Problem

Ambition 1 of the End of Life Ambitions Framework 2015–2020 is that each person is seen as an individual.1 Within this, the first building block is honest conversations: ‘Everybody should have the opportunity for honest and well informed conversations about dying, death and bereavement.’ Long-standing problems exist with both Do Not Attempt Cardio-Pulmonary Resuscitation (DNACPR) discussions and documentation. Hospitals suffer repeated complaints, with poor recollection of discussions and poor documentation. The long-term outcomes of bereavement are significant, with increased morbidity, mortality, suicide and depression. DNACPR complaints are part of the picture of severe bereavement reactions.2 Experience of one of the authors (JA) in investigating complaints about DNACPR is one of trauma for those involved, with long-standing issues remaining unresolved for many months and poor understanding of the role of DNACPR among executive level staff, with inappropriate referral of senior doctors to the General Medical Council for investigation. Current DNACPR documentation has the hallmarks of a top-down approach to change with lack of quality improvement methodology in the design and implementation of both forms and guidelines. Documentation tends to be tick box in nature and does not enable a good process of discussion.

In the light of the above problems, we developed a quality improvement project with the main outcome being to reduce DNACPR complaints. Other primary outcome measures were to design a form that enabled good quality conversations, have good feedback from patients, relatives and the doctors who used them and to have an effective handover document in the out-of-hours setting.

Our quality improvement project was undertaken at a small district general hospital in southwest England, where the age of the local population is above the national average. The project was undertaken by four Foundation Year 1 doctors under the supervision of a palliative care consultant.

Previous junior doctors had highlighted problems with inadequate plans for treatment escalation, poor documentation about ceilings of care and unsatisfactory communication with patient and families.

These factors were reflected in the numbers of DNACPR-related complaints and feedback from doctors that they felt ill-equipped in making treatment escalation decisions out of hours with unfamiliar patients.

A lack of a clear structure for resuscitation conversations and sporadic documentation regarding care plans was adversely affecting out-of-hours staff when trying to care for deteriorating patients, without prior knowledge of their wishes or concerns. In turn this resulted in distressing and unnecessary investigations and management for patients.

We set out to develop a user-friendly form to guide and document treatment escalation discussions and plans encompassing all elements of medical care. Through this we hoped to inspire doctors to hold such conversations with patients and their relatives in order to identify and understand their wishes, priorities and concerns. In addition, with evidence to suggest that advance planning helps to reduce inappropriate out-of-hours investigations,2 we hoped that through uniform documentation and filing of such discussions we could improve out-of-hours care of deteriorating patients by equipping doctors with an understanding of their agreed ceiling of care.

Background

DNACPR decisions formed the main basis for treatment escalation decisions. However, these forms do not encompass the breadth of medical treatments available to patients and do not encourage good communication with patients or their families.

Evidence suggests that DNACPR decisions alone have the potential to be misinterpreted and result in the withholding of other medical treatment options which may be appropriate for and desired by patients.3 They also have the potential to be viewed as negative by patients as they focus on treatment being withheld and do not use the opportunity to explore what other treatment modalities would continue to be offered.4 Therefore, the current DNACPR process does not provide a holistic and positive approach to advanced care planning.

Furthermore, evidence currently suggests that communication with patients about DNACPR and treatment escalation decisions is poor and doctors often feel uncomfortable or ill-equipped to hold such discussions.4–8 This is often incongruent with patients’ wishes, many of whom want to be involved with discussions about their ceiling of care and would welcome the opportunity to explore such matters.7 In addition, contrary to the beliefs held by some doctors, evidence suggests that patients rarely find being given information about their ceiling of care upsetting or distressing.9 This highlights a further problem with the current DNACPR process, in that the current design of the form does not encourage or support discussion with patients and families, which they would welcome.

Finally, since the current DNACPR form only documents decisions relating to Cardio-Pulmonary Resuscitation (CPR), where treatment escalation discussions are held with patients or their family, it is often difficult to locate this information in patient notes. Since individualised advance care planning should be encouraged, it is vital that this is documented in a way that is readily accessible and communicated effectively among all healthcare professionals.10 11 This is particularly beneficial out of hours, where doctors may be reviewing patients with whom they may be unfamiliar. Treatment escalation plans (TEPs) have been found to be helpful in prompting discussion with patients and relatives and providing an opportunity to explore their wishes.3 Previous TEP projects have demonstrated that TEP forms improve documentation about treatment escalation decisions and decrease uncertainty about ceilings of care.3

Unsurprisingly, these factors have contributed to TEPs being popular with junior doctors reducing anxiety and improving decision-making about deteriorating patients, especially out of hours.4 11

Baseline measurement

The main emphasis of the baseline measurement of the TEP forms was to monitor the output and aim for areas of further improvement. Once the initial baseline measurement was obtained, the collected data were reviewed and analysed, with a view to developing theories to be tested in plan-do-study-act (PDSA) cycles (see figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Driver diagram. ED, emergency department; MAU, Medical Admission Unit; TEP, treatment escalation plan.

We used DNACPR-related complaints as our primary outcome measure. The previous Foundation Year 1 doctors had collected data and patient feedback, which we compared with our baseline results. DNACPR complaints related  mainly (83%) to poor communication: lack of communication between the patients/relatives regarding the DNACPR, families not being informed of decisions made by the medical team, incorrect information regarding comorbidities being documented and discharge issues such as a copy of the DNACPR not being sent home with the patients or to their general practitioner.

Design

The initial TEP form used a three-step process, which consisted of a medical plan, but no ceiling of care or space for documentation.

Using the PDSA Quality Improvement methodology, the initial TEP forms were modified and moulded into a new two-paged form. The concepts of the forms were discussed with multiple specialties to accommodate the various levels of ceiling of care, that is, medical, surgicaland Intensive Care.

The TEP form consisted of three sections:

1. The initial decision made by the medical team centred on the patient’s clinical condition. This decision was made by the most senior doctor to see the patient and formed the basis of the TEP discussion. This meant that the doctor who conducted the TEP discussion had the most senior available opinion on what was good patient care.

2. Capacity assessment followed by recording of the TEP discussion with the patient and family. The discussion included decisions regarding non-invasive interventions, options concerning levels of appropriate medical and surgical interventions, artificial feeding and cardiopulmonary resuscitation treatment at the end. It was felt that a graded form of discussion was more likely to be conducted in a sequential and sensitive fashion.

3. Page 2 of the form was bordered red and was the final outcome of the discussion, including DNACPR decisions, countersigned by the consultant. A copy of the completed form was given to the patient and family where desired.

Changes were made to the initial form: the date of admission was included, ‘high care’ was removed as an option in the ceiling of care due to ward changes, removal of tick boxes which were poorly read and filled out, and prompts were included into the new forms to help aid the discussions. As mentioned above, we monitored these on a regular basis, making note of the positives and negatives, gaining feedback from colleagues, and installed all this information together and created the final version, which we rolled out into all the wards.

Strategy

We met weekly as a group to discuss the improvements or data we had gathered the previous week to suggest changes and how to measure them. Meeting weekly allowed us to make small, incremental changes in the design of the form and its roll-out across initially the medical wards followed by the surgical wards. We used the methodology of ‘Adopt, Adapt, Abandon’ as the basis for discussing results, with further formulation of theories which were tested in the forthcoming week.

Overall we carried out 11 PDSA improvement cycles over the course of the year August 2015 to August 2016.

These involved collecting numbers of TEP forms, changing the form itself to improve the quality, collecting feedback from our colleagues, patients and relatives, and enlisting the help of junior doctors across all the wards in the hospital to gather support for our project, increasing the sense of ownership.

Examples of PDSA cycles are given in figure 2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Plan-do-study-act (PDSA) cycle highlights. MAU, Medical Admission Unit; TEP, treatment escalation plan.

Our first PDSA cycle on the care of the elderly ward was to count the number of patients with DNACPR and TEP forms. Many patients had DNACPR forms but had not had TEP discussions. We added a column onto the ward handover sheet, allowing us to see quickly which patients had DNACPR but had not had a TEP discussion. The increase in patients who had completed both in a 28-bedded ward was 4/28 to 24/28. This increase in TEP numbers was maintained as a consequence of this change.

Our fourth PDSA cycle added the date of admission, removing the tick boxes on the front of the form relating to advanced care planning decisions that were never filled in, and adding prompts for both the discussion and treatment plan boxes. The latter change we hoped would improve quality of documentation, as well as assisting in communication. We presented our TEP changes at the junior doctor teaching, and in addition went to talk to all doctors on the Medical Admission Unit about the TEP form changes to educate them on the changes. As a group we created a scale on which to judge the quality of the discussion documentation and of the treatment plan documentation. We also measured how much of the form was filled in. The quality of the discussion and treatment plan improved with the new form, so we adopted the changes.

With a new, improved form, our next focus was on how to roll out the TEP form on a large scale to all the wards. We decided on individual ownership as an important principle. To track the progression and uptake of the new forms, we counted the number of patients with DNACPR with/without TEPs across initially all the medical wards. With this in mind we presented our data about the new TEP forms to our colleagues and asked for ‘TEP Champions’ for each ward. With volunteers, we then went round the individual wards, explaining the changes and showing the ‘TEP Champions’ the spreadsheet on the intranet where they could record data they collected.

We continued to monitor the quality of performance of each section of the form and found the quality of discussion had improved, but the treatment plan quality had slightly decreased. We went back to our colleagues to gather informal feedback on whether there were any more prompts that could be helpful.

We finally surveyed the junior doctors, again raising the profile of our project in the hospital and among our colleagues, and gathering important information through white space questions, on barriers to implementation and what we could work on in the future.