Article Text
Abstract
Problem The need to develop a patient pathway for emergency admissions who have a previously undiagnosed cancer.
Design The existing patient pathway was audited and process-mapped to identify delays and areas for improvement. Discussions with key stakeholders were held to identify their needs from an acute oncology service.
Strategies for change A new patient pathway was developed, and a new online referral process was implemented. The publicity and education campaign was repeatedly aimed at referring physicians at consultant and junior level, and took the form of emails and presentations with handouts at all teaching sessions, multidisciplinary team meetings, the Consultants Committee and Junior Doctors' induction.
Effects of change The new system was piloted for 6 months. 12/18 patients were referred via the new pathway. 15/18 patients were referred via the new online system. Length of stay, endoscopies, biopsies and blood tests were all statistically significantly reduced during the study period compared with the original audit.
Lessons learnt The challenge was to convince the referring general physicians to use the new patient pathway and referral method. Incorporating their ideas for improvement and implementation made it more likely that they would take up the new ideas. Education and publicity were also extensive, often repeated, and at both junior and senior level. Having the Lead Cancer Clinician and Acute Medicine Consultant engage directly with consultant colleagues, as well as strong support from the Medical Director, was also crucial to the project's success.
- Acute oncology
- length of stay
- unknown primary
- healthcare quality improvement
- patient outcomes
Statistics from Altmetric.com
Background
The number of patients being diagnosed as having cancer in the UK continues to rise, with over 293 000 new cases of cancer diagnosed each year. Between 1977 and 2006, incidence rates for cancer in Great Britain increased by 25%; more than one in three people will develop some form of cancer during their lifetime. Professor Sir Mike Richards, the National Cancer Director, has set a target of reducing the length of stay for oncology patients by ‘1 million bed days.’1 One of the proposed mechanisms for achieving this target is to ensure that all oncology patients are on a defined patient pathway for their admission.
The National Chemotherapy Advisory Group report2 has recommended that all hospitals with an emergency department should establish an acute oncology service: ‘Acute oncology encompasses both the management of patients who develop severe complications following chemotherapy or as a consequence of their previously diagnosed cancer, as well as the management of patients who present as emergencies with previously undiagnosed cancer.’ This last statement poses a significant challenge as oncologists are usually only involved once a diagnosis of cancer is established. Patients who are diagnosed as having cancer having been admitted via the emergency department represent approximately 10% of all cancer diagnoses in our unit. They tend to present late, when very symptomatic. Such patients lack a standardised, well-understood patient pathway.
International efforts to redesign health systems have placed strong emphasis on standardising patient care in order to improve quality and patient safety.3 4 Care pathways are a commonly used technique by which this is achieved in a patient-centred manner, and have been shown to improve outcomes in patients with cancer.5 6 However, in the UK, cancer patient pathways to date have concentrated on the outpatient pathway for diagnosis and treatment.7–9 The inpatient cancer pathway has not been addressed until now.
This work took place at the Whittington Hospital NHS Trust in London, a teaching hospital set in a local community and serving a population of 300 000. The Whittington is a cancer unit with one full-time Consultant in Medical Oncology with gastro-intestinal and lung-cancer interests, one full-time specialist-grade doctor and a separate Palliative Care Team. Different external oncologists attend weekly breast, urological, gynaecology and skin-cancer clinics or multidisciplinary team meetings. There were 583 new oncology diagnoses in 2008, of whom the majority were referred via outpatient pathways: 9.9% of patients were diagnosed following emergency admission.
Outline of problem
There is no formal care pathway for those individuals who are diagnosed as having cancer during a hospital admission. These patients have long hospital stays during which they are subject to multiple investigations and blood tests in an effort to ‘hunt the primary,’ often overlooking whether the patient is fit enough to undergo subsequent treatment, or whether the patients would accept any treatment offered. For those patients who are not fit for active treatment or who decline active interventions, quality-of-life considerations should be paramount.10 Patients should not be subject to investigations, many of which are invasive or have risks associated with them, unless the results of these investigations will alter their management.11–13 Length of stay should be kept to a minimum.
Key measures for improvement:
reduced investigations;
reduced blood tests;
reduced length of stay.
Assessment of problem and analysis of its causes (summary box 1):
The ACCESS database of all new cancer diagnoses was used to identify all patients at the Whittington Hospital NHS Trust who had been diagnosed as having cancer following an admission via the emergency department in 2008 (1 January to 31 December inclusive). A retrospective audit of clinical notes was conducted to assess the patient pathway and identify areas for improvement.
Assessment of problem and analysis of its causes
Problem
Lack of care pathway for inpatients diagnosed as having cancer
These patients undergo invasive tests to confirm their cancer diagnosis (which prolongs their length of stay), despite not being fit for active oncological treatment
Assessment of problem and analysis of causes
Retrospective audit of all inpatient cancer diagnoses in 2008
Patient pathway mapping
Interviews with key stakeholders
Analysis and interpretation
29/34 patients were treated palliatively
31/34 patients had histologically confirmed cancer
Delays in pathway (prior to oncology referral):
1. Waiting for biopsy to be performed
2. Waiting for specimen to be processed
3. Waiting for weekly multidisciplinary team meetings
Referral to oncology only after histological diagnosis of cancer confirmed
Strategy for change
New care pathway: suspected cancers to be referred to Oncology Team when radiology report was suggestive of malignancy
New online referral process, piloted by end users
Engagement, education and publicity campaign
Effects of change—measurement of improvement
Prospective data collection/reaudit
Compared with initial audit, reduction in biopsies, endoscopies, blood tests, length of stay and cost of admission
The initial audit identified 58 patients who were diagnosed as having cancer during an admission in 2008. Of these, 20 had presented surgically and 38 medically. Of the medical group, data were available for 34. This group had a median length of stay of 19 days (range 4–97) and underwent a median of three investigations (defined as all endoscopies, biopsies and imaging excluding x-rays (range 1–9)) and 42 blood tests (range 6–215). These patients had a poor prognosis, with 29/34 receiving palliative treatment only; 31/34 patients had a histologically confirmed diagnosis of cancer, with 23 patients undergoing an endoscopic procedure to obtain a tissue diagnosis and eight patients having a radiologically guided biopsy.
Interviews were also held with key stakeholders, including consultants and junior doctors in the emergency department, Acute and General Medicine and Radiology to ascertain what they wanted from an acute oncology service and how they felt the current situation could be improved. Several key points emerged from interviews with other stakeholders:
Referrals for oncological advice were made via multidisciplinary team meetings (MDT). This is the formal meeting where all new cancer diagnoses are discussed, which incurred delays, as an MDT is held only once a week. With several Oncologists visiting on different days of the week, it was easier for junior doctors to refer the patient to the MDT than the relevant Oncologist. Physicians wanted Oncology input to be more easily accessible.
Referrals to Oncology were usually made following histological evidence of malignancy (31/34), which incurred delays while biopsy specimens were being processed. The median delay between histological confirmation of cancer and referral to Oncology (via the MDT) was 1.6 days. In contrast, the median delay between radiological evidence of malignancy and referral to Oncology was 9 days.
Discharge planning was generally initiated after the MDT decision. Physicians were often reluctant to discharge patients without an MDT agreed management plan, even when the patient was medically fit for discharge.
Junior doctors wanted a more robust, efficient referral method which was not reliant on paging oncologists who were often unavailable/off site. Their preferred method of referring patients was via the existing online system for requesting blood tests and radiological investigations, which was already being successfully used for cardiology referrals.
Analysis and interpretation
The initial audit data demonstrated that patients who are admitted as a medical emergency and subsequently diagnosed as having cancer have a poor prognosis. As such, the emphasis of management should be on maintaining their quality of life and timely transfer to their preferred place of care.14 Process mapping of the patient pathway revealed that the main delays were: (1) waiting for biopsies to be performed, (2) waiting for histology reports and (3) waiting for an MDT decision (figure 1A). Histological evidence of malignancy is an essential prerequisite before chemotherapy or radiotherapy can be safely offered. However, when performance status is poor, or comorbidities exclude the safe giving of systemic chemotherapy, pursuit of a tissue diagnosis is superfluous to clinical need. The average delay between radiological evidence of cancer and referral to Oncology was 9 days. We proposed that Oncologists could play a role at the point at which cancer is suspected radiologically in assessing whether a patient is fit for oncological intervention—if not, then the benefit of further investigations and pursuing a histological diagnosis should be questioned. Cancelling unnecessary biopsies would not only reduce the length of stay, but also free up resources for more appropriate patients. In addition, better access to oncology advice outside the MDT was required, using a more reliable referral method than the current pager system.
Strategy for change
A new referral pathway for all inpatients with radiological signs of suspected cancer: referral for oncological review to be made as soon as the radiology report was suggestive of cancer (figure 1B)
New online referral procedure. This was overseen by the Oncology Consultant/Lead Cancer Clinician and Darzi Fellow, and facilitated by the Information Management and Technology department, who had set up a similar process for Cardiology referrals. The system was piloted by junior doctors before it went live. A rota was set up so that referrals were checked at least twice daily by a member of the Acute Oncology team.
Engagement, education and publicity: the new pathway and referral process represented a change in practice on behalf of the referring clinicians, and engaging with them was crucial. As two-thirds of all emergency medical admissions are initially admitted to the Medical Admissions Unit, having an Acute Medicine Consultant as a joint lead on the project was vital in embedding changes to practice. Education and publicity were also central to engaging with the physicians: there were presentations with handouts at all MDTs, junior doctor teaching sessions, Grand Round and the Consultants Committee meetings, followed up with repeated publicity and reminder emails. In liaison with the Royal College tutors and Director for Medical Education, the new referral pathway was also included in the Junior Doctors' handbook and induction process.
Effects of change: measurement of improvement
Following a publicity and education campaign, the new pathway was launched. A prospective audit of patients diagnosed as having cancer as inpatients was performed for 6 months (June–December 2009). Patients were identified via the online referral system. In addition, all MDT lists were cross-checked to identify new inpatients with cancer diagnoses who had not been referred to the Acute Oncology Team.
There were 20 inpatient cancer diagnoses, of which there were complete data for 18. Of these, 12 were referred to the Acute Oncology Team within 24 h of a radiology report which was suggestive of cancer. Referrals were received from 14 different teams, and 15/18 referrals were received via the new online referral procedure.
All referrals were seen by the Acute Oncology Team within 1 working day. As a result, seven biopsies and four CTs were cancelled in patients who were not fit for oncological intervention; 11/18 (61%) patients had biopsies, compared with 31/34 (91%) in the original audit, a statistically significant reduction (p<0.001 Fisher exact test) (table 1). The number of endoscopies performed was also significantly reduced: 0/18 versus 20/34 (59%) for the original audit (p<0.001, Fisher exact test). The average length of stay was significantly reduced at a median of 10.6 days, compared with 19 days in the original audit (p<0.05 Mann–Whitney U test). This is the equivalent of 200 bed days saved per year. The average number of blood tests was also significantly reduced from 42 to 29 (p<0.05 Mann–Whitney U test). The median average cost per admission decreased from £7027 to £4462 (p=0.051 Mann–Whitney U test). All 18 patients were treated palliatively: one palliative chemotherapy, four palliative radiotherapy, one palliative endocrine therapy and 13 best supportive care.
Case history: reducing length of stay
Audit 1
A 61-year-old male investigated for weight loss, anaemia & dysphagia. CT and endoscopy were suggestive of oesophageal cancer. The patient was fit for discharge, but the team wanted an MDT plan in place first. Due to the MDT falling on a Bank Holiday, the patient was not discharged for a further 13 days.
Reaudit
A 30-year-old male investigated for headaches. CT demonstrated brain and liver metastases from an unknown primary. The patient was seen by the Acute Oncology Team while awaiting a tissue diagnosis and discharged with outpatient follow-up. The formal histology was not available for a further 12 days as extensive immunohistochemistry and a second opinion was sought.
Case history: reducing unnecessary investigations
Audit 1
An 84-year-old female, bedbound with dementia, was investigated for an abdominal mass. She underwent two CTs, MRI and colonoscopy (to obtain tissue), subsequently for palliative care.
Reaudit
An 85-year-old female, bedbound with dementia, was referred to Acute Oncology Team with ultrasound scan demonstrating multiple liver metastases. The patient was assessed and not suitable for active treatment, and so the CT and biopsy were cancelled and the patient referred for palliative care
It should be noted that one limitation of this study is that it does not demonstrate directly that the new pathway and other changes made were responsible for the subsequent reduction in length of stay. It is therefore possible that these results occurred owing to another external factor, although the authors are not aware of any alternative efficiency measures introduced over this time period, or other factors which could otherwise explain these results. There were several documented instances where investigations were cancelled (seven biopsies and four CT scans) on the advice of the Oncology team in patients who were not fit for, or who did not want, subsequent treatment. After the introduction of the new system, the average reduction in length of stay seen was 8.5 days, which is what we would have predicted given the 9-day delay seen in the original pathway.
Lessons learnt
Prior to the new system being introduced, the admitting medical team made key decisions about whether investigations were appropriate for patients with suspected cancer. The patient's case was presented at the relevant MDT after a full work-up and then referred to the visiting Oncology team when histological confirmation of cancer was available. Many Trusts throughout the UK rely on this model of care.
The novel appointment of a full-time medical oncologist to this cancer unit brought many benefits to this project and the overall delivery of care to patients with suspected cancer. This consultant had prior experience of having worked in a cancer centre in addition to being a visiting consultant to a cancer unit. This perspective was valuable in shaping the new service locally. Also holding the role of lead cancer clinician for the Trust was a critical factor in enabling the appointee to authorise the changes needed to develop a new pathway and service.
The performance status of patients who present as unwell through the emergency department, and are found after investigation to have an underlying cancer, is very different from the majority of outpatients who are referred with suspected cancer. These data showed that if a patient is sick enough to be admitted as an emergency, they are often too sick to be fit for any active treatment of their cancer, including palliative chemotherapy. Appropriate management of such patients and their families requires enormous skill and sensitive communication. Being able to explain to a patient why further tests may not be helpful, as the results would not change their overall management, is not an easy conversation. However, we found that this was among the most important learning in developing an Acute Oncology service. Doctors are trained to diagnose and investigate patients, and may be unaware that this can falsely raise expectations regarding treatment options in patients and their families. By involving an experienced oncologist who is able to assess a patients' fitness for all possible treatments (including some of the newer, better-tolerated oral medications) an individualised treatment plan can be formulated first—which may be a direct referral to palliative care, bypassing any superfluous investigations.
Any new service or pathway will ultimately have advantages for some clinical teams, but also some perceived disadvantages for others. Extensive involvement was sought from all clinical teams involved in the emergency care of admitted patients with suspected cancer. The Lead Cancer Clinician first spent time engaging with the referring physicians to ascertain what they wanted from the Acute Oncology Team. The new patient pathway represented a change in practice for referring physicians, who were used to pursuing a tissue diagnosis of cancer. It was therefore important to engage with physicians repeatedly to address their questions or concerns. One of the Consultant Physicians who worked on the Acute Admission Unit was invited to become a member of the Acute Oncology team, as 66% of all medical patients pass through the Medical Admissions Unit.
The new referral system encouraged all clinicians to refer patients when cancer was suspected from a radiological investigation. It was essential therefore to educate the Consultant Radiologists on how to refer suspicious scans directly to the acute oncology team, without duplicating existing pathways for suspected lung, colorectal or breast cancer.
Engaging with the juniors who are responsible for requesting investigations was critical to the dissemination of this new service and its implementation. Piloting the new referral system with a small group of representative juniors and gaining their feedback gave us confidence that the junior staff found it easy to use. Placing Oncology referrals on an existing order system used for the routine requests meant the new system was taken up quickly, as it was just an extension of what was already in use. The Acute Oncology team devised a daily rota so all referrals were seen within 24 h (Monday to Friday). The prompt review of referrals and excellent clinical management of patients instilled faith in the new system. Feedback from the consultant body on the new service was overwhelmingly positive, with more than 80% of those who returned their feedback questionnaire stating they were very satisfied with the service. However, junior medical staff change rapidly at any Trust, so increased awareness of the service through publicity and education had to be repeated to ensure it was effectively embedded into the investigation culture.
Next steps
The most critical part of ensuring a successful referral pathway is introducing a referral mechanism that is easily accessible and well understood for all clinicians. However, there is still a need to integrate diagnostic algorithms for the management of patients with previously undiagnosed cancer who present in predictable ways (eg, hypercalcaemia, brain metastases). This is important to ensure not only an agreed approach, but also to maintain the clinical skills of the admitting physicians and continue the ethos of tailoring investigations to the individual patient. It is hoped that such treating algorithms will be developed by a local operational group encompassing other specialist teams.
Many of the lessons we have learnt from the pilot project are generic and transferable to other healthcare providers. We are keen to disseminate our learning to the wider Cancer Network to allow other providers to learn from our efforts and introduce similar services locally. There is so much common learning that we hope to collaborate with other Acute Oncology Teams nationally to develop standardised protocols, audit tools and patient pathways for oncology patients presenting to the emergency department.
Discussion
Care pathways have been promoted internationally in response to concerns for patient safety, variability in care and the increasing costs of healthcare. Examples include the 4 h target for assessing and managing all emergency patients in the UK NHS and the 100 000 Lives Campaign launched by the Institute for Healthcare Improvement in the USA.15 16 The current economic climate brings a new set of challenges for healthcare providers. Significant efficiency savings are required to meet the increasing demands for health services around the globe, while maintaining standards of care. One route by which this may be achieved is for specialists to review patients prior to MDT discussion to ensure that unnecessary diagnostic tests are not performed in patients who are not fit for subsequent treatment. This will, however, require doctors to reappraise long-held views on their traditional role in the patient pathway.
Patients and their families also have expectations of treatment. The explosion of new technologies and treatments has given both patients and healthcare professionals the impression that all cancers are treatable. Challenging these attitudes and behaviours is not easy, especially when physicians are trained to leave no stone unturned and explore all possible diagnoses and treatments. This, along with many high-profile stories in our popular press regarding new wonder drugs and lack of availability owing to cost, may have a negative impact on the value of an acute oncology service. Denying a patient a biopsy may be viewed as making a decision without really knowing what the cancer is, especially with the advent of targeted treatment for selected groups of patients. In our work, we have learnt that what patients and their families value are honest clear conversations about all possible treatments and their likely benefits. If they have confidence in that opinion, they can make an informed choice. It was not an aim of our pilot project to prospectively collect patient feedback, as our first aim was to see if we could engage with clinicians to use the service. Patients presenting for the first time to hospital via an emergency department who undergo a series of tests for suspected cancer may not have a comparable experience to judge their experience against. We did, however, collect anecdotal feedback about how it felt to be seen in an efficient way and how much they valued being discharged if able to go home once they had a definite follow-up plan (where and when they would next be seen, by whom and a contact number for the interim). Frequently, we found it was clinicians' strongly held views that kept patients in hospital longer than needed, for example: ‘you will get the results more quickly,’ ‘you are more likely to start treatment earlier’ and ‘it is only right you see an oncologist before discharge.’ In addition to managing the expectations of patients and their families, these traditional views and ways of working needed to be addressed when developing an acute oncology service.
Care pathways for patients with cancer have had mixed results regarding clinical outcomes, and not all established pathways are evidence-based.6 17–21 Previous models which have actively involved Oncologists in the leadership and development of new pathways have yielded positive results, in terms of both patient outcomes and financial savings.22 23 Nonetheless, prospective data collection with qualitative and quantitative analysis will be required to demonstrate the real value of Acute Oncology models such as that presented here, in terms of both value for money and value for patients. Outcome measures also need to be tailored to the patient population being studied—patients with cancer treated with curative intent will have different priorities to palliative patients, where quality of life will be paramount. These values must be reflected in the care pathway, while at the same time delivering a cost-effective service.
References
Footnotes
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.