Conducting Goals-of-Care Discussions Takes Less Time Than Imagined

Goals of care documentation by medical oncologists and oncology patient end-of-life care outcomes

BACKGROUND

Goals of care (GOC) documentation is important but underused. We aimed to improve oncologist GOC documentation and end-of-life (EOL) care.

METHODS

In April 2020, our cancer center launched a GOC note template, including optional fields for documenting discussion with the patient about: cancer natural history, goals, and/or EOL (resuscitation preferences, hospice receptivity). Associations between GOC notes and EOL care were evaluated.

RESULTS

Among 1721 patients dying between June 1, 2020 and June 30, 2021, median days from first GOC note (± with documentation of EOL discussion) to death was 92, whereas a GOC note including EOL discussion ("GOC EOL note"), specifically, was 31. Patients with a first GOC note >60 days before death spent fewer days inpatient (6.7 vs 10.6 days, p < .001). Among patients with GOC EOL notes, those with such documentation >30 days before death had fewer inpatient (5 vs 11, p < .001) and intensive care unit days (0.5 vs 1.5, p < .001), more hospice referrals (57% vs 44%, p = .003), and less chemotherapy ≤14 days before death (6% vs 11%, p = .010). Of 925 admissions of patients dying within ≤30 days, those with GOC EOL notes were shorter (7 vs 9 days, p = .013) but not associated with more hospice discharge (30% vs 25%, p = .163). Oncologist (vs nononcologist) GOC documentation and earlier documentation of EOL discussion were associated in subset analyses with less inpatient care and more hospice referrals.

CONCLUSIONS

Documentation of GOC, including EOL discussions, is associated with favorable performance on accepted indicators of quality EOL care.

Interventions for interpersonal communication about end of life care between health practitioners and affected people

BACKGROUND

Communication about end of life (EoL) and EoL care is critically important for providing quality care as people approach death. Such communication is often complex and involves many people (patients, family members, carers, health professionals). How best to communicate with people in the period approaching death is not known, but is an important question for quality of care at EoL worldwide. This review fills a gap in the evidence on interpersonal communication (between people and health professionals) in the last year of life, focusing on interventions to improve interpersonal communication and patient, family member and carer outcomes.

OBJECTIVES

To assess the effects of interventions designed to improve verbal interpersonal communication about EoL care between health practitioners and people affected by EoL.

SEARCH METHODS

We searched CENTRAL, MEDLINE, Embase, PsycINFO, and CINAHL from inception to July 2018, without language or date restrictions. We contacted authors of included studies and experts and searched reference lists to identify relevant papers. We searched grey literature sources, conference proceedings, and clinical trials registries in September 2019. Database searches were re-run in June 2021 and potentially relevant studies listed as awaiting classification or ongoing.

SELECTION CRITERIA

This review assessed the effects of interventions, evaluated in randomised and quasi-randomised trials, intended to enhance interpersonal communication about EoL care between patients expected to die within 12 months, their family members and carers, and health practitioners involved in their care. Patients of any age from birth, in any setting or care context (e.g. acute catastrophic injury, chronic illness), and all health professionals involved in their care were eligible. All communication interventions were eligible, as long as they included interpersonal interaction(s) between patients and family members or carers and health professionals. Interventions could be simple or complex, with one or more communication aims (e.g. to inform, skill, engage, support). Effects were sought on outcomes for patients, family and carers, health professionals and health systems, including adverse (unintended) effects. To ensure this review's focus was maintained on interpersonal communication in the last 12 months of life,  we excluded studies that addressed specific decisions, shared or otherwise, and the tools involved in such decision-making. We also excluded studies focused on advance care planning (ACP) reporting ACP uptake or completion as the primary outcome. Finally, we excluded studies of communication skills training for health professionals unless patient outcomes were reported as primary outcomes.

DATA COLLECTION AND ANALYSIS

Standard Cochrane methods were used, including dual review author study selection, data extraction and quality assessment of the included studies.

MAIN RESULTS

Eight trials were included. All assessed intervention effects compared with usual care. Certainty of the evidence was low or very low. All outcomes were downgraded for indirectness based on the review's purpose, and many were downgraded for imprecision and/or inconsistency. Certainty was not commonly downgraded for methodological limitations. A summary of the review's findings is as follows. Knowledge and understanding (four studies, low-certainty evidence; one study without usable data): interventions to improve communication (e.g. question prompt list, with or without patient and physician training) may have little or no effect on knowledge of illness and prognosis, or information needs and preferences, although studies were small and measures used varied across trials.  Evaluation of the communication (six studies measuring several constructs (communication quality, patient-centredness, involvement preferences, doctor-patient relationship, satisfaction with consultation), most low-certainty evidence): across constructs there may be minimal or no effects of interventions to improve EoL communication, and there is uncertainty about effects of interventions such as a patient-specific feedback sheet on quality of communication.  Discussions of EoL or EoL care (six studies measuring selected outcomes, low- or very low-certainty evidence): a family conference intervention may increase duration of EoL discussions in an intensive care unit (ICU) setting, while use of a structured serious illness conversation guide may lead to earlier discussions of EoL and EoL care (each assessed by one study). We are uncertain about effects on occurrence of discussions and question asking in consultations, and there may be little or no effect on content of communication in consultations.  Adverse outcomes or unintended effects (limited evidence): there is insufficient evidence to determine whether there are adverse outcomes associated with communication interventions  (e.g. question prompt list, family conference, structured discussions) for EoL and EoL care. Patient and/or carer anxiety was reported by three studies, but judged as confounded. No other unintended consequences, or worsening of desired outcomes, were reported. Patient/carer quality of life (four studies, low-certainty evidence; two without useable data): interventions to improve communication may have little or no effect on quality of life.  Health practitioner outcomes (three studies, low-certainty evidence; two without usable data): interventions to improve communication may have little or no effect on health practitioner outcomes (satisfaction with communication during consultation; one study); effects on other outcomes (knowledge, preparedness to communicate) are unknown. Health systems impacts: communication interventions (e.g. structured EoL conversations) may have little or no effect on carer or clinician ratings of quality of EoL care (satisfaction with care, symptom management, comfort assessment, quality of care) (three studies, low-certainty evidence), or on patients' self-rated care and illness, or numbers of care goals met (one study, low-certainty evidence). Communication interventions (e.g. question prompt list alone or with nurse-led communication skills training) may slightly increase mean consultation length (two studies), but other health service impacts (e.g. hospital admissions) are unclear.

AUTHORS' CONCLUSIONS

Findings of this review are inconclusive for practice. Future research might contribute meaningfully by seeking to fill gaps for populations not yet studied in trials; and to develop responsive outcome measures with which to better assess the effects of communication on the range of people involved in EoL communication episodes. Mixed methods and/or qualitative research may contribute usefully to better understand the complex interplay between different parties involved in communication, and to inform development of more effective interventions and appropriate outcome measures. Co-design of such interventions and outcomes, involving the full range of people affected by EoL communication and care, should be a key underpinning principle for future research in this area.

Expected clinical utility of automatable prediction models for improving palliative and end-of-life care outcomes: Toward routine decision analysis before implementation

Objective

The study sought to evaluate the expected clinical utility of automatable prediction models for increasing goals-of-care discussions (GOCDs) among hospitalized patients at the end of life (EOL).

Materials and Methods

We built a decision model from the perspective of clinicians who aim to increase GOCDs at the EOL using an automated alert system. The alternative strategies were 4 prediction models—3 random forest models and the Modified Hospital One-year Mortality Risk model—to generate alerts for patients at a high risk of 1-year mortality. They were trained on admissions from 2011 to 2016 (70 788 patients) and tested with admissions from 2017-2018 (16 490 patients). GOCDs occurring in usual care were measured with code status orders. We calculated the expected risk difference (beneficial outcomes with alerts minus beneficial outcomes without alerts among those at the EOL), the number needed to benefit (number of alerts needed to increase benefit over usual care by 1 outcome), and the net benefit (benefit minus cost) of each strategy.

Results

Models had a C-statistic between 0.79 and 0.86. A code status order occurred during 2599 of 3773 (69%) hospitalizations at the EOL. At a risk threshold corresponding to an alert prevalence of 10%, the expected risk difference ranged from 5.4% to 10.7% and the number needed to benefit ranged from 5.4 to 10.9 alerts. Using revealed preferences, only 2 models improved net benefit over usual care. A random forest model with diagnostic predictors had the highest expected value, including in sensitivity analyses.

Discussion

Prediction models with acceptable predictive validity differed meaningfully in their ability to improve over usual decision making.

Conclusions

An evaluation of clinical utility, such as by using decision curve analysis, is recommended after validating a prediction model because metrics of model predictiveness, such as the C-statistic, are not informative of clinical value.