Systems Analysis of Clinical Incidents The London Protocol
Systems Analysis of Clinical Incidents The London Protocol
Systems Analysis of Clinical Incidents The London Protocol
INCIDENTS
THE LONDON PROTOCOL
Contributors
Principal Authors
Sally Taylor-Adams, PhD is Assistant Director of Patient Safety (Midland and
East Region) at the National Patient Safety Agency, UK (NPSA)
Charles Vincent, PhD is Director of the Clinical Safety Research Unit and the
Smith and Nephew Foundation Professor of Clinical Safety Research at Imperial
College, London.
Other Contributors
Stephen Rogers, FRCGP, University College London
Maria Woloshynowych, PhD, Imperial College London
ALARM Research Group
David Hewett
Jane Chapman
Sue Prior
Pam Strange
Ann Tizzard
Mental Health Research Group
Alison Prizeman
Yvonne Connolly
Mohamed Sheikh
Ann Rozier
Valli Agbolegbe
Primary Care Research Group
Kathy Caley
Louise Worswick,
Janet Cree
Greg Cairns
Andrew Harris
Juliet Swanwick
Kim Allen
Sarah Raymond
Amee Fairburns
Pauline Grace
CONTENTS
1.
2.
3.
4.
5.
INTRODUCTION
page 1
1.1
page 1
1.2
page 2
1.3
page 2
1.4
page 3
RESEARCH FOUNDATIONS
page 4
2.1
page 4
2.2
page 5
2.3
page 6
ESSENTIAL CONCEPTS
page 7
3.1
page 7
3.2
Clinical Context
page 7
3.3
Contributory Factors
page 7
page 8
Section A
page 9
Section B
page 9
Section C
page 10
Section D
page 13
Section E
page 14
Section F
page 15
Section G
REFERENCES
page 18
page 4
page 8
page 13
page 16
page
Table 2
page 17
INTRODUCTION
The London Protocol is the revised and updated version of our original Protocol for
the Investigation and Analysis of Clinical Incidents1. The protocol outlined a process
of incident investigation and analysis developed in a research context, which was
adapted for practical use by risk managers and others trained in incident investigation.
This approach has now been refined and developed in the light of experience and
research into incident investigation both within and outside healthcare.
The purpose of the protocol is to ensure a comprehensive and thoughtful investigation
and analysis of an incident, going beyond the more usual identification of fault and
blame. A structured process of reflection is generally more successful than either
casual brainstorming or the suspiciously quick assessments of experts. The
approach described does not supplant clinical expertise or deny the importance of the
reflections of individual clinicians on an incident. Rather the aim is to utilise clinical
experience and expertise to the fullest extent. The approach we describe assists the
reflective investigation process because:
1.1
The first edition of the protocol was primarily aimed at the acute medical sector. The
present edition can be applied to all areas of healthcare including the acute sector,
mental health, ambulances and primary care. We have found the basic method and
concepts to be remarkably robust when tested in these different contexts.
Those familiar with the first edition will find that the basic process is unchanged,
though there is more emphasis on following through with recommendations and
action. We have endeavoured to simplify both the structure and the language of the
protocol where possible. We have abandoned the absolute distinction between
specific and general contributory factors as unworkable, although the importance
of identifying contributory factors that are of wider significance remains. Finally, we
have removed the forms used for recording data in this edition, to allow teams and
individuals more flexibility when producing case summaries. However, we have
1.2
The term root cause analysis originates from industry, where a group of tools are
used to identify root causes from the investigation and analysis of incidents. To us the
term root cause analysis, while widespread, is misleading in a number of respects. To
begin with it implies that there is a single root cause, or at least a small number.
Typically however, the picture that emerges is much more fluid and the notion of a
root cause seems a gross oversimplification. Usually there is a chain of events and a
wide variety of contributory factors leading up to the eventual incident. The
investigation team needs to identify which of these contributory factors have the
greatest impact on the incident and, more importantly still, which factors have the
greatest potential for causing future incidents2.
A more important and fundamental objection to the term root cause analysis relates to
the very purpose of the investigation. Surely the purpose is obvious? To find out
what happened and what caused it? We believe that this is not the most penetrating
perspective. Certainly it is necessary to find out what happened and why in order to
explain to the patient and family and others involved. However, if the purpose is to
achieve a safer healthcare system, then finding out what happened and why is only a
way station in the analysis. The real purpose is to use the incident to reflect on what it
reveals about the gaps and inadequacies in the healthcare system. This proactive,
forward-looking approach is more strongly emphasised in this second edition.
Because of this orientation we have called our approach a `systems analysis, by
which we simply mean a broad examination of all aspects of the healthcare system in
question. We emphasise that this includes the people involved throughout the system
(from management to those working at the sharp-end), and how they communicate,
interact, work as a team, and work together to create a safe organisation.
1.3
The original protocol was designed at a time when investigations were generally
carried out by individual risk managers. It was therefore investigator led, in that the
description and format assumed that one or two individuals would assemble and
collate the information, carry out interviews and then report back to the board or the
clinical team to consider what action should be taken. However, many organisations
now prefer to assemble a team of individuals with different skills and backgrounds.
Serious incidents are certainly likely to require a team of people using both interviews
and other documents as their sources of information. This version of the protocol can
be used by either individuals or teams.
This document describes a full investigation, but we wish to emphasise that much
quicker and simpler investigations can also be carried out using the same basic
approach. Experience has shown that it is possible to adapt the basic approach of the
protocol to many different settings and approaches. For instance it can be used for
quick 5 or 10-minute analyses, just identifying the main problems and contributory
factors. The protocol can also be used for teaching, both as an aid to understanding
the method itself and as a vehicle for introducing systems thinking. While reading
about systems thinking is helpful, taking an incident apart in a structured manner
brings the approach alive for a clinical team.
1.4
RESEARCH FOUNDATIONS
The theory underlying the protocol and its application is based on research in settings
outside healthcare. In the aviation, oil and nuclear industries for instance, the formal
investigation of incidents is a well established procedure. Researchers and safety
specialists have developed a variety of methods of analysis, some of which have been
adapted for use in medical contexts though few have been explored in depth3-5. These
and other analyses have illustrated the complexity of the chain of events that may lead
to an adverse outcome6-10.
O R GAN IS A T IO N
&
M AN A GE M E N T
C UL T UR E
C O N TR IB UTO R Y
F AC TO R S
IN F L UE N C IN G
P R AC TIC E
W ork /
E nv ironm en t
F actors
M an ag em en t
D ecis ions
an d
O rg an is ation al
P rocesses
C AR E
D E L IV E R Y
P R O B LE M S
U ns afe A cts
T eam F actors
E rrors
In d iv id u al
(s taff) F actors
In c iden t
T as k F actors
V iolations
P atien t F actors
L AT E N T
FA IL U R E S
2.1
D E F E NC ES /
B AR R IE R S
ERROR &
V IO L AT IO N
PR O D U C IN G
C O N D IT ION S
A C T IV E
FA IL U R E S
Studies of accidents in industry, transport and military spheres have led to a much
broader understanding of accident causation, with less focus on the individual who
makes the error and more on pre-existing organisational factors. Our approach is
based on James Reasons model of organisational accidents (Figure 1). In this model
fallible decisions at the higher echelons of the management structure are transmitted
down departmental pathways to the workplace, creating the task and environmental
conditions can promote unsafe acts of various kinds. Defences and barriers are
designed to protect against hazards and to mitigate the consequences of equipment
and human failure. These may take the form of physical barriers (e.g. fence), natural
barriers (e.g. distance), human actions (e.g. checking) and administrative controls (e.g.
training). In the analysis of an incident each of these elements is considered in detail,
starting with the unsafe acts and failed defences and working back to the
organisational processes. The first step in any analysis is to identify active failures unsafe acts or omissions committed by those at the `sharp end' of the system (pilots, airtraffic controllers, anaesthetists, surgeons, nurses, etc) whose actions can have immediate
adverse consequences. The investigator then considers the conditions in which errors
occur and the wider organisational context, which are known as contributory factors.
These conditions include such factors as high workload and fatigue; inadequate
knowledge, ability or experience; inadequate supervision or instruction; a stressful
environment; rapid change within an organisation; inadequate systems of
communication; poor planning and scheduling; inadequate maintenance of equipment
and buildings. These are the factors which influence staff performance, and which may
precipitate errors and affect patient outcomes.
We have extended Reasons model and adapted it for use in a healthcare setting,
classifying the error producing conditions and organisational factors in a single broad
framework of factors affecting clinical practice11, see Table 1.
Table 1: Framework of Contributory Factors Influencing Clinical Practice
FACTOR TYPES
Patient Factors
Team Factors
Organisational
Factors
&
Management
2.2
At the top of the framework are patient factors. In any clinical situation the patients
clinical condition will have the most direct influence on practice and outcome. Other
patient factors such as personality, language and psychological problems may also be
important as they can influence communication with staff. The design of the task, the
availability and utility of protocols and test results may influence the care process and
affect the quality of care. Individual factors include the knowledge, skills and
experience of each member of staff, which will obviously affect their clinical practice.
Each staff member is part of a team within the inpatient or community unit, and part of
the wider organisation of the hospital or mental health service. The way an individual
practises, and their impact on the patient, is constrained and influenced by other
members of the team and the way they communicate, support and supervise each other.
All members of the team are influenced by the working environment, both the physical
environment, (light, space, noise) and factors which affect staff morale and ability to
work effectively. The team is influenced in turn by management actions and by
decisions made at a higher level in the organisation. These include policies for the use of
locum or agency staff, continuing education, training and supervision and the availability
of equipment and supplies. The organisation itself is affected by the institutional
context, including financial constraints, external regulatory bodies and the broader
economic and political climate.
Each level of analysis can be expanded to provide a more detailed specification of the
components of the major factors. For example, team factors include verbal
communication between junior and senior staff and between professions, the quality of
written communication such as the completeness and legibility of notes, and the
availability of supervision and support. The framework provides the conceptual basis for
analysing adverse incidents. It includes both the clinical factors and the higher-level,
organisational factors that may be influential. In doing so, it allows the whole range of
possible influences to be considered and can therefore be used to guide the investigation
and analysis of an incident.
2.3
Active failures in health care come in various forms. They may be slips, such as picking
up the wrong syringe, lapses of judgement, forgetting to carry out a procedure or, rarely,
deliberate departures from safe operating practices, procedures or standards. In our work
we have substituted the more general term `care delivery problems (CDP) for unsafe
acts. This is because we have found, in healthcare that this more neutral terminology is
helpful and because a problem often extends over some time and is not easily described
as a specific unsafe act. For instance a failure of monitoring of a patient may extend
over hours or days.
Having identified the CDP, the investigator then considers the conditions in which errors
occur and the wider organisational context, which are known as contributory factors.
These are the factors which influence staff performance, and which may precipitate
errors and affect patient outcomes.
ESSENTIAL CONCEPTS
Reasons model and our framework provide the conceptual foundations of the
investigation and analysis process. However, before incident investigation can be
undertaken, key essential concepts need to be defined.
3.1
CDPs are problems that arise in the process of care, usually actions or omissions by
members of staff. Several CDPs may be involved in one incident. They have two
essential features:
Care deviated beyond safe limits of practice
The deviation had at least a potential direct or indirect effect on the eventual
adverse outcome for the patient, member of staff or general public.
Examples of CDPs are:
Failure to monitor, observe or act
Incorrect (with hindsight) decision
Not seeking help when necessary
3.2
Clinical Context
Salient clinical events and the clinical condition of the patient at the time of the CDP
(e.g. bleeding heavily, blood pressure falling). The essential information required to
understand the clinical context of the CDP.
3.3
Contributory Factors
The accident investigation and analysis process flowchart (see figure 2) provides a
overview of all the stages of the incident investigation and analysis process. The
flowchart shows the objectives of each stage and how each objective is achieved.
The basic process of incident investigation and analysis is relatively standardised, and
will be followed whether investigating a minor incident or a very serious adverse
outcome; the process is essentially the same where an individual or a large team are
responsible for the investigation. However, the team can choose whether to quickly
run through the main issues in a short meeting or to carry out a full, detailed
investigation over several weeks, making full use of all associated techniques to
comprehensively examine the chronology, CDPs and contributory factors. The
decision on the time taken will depend on the seriousness of the incident, potential for
learning and the resources available.
Figure 2 Accident Investigation and Analysis Process Flowchart
Go to Section
Identification and
Decision to Investigate
Organisation and
Data Gathering
Determine Incident
Chronology
Identify CDPs
Identify Contributory
Factors
The protocol can also be used to investigate less serious incidents and near misses. In
this situation it might be that a departmental or ward manager with appropriate
training would facilitate the incident investigation and analysis. They would lead the
process, but would call on relevant clinical and other expertise as necessary.
Information/Data Source
Date requested
Date received
Stored?
Case 25/02
Case 25/02
Case 25/02
Case 25/02
24/10/01
24/10/01
24/10/01
24/10/01
24/10/01
25/10/01
26/10/01
26/10/01
Cabinet A RM Office
Cabinet A RM Office
Cabinet A RM Office
Cupboard G Legal Office
10
11
12
Interviews to take place in a relaxed and private setting, away from the ward
Allow interviewee to be supported by someone else if they wish
EXPLAIN PURPOSE OF THE INTERVIEW
Find out what happened
13
contributory factors can be added once the chronology is complete. There are various
ways of doing this.
Timeline tracks the incident and allows the investigators to discover any parts of
the process where problems may have occurred. This approach is particularly
useful when a team works together to generate the chronology.
Pre-prepare drugs
12.00noon
Respiratory Arrest
1.30pm
Patient dies
1.45pm
Time Person Grids allows you to track the movements of people before during
and after an incident.
SHO
Ward Manager
Nurse
9.02am
With patient
In office
With patient
9.04am
At Drs station
In office
With patient
9.06am
At Drs station
With patient
With patient
9.08am
With patient
With patient
With patient
14
Ensure that all CDPs are specific actions or omissions on the part of the staff, rather
than more general observations on the quality of care. It is easy for example to put
down poor teamwork as a CDP which maybe a correct description of the team, but
should be recorded as a contributory factor as it was likely that poor teamwork
influenced the CDP. Although in practice CPDs and contributory factors may engage
together, it is best not to explore the contributory factors until the team is sure they
have a complete list. A variety of techniques are available to both an individual
investigator or team to tease out the CDPs, such as brainstorming, brain writing and
failure modes and effects analysis.
TIME
CDPs
ContriButory
Factors
Recommendations
15
Team
Task
CDP
Patient
Individual
Environment
16
Actions
to
Address
Factors
Level of
Recommendation
(Individual, Team,
Directorate,
Organisation
By
Whom
By
When
Resource
Requirements
Evidence of
Completion
Completion
Sign-off
17
Contributory
Factors
REFERENCES
1. Vincent, C., Taylor-Adams, S., Chapman, E.J., Hewett, D., Prior, S., Strange,
P. et al. How to investigate and analyse clinical incidents: Clinical Risk Unit
and Association of Litigation and Risk Management Protocol, Br Med J.
2000;320:777-81
2. Vincent, C.A. Understanding and responding to adverse events, N Engl J Med.
2003; 348: 1051-56
3. Eagle, C.J., Davies, J.M. and Reason, J.T. Accident analysis of large scale
technological disasters: applied to anaesthetic complications. Can J Anaesth.
1992; 39: 118-22
4. Reason, J.T. The human factor in medical accidents. In Vincent C.A. editor.
Medical Accidents. Oxford: Oxford Medical Publications; 1993
5. Reason, J.T. Understanding adverse events: human factors. In Vincent C.A.
editor. Clinical Risk Management. London: BMJ Publications; 1995
6. Cooper, J.B., Newbower, R.S. and Kitz, R.J. An analysis of major errors and
equipment failures in anaesthesia management considerations for prevention
and detection. Anesthesiology, 1984; 60: 34-42.
7. Cook, R.I. and Woods, D.D. Operating at the sharp end: the complexity of
human error. In: Bognor M.S. editor. Human Error in Medicine. Hillsdale,
New Jersey; Lawrence Erlbaum Associates Publishers: 1994
8. Vincent, C.A., Bark, P. Accident analysis. In Vincent CA editor. Clinical Risk
Management. London; BMJ Publications: 1995.
9. Stanhope, N, Vincent, C.A., Taylor-Adams, S., OConnor, A., Beard, R.
Applying human factors methods to clinical risk management in obstetrics.
BJOG. 1997; 104: 1225-32.
10. Taylor-Adams, S.E., Vincent, C., and Stanhope, N. Applying Human Factors
Methods to the Investigation and Analysis of Clinical Adverse Events. Safety
Science. 1999; 31: 143-159.
11. Vincent, C.A., Adams, S. and Stanhope, N. A framework for the analysis of
risk and safety in medicine. Br Med J. 1998; 316: 1154-7
Acknowledgements
Over the years the research that led to this protocol has been supported by a number
of organisations and charities. In particular, we thank BUPA Foundation, UK
Department of Health Patient Safety Programme and the Nuffield Trust and the Smith
and Nephew Foundation.
18