The Clinical Governance Unit in an adult acute hospital reviewed the minutes of all the morbidity and mortality meetings from the previous year. The review noted that one-third of patients who died had been seen by the rapid response team once or more before they died and, that of these patients, 40% had a new treatment limitation put in place by the rapid response team. Ideally, treatment limitations for end-of-life care should be put in place by the admitting team, after discussion with the patient and family, to avoid unnecessary and potentially distressing interventions. 

The morbidity and mortality minutes review prompted the Clinical Governance Unit to review their policies in light of the National Consensus Statement: essential elements for safe and high-quality end of life care. This statement, and the NSQHS Standards, require health service organisations to measure and improve the quality of end-of-life care. The Clinical Governance Unit asked the Safety and Quality manager to lead this work. 

The Safety and Quality manager assembled a team with medical, nursing, palliative care, consumer, and rapid response team representatives. The team decided to use the End of life Care Audit Toolkit, because it allows comparison with similar services and the rest of the hospital network was already using it. The End-of-Life Care Audit Toolkit includes:

• Audit tool for data collection from patient records
• Clinician survey tool 
• Hospital demographic information survey tool for benchmarking
• Data dictionary
• Analysis advice with a plan and table templates
• Presentation template.

The team audited the records for 200 randomly selected patients who died on an inpatient ward more than 48 hours after admission. Patients on the mental health ward or maternity ward were excluded.

The results showed:

• 20% of patients had an advance care plan documenting their goals of care
• 53% of patients were seen by the rapid response team in the last 48 hours of life (most more than once) and 90% had interventions like blood tests performed during this time
• 40% of patients were referred to specialist palliative care services, but 32% were referred within 24 hours prior to death. 

All clinicians from the relevant wards were asked to complete the audit survey. Responses from 109 clinicians found that 68% had not had any training about end of life conversations, and 80% of clinicians thought they were able to identify patients nearing end-of-life.

The data highlighted minimal use of advance care planning, late recognition of dying and that triggers to recognise patients who may be at the end of life were not always acted on, resulting in unnecessary, and potentially distressing, interventions for dying patients. Clinicians’ overestimated their ability to identify dying patients.

When compared to similar wards in the network, results were not significantly different. However, the team identified these areas for improvement because they varied significantly from the guiding principals of the National Consensus Statement: essential elements for safe and high-quality end-of-life care.

The team thought likely reasons for the poor results included:

• Inconsistent recording and storage of advance care plans across services in the facility
• Inconsistent documentation of goals of care on comprehensive care plan
• Insufficient clinician training
• Lack of a specific trigger tool for identifying end of life.

The team met again with several departments to discuss potential improvements, resulting in the following changes being introduced over the subsequent three months:

• Use of disease-specific assessment tools for all admitted patients in three wards, to help identify end of life
• Asking all patients “what is important to you?” as part of admission processes to support goal setting and care planning 
• Promotion of online training for clinicians (End-of-life essentials) by the clinical governance unit and palliative care service
• An audit of comprehensive care plans to ensure they were being developed using a multidisciplinary process.

The team conducted a follow up end-of-life care audit six months later, finding:

• 42% of patients admitted to the facility had an advance care plan (increased from 20%)
• 24% of patients were seen by the rapid response team in the last 48 hours of life (most more than once) and 72% had interventions like blood tests performed during this time (down from 53% and 90%, respectively)
• 73% of clinical staff had completed selected online training modules
• The specialist palliative care service saw an increase in telephone consultation although there was not a significant increase in referral or transfer of care. 

The team considered the successes of the program and looked at measures to refine and share the learning across the network to encourage further improvement. They conducted regular audits of end of life care every six months and provided the results to the Clinical Governance Unit. 

The team introduced further initiatives to address needs identified in the regular audits. Assessing the quality of end-of-life care was also introduced as a regular item in mortality review meetings in the network to maintain the focus and sustain the improvements.

Senior clinicians within the hospital identified the issue of long waiting times for colonoscopy for patients who have had a positive faecal occult blood test (FOBT) as one of the top priorities due to both high risk and high volume: 

• Almost 5000 colonoscopies are performed in the health service each year.
• Gastroenterology and surgical clinical staff raised concerns about delays to colonoscopy services following the introduction of the National Bowel Cancer Screening Program.  
• Root cause analysis of a patient death identified the long waiting time between positive FOBT and colonoscopy as an important contributing factor.
• Feedback from colorectal cancer patients invited by Cancer Services to talk about their experiences raised the problem of delays between referral and treatment.
• A group medical student project on waiting lists for surveillance colonoscopy, presented at a surgical and medical meeting, found variation in waiting times.

A clinical nurse consultant (project manager), two gastroenterologists, and two surgeons worked together to determine what the target waiting time should be, and found the following:

a. The National Bowel Cancer Screening Program monitoring reports include the national median time between a positive screening test and diagnostic assessment, which was 54 days. No relevant clinical quality registries or audits were identified.

b. Two recommendations were identified for waiting time between a positive faecal occult blood test (FOBT) and colonoscopy: 

• The National Bowel Cancer Screening Program performance framework recommends a maximum of 30 days. 
• Cancer Council Australia guidelines state that colonoscopy should be performed as promptly as possible after a positive FOBT to minimise psychological harm, although there is no evidence that prognosis is worsened within 120 days if cancer is present. 

The project team decided to aim for a maximum of 30 days between referral and colonoscopy.

The team decided to regularly monitor waiting times for colonoscopy, and compare it with the national median, the benchmark waiting period.

The project manager reviewed records for all patients referred for colonoscopy due to a positive faecal occult blood test over a one-month period. The median waiting time was 82 days, and the range was 28 – 435 days. The project manager presented the findings to medical, nursing and administrative staff at a department meeting. 

The long median waiting time showed substantial variation from the chosen benchmark of a maximum of 30 days, and the wide range showed large variation within the service. The team decided the risk of harm to patients was high, and warranted immediate action.

The project team identified the key relevant staff, including the endoscopists, cancer services, facility managers, bookings administrators, and information technology experts. The team met with these groups to discuss the causes of the long waiting times, as well as challenges and potential improvements. The discussions identified two major reasons for delays:

• Waiting for appointments at the gastroenterology outpatient clinic for screening for suitability for colonoscopy, which could be done by phone.
• Referrals to individual clinicians for colonoscopy rather than for the next available list vacancy in the service.

Surveillance colonoscopies accounted for a large proportion of procedures, and after review of records, the team found that many patients underwent surveillance more frequently than guidelines recommend.

The team also noticed a high proportion of patients were having both gastroscopy and colonoscopy. They accessed gastroscopy data for their local hospital network area using the interactive Australian Atlas of Healthcare Variation which showed a markedly higher rate than the national average. They also reviewed the relative rates of upper and lower gastrointestinal cancer in Australia (Figure). A review of the outcomes of gastroscopies performed in the hospital on patients without indications showed no serious pathology was found over one year.

Redesigning the referral and booking system was agreed as the best way to reduce delays. The health service executive allocated a position to set up a direct access colonoscopy model.  

A process for rapid access with phone screening was developed, with suitable patients proceeding to colonoscopy and bypassing outpatient clinics. The allocation to colonoscopy lists was based on patient factors, colonoscopy wait list size and time. 

A HealthPathway for colorectal and positive FOBT referrals was developed, including a standardised referral form and a central referral point for all patients. Education sessions were delivered to referring GPs on FOBT screening and the new referral system.

The team also put in place referral criteria for gastroscopy to reduce unnecessary procedures and to free up theatre time for colonoscopies for people with positive FOBTs. They also put in place systems to alert clinicians to the recommended follow up intervals for surveillance colonoscopy.

A review of the six-month pilot of the rapid access colonoscopy model showed that the median waiting time was reduced by 38 days (46%) compared with the previous data, and the range narrowed to 11-188 days. The service is continuing to monitor waiting time each quarter to check whether the improvements are sustained, and reports the results to the hospital board.

The service is continuing the quality improvement cycle by trialling further system changes, with the aim of achieving a maximum 30-day waiting time. The changes and corresponding data are reported back to clinicians and the board regularly.