1 Minimum Standards for Investigation Reports (NHS East Midlands Incident Policy) General The report must: • • • •
State who the authors/ investigation team are Be dated Have page numbers Have a contents page
Body of report must contain: • Summary
Background to the incident, including the outcome for the patient (including relevant medical, psychiatric or social history), details of incident and findings.
• Process
of investigation Who commissioned the investigation, the documentation reviewed, interviewers, and methodology interviewees, tools and techniques used during the investigation processes.
• Chronology of events
Use of a timeline, e.g. tabular timeline (to include where appropriate details of referral, admission, day services, community services, primary care)
• Analysis and findings
Care and service delivery problems, contributing factors and root causes identified Reference where appropriate to the NPSA Contributory Factor Classification Framework: • • • • • • • • •
• Recommendations
Individual Factors Team and Social Factors Communication Factors Task Factors Education and Training Factors Equipment and Resource Factors Working Conditions Organisational and Strategic Factors Patient Factors
Solutions identified and recommendations made to try and prevent recurrences, based on analysis and findings
•
Sharing good practice
•
Appendices
Detail particular areas of review that went well areas of good practice that can be shared and learned from List of documents reviewed Root Cause Analysis tools; timeline, fishbone diagrams, 5 whys etc
3.6.2 Minimum Standards for Action Plans An action plan is used to ensure the risks identified from the RCA are addressed. This action plan must be reviewed by the Trend Review Group and appropriate committees. General The action plan must define: • • • •
Who has agreed the action plan and timescales Who will oversee the delivery of the action plan How often the action plan will be reviewed Who will sign off the action plan when all actions have been completed
The action plan must contain (see Appendix 3): 1. Root causes/ care or These should be drawn from the analysis and service delivery problems/ findings of the investigation contributing factors 2. Action agreed These should be the actions the Trust needs to take to resolve the root cause. 3. Level of recommendation
Does this action need to be taken at: • Individual level • Team level • Directorate level • Organisation level
4. By who
Who in the Trust will ensure the action is completed
5. By when
Target date for completion of the action.
6. Resource requirements
To be able to complete the action, what resources are required?
7. Evidence of completion
What evidence will be available to demonstrate that the action has been completed?
8. Sign-off
Date when the action has been completed.
4.
Sharing Safety Lessons
It is the role of the Trend Review Group to analyse and review investigations and ensure that lessons are disseminated appropriately via the quarterly reviews with localities and PCT News. Sharing of lessons further a field are done via membership of the SHA SUI Network Group. 5.
Staff training
Staff training has been ongoing to ensure that people have the relevant and necessary skills to undertake investigations and apply root cause analysis tools and techniques. If further support is required in terms of undertaking an investigation there is the potential for a ‘buddy’ to be put in place to ensure there is the relevant support. Contact should be made via the Corporate Governance Manager on 0116 295 7543. 6.
Support Mechanisms for patients, carers and staff
The PCT recognises that the experience of those in the service involved with investigations can be very stressful. The PCT will treat all staff in these circumstances with sensitivity. The culture of the PCT is to follow the principles of “being open”. Any staff involved in traumatic or stressful incidents, complaints or claims will be supported by management, offered access to occupational health support, if appropriate, and offered appropriate legal advice and expertise in the event of full legal proceedings. 7.
Accountability and Performance Management
All Trusts (according to NHS East Midlands) must have systems in place to review serious incidents to: • • • • • •
Discuss them as soon as possible after the event. Establish terms of reference for investigations Review and sign off investigation reports Agree action plans Monitor implementation of actions Monitor effectiveness of actions
Recommendations from investigations will be reviewed and considered by the relevant governance structures. Ownership in terms of accepting recommendations and ensuring action plans are delivered will be the responsibility of the PCTs Trend Review Group and the Provider Governance Group/appropriate governance committee
Appendix 1 Five Whys
ABOUT THIS DOCUMENT The Five Whys is a technique to help you to drill down into a particular issue through the various layers of cause to find the fundamental cause of the problem. This document describes its nature and application.
Five Whys What is the Five Whys Technique? Five whys is also known as the why-why chart (Ammerman 1998) and its focus is to enable the RCA Practitioner to delve more deeply into the causes of a patient safety incident. The main purpose of this technique therefore is to constantly ask Why? through the various layers of cause thus progressing towards the true root cause of the identified problem or issue. When to Use Five Whys • to question each identified primary cause of a problem and to identify whether it is a symptom, an influencing factor or a root cause • to continue the search for true root causes, even after finding a possible cause. How to Use Five Whys First decide upon your starting point. This will usually be the identified CDP or SDP (i.e. problem), or an already identified contributory factor to a CDP or SDP that should be further analysed. 1. Use brainstorming, brain writing, or another appropriate approach to encourage the participants to volunteer some causes of the problem identified. 2. If more than one of these is identified then each should be analysed separately. 3. For each primary cause ask the group “Why is this a cause of the original problem?” This should generate further and deeper reasons as to why the problem exists. 4. Map the process on a flip chart or white board. (see Figure 1) 5. For each new ’cause’ generated repeat steps 4 and 5 until no new answers are forthcoming. 6. At this point you can be reasonably confident that you have identified the core (i.e. root) of the problem. 7. As a rule of thumb this method normally requires five rounds of the question “Why?” to elicit the root of a problem. However this is not cast in stone and you will frequently find the numbers of rounds of WHY varies. (see Figure 1).
Positives Attributes of the Five Whys Technique
Appendix 1 • •
it allows the individual or group to drill down through the causal pathway specifically it is a simple and effective tool that works well in groups and when undertaken by an individual.
Negative Attributes of the Five Whys Technique • causal analysis could be constrained by mind set and lack of breadth and depth of thought • why can be asked as frequently as possible to get to the real cause of the problem, therefore asking why 5 times may not be appropriate • a specific and contained issue to drill down on is important if this technique is to work. The Five Whys Template and Example A patient had the wrong leg amputated 1. Why Patient gave consent for amputation the night before the proposed surgery to Registrar (who was not going to undertake procedure).
2. Why Amputation site marked with a biro (wrong leg).
3. Why Registrar unaware of hospital policy on amputation sites being marked with a skin pencil and with bodily part being fully visible to Doctor.
4. Why the department had no induction procedures for new medical staff working in the department.
5. Why because “we’ve never been asked to”.
Five Whys - Figure 1 Please note that there is a template for the Five Whys tool available for you to use in the Template section of the Resource Centre.
References
Appendix 1 Adams, S. et al, (2002) Long Version of the CRU/ALARM Protocol: Successful Systems Event Analysis (In print, 2002) Ammerman, M., The Root Cause Analysis Handbook: A Simplified Approach to Identifying, Correcting and Reporting Workplace Errors (Quality Resources, New York, 1998)
Dineen, M., Six Steps to Root Cause Analysis 18 September Consequence (Oxford, 2002, ISBN 0-9544328-0-0) Joint Commission on Accreditation of Healthcare Organisations, Root cause analysis in healthcare: Tools and techniques (2000) Warren, I. Events and Causal Factor Charting for Healthcare. Conference (Presentation at The St Paul Conference on Root Cause Analysis, London, 3 October 2001)
Appendix 2
Representing Contributory Factors: Fishbone Diagrams and other formats ABOUT THIS DOCUMENT When contributory factors have been identified for a specific problem, it may be helpful to use a classification framework to classify and group them. Different methods exist to do this graphically and the fishbone diagram is a particularly popular method. This document describes the nature and application of fishbone diagrams and other formats. A common way of representing contributory factor information: the fishbone diagram A common approach is to draw a horizontal arrow on a large sheet of paper or white board. At the head of the arrow is the problem to be explored (the CDP or SDP); spines are then added to the arrow in a fishbone arrangement. Each spine is given a classification heading (from the NPSA Contributory Factors Classification Framework) which represents the main areas under which you may want to explore the influencing factors which contributed to the identified issue or problem (Care Delivery Problem - CDP or Service Delivery Problem – SDP), see figure 1. It is important to remember that each fish should explore the contributory factors associated with just one CDP / SDP. Organisational and Strategic Factors
Task Factors
Equipment & Resource Factors
Working Condition Factors
Individual Factors
Problem or Issue to be explored CDP/SDP
Communication Factors
Education and Training Factors
Team and Social Factors
Patient Factors
Diagrams and other formats - Figure 1 It is important to remember that investigators may not find factors for each heading and that they should not “force” factors merely to have something in every category. Factors from the strategic / policy heading may be more relevant for an issue where service delivery is the key problem rather than care delivery. Figure 2 illustrates a Care Delivery Problem from one of the case studies used in the e-Learning package and shows how the contributory factors can be categorised according to the framework.
Appendix 2
Organisational and Strategic factors
Tolerance of persistent low staffing levels.
Equipment and resource factors
Task factors
Working condition factors
Individual factors
Patient admitted to ward after ward round.
No written guidance for who should see patients presurgery.
SHO has limited access to senior medical advice.
Consultant has no fixed time for ward rounds.
Delay in patient's admission to ward.
Ward busy - 2 patients requiring 1:1 nursing.
Marking covered by sock SHO inexperienced in orthopaedics.
Nurse applying sock unsure of her responsibilities.
SHO unaware of procedure for marking.
Communication factors
Education and training factors
Team and social factors
Patient factors
Example of a Fishbone Diagram - Figure 2 You will also find a template Fishbone to use for this type of analysis in the Resource Centre. Again, there may be different ways for presenting data during this stage of the process and the methodology and template described in the Protocol for Investigating Clinical Incidents (CRU/ALARM Protocol, Vincent et al 1999) has also been used widely within the NHS. Experience has shown that many people find that mapping the contributory factors via fishbones is a helpful tool that forces them to consider the variety of issues that can influence performance. However, not everyone likes this format - therefore it is also perfectly acceptable to consider the contributory factors either in list form under each of the relevant headings, or in tabular format. See Figure 3 for a tabular example linked to a timeline. As long as all the relevant contributory factors have been identified, it does not make a significant difference which format is used. The important thing is to use the same classification system in your analyses to ensure consistency and reliability, rather than to worry about formatting arrangements
Appendix 2 CHRONOLOGY
TIME Problem/Issue CDPs/SDPs
Contributory Factors: causal or influencing* Main Causal Factors (Fundamental Issues/ Roots) Recommendations
Figure 3 Chronological Mapping of Problems/issues (CDPs/SDPs) and associated Contributory Factors (Taylor-Adams (2002)) You will also find a Template for the Chronological Mapping of Problems in the Resource Centre. How to Identify Contributory Factors The easiest way to identify the contributory factors for each problem identified when you are a novice is to consider each classification heading in turn and identify whether or not there were any issues or influences that map under it. The NPSA classification system can therefore be used in checklist format to facilitate this process. Brainstorming can help a group identify all the relevant contributory factors, but if this is being undertaken in a group where there are sensitivities then the brainwriting technique can easily be applied. Indeed, if time is tight, this approach can be a distinct advantage as it keeps the group very focused. As previously mentioned, both the positive and the negative contributory factors associated with a Problem/Issue (CDP/SDP) should be identified. We normally suggest using different colours, or placing a + next to the positive factors. Verification of contributory factors is an important component in the causal analysis process. It is tempting when investigating a case to use it as an opportunity to identify all the issues that you know have influenced this case or similar cases in the past. It is therefore essential that all contributory or causal factors identified are verified as being significant to the problem/issue (CDP/SDP) being analysed. Failure to do this will lead to a reduction in the credibility of the analysis. Failure to authenticate your
Appendix 2 causal data may lead the investigation team/RCA participants to focus on an improvement strategy which does not address the fundamental causes of the incident and therefore leaves the department and organisation open to the possibility of a similar incident occurring in the future, Dineen 2002. Positive Attributes of Causal Analysis using a Contributory Factor Classification System and Fishbone Diagrams • it provides a structured and semi-comprehensive system for considering those influences on performance in an incident • fishbone diagrams are easily constructed and understood by the novice investigator • it allows more reliable improvement strategies to be developed, as they are based on verified causal information. Negative Attributes of Causal Analysis using a Contributory Factor Classification System and Fishbone Diagrams • contributory classification systems have not been formally validated, the data which has not been verified may lead to inappropriate improvement strategies being implemented • not all investigators feel comfortable with the fishbone diagram format. Influencing Factors An influencing factor is something that influenced the occurrence of, or outcome of an incident. Generally speaking the incident may have occurred in any event, and the removal of the influence may not prevent incident recurrence but will generally improve the safety of the care system Causal Factors A causal factor is something that led directly to an incident. Removal of these factors will either prevent, or reduce the chances of a similar type of incident from happening in similar circumstances in the future. Causative factors tend to be more closely related to the incident being analysed.
References Dineen M., Six Steps to Root Cause Analysis. Consequence. (Oxford,2002, ISBN 09544328-0-0)
Appendix 2 Taylor-Adams S.E., (2002) Long Version of the CRU/ALARM Protocol. (In Press, 2002) Vincent, C .Taylor-Adams, S.E., Stanhope, N. (1998). Framework for analysis risk and safety in clinical medicine, (British Medical Journal, 1998) pp.316, 1154-7. Vincent C.A, Adams S, Hewett D, Chapman J et al,. A Protocol for the Investigation and Analysis of Clinical Incidents. Clinical Risk Unit in association with the Association of Litigation and Risk Managers (Royal Society of Medicine Press Ltd, 1999)
Appendix 3 ACTION PLAN EXAMPLE ACTION PLAN AGREED BY _______________________________
se/ ng
Level of Risk
Agreed Action
Level of Recommendation Individual, Team, Directorate, Organisation
By Whom
By When
Resources Required
Expected Outcome
Evidence Completio
