Anticipating death among people with complex needs: a longitudinal mixed methods study comparing nurses' predictions with mortality risk indices
Publication Date
July 19, 2019
Creator
Abstract
Background:
Identifying patients entering their last year of life can facilitate advance care planning and align future care to patients’ preferences. In the United Kingdom, supporting patient choice in relation to end of life care is considered a key responsibility of community matrons, nurses who are based in the community and who act as case managers for predominantly older people with multiple comorbidities and complex needs. Yet recognising dying among patients with multiple and fluctuating long-term conditions is challenging.
Aims:
This mixed methods study aimed to (1) evaluate the accuracy of community matrons’ predictions of mortality among caseloads of people with multiple long-term conditions; (2) compare these with the Revised Minimum Dataset Mortality Rating Index (MMRI-R) and Walter’s Prognostic Index; (3) understand the extent to which the prediction of mortality influences advance care planning; and (4) explore the role and experiences of the nurses in identifying patients approaching the end of their lives.
Methods:
The study sought to recruit all community matrons employed by two large NHS Trusts providing community healthcare services to a population of over a million people living within Derbyshire and Derby City. Questionnaires were administered prospectively to gather demographic, comorbidity, symptom and laboratory information from community matrons for each of the patients they case-managed. Information pertaining to the prevalence of advance care planning was also collected. Community matrons were asked to estimate the likelihood of death in the next 12 months for each patient. Patients were followed up for a minimum of 12 months at which point mortality status was determined. Cox proportional hazard models were developed for each mortality index and nurses’ predictions individually and in combination. Measures of discrimination and calibration were evaluated. Logistic regression was used to determine which patient characteristics were associated with advance care planning.
A subgroup of community matrons was invited to participate in semi-structured interviews. Thematic analysis of their interview transcripts was undertaken to identify key themes using a grounded theory approach. Validation of qualitative findings was sought from the community matrons via a focus group.
Results:
Questionnaires pertaining to 867 patients were included in the study of which 249 patients (29%) died. Walters Prognostic Index demonstrated modest discrimination (Harrell’s c-index 0.66), MMRI-R had good discrimination (Harrell’s c-index 0.70) and nurses’ predictions modest discrimination (Harrell’s c-index 0.68). The addition of nurse’ predictions of death had a significant effect on the performance of Walter’s Prognostic Index (Harrell’s c-index 0.70) but only marginal improvement in the performance of the MMRI-R (Harrell’s c-index 0.72).
Fewer than 40% of the matrons’ patients had evidence of advance care planning. Patients who were older, had a cancer diagnosis, were functionally dependent and whose matron gave a positive response to the ‘surprise’ question were more likely to have a form of advance care plan.
Semi-structured interviews were conducted with 13 community matrons and highlighted the autonomous role of the nurses regarding the initiation of end of life care discussions and planning supportive care. Nurses’ confidence in clinical decision making at the end of life was variable and accountability was readily deferred to the GP. An overarching theme of reciprocal interdependence emerged. This was characterised by the interconnections between the community matron and the general practitioner and the coordination, interactions and complexity involved in caring for a person with multiple comorbidities approaching the end of their life in the community.
Conclusions:
The study found no evidence of a difference between MMRI-R and nurses’ predictions in identifying patients who subsequently died indicating that nurses’ judgement can be trusted to initiate anticipatory care interventions at an early stage whilst avoiding an over-reliance on objective measures. However, few patients had evidence of advance care planning interventions in practice.
The study reinforces the importance of teamwork and a collaborative approach to the provision of end of life care. The interdependent nature of the relationship between community matron and the GP reflects a redistribution of medical work within primary care and the growing contribution of nurses to end of life decision making and advance care planning. Community matrons play an active role in facilitating end of life care for people with multiple comorbidities and improving their confidence in anticipatory care interventions could strengthen their contribution further.
Item Type
ethesis
Thesis Type
PhD
Supervisors
Subjects (LC)
Associated Schools / Departments
School of Health Sciences
eprints ID
56988
UoN Repository URI
Except where otherwise noted, this item's license is described as
File(s)![Thumbnail Image]()
Name
Andy Cole PhD Thesis 2019.pdf
Type
Full-text
Description
Examined
Size
3.18 MB
Format
Adobe PDF
Checksum (MD5)
3a80e4e08f47c181e9b514795cdda7ce