7. Retrospective chart audit
This Toolkithas attempted to synthesise the evidence available on each of these activities in order to provide a better practice guide for each.The bibliography contained in Section 5 provides further information on each of the activities described.
The activities described in this section should be used in the following context:
•
Whenever possible, a multidisciplinary approach should be taken in applying concepts and strategies.•
The processes undertaken by clinicians in applying the concepts in this Toolkit must be transparent and accountablenot only to other clinicians but also to health service managers and patients.•
Clinicians should be encouraged to use de-identified informationto ensureconfidentiality of the patient and the clinician and to enable open and frank discussions of the issues raised around specific events.
•
Evaluation or criticisms of events should look at system issues in the first instance, rather than seeking to blame the individual for errors or perceived errors that may have occurred.•
Involvement of junior staff and studentsshould be facilitated and supported.2
on the quality of clinical care
Introduction
Research studies7 have validated an epidemic of grossly under reported, preventable injuries due to medical management. Recent policy documents have placed high priority on improving incident reporting as the first step in addressing patient injuries, and have called for translation of lessons from other high risk industries.8Complex non-medical industries have evolved incident reporting systems that focus on near misses, provide incentives for voluntary reporting, ensure confidentiality (as distinct from anonymity) and emphasise perspectives of systems in data collection, analysis and improvement.
Definitions
An incidentis an unplanned event resulting in or having the potential for injury, ill health, damage or other loss.9Any event that could have had adverse consequences but did not, and is indistinguishable from fully fledged adverse events in all but outcome10 should also be considered to be an incident.These are known as ‘near misses’.
Incident monitoringis a system for identifying, processing, analysing and reporting incidents with a view to preventing their recurrence.11
The process of incident monitoring has, in the past, been criticised because of the limited ability it has to identify the systemic problems underlying adverse event occurrence. Some important categories of incidents are unlikely to be identified using this system (eg. incidents involving errors of omission rather than comission) unless the process for identification is facilitated.
Incident monitoring has also been criticised for its unidisciplinary nature, having become the domain of nursing staff. If incident monitoring and management is to be effective, it must be team based, multidisciplinary and involve both senior and junior staff.
Effective incident monitoring is also dependent on a commitment to act upon the information that arises from the process for improvements in the systems of care.This also involves the reporting of incidents or processes that require action at the facility level. Such action is the responsibility of the peak quality of care committee in that facility.
Facilitated incident monitoringis the use of current incident reporting mechanisms, enhanced by the opportunistic identification of a greater range of incidents than can be expected from the current voluntary methodology.
Objective
The objective of facilitated incident monitoring is to strengthen the current incident reporting system, to facilitate the identification of a greater proportion of incidents that occur in clinical practice and to act upon the information derived from these reports to improve care.
Method
1. Each clinical team or ward-based unit will identify an appropriate time to discuss the incidents occurring in their clinical area in the previous time period, eg. the past week.This should not be a separate ‘incident meeting’ but should instead be the ‘normal’ clinical team meeting.
2. The team will take those incidents which have been voluntarily reported using the current reporting forms and add to them any other incidents identified by asking a set of (approximately eight) appropriate questions.
The questions should be based on local knowledge of incidents which could occur in that clinical setting and should be based on the six dimensions of quality: safety, effectiveness, appropriateness, consumer participation, access and efficiency.12 The questions will be different for each clinical team and the incidents being considered.
1 Facilitated incident monitoring
For example an acute clinical care team may ask the following questions.“In the past week, have there been...
•
any drug errors?•
any intravenous line infections?•
any unanticipated admissions to ICU?•
any falls?•
any wound infections?•
non-compliance with (identified) guidelines?•
inappropriate admissions/treatments?•
unreported results?•
reports not acted upon in a timely fashion?•
delayed, premature or inadequate discharges/transfers?•
complaints?•
any pressure ulcers? or•
any gaps in care?”A community care team would ask a different set of questions more relevant to their practices and expected incidents.
Such incidents should be identified prior to the ward/team meeting and an appropriate person nominated to follow up on the relevant details of the incident prior to discussion at the meeting.
3. The facts about the incidents should be presented by a team member and discussed:
•
Patient and provider information should, when possible, be de-identified•
Discussion should be robust, but the approach should always be educational rather than fault-finding•
Discussion should be focussed around identifying the system issues in the care delivered.4. To assist the discussions the following questions should then be asked.
•
What did we do or what did we forget to do that contributed to these incidents? (It should be recognised that errors of omission are far more common than errors of commission.)•
What needs to be done at this level to prevent this incident from occurring again?•
Who is responsible for follow-up action?•
Who else needs to know about this? For example, does it need to be reported to the facility’s quality committee, either for action or information?5. Possible actions/outcomes:
•
If insufficient information is available regarding an incident, a person should be assigned to follow up and re-present the issue at the following meeting.•
If a deficiency in the system is identified, any of the following may be appropriate:– a new or revised practice – a new or revised protocol
– improved lines of communication.
•
If there are concerns about an event in another department/grouping– a letter could be written – a meeting could be arranged
•
If a broader system/facility issue is identified it should be reported to the facility’s peak quality committee.In conclusion
Facilitated incident monitoring should be a continuous, ongoing fundamental activity for every clinician and clinical team. Incidents and near misses will always occur. Non-punitive, voluntary incident reporting systems in high risk non-medical domains have grown to produce large amounts of essential process
information that is unobtainable by other means.
Health care professionals need also to report and examine incidents and near misses in the same way.
Definition
“A sentinel event is an unexpected occurrence involving death or serious physical or psychological injury and includes any process variation for which a recurrence would carry a significant chance of a serious adverse outcome.”14 A sentinel event can be further defined as an event that happens rarely or one which represents an adverse event of such significance that it warrants individual investigation. Such events are called ‘sentinel’ because they signal the need for investigation and response.
Objectives
The objectives of sentinel event monitoring are to:
1. Have a positive impact in improving patient care.
2. Focus the attention of a team/facility that has experienced a sentinel event on understanding the causes underlying the event, and on making changes in the care delivery systems and processes to reduce the probability of such an event in the future.
3. Increase the general knowledge about sentinel events, their causes, and strategies for prevention.
4. Improve the safety of health care for the consumer and to maintain the confidence of the public in the care provided.
When a sentinel event occurs in a health care
organisation, it is necessary that appropriate individuals within the organisation are made aware of the event;
investigate and understand the causes that underlie the event; and make changes in the organisation’s systems and processes to reduce the probability of such an event occurring in the future.
Identification of sentinel events
Sentinel events can occur and be identified in three ways:
1. There are sentinel events that can be identified by teams prospectively, that is, anticipated. It will be known, that if certain outcomes eventuate, they will constitute sentinel events. If such an event does occur, it could flag a major quality of care issue.These events will be different for each clinical area.Examplesof sentinel events that can be identified prospectively are:
•
an unexpected neonatal death•
surgery on the wrong patient or the wrong limb•
haemolytic transfusion reaction involving administration of blood or blood products having major blood group incompatibilities•
cannulation of a peripheral artery for administration of IV drugs•
loss of digits or limbs as a result of health care management•
infection of children who suffer from cystic fibrosis and who are Pseudomonas cepacia negative by children who are Pseudomonas cepacia positive•
significant adverse drug reactions•
significant medication errors•
death of a renal transplant patient within three months of transplant•
renal transplant graft failure within 3 months, including unsuccessful intraoperativecompletion of transplantation.
2. Other sentinel events will only be recognised at the time they occur, as the event could not be anticipated. For example:
•
death of a conscious, previously fit, 19 year old male, admitted to ICU overnight for observation of a possible head injury, following an MVA.2 Sentinel event management
133. Other events may initially appearas a series of incidents, which may only be retrospectively identified as a sentinel event. Staff must be taught to recognise the appearance, or development of, a sentinel event. For example:
•
the prolonged length of stay (LOS) of one patient following an endoscopic retrograde cholangeo pancreatography (ERCP) because of unexpected pancreatitis and followed soon after by the prolonged LOS of another patient and then perhaps another patient, following ERCP, all of whom have suffered unexpectedpancreatitis.
Each event in itself would not constitute a sentinel event, but the cluster of events should alert
clinicians and managers to the possibility of a significant problem.
Method
When a clinical team detects or suspects significant undesirable performance or variation, which has, or could result in, a sentinel event, an intense analysis must be initiated to determine where best to focus changes for achieving improvement.
All sentinel events require investigation and analysis and such analysis should be conducted under the six dimensions of quality of care.
The investigation should focus primarily on systems and processes, not individual performance. It should progress from special causes in clinical processes to common causes in organisational processes.The analysis should identify potential improvements in processes or systems that would tend to decrease the likelihood of such events in the future, or it may be determined after analysis, that no such improvement opportunities exist. Such an investigation is referred to as a Root-Cause Analysis.
Action required
The product of the investigation and analysis of a sentinel event is an action plan that identifies the strategies that the clinical team intends to implement to reduce the risk of similar events occurring in the future.The plan should address responsibility for implementation, mechanisms for oversight, time lines, and strategies for measuring the effectiveness of the actions.
Reporting
The investigating team leader should report sentinel events to both the clinical director and the peak quality committee in the facility in which the event occurred.That committee is responsible for
undertaking the actions required at the facility level to ensure that the risk of a repeat event is reduced. Area Health Services need to clearly define the lines for reporting sentinel events to the Area Executive, the Area Quality Council and beyond.
Such reports should include:
•
a brief summary of the event•
a brief summary of the investigation and analysis that has taken place•
actions (to be) taken•
recommendations for other teams, other facilities and/or the Area Health Service and/or the health system as a whole.A statewide Quality Improvement Incident
Management Strategy for the NSW health system has identified the need for mandatory reporting of sentinel events and selected frequent incidents at the State level to ensure that lessons learned from each event are communicated to the entire health system.
Managers will be required to report, investigate and analyse such events in the future.
Definition
Clinical indicatorsare indicators or measures that relate to specific clinical conditions, or measures of function that have particular significance for particular conditions.15A clinical indicator is also defined by the Australian Council on Healthcare Standards (ACHS) as a measure of the clinical management and outcome of care. It is an objective measure of either the process or outcome of patient care in quantitative terms.
Objectives
Clinical indicators are not exact standards; rather, they are designed to be ‘flags’ which, through the collection and analysis of data, can alert clinicians to possible problems and/or opportunities for improvement in patient care.These areas can then be further
investigated within the clinical team or referred to the peak quality committee within that facility for action at the organisational level.
The benefits to be gained from the use of clinical indicators do notlie in the collection of the data, but in how those data are used; that is, in the data analysis and the actions taken to achieve sustained improvements in clinical practice. Clinical indicators do not ‘work’ unless used effectively by clinicians and managers to bring about improvements.
The following should be considered when selecting indicators:
•
Identify those indicators which need to be monitored and reported.•
Indicators should be chosen for the value they have in providing the information required by clinicians to measure and improve the quality of care, not simply for the purposes of reporting.•
The indicators may be based on ACHS or other established and validated indicators.•
The list of indicators collected and reported does not necessarily need to be lengthy. A smaller number of indicators, that provide useful and relevant information about the quality of clinical care and upon which action can be taken, is far better than a large number of indicators that do not fulfil these needs.•
The identification of appropriate indicators should be an iterative process and should involve an assessment of issues such as the usefulness of the data, availability of existing collection mechanisms and resources required for collection.•
These indicators may be ACHS, State, national, College, or locally developed indicators. It should be noted that the ACHS clinical indicators are developed and endorsed by the medical colleges and the use of national or State based clinical indicators there is the added advantage of gaining aggregated comparative values.Method
•
The indicator data collected will give clinical teams information about the care that is delivered.•
Teams should discuss the data collected on each indicator chosen and identify those areas ofpractice variation that require further investigation.
Teams should incorporate these discussions into other team/clinical unit meetings.
•
Such data should be used to ‘flag’ areas for possible investigation.The absolute numbers collected may not give teams a great deal of information initially.Graphed trends in the data will give teams more valuable information and will alert teams to the need for investigation.
•
These investigations should be undertaken using the scientific, clinical practice improvement method described in Section 3 of this Toolkit.3 The effective use of clinical indicators
Variationsin the data collected can be found in almost every indicator, whether they be clinical, facility or Area level data.There will almost always be a range in the levels of performance identified.There are many reasons for these variations. It is not always possible to immediately identify the cause(s) of
variations in data and therefore what needs to be done to reduce it. If variation is found however, it is
essential that health services thoroughly examine the cause and institute action when such an examination reveals opportunities for improvement.
It is recommended to initially identify the following information:
•
the reason for collecting the data•
the definition of the numerator, denominator and standard/benchmark (if available)•
the reporting period•
the numbers•
the action required to improve the quality of care as a result of the data collection•
comments on the data, eg. an explanation of any unusual results.A model collection and action format can be found at Appendix 2.
Definitions
Peer reviewis the process of reviewing one’s peers.
The difficulty arises in defining ‘one’s peers’.The term
‘peer’ will therefore be deliberately left undefined, but medical practitioners are encouraged to include nurses and allied health professionals in peer review where possible and appropriate.
Peer review is accepted as an important part of the quality improvement processes intrinsic to health care, and the practice of peer review is as diverse as medicine itself. Peer review is not necessarily a distinct entity, but can and does form a major part of a number of quality improvement strategies, as described in this Toolkit.
A Peer Review meetingis but one way of undertaking peer review. It is a meeting in which clinicians seek to improve their treatment of patients, and to maintain the currency of their practice by focussing on recent events and outcomes (individual or collected) of the patients under the care of the group forming the meeting.
The benefits of peer review, in its most narrow definition (ie. a meeting of senior medical
practitioners only), are not easily able to be quantified, but it still plays an important role in the quality improvement efforts of a number of clinicians, particularly in some disciplines.
Objective
The primary role of peer review is to inform those present about the status of their own practices against their peers.
Guidelines
Clinicians who engage in peer review meetingsneed to establish guidelines about the form, content and documentation of meetings, with respect for the existing diversity across health disciplines. A recent report16on peer review suggests the following
1. A Peer Review meeting should be held at least four times per year in most disciplines, and should be constituted so that the minimum quorum is three participants.There should be an elected Chair, and a member should be nominated to take note of key findings.
2. The report that results from a Peer Review meeting will depend upon the discipline involved and whether the meeting has been granted Qualified Privilege under the NSW Health Administration Act 1982(as amended).17This legislation itself is being reviewed, and it is hoped that any changes will make the process for obtaining privilege clearer. Most current peer review committees are not privileged, and may not need to be.
The report needs to be in a form that recognises the essential components of Peer Review.
The report needs to be in a form that recognises the essential components of Peer Review.