skip to navigation | skip to content

Clinical Incident Management The Clinical Excellence Commission Clinical Incident Management in the NSW public health system


Clinical Incident Management in the NSW public health system
 

Clinical Incidents

A clinical incident is any unplanned event which causes, or has the potential to cause, harm to a patient.

NSW Health staff are required to report all incidents (both clinical and corporate), near misses, and complaints so that risks to patient safety are recognised and action is taken to prevent recurrence. This is supported by NSW Health Incident Management Policy (PD2014_004).

The incident reporting system in the NSW public health system is called the Incident Information Management System (IIMS). St Vincent's Health Network uses Riskman.

Since 2005, the IIMS has been used to notify and manage incidents from across NSW public health facilities. Thematic analysis of incident data enables significant issues, risks and trends relating to clinical care to be identified. This facilitates improvement work, ultimately improving the reliability and safety of clinical care being provided to patients. Dissemination of the lessons learned to staff is through regular reports on key findings, and in-depth analysis of clinical issues, presented as Clinical Focus Reports and focused summary reports and Patient Safety Watches.

Incidents reported in the IIMS and Riskman are classified according to a list of Principal Incident Types (PITs), and then further stratified against a Severity Assessment Code (SAC).  The key purpose of the SAC is to direct the level of investigation and action required for a particular event. There are four SAC ratings, ranging from SAC1 (extreme risk) to SAC4 (low risk). All SAC1 incidents are subject to a thorough investigation known as a Root Cause Analysis (RCA). This investigation method identifies root causes, contributing factors and/or systems issues, and recommendations are made to mitigate the risks identified.

Following the identification of a SAC1 incident there is a defined process which takes place, from point of notification and investigation, to open disclosure and analysis of findings, and dissemination of lessons learned. (Figure 1).

The NSW Ministry of Health is notified of SAC1 and other significant clinical events through the Reportable Incident Brief (RIB) system. All SAC1 incidents, including national sentinel events, as outlined in the NSW Health Incident Management Policy PD2014_004, must be notified via the RIB process. National Sentinel Events occur infrequently and occur due to system and process deficiencies in our healthcare system. These sentinel events result in death or serious harm. There are eight Nationally agreed sentinel events, these are:


  1. 1. Procedures involving the wrong patient or body part resulting in death or major permanent loss of function

  2. 2. Suicide of a patient in an inpatient unit

  3. 3. Retained instruments or other material after surgery requiring re-operation or further surgical procedure

  4. 4. Intravascular gas embolism resulting in death or neurological damage

  5. 5. Haemolytic blood transfusion reaction resulting from ABO incompatibility

  6. 6. Medication error leading to the death of a patient reasonably believed to be due to incorrect administration of drugs

  7. 7. Maternal death associated with pregnancy, birth and the puerperium

  8. 8. Infant discharged to the wrong family

The Clinical Excellence Commission (CEC) collates and analyses all clinical RIBs. The following report is informed by IIMS, CEC RIB and RCA databases.

CEC's role in incident reporting

As the lead agency to support patient safety and clinical quality improvement in the NSW public health system, the CEC has a key role to play in analysing and reporting on the information provided from the IIMS. The CEC has developed a number of projects and programs which have resulted from the analysis of incident data in the IIMS, including those focused on the recognition and management of the deteriorating patient and addressing patients with sepsis.

NSW Health was the first Australian jurisdiction to publically report healthcare incident data. The first report was released in 2005.  The CEC published its first web-based clinical incident management report in 2013. Previous reports are available on the CEC website. This regular publication of data is part of the CEC's  commitment to be transparent and support the NSW health system in providing the safest and highest quality care for every patient.

For more information about the Clinical Excellence Commission, its programs, resources and publications see http://www.cec.health.nsw.gov.au


Included in this report

The information in this report includes data extracted from the IIMS, the CEC RIB and RCA databases, Riskman and Health Information Exchange (HIE). For more specific extract information, please contact the CEC Patient Safety Team CEC-patientsafety@health.nsw.gov.au


This report contains:
  • How to interpret the data and information
  • Clinical incident notification data, such as principal incident types and the severity of incidents
  • Systems factors in serious clinical incidents identified through RCA
  • Specific Clinical incidents involving:
    • Patient identification
    • Falls
    • Medication Safety
    • Paediatrics
    • National Sentinel Events
  • Complaints notified in the incident information management system and how they are resolved

Data Interpretation

The IIMS narrative and analysis may provide greater insight into incidents, provide context, and highlight issues, and system-related opportunities for improvement.

Given the wide variation between services and facilities, accurate comparisons based on notification numbers alone cannot be made. Caution is advised if using IIMS reporting counts or rates as the single source of benchmarking data for a project or program, as many variables influence incident reporting. Lower rates of reporting are not a reliable indicator of safer care, therefore, further qualitative, rather than quantitative, interpretation of the data is recommended..


Figure 1: Serious clinical incidents requiring RCA investigation
Clinical Incident Management

Severity of Clinical Incidents

Clinical incidents notified in IIMS and Riskman are allocated a Severity Assessment Code (SAC) rating in accordance with NSW Health Incident Management Policy PD2014_004. The most serious types of clinical incidents are rated as SAC1 (the other possible scores are SAC2, SAC3 or SAC4 in declining order of severity). The key purpose of the SAC is to determine the level of investigation and action required. While a notifier may assign an initial SAC, the actual SAC must be confirmed by a manager within five days of the incident notification. All SAC1 incidents and National Sentinel Events (NSE) require a Reportable Incident Brief (RIB) to be submitted to the NSW Ministry of Health.

While the overall number of incidents is increasing in line with increased activity in NSW hospitals, the data shows that serious incidents in healthcare are extremely rare. Less than two per cent of incidents notified were rated as SAC1 or SAC2, indicating that rates of serious clinical incidents are very low across NSW, and most incidents resulted in little or no harm to the patient involved.

Most reported incidents (more than ninety-eight per cent), were rated as SAC3 or SAC4, and resulted in minimal or no patient harm.

Table 1: Clinical Incidents notified by Actual SAC rating, January 2013 - June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, SAC2-4 obtained from IIMS/SVHN Riskman

There has been a slight increase (0.35 per cent) in the overall number of clinical incident notifications for January – June 2017 when compared to the previous reporting period. Table 1 identifies that the number of incidents reported in the months July to December are consistently higher than incidents reported during the months of January to June in each calendar year.

Figure 2 and 3: Clinical SAC1 & SAC2 and SAC3 & SAC4 incident notifications, January 2013 - June 2017

Clinical Incident Management

* SAC1 data obtained from CEC RIB database, SAC2-4 obtained from IIMS/SVHN Riskman

Clinical Incidents per 1,000 Bed Days

Reporting the number of clinical incidents in relation to activity i.e. per 1,000 bed days, provides greater insight than the number of incidents alone. The rate of SAC1 and SAC2 incidents from 2013 – 2017 has remained stable. SAC3 and SAC4 incidents that result in little or no harm have increased slightly during the first half of 2017. This demonstrates NSW hospitals have a robust reporting culture and high reliability. The rate of incidents per 1,000 acute bed days is slightly higher for the January – June 2017 reporting period compared with July – December 2016.

Table 2: Clinical Incident notifications by SAC per 1,000 acute bed days, January 2013 - June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, SAC2-4 data obtained from IIMS/SVHN Riskman
** Upon advice received from the NSW Ministry of Health, denominator data has been amended for 2015 and 2016 reporting periods. As a result, rate based data has also been updated. An increase in the rate per 1,000 bed days is noted from July – December 2015 to July – December 2016 reporting periods.

Figure 4 and 5: SAC1 & SAC2 and SAC3 & SAC4 clinical incidents per 1,000 acute care bed days, January 2013 – June 2017

Clinical Incident Management


Principal Incident Type

When notification of a clinical incident is made, a Principal Incident Type (PIT) is recorded in the IIMS which enables the notifier to further categorise the nature of the incident. There are 19 PITs. Table 3 and Figure 6 display clinical incidents by PIT.

During the January – June 2017 reporting period, the top four PITs have remained consistent with previous reporting periods: Clinical Management, Medication/IV fluid, Fall and Pressure Ulcer (also known as Pressure Injury). Since January 2015, Clinical Management has replaced Fall as the most frequently reported PIT.

Reporting of pressure injuries demonstrates changes which could be due to the release of the Pressure Injury Prevention and Management policy PD2014_007, along with the National Safety and Quality Health Service Standards, the Pan Pacific Guideline for the Prevention and Management of Pressure Injury (2012) , the International Prevention and Treatment of Pressure Ulcers: Clinical Practice Guideline (2014), and the establishment of a statewide working party in January 2013.

Table 3: Clinical Incidents by PIT as a percentage*, January 2013 – June 2017

Clinical Incident Management
Data obtained from IIMS, excludes St Vincent’s Health Network
* PIT expressed as a % of total IIMS incidents for the reporting period
** Other includes categories of blood/blood product, building/fitting/fixture/surround, complaint, nutrition, oxygen/gas/vapour, pathology/laboratory and security

Figure 6: Clinical incidents by top four principal incident types (PITs) as a percentage of total clinical incident notifications,
January 2013 – June 2017

Clinical Incident Management

SAC1 Reportable Incident Briefs (RIBs)

The total number of SAC1 notifications has decreased by five per cent over the January – June 2017 reporting period. The rate of SAC1 incidents has remained stable for a number of consecutive reporting periods.


The most frequently notified SAC1 incidents from January 2013 to June 2017 continued to be categorised under the PIT of Clinical Management. This includes incidents associated with diagnosis, patient identification, and treatment of patients in any inpatient care setting (Table 4).

Table 4: SAC1 incidents by PIT, January 2013 – June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database
** All clinical streams, includes patient identification errors (see 'Definitions' TAB)
*** Patient identification reporting requirements changed on 10th February 2014
**** Expressed Breast Milk (EBM) excluded in Jul-Dec 2013 and reported in Incorrect Person Procedure Site data
† Other includes Blood/Blood Products, Health care associated infection, Medication/IV Fluid, Undetermined cause of death and RCAs not received
# Includes patient identification incidents

SAC2-4 IIMS Data


Table 5: SAC2 incidents by principal incident type, January 2013 – June 2017

Clinical Incident Management
SAC2 data obtained from IIMS, excludes St Vincent's Health Network
* Other includes the categories of Accident/occupational health and safety, Aggression-Victim, Blood/blood product, Building/fitting/fixture/surround, Complaint, Documentation, Medical device/equipment/property, Nutrition, Organisation management/service, Oxygen/gas/vapour, Pathology/laboratory and Security

Table 6: SAC3 incidents by principal incident type, January 2013 – June 2017

Clinical Incident Management
SAC3 data obtained from IIMS, excludes St Vincent's Health Network
* Other includes the categories of Accident/occupational health and safety, Aggression-Victim, Blood/blood product, Building/fitting/fixture/surround, Complaint, Documentation, Health care associated infection/infestation, Medical device/equipment/property, Nutrition, Organisation management/service, Oxygen/gas/vapour, Pathology/laboratory and Security

Table 7: SAC4 incidents by principal incident type, January 2013 – June 2017

Clinical Incident Management
SAC4 data obtained from IIMS, excludes St Vincent's Health Network
* Other includes the categories of Accident/occupational health and safety, Aggression-Aggressor, Aggression-Victim, Behaviour/human performance, Blood/blood product, Building/fitting/fixture/surround, Complaint, Health care associated infection/infestation, Medical device/equipment/property, Nutrition, Organisation management/service, Oxygen/gas/vapour, Pathology/laboratory and Security


System Factors in Clinical Incidents - Root Cause Analysis

A Root Cause Analysis (RCA) is required to investigate every SAC1, and selected SAC2-4, clinical incidents in the NSW Health public health system. The RCA method is used to identify how organisational systems can cause or contribute to clinical incidents. The RCA report findings, and state-wide aggregated analysis, inform system improvements which could prevent similar incidents from occurring in the future. Examples of these include the development of programs, reports and quality tools including Between the Flags, Sepsis Kills, Patient Safety Watches and Clinical Focus Reports.

The CEC reviews all clinical RCA reports through four RCA review committees; Clinical Management, Maternal and Perinatal, Mental Health/Drug & Alcohol (MHDA) and Child and Young Person. The RCA Review Committees classify each RCA report using a standard taxonomy. The classification taxonomy is revised as emerging issues and clinical practice changes are identified.

During the reporting period of January – June 2017, the top two system factors identified by the Clinical Management, Maternal and Perinatal RCA review committees and the Mental Health/Drug and Alcohol Classification Group were Communication and Care planning.

The system factor communication relates to both verbal and written forms of communication. This includes: handover between clinicians or shifts; clinical documentation (including electronic medical record), communication with patients, families and carers and informed consent. Inadequate communication is also a recognised contributor to incidents in other industries. Human factors is identified as a key element for improving communication and the CEC has developed human factors training which was launched in 2015.

The system factor care planning pertains to incidents where there may have been gaps or failures in collaborative planning for patients receiving care from more than one team. This includes: care continuity and care co-ordination within a facility or between health care facilities, including private providers, inpatient and community-based services. Care planning also covers incidents which occur when a patient's risk factors including co-morbidities, falls risk, or the capacity of their carers to manage ongoing care, have not been adequately assessed or managed.

System Factors identified in the Clinical Management RCA Review Committee

The top three system factors identified by the Clinical Management RCA Review Committee (Table 8) relate to Care Planning, Communication, and Policy and Guidelines. During the January to June 2017 reporting period the system factor Policy and Guidelines has replaced Assessment as the third most identified system factor.


The system factor Policy and Guidelines is applied to incidents where it is identified that there is no policy or guideline; existing policies and guidelines have not been implemented by a staff member or organisation; when a policy or guideline is not in line with NSW Health policy or evidence based practice; if it is not available, unclear or unworkable.


Table 8 outlines the system factors identified in Clinical Management RCAs January 2013 to June 2017 and figure 7 outlines the top three system factors identified through Clinical Management RCAs.

Table 8: System factors identified through Clinical Management RCA reports, January 2013 – June 2017

Clinical Incident Management

Private health facility RCAs are included, when provided by the private facility and represents RCA reports received during the specified reporting period
* ‘No factors identified’ was added as a system factor in January - June 2017

Figure 7: Top three system factors identified through Clinical Management RCA reports, January 2013 – June 2017

Clinical Incident Management


System Factors identified in the Mental Health/Drug and Alcohol RCA Review Committee

A revised MHDA classification taxonomy was implemented in July 2016. It was developed with the assistance of mental health clinical experts, to ensure that the classification categories reflect current clinical care practices. This has been an iterative process with refinement to the classifications occurring over time.


The data from MHDA RCAs has shown similar systems factors to Clinical Management RCAs. The top three identified issues are Care Planning, Communication and Assessment. The system factor Assessment is applied to incidents where deficiencies or gaps in the physical, cognitive or mental health assessment of a patient have been identified, when risk assessments are inadequate or not completed, or when patient or carer concerns are not considered during assessment.

Table 9 outlines the system factors identified in MHDA RCAs for the period January 2013 to June 2017, and figure 8 outlines the top three system factors identified through analysis of MHDA RCAs for the period January 2013 to June 2017.

Table 9: System factors identified through review of Mental Health/Drug and Alcohol RCA reports, January 2013 – June 2017

Clinical Incident Management

Private health facility RCAs are included, when provided by the private facility and represents RCA reports received during the specified reporting period.
* Changes were made to Mental Health/Drug & Alcohol RCA System Factors in January - June 2017. New categories are now included


Figure 8: Top three system factors identified through review of Mental Health/Drug and Alcohol RCA reports
January 2013 – June 2017

Clinical Incident Management

System Factors identified in the Maternal and Perinatal RCA Review Committee

In the January – June 2017 reporting period, Care Planning, Communication, and Policy and Guidelines were most frequent system issues identified by the Maternal and Perinatal RCA Committee.


Table 10 outlines the system factors identified in Maternal and Perinatal RCAs for the period January 2013 to June 2017 and Figure 9 outlines the top three system factors identified through Maternal and Perinatal RCAs for the period January 2013 to June 2017.


Table 10: System factors identified through Maternal and Perinatal RCA reports, January 2013 – June 2017

Clinical Incident Management
Private health facility RCAs are included, when provided by the private facility & represents RCA reports received during the specified reporting period.

Figure 9: Top three system factors identified through Maternal and Perinatal RCA reports, January 2013 – June 2017

Clinical Incident Management

System Factors identified in the Child and Young Person RCA Review Committee

The Children and Young Person RCA Review Committee was established in 2016. In the January – June 2017 reporting period, Care Planning, Policy and Guidelines, and Communication were the most frequent system issues identified (Table 11). This is consistent with the system factors identified in the Clinical Management and Maternal and Perinatal RCA Review Committees

Feedback reports in the form of Paediatric Watches have been developed to share the learnings from incident notifications with clinicians across the state.


Table 11 outlines the system factors identified in Child and Young Person RCAs for the period January 2013 to June 2017.

Table 11: System factors identified through Child and Young Person RCA reports, January 2013 – June 2017

Clinical Incident Management
* The system factor of Investigations was added as a system factor in July 2016
** As from July - December 2016, RCA reports involving Paediatric Mental Health are included
*** All RCAs reviewed at the Child and Young Persons RCA Review Sub-Committee are included, regardless of Primary RCA Committee
Risk Factors in RCA Reports

The clinical risk factors identified by the RCA Review Committees relate to the conditions or situations that were identified as being a direct cause, or contributing factor, to the incident. In August 2013, the taxonomy for reviewing risk factors was refined to be more detailed and provide improved analysis. Tables 12, 13, 14 and 15 highlight the top five risk groups that were identified in the RCA Review Committees between January 2013 to June 2017 reporting periods.


Table 12: Top five risk factors identified through Clinical Management RCA reports, January 2013 – June 2017*

Clinical Incident Management
* Top five risk factors for January - June 2017
** New category collected from January - June 2017

Table 13: Top five risk factors identified through Mental Health/Drug and Alcohol RCA reports, January 2013 – June 2017*

Clinical Incident Management
* Top five risk factors for January - June 2017
** Definitions revised in 2016
*** New category collected from January - June 2017

Through close collaboration with mental health and drug and alcohol clinical experts, the MHDA classification taxonomy has continued to be strengthened over this period of time. The MHDA risk factors (Table 13) are reflective of these changes as shown in the January – June 2017 reporting period.

Table 14: Top five risk factors identified through Maternal and Perinatal RCA reports, January 2013 - June 2017*

Clinical Incident Management
* Top five risk factors for January – June 2017

Table 15: Top five risk factors identified through Child and Young Person RCA reports, January 2015 - June 2017*

Clinical Incident Management
* Top five risk factors for January - June 2017
** As from July - December 2016, RCA reports involving Paediatric Mental Health are included
*** All RCAs reviewed at the Child and Young Persons RCA Review Sub-Committee are included, regardless of Primary RCA Committee
**** Included as a risk factor in July - December 2016

The failure to recognise and escalate deterioration is a common risk factor identified by all RCA committees. To address the problems associated with these risk groups, the CEC's Between the Flags program provides a suite of standard observation charts which incorporate standard calling criteria which maps clinical parameters to assist in the recognition of deterioration, and when to escalate care of patients. The program also includes minimum standards for escalation including processes for both Clinical Review and Rapid Response in all NSW Health facilities. The CEC continues to reinforce the importance of early recognition and response to patients who are clinically deteriorating through the SEPSIS KILLS and REACH programs.

Local Health Districts have adopted the implementation of Between the Flags into the electronic medical record. This initiative will assist staff in recognising abnormal clinical observations and deterioration earlier, and subsequently escalate care requirements in accordance with policy.


Patient Identification Clinical Incidents


Patient identification incidents refer to incidents associated with the matching of the correct patient, site and procedure. The NSW Health Policy – Clinical Procedure Safety (PD2017_032) describes the steps that must be taken to ensure that an intended invasive or diagnostic procedure (including surgical operations, endoscopy, dentistry, radiology, nuclear medicine, chemotherapy and radiation therapy) is performed on the correct patient, on the correct site and, if applicable, with the correct implants/prostheses and equipment.

Prior to 10 February 2014, all incidents which involved the incorrect patient, procedure, body part or surgical implant were classified as serious incidents (SAC1) in the Incident Information Management System (IIMS), and subsequently underwent Root Cause Analysis (RCA) investigation. In 2014, the NSW Health Incident Management Policy PD2014_004 was amended and mandated that incorrect patient, site, and/or procedure incidents be classified according to actual harm, and those not resulting in harm were no longer automatically recorded as a SAC1 incident.

Procedures involving the wrong patient or body part, regardless of the outcome, requires notification to the NSW Ministry of Health as a Reportable Incident Brief (RIB). These incidents are referred to as patient identification incidents in this report.

In January – June 2017, there were two SAC1 incidents reported, both occurring in the Operating Theatre or Anaesthetics area. The number remains consistent compared to the previous reporting period (Table 16).

During this reporting period there has been a six per cent decrease in SAC2-4 notifications when compared with the previous reporting period. These incidents involve mismatching / failure to correctly identify patients prior to intended clinical intervention which may include diagnostic imaging and pathology tests. In most cases, these events have not resulted in actual harm to the patient (Table 17).

Table 16: Location of SAC1 incidents involving patient identification where clinical procedure has occurred
January 2013 – June 2017
Clinical Incident Management
* Incorrect patient / procedure / site reporting requirements changed 10 February 2014
** Incidents already accounted for in SAC1 data

Table 17: Location of SAC2, SAC3 and SAC4 incidents involving patient identification where clinical procedure has occurred
January 2014 – June 2017

Clinical Incident Management

* Incorrect patient / procedure / site reporting requirements changed 10 February 2014


Fall Related Incidents


Patient falls in hospital can be a major cause of harm, and may result in increased hospital length of stay.  Hospitalised patients older than 75 years of age have a significantly higher risk of falling (Figure 10). They often are frail, have poor mobility and may be confused. The confusion can be multifactorial and be secondary to dementia and or delirium, an unfamiliar hospital environment, acute illness and/or multiple medications.

The CEC Falls Prevention Program supports the implementation of the National Safety and Quality Health Standard 10: Preventing falls and harm from falls. A system-wide approach to fall risk screening and assessment has been introduced in NSW hospitals which includes a range of initiatives and resources to prevent falls and the harm associated with falls.

Figure 10: Falls by age, January 2013 – June 2017

Clinical Incident Management

The Severity of Inpatient Fall


During the reporting period January – June 2017, the number of incidents with a Principal Incident Type of Fall was 13,192. Of these incidents, 16 were classified as SAC1 and 246 were classified as SAC2 incidents. The SAC1 and SAC2 incidents account for 2 per cent of all falls notified. It is important to note that the majority of fall events were categorised as SAC3 or SAC4 events which did not result in serious patient harm.


Table 18: Falls by SAC rating, January 2013 – June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, SAC2-4 data obtained from NSW IIMS

Figure 11 and 12: Falls by SAC1 & SAC2 and SAC3 & SAC4,January 2013 – June 2017

Clinical Incident Management* SAC1 data obtained from CEC RIB database, SAC2-4 data obtained from IIMS

Type of Fall and Activity at Time of Fall

The most frequent type of fall notified during the January – June 2017 reporting period was loss of balance and slips (Figure 13). Forty-two per cent of patients with this mechanism of fall were recorded as moving and twenty per cent were standing when the fall occurred (Figure 12). A further fourteen per cent of patients were undertaking activities of daily living, such as unassisted toileting and showering. This is consistent with data from previous reporting periods.


Figure 13: Type of fall, January 2013 – June 2017

Clinical Incident Management* 'Fall type' field is not mandatory for completion

Figure 14: Activity at time of fall, January 2013 – June 2017

Clinical Incident Management* Multiple responses are allowed in the 'activity' field and is not mandatory for completion

Time of Patient Falls

Falls occur most frequently between 09:00 and 11:59hours. During these times, patients are more likely to be mobilising and attending to personal care activities. Significant activity also occurs on wards during these times with clinical handover, ward/medication rounds and diagnostic testing. Other peak times for falls are 13:00 – 14:59hours and 18:00-18:59hours which coincide with increased patient activity and toileting following meal times. It is unclear as to the effect medications may have in contributing to falls at this time e.g. effect on patient taking night sedation.


Figure 15: Time of patient falls, January 2013 – June 2017

Clinical Incident Management* 'Fall type' field is not mandatory for completion

Outcomes Following a Fall

Of the fall incidents reported during the January – June 2017 reporting period, forty-eight per cent of falls resulted in no harm to the patient. Twenty-seven per cent resulted in an injury to the patient. These injuries often require intervention. There were eleven (11) fall incidents (0.09%) recording death as an outcome for this period.

Table 19: Outcome of fall, January 2013 – June 2017

Clinical Incident Management
* Multiple responses are allowed in the 'outcome' field and is not mandatory for completion. This dataset is generic in nature as it is consistent across all clinical incidents and is not specific to fall related incidents.

Caution is advised if using the Incident Information Management System (IIMS) reporting counts or rates as the single source of benchmarking data for a project or program, as many variables influence incident reporting.  Lower rates of reporting are not an indicator of safer care, therefore, further qualitative, rather than quantitative, interpretation of the data is recommended.

Clinical Incident Management


Medication Related Incidents


The prescription and administration of medication continues to be the most common intervention in the NSW public health system. During the reporting period January – June 2017, medication incidents were the second most commonly reported clinical incident. It is important to note that 99.9% of these incidents resulted in either little or no harm to patients. The information notified in the Incident Information Management System (IIMS) provides details of the type and classification of medication incidents and the medications involved.

The Severity of Medication Incidents

During the reporting period January – June 2017, less than half of one per cent (0.38%) of medication incidents notified were rated as SAC1 or SAC2 incidents. Sixty-four per cent of all medication incidents received the lowest severity rating (SAC4), a further thirty-two per cent were identified as SAC3, and the remaining three per cent represented incidents with no SAC score applied (Table 20). This finding is consistent with the previous reporting period.

Table 20: Medication incidents by SAC rating, January 2013 – June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, SAC2-4 data obtained from IIMS
** Includes St Vincent's Health Network data  for SAC2-4 incidents from January - December 2013

Figure 16 and 17: Medication incidents by SAC1 & SAC2 and SAC3 & SAC4 rating, January 2013 – June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, SAC2-4 data obtained from IIMS
** Includes St Vincent's Health Network data  for SAC2-4 incidents from January - December 2013

Table 21: Medication incidents by SAC rating per 1,000 acute care bed days, January 2013 – June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, SAC2-4 data obtained from IIMS
** Includes St Vincent's Health Network data  for SAC2-4 incidents from January - December 2013


Type of Medication Incident

During the reporting period January – June 2017, there were 14,151 medication-related notifications (Table 20). The IIMS captures where in the medication management cycle the incident occurred. Medication administration continues to be the most frequently reported medication process involving an incident, followed by prescribing, which is consistent with international findings.

The CEC Medication Safety and Quality Program aims to reduce administration and prescribing incidents by assisting health care teams to improve their local medicine-use systems. Table 22 outlines the medication incidents by type and reporting period. Forty per cent of medication related incidents include medications being administered incorrectly, for example an oral medication administered instead of intravenous, the wrong amount given, or at the wrong time. Twenty per cent of notified incidents were prescribing errors which includes illegible writing, incomplete prescriptions or errors made on the prescription (Table 22).

Table 22: Medication incidents by type, January 2013 – June 2017

Clinical Incident Management

em>* Multiple responses are allowed in the 'medication problem type' field and is not mandatory for completion
**Includes St Vincent's Health Network data January - December 2013
*** A dispensing incident is classified to include those medication incidents that relate to the pharmacy dispensing medication. An administration incident is classified to include those medication incidents that relate to the provision of the medication to the patient. However it is accepted that some staff may be inadvertently using the classifications incorrectly by including administration errors in the dispensing classification.


Type of Medications Involved

The most frequent medications involved in incidents during the reporting period January – June 2017 included opioids (such as oxycodone, morphine, fentanyl, hydromorphone and methadone) paracetamol, insulin and anticoagulant medicines (such as enoxaparin, heparin and warfarin) (Table 23).

Opioids, insulin and anticoagulants are recognised as high-risk medicines, i.e. those that have a high risk of causing injury or harm if they are misused or used in error. The CEC’s High-Risk Medicines Program aims to heighten awareness of the potential harm that could be caused by these medicines and provides action-oriented information to assist clinicians in improving their management of high-risk medicines.


Table 23: Top 15 Medications involved in Clinical Incidents, January 2013 – June 2017

Clinical Incident Management
* Multiple responses are allowed in the 'medication involved' field and is not mandatory for completion
**Includes St Vincent's Health Network data January - December 2013
*** Includes Oxycodone and Oxycodone Hydrochloride
**** Includes Morphine and Morphine Sulfate
n/a - Data not available

Time of Medication Incident


During the reporting period January – June 2017, the peak time for medication related incidents continued to be between 08:00-08:59hours, when fifteen per cent of incidents occurred, and between 20:00-20:59hours when a further nine per cent of incidents occurred. The most frequent time of medication related incidents is consistent with previous reporting periods (Figure 18) and corresponds with the literature (Roughead and Semple 2009).


Figure 18: Time of Medication incident as a percentage, January 2013 – June 2017

Clinical Incident Management
* 'Time of Incident' field is not mandatory for completion

The Outcome of Medication Incidents


During the reporting period January – June 2017, fifty-one per cent of medication related incidents did not result in patient harm (SAC4). In twenty-four per cent of incidents, the outcome was unknown or unspecified. Less than six per cent of incidents resulted in pathophysiological outcome / disease related factors (for example asthma, reddened area, confusion or disorientation), procedural complication or patient injury (Table 24).


Table 24: Outcome of Medication Incidents, January 2013 – June 2017

Clinical Incident Management
* Multiple responses are allowed in the 'outcome' field and is not mandatory for completion. This dataset is generic in nature as it is consistent across all clinical incidents and is not specific to medication related incidents

Caution is advised if using the Incident Information Management System (IIMS) reporting counts or rates as the single source of benchmarking data for a project or program, as many variables influence incident reporting. Lower rates of reporting are not a reliable indicator of safer care. Further qualitative, rather than quantitative, interpretation of the data is therefore recommended.


Paediatric Quality Care in NSW


There are almost 1.7 million children and young people aged 0-17 years who live in NSW, comprising 22 per cent of the state’s population. During 2013/14 there were 2.5million visits made to NSW public hospital emergency departments. Children and young people accounted for 25 per cent of these visits (Healthcare performance across the lifespan, Volume 2, Bureau of Health Information). Approximately 1 in 10 clinical incidents reported in the Incident Information Management System (IIMS) involved children and young people under 20 years of age.

The Paediatric Patient Safety Program works across a range of areas to improve the quality and safety of health care for children and young people in NSW. The program is the centralised point of information for paediatric quality and safety for clinicians in the NSW public health system.

The program works in collaboration with other programs including Sepsis Kills, Between the Flags, Medication Safety, Pressure Injury Prevention, End of Life Care and Falls Prevention to improve the quality and safety of health care for children and young people in NSW.

During the reporting period January – June 2017, two per cent of all paediatric incidents notified were rated as SAC1 or SAC2. Fifty-two per cent of all paediatric incidents received the lowest severity rating (SAC4), a further forty-two percent were allocated a SAC3 and the remaining four per cent represented incidents with no allocated SAC score. Table 25 outlines the final SAC allocation for all incidents involving patients aged between 0 and 16 years of age.

The SAC1 data includes the age group of 0-28 days. This incorporates all babies admitted to neonatal units and special care nurseries post-delivery and is not included in paediatric service provision. Included in the paediatric IIMS data set (SAC1-4) are those neonates in Neonatal Intensive Care Units (NICUs), neonatal high dependency and birthing units.  The ‘0 to 28 days’ age band have a larger number of incidents due to the acuity and complexity of this group of patients. A significant number of babies within this age group are under the care of maternity and neonatal services.


Table 25: Clinical incidents 0-16 years by SAC rating, January 2013 – June 2017

Clinical Incident Management
* SAC1 data obtained from CEC RIB database, and includes notifications for all patients aged between 0 and 16 years of age, and excludes stillbirths. SAC2-4 data obtained from IIMS where the ‘age band’ field identified the age of the patient between 0 and 16 years. The ‘age band’ field is not mandatory for completion.
Note: Included in the paediatric data set (SAC1-4) are those neonates in NICUs, neonatal high dependency and birthing units, which sits outside the scope of the Paediatric Patient Safety Program.

Figure 19 and 20: Clinical incidents 0-16 years by SAC1 & SAC2 and SAC3 & SAC4, January 2013 – June 2017

Clinical Incident Management* 2016 SAC1 data obtained from CEC RIB database, and includes notifications for all patients aged between 0 and 16 years of age, and excludes stillbirths.
SAC 2-4 data obtained from IIMS

Incidents by age and principal incident type


The '0 to 28 days' age band consistently reports the largest number of incidents. This can be attributed to the inclusion of maternal and birth-related incidents often being reported as being within the '0 to 28 days' age band (Figure 21).


Figure 21: Clinical incidents 0-16 years by age group January 2013 – June 2017

Clinical Incident ManagementNote: Included in the paediatric data set (SAC1-4) are those neonates in NICU’s, neonatal high dependency and birthing units, which sits outside the scope of the Paediatric Patient Safety Program.


Clinical Management continues to be the most prevalent Principal Incident Type (PIT) amongst paediatric related incidents and includes incidents associated with diagnosis and treatment. The Paediatric Safety & Quality Network has developed resources to assist clinicians in early recognition of the deteriorating patient. These communication tools will support consistency in the communication of clinical concern. Medication/IV fluid continues to be the second most prevalent PIT (Figure 22) and includes incidents associated with the administration, prescribing, dispensing, drug count discrepancies, delivery issues, and storage of medication/IV fluids. The Paediatric Patient Safety Program has recently developed the Safe Gentamicin Prescribing in Paediatrics resource which provides drug dosing and therapeutic drug monitoring information to assist with the safe prescribing of gentamicin in the paediatric patient population.  Accompanying this resource is two short medication safety animations to improve medication safety practices across the state.

Figure 22: Clinical incidents 0-16 years by top four principal incident types (PIT), January 2013 – June 2017

Clinical Incident Management

At a state level, the CEC has established the Children and Young Person RCA Review Committee to provide specialty focused classification of RCA reports involving paediatric patients. In the period July - December 2016 there were 10 RCAs reviewed by the Children and Young Person RCA Review Committee and in the period January - June 2017 there were 13 RCAs reviewed.


Further information regarding the outcomes of this committee can be viewed on the RCA tab of this report.


Paediatric Watch Newsletters have been developed to share the learnings from incident notifications with clinicians across the state. These newsletters can be located on the CEC website.


National Sentinel Event (NSE)


In 2002, Australian states and territories agreed to ongoing contribution to a set of eight core Sentinel Events which have been reported nationally since 2004. Sentinel Events are adverse events that result in death or serious harm to a patient and are considered to be preventable. Public reporting against these Events was considered to be an opportunity for jurisdictions to share learnings about these Events, and to reduce the risk of their recurrence. It is important to note that these Sentinel Events occur infrequently, and, are often due to system and process deficiencies in our healthcare system, and result in the death of, or serious harm to the patient.


The eight agreed Sentinel Events are:


  1. 1. Procedures involving the wrong patient or body part resulting in death or major permanent loss of function
  2. 2. Suicide of a patient in an inpatient unit
  3. 3. Retained instruments or other material after surgery requiring re-operation or further surgical procedure
  4. 4. Intravascular gas embolism resulting in death or neurological damage
  5. 5. Haemolytic blood transfusion reaction resulting from ABO incompatibility
  6. 6. Medication error leading to the death of a patient reasonably believed to be due to incorrect administration of drugs
  7. 7. Maternal death associated with pregnancy, birth and the puerperium
  8. 8. Infant discharged to the wrong family.

It is important to note that Australian jurisdictions interpret these sentinel events differently. NSW has adopted a broad interpretation of these events and therefore caution is required in interpreting this data.

Commencing in 2007, sentinel events have been reported by each Australian jurisdiction for inclusion in the Productivity Commission’s Report on Government Services (ROGS). The ROGs report provides information on the effectiveness and efficiency of government services in Australia, and contains data which are published annually on the equity, efficient and cost effectiveness of government services. NSW sentinel event data is displayed at Table 26 and Australian sentinel event data is displayed in Table 27.

Suicide of a patient in an inpatient unit

There has been a decrease of twenty per cent in inpatient suicides from 18 in 2013/14 to 15 in 2014/15. It is important to note that the definition of events in this category are aligned with the Australian Commission for Safety and Quality in Healthcare (ACSQHC) definition which includes patients on approved leave and those who have absconded from care.

Retained material

There has been an increase by ten per cent in retained instruments or other material requiring re-operation or further surgical procedure. Note, eight out of 20 of these events were reported to have occurred in private health facilities.

Medication Safety

Since 2013/14 NSW public hospitals have responded to the need to drive improvements in medication safety with support from the Clinical Excellence Commission by establishing and implementing dedicated improvement programs focussed on high risk medicines and anticoagulants.  Some of the strategies include:

  • Released Newer Oral Anticoagulants (update) Safety Notice, March 2014
  • Released Ten-fold Medication Dosing Errors Patient Safety Watch, April 2014
  • Established the CEC Anticoagulant Working Party, June 2014
  • Updated the High-Risk Medicines Management Policy PD2015_029, August 2015
  • Safety Notices on newer oral anticoagulants and ten-fold medication dosing errors have been released
  • Additionally, a CEC anticoagulant working party has been established, and updates and re-release of NSW Health High-Risk Medicines Management Policy PD2015_029 includes the management of anticoagulants

For the 2014-15 period, there has been a decrease (76%) of medication errors resulting in death due to incorrect administration of drugs, which may be attributed to NSW public hospitals’ uptake of the improvement strategies and resources.

Maternal deaths

There has been an increase in the number of maternal deaths associated with pregnancy, birth or the puerperium in the 2014/15 reporting period. However, the definition was altered for the 2014/15 reporting period to include both the antenatal and post natal periods, whereas the previous definition included delivery only. Subsequently, the 2014/15 number is not comparable to any results previously published.

Note, during 2016/17, the Australian Commission for Safety and Quality in Health Care (ACSQHC) have been facilitating a review of the existing eight sentinel events in consultation with clinicians and healthcare consumers.  Finalisation of the revised list of sentinel events is expected to occur late 2017.

Table 26: NSW selected Sentinel Events (*)

Clinical Incident Management
* Sentinel Events definitions can vary across jurisdictions.
National Sentinel Event data for FY2015/16 will be provided early 2018.
(b) Data are sourced from the NSW Maternal and Perinatal Mortality Review Committee. Data for all prior years included in the table have been updated to reflect the modified definition, and are not comparable to results published in the 2014 Report and prior versions.
Source: Report on Government Services 2017, Chapter 12, Volume E available at
http://www.pc.gov.au/research/ongoing/report-on-government-services/2017/health/public-hospitals/rogs-2017-volumee-chapter12.pdf

Table 27: Australian Total selected Sentinel Events (*)

Clinical Incident Management
* Sentinel Events definitions can vary across jurisdictions.
The total includes Sentinel Events for the ACT which are not reported in the 8 sub categories of Sentinel Events due to confidentiality issues
National Sentinel Event data for 2015-16 will be provided early 2018.
Source: Report on Government Services 2017, Chapter 12, Volume E available at

http://www.pc.gov.au/research/ongoing/report-on-government-services/2017/health/public-hospitals/rogs-2017-volumee-chapter12.pdf


Complaints and their resolution

The key priority of the NSW public health system is its focus on patient-centred care. Feedback from consumers, their families and carers about their health care experiences is actively encouraged. Complaints received are entered into the Incident Information Management System (IIMS).

Encouraging staff to engage with patients and families during care delivery is known to improve communication, and results in a better experience of care. The CEC’s Partnering with Patients program was established in 2010 to work with local health districts to help include patients and family as care team members, improve consumer engagement and promote safety and quality in health care.

The number of consumer complaints has seen a small decrease of 2.1 per cent over the reporting period of January – June 2017.

When reviewing clinical incident and complaint notifications against occasions of service, the proportion of both clinical incidents and complaints has remained consistent over time (Figure 23).

Figure 23: Notification of clinical incidents and complaints by NSW separations, January 2013 – June 2017

Clinical Incident Management

During the reporting period, less than one per cent of complaints notified were classified as a SAC1 or SAC2 incident. The majority, seventy-five per cent, of complaint notifications were categorised as a SAC4 severity rating, a further twenty three per cent were allocated a SAC3 rating. One per cent of complaints received were not assigned a SAC score at the time of data extraction (Table 28).

Table 28: Complaints by SAC rating, January 2013 – June 2017*

Clinical Incident Management

* All data obtained from IIMS and excludes St Vincent's Health Network


Complaint by issue type

The most frequently reported complaint type for January – June 2017 relates to treatment, followed by communication, and then access to a provider, service or hospital bed (Table 29). This is consistent with themes identified in clinical incidents reported by staff.


Table 29: Complaints by issue type*, January 2013 – June 2017

Clinical Incident Management
* Multiple responses are allowed in the 'issue type' field and is not mandatory for completion
The determination of the issue type is made following review of the complaint


Figure 24: Top three complaints by issue type, January 2013 – June 2017

Clinical Incident Management

The nature of complaints


The nature of the complaint identifies from the consumer's perspective, what the care delivery issues and concerns were about. Treatment, communication and access to discharge transport were the three most frequently reported complaint issues for this reporting period. This is consistent with data from previous reporting periods.

Treatment

Where clinical treatment was the primary issue type reported, the nature of these complaints related to inadequate treatment, coordination of treatment, and medication concerns. Table 30 displays that inadequate treatment was more than double that of the next highest nature of complaint in treatment.


Table 30: Nature of complaint about treatment*, January 2013 – June 2017

Clinical Incident Management
* This is not a mandatory field and multiple responses are allowed

Communication

Where communication was the primary issue type reported, the nature of these complaints related to attitude of healthcare staff, inadequate information provided to the patient and/or their carer, and wrong or misleading information provided to the patient and/or their carer (Table 31).

Table 31: Nature of complaint about communication*, January 2013 – June 2017

Clinical Incident Management
* This is not a mandatory field and multiple responses are allowed

Access

Where access was the primary issue type reported, the nature of these complaints related to delay in admission or treatment, followed by discharge or transfer arrangements, waiting lists and resources/service availability. Access complaints highlight consumers’ concerns about demands on the health care system (Table 32).


Table 32: Nature of complaint about access*, January 2013 – June 2017

Clinical Incident Management
* This is not a mandatory field and multiple responses are allowed

Resolution of complaints


The top five most common forms of complaint resolution remains consistent with previous reporting periods and includes; giving an apology, providing an explanation, and providing feedback to the clinician who was involved in the complaint (Table 33).


Table 33: Complaint by resolution type*, January 2013 – June 2017

Clinical Incident Management* Multiple responses are allowed in the 'resolution type' field and is not a mandatory reporting field

Figure 25: Complaints by top five resolution types, January 2013 – June 2017

Clinical Incident Management

How the complaint was received

During the January – June 2017 reporting period, complaints were most commonly communicated directly by telephone (33%), by letter/email (22%) and in person (20%). This has been a constant theme across all previous reporting periods. A small number were received via other entities, including the Health Care Complaints Commission (8%), NSW Minister and Members of Parliament (5%), Local Health Districts (2%).  The remaining 10 per cent comprises complaints received by the NSW Ombudsman, Official visitors, NSW Ministry of Health, Other State Government Department, Commonwealth Government Department and Health Insurance Commission.


Table 34: How the complaint was received*, January 2014 – June 2017

Clinical Incident Management
* Multiple responses are allowed in the 'complaint received' field and is not mandatory for completion

Glossary

Acute bed day data

Acute bed day data has been provided to the CEC from the Health System Information and Performance Reporting Branch of NSW Health. The following exclusions have been applied for the reports:
1) Care type is 0 (Hospital Boarder).
2) Bed types are 25 (Hospital in Home - General), 66 (Delivery Suite), or 67 (Operating Theatre/Recovery).


(reference for bed types can be found in PD2012_054 Appendix 2)

Clinical incident/Incident

Refers to any unplanned event resulting in, or having the potential to result in, harm to a patient.


Complaint - Issue type

The manager’s determination of identified issues after review of the complaint notification. A list of possible types is available within the incident information management system (IIMS) for selection. Multiple responses are allowed. This is not a mandatory reporting field.


Complaint - Nature of complaint

A further analysis of the complaint issue type from the complainants perspective e.g. the issue was communication and the nature was the attitude of staff. A selection list is available within the incident management system for selection. Multiple responses are allowed. This is not a mandatory reporting field.


Human Factors

The field of study concerned with the design of systems or processes to take proper account of the capabilities and limits of the people working within that system.


Incident Information Management System (IIMS)

An online incident reporting and management system developed in Australia for NSW Health. The St Vincent's Health Network uses a different system called Riskman.


Incident management

The cycle of activities required to recognise, report, understand and reduce the risk of unplanned events occurring. In the health system, feedback to the notifier and sharing of learnings are essential components of this cycle.


Near-miss

An unplanned event, that did not result in injury, illness, or damage but had the potential to do so. A break in the chain of events prevented harm, due to either staff recognition and action, or a fortuitous event.


Notification

The initial report within IIMS that an incident or near miss may have occurred. All staff are required to report incidents in IIMS and must complete the mandatory fields within the system. Notifications can be anonymous and reflect the information known by the reporter at the time.


Patient Safety Watch

Is a series of focussed summary reports based predominantly on incidents which have been subjected to root cause analysis or other investigative methodologies. The aim is to provide the lessons learned back to local health districts and specialty networks, highlighting key risks and recommending preventative actions for local implementation.


Perinatal

The period shortly before, during and after, the birth of a baby.


Principal Incident Type (PIT)

The classification system within IIMS which assists the incident notifier to describe the incident. This term is often abbreviated to PIT.


Retained accountable items

Unintended material (such as a swab) which may require patients to undergo a further procedure to remove the retained item.


Reportable incident brief (RIB)

A document used to notify NSW Health of a reportable incident. RIBs are subject to statutory privilege under section 23 of the Health Administration Act. For more information refer to the NSW Health Incident Management Policy PD2014_004.


Severity Assessment Code (SAC)

The system by which the severity of an incident is rated and the required response is directed across NSW Health services. More information is contained in the NSW Health Incident Management Policy PD2014_004.

Abbreviations
CEC Clinical Excellence Commission PIT Principal Incident Type
EBM Expressed Breast Milk RCA Root Cause Analysis
eMR Electronic Medical Record RIB Reportable Incident Brief
IIMS Incident Information Management System SACSeverity Assessment Code

Principal Incident Type (PIT) Descriptors

Accident/Occupational Health and Safety

This is used to classify incidents related to accidents, occupational health and safety, or the physical environment and staff incidents. Examples are a needle stick injury; exposure to a hazardous substance; a staff member sustains a burn after spilling a hot drink over their arm; and a wet or slippery floor surface.


Aggression

There are two incident types for reporting aggression experienced during health care. These most commonly relate to instances where a patient's verbal communication and/or behaviour to staff or other patients are perceived to be agitated or aggressive in nature. This may be due to the patient's underlying condition, such as confusion, mental illness or physical discomfort. Staff are asked to report about the "aggressor", as this is often an indication of how well the underlying condition is, or can be managed. They are also asked to report all instances where patients, staff or visitors are "victims" of such behaviours. Incidents about patient or staff assaults are also reported under these incident types.


Aggression - Aggressor

This is used to classify the details of the aggressive incident, in the context of the aggressor. Examples are a patient punching another person or a person making physical or verbal threats.


Aggression - Victim

This is used to classify any harm to the victim of an aggressive episode. Examples are a patient being punched by another individual, a victim of a physical or verbal threat.


Behaviour/Human Performance

This is used to classify the details of behaviour or human performance incidents. Examples are a patient exhibiting self-harming behaviour or absconding from hospital grounds.


Blood/Blood Product

This is used to classify the details of incidents related to blood/blood product transfusion processes, dispensing or quality problems. Examples are a patient suffers an anaphylactic reaction to a blood transfusion, a blood unit is mislabelled, blood is stored at the incorrect temperature or an incorrect blood pack is dispensed from the transfusion service.


Building/Fitting/Fixture/Surround

This is used to classify the details directly related to a building, including fittings within a building, the fixtures attached and the external surrounds. Examples are poorly designed building/room for its intended purpose, leaky plumbing, loose or insecurely fixed wall mounted appliance, cracked or uneven pathways and power failure.


Clinical Management

This is used to classify the details related to the clinical management of a patient. This includes diagnosis, treatment planning and delivery and ensuring the correct identification of each patient and procedure. Examples are unintended injury during a medical/surgical procedure, procedure performed on the wrong body part or side and delay in diagnosis of patient's condition.


Complaint

This is used if a consumer expressed dissatisfaction about health care services. Examples include a complaint about the care provided or the manner in which it is delivered.


Documentation

This is used to classify the details of an incident involving a problem with any written, typed, drawn, stamped or printed text/information and/or any document into which it has been entered. Examples are a patient's medication chart filed into another patient's medical record, a treatment order is ambiguous or difficult to read and incorrectly labelled specimens.


Fall

This is used to classify details related to a fall. Examples are a patient found on the floor is suspected of having fallen or a disorientated patient fell after forgetting to use his walking frame.


Health Care Associated Infection/Infestation

This is used to classify the details of infections or infestation acquired during hospitalisation. Examples are a post-operative wound infection or an infected IV (intra-venous) cannula site.


Medical Device/Equipment/Property

This is used to classify the details directly related to medical devices, equipment or property. Examples are routine maintenance not performed on an autoclave, no diathermy earthing plates available for a theatre procedure or a damaged or faulty patient lifter.


Medication/IV fluid

This is used to classify the details related to medication or intravenous fluid incidents. Examples are medication prescribing errors or incorrect intravenous fluid infusion rates.


Nutrition

This is used to classify the details of nutrition incidents. Examples are a diabetic patient received a non-diabetic meal, the wrong TPN (Total Parenteral Nutrition) formula infused or a patient's nasogastric feed given at 80 ml/hour instead of 40 ml/hour.


Organisation Management/Service

This is used to classify the details of any incident involving the provision of patient, staff and visitor services, or the organisational management of the health care institution. Examples are no hospital beds available, inadequate staff supervision, insufficient staff for workload, inadequate staff facilities or no after-hours kitchen service available.


Oxygen/Gas/Vapour

This is used to classify the details of incidents involving both therapeutic and non-therapeutic use of oxygen and/or other gas. Examples are oxygen administered at 4L/min when it should have been 8L/min, or medical air administered instead of oxygen.


Pathology/Laboratory

This is used to describe issues associated with the collection, transport and processing of specimens.


Pressure Ulcer

This is used to classify details of either new pressure injuries or the worsening of pre-existing pressure injuries which occur during clinical care. An example is when a bed-bound patient develops a pressure injury.


Security

This is used to classify the details of incidents directly related to the security of the organisation. Examples are theft of personal property and bomb scare.

Clinical Incident Management