Interventions to Improve Prognostic Understanding in Advanced Stages of Life-Limiting Illness: A Systematic Review

Abstract

Context.

Among patients with advanced life-limiting illness, an inaccurate understanding of prognosis is common and associated with negative outcomes. Recent years have seen an emergence of new interventions tested for their potential to improve prognostic understanding. However, this literature has yet to be synthesized.

Objectives.

To identify and characterize tested interventions, summarize intervention findings, and outline directions for future research.

Methods.

Systematic searches were conducted in five databases — PubMed/MEDLINE (NLM), EMBASE (Elsevier), CINAHL (EBSCO), PsycINFO (OVID), and Cochrane Central Register of Controlled Trials (Wiley) — to identify interventions evaluated within a randomized controlled design for their impact on prognostic understanding.

Results.

Of the 2354 initial articles, 12 were selected for final inclusion, representing nine unique interventions. Intervention types included decision aids accompanying medical consults, palliative care consultations, communication training for patients and physicians, and targeted discussions regarding prognosis and treatment decision making. Common components of interventions included provision of prognostic information, assistance with end-of-life care planning, and provision of decisional and emotional support during discussions. Most interventions were associated with some indication of improvement in prognostic understanding. However, even after intervention, inaccurate prognostic understanding was common, with 31−95 percent of patients in intervention groups exhibiting inaccurate perceptions of their prognosis.

Conclusion.

Prognostic understanding interventions hold the potential to improve patient understanding and thus informed decision making, but limitations exist. Future research should examine why many patients receiving intervention may continue to maintain inaccurate perceptions, and identify which intervention components can best enhance informed, value-consistent decision making.

Introduction

An accurate understanding of prognosis is critical for informed treatment decision making and planning at the advanced stages of life-limiting illness.^1–3 Yet, research shows that inaccurate, overly optimistic prognostic understanding is widely prevalent.⁴ Many patients misperceive their palliative treatments as curative, do not consider themselves as having a terminal illness, and overestimate their life expectancy.^5–7 More concerning, such inaccurate expectations of prognosis are associated with negative outcomes such as less advance care planning and higher receipt of value-inconsistent, minimally beneficial care in the final week of life (e.g., chemotherapy, mechanical ventilation).^8–11

Interventions to improve prognostic understanding are therefore critical, offering the potential to improve patient well-being, while reducing minimally beneficial care at the end of life.^2,12,13 Such interventions may facilitate more informed, value-based decision making. In advanced illness, patients may have to weigh in on treatment decisions such as discontinuing further disease-directed treatments or enrolling in hospice. In these decisions, patients must weigh the relative benefits and costs of available care options, which is based on patients’ perceptions of their prognosis. When prognostic understanding is inaccurate, patients’ ability to make informed, value-based decisions regarding their care is impaired.

Fortunately, recent years have seen an emergence of new interventions, tested for their potential to improve prognostic understanding.^14–16 There is a need however to synthesize what has been found and identify gaps in current knowledge and areas for future research. We therefore conducted a systematic review of interventions, evaluated within a randomized controlled design, for their impact on prognostic understanding (PROSPERO registration: CRD42020179203). We included in our search all patient populations with incurable, life-limiting malignant and non-malignant disease (e.g., advanced cancer, heart failure, renal disease, chronic obstructive pulmonary disease) given the documented challenges with accurate prognostic understanding across multiple disease contexts.^17–19

Methods

Literature Search

A systematic literature search was conducted (originally on August 10, 2019 and updated on May 22, 2021) in five electronic databases: PubMed/MEDLINE (NLM), EMBASE (Elsevier), CINAHL (EBSCO), PsycINFO (OVID), and Cochrane Central Register of Controlled Trials (Wiley). A combination of relevant keyword terms and database-specific subject headings were used to design the search strategies for the PubMed/MEDLINE (Medical Subject Headings, MeSH), EMBASE (EMTREE), CINAHL (CINAHL Subject Headings) and PsycINFO (Thesaurus of Psychological Index Terms) databases and a keyword-only search strategy was used in the Cochrane Central Register of Controlled Trials.

The overall search strategy was developed first using PubMed/MEDLINE and then translated to the other databases. The search strategy consisted of three main component concepts that were first searched upon individually (combining relevant synonyms with Boolean operator “OR”) and then combined altogether using the Boolean operator “AND”: 1) Prognostic understanding/awareness, 2) Advanced, life-limiting disease, and 3) Intervention/experimental study design (see Appendix 1 for further details). The final search results were restricted to English language publications, animal studies were excluded, and no date limits were applied.

The final database search results were harvested/managed and the duplicate citations were removed using the Endnote citation management software program (Clarivate Analytics). The Covidence systematic review software (Veritas Health Innovation) was used to carry out the record screening.

Eligibility, Review and Abstraction Process

Eligibility criteria included the following: 1) a study sample of adult (≥18 years of age) patients incurably ill with life-limiting malignant and non-malignant diseases (e.g., advanced cancer, heart failure, chronic obstructive pulmonary disease), or their caregivers, 2) the study utilized a randomized controlled design, and 3) the reference featured evaluation of the impact of an intervention on prognostic understanding. Prognostic understanding was defined as patient or caregiver awareness of prognosis, including, but not limited to, terminal nature of illness, life-expectancy, and likelihood of curability. Studies measuring treatment understanding were deemed eligible to be included if the operationalization was broad enough to reflect overall prognosis with illness (e.g., perceiving palliative chemotherapy as curative).

Article screening and review were conducted through several iterative phases: 1) title screening, 2) abstract screening, 3) full text review, and 4) data extraction and quality assessment. Each article was reviewed by two coders. In the title and abstract screening phases, coders evaluated if articles met eligibility criteria based on the information available in the title and abstract respectively (if sufficient information was not available to make a determination, references were forwarded to the next stage for further evaluation). Articles that were coded as relevant by one or both coders were forwarded to the next stage. In the full text review phase, discrepancy in inclusion-exclusion decision were resolved via discussion. This process resulted in identification of 12 articles for inclusion in the final review. Using a structured data abstraction form, data were extracted from each article by two coders, including information regarding intervention content and structure, sample characteristics, prognostic understanding operationalization, and findings reported.

Study quality was also assessed and considered in inclusion decisions. Quality was assessed using a modified version of the Downs and Black Study Quality Checklist (25 of the most relevant items from the original 27-item checklist used).^20,21 References were to be excluded if they did not meet at least 33% of the modified Downs and Black Quality Checklist indicators.

Results

The literature search identified 2574 references, of which only 177 were selected for full text review, and only 12 identified for final inclusion (see Fig. 1 for PRISMA flowsheet). As some of the final references pertained to the same studies, our final 12 included articles reflected nine unique intervention studies. All final references showed adequate scores on the quality assessment and none were excluded based on study quality (references possessed between 65%–100% of the quality indicators).

Fig. 1.

PRISMA flow diagram.

Study Characteristics

Of the nine included randomized controlled trials, two were pilot studies (Table 1). All studies were conducted among patients with advanced cancer, except for one study conducted among patients with advanced heart failure.²² Five of the nine studies were conducted in the United States, and the others were conducted in Taiwan, Belgium, Saudi Arabia, Australia and Canada. Several studies had multiple operationalizations of prognostic understanding. The most common operationalization was perception of cancer as curable (n = 7); other operationalizations of prognostic understanding included understanding the non-curative nature of palliative chemotherapy, perception of life expectancy, and recognition of the terminal nature of illness.

Table 1.

Characteristics of Included Studies

Study	Sample Size	Sample characteristics	Intervention	Prognostic understanding conceptualization	Effect on prognostic understanding	Effect on decisional & care-received outcomes
Chen et al., 2019, 202023,24 Wen et al., 202025	460	Patients with terminal cancer, and their caregivers, from medical center in Taiwan	Multi-component advanced care planning (ACP) intervention Masters level-nurses met with patients monthly (outpatients) or weekly (hospitalized) until death. Intervention consisted of: -Repeated assessment of patient and caregiver readiness for ACP -Tailoring of interaction based on stage of readiness -Facilitation of prognosis and end-of-life care discussions among patients, caregivers, and physicians. -Use of ACP booklet and video educational aid -Discussing participant concerns and providing comfort Control group received sham treatment focused on symptom management	Perception of curability	Intervention group had higher odds of accurate prognostic understanding earlier in the illness trajectory relative to control group	Intervention group patients with accurate prognostic understanding had lower likelihood of receiving CPR in the last month of life relative to control group participants without accurate prognostic understanding No effect on likelihood of life-sustaining treatments at endof-life, or preferenceconsistency of received life-sustaining treatments26
Leighl et al., 201114	207	Patients with incurable, metastatic colorectal cancer considering first line chemotherapy from Australia and Canada	Decision aid as part of medical consultation (booklet and audio recording) Decision aid included information on: -Goals of palliative treatment with or without chemotherapy -Treatment options including supportive care alone and adding first-line systemic therapy and clinical trial participation. -Treatment administration, benefit, and risks -Seven steps of decision making were reviewed, along with a values clarification exercise (a “weigh scale” of pros and cons of systemic therapy versus supportive care alone) Control group received standard medical oncology consult; intervention group received consultation in which decision aid was used	Illness/treatment understanding	No significant differences in prognostic understanding immediately post consultation. At 1–2 weeks post consultation, prognostic understanding increased in both groups, but more so among those in the intervention group.	No effect on patient decision to undergo chemotherapy, and on satisfaction/conflict regarding decision
Epstein et al., 201716	265 Patients 194 Caregivers 38 Oncologists	Patients with advanced nonhematologic cancers, their oncologists, and when available, their caregivers, from community-and hospital-based cancer clinics in America.	Communication training for patients, oncologists, and caregivers Patients and caregivers underwent 1-hour coaching session incorporating a question prompt list, to help patients raise their concerns during an upcoming office visit, plus up to 3 follow-up phone calls prior to subsequent visits. Oncologists underwent 2-session in-office training (1.75 hours) using a brief video feedback from standardized patients portraying roles of patients with advanced cancer who also critiqued up to 2 audio recorded study patient visits Both trainings focused on engaging patients in consultations, responding to emotion, informing patients about prognosis and treatment choices and balanced framing of information. Control participants received no training	Perception of curability & life expectancy alignment with treating oncologist	Intervention had no effect on improving alignment of patients’ prognostic understanding with that of their oncologists	No effect on hospice utilization or aggressive care in the last 30 days of life
Temel et al., 201115	151	Patients with newly diagnosed metastatic non-small-cell lung cancer from academic medical center in America	Early palliative care integrated with standard oncology care Patients met with a palliative care team member (physician or advanced-practice nurse) for initial consultation and at least monthly thereafter until death. Encounters focused on addressing illness understanding and education; symptom management; treatment decision making; coping with a life-threatening illness for patients and families; and development of care plans for future appointments and referrals to other care providers. Control group received standard oncology care	Perception of curability and treatment understanding	Across study duration, intervention group patients were more likely to either remain or become accurate in their perception of curability relative to control group. However, intervention had no effect on treatment misunderstanding	Intervention associated with lower likelihood of aggressive end-of-life care and higher likelihood of documented resuscitation preferences27
Step et al (2019)25	28 (pilot)	Women with advanced breast cancer from academic medical center in America	Prognostic/treatment discussion 60-minuted structured discussion, with consecutive 15-minute consultations each, with oncologist, oncology nurse, and social worker, and 15 minutes for patient questions. Discussion centered on disclosure of patient’s likely prognosis and elicitation of preferences for prognosis related information, treatment preferences, and identification of individualized care needs. Control group received usual care; intervention group had usual care plus a prognostic/treatment discussion	Perception of curability	Relative to control, intervention associated with greater proportion of patients showing accurate changes in perception of curability.	No effect on patient preference for aggressive care
O’Donnell et al., 201822	50 (pilot)	Recently hospitalized advanced heart failure patients at high risk of mortality from American hospital	Social worker-led structured goals of care conversation Intervention consisted of a structured goals of care conversation with palliative care social worker. Conversation based on the Serious Illness Conversation Guide and education provided on advance care planning. Conversation addressed patient’s prognostic understanding, information preferences, values, goals of care, hopes and fears, unacceptable function, tradeoffs, and family prognostic understanding. Results of conversation documented in electronic health record and communicated to treating clinician. Following conversation: -Social worker reviewed cases with a palliative care physician for further treatment recommendations. -Follow-up encounters scheduled with patients as needed and palliative care needs monitored Control group received usual care	Life-expectancy alignment with treating physician	Intervention group patients more likely to revise baseline life expectancy estimate in a direction consistent with treating physicians’	Intervention associated with higher likelihood of documented advanced care preferences
AlSagheir et al., 202029	92	Patients with metastatic colorectal cancer from oncology centers in Saudi Arabia	Decision aid (informational video) accompanied usual physician consultation Video provided information on treatment benefit, response rates, and survival, with and without chemotherapy, using graphical illustrations and numeric estimates. Control group received physician consultation as usual	Perception of cancer as curable with chemotherapy	No difference in prognostic understanding between intervention and control group (5% vs. 8%)	None assessed
Vanbutsele et al., 201830	186	Patients with advanced cancer from a university hospital in Belgium	Early integration of palliative care in oncological care Patients received consultation from palliative care nurse and monthly meetings thereafter, focused on, symptom assessment, illness understanding, coping, and medical decision-making. Nurses participated in weekly multidisciplinary oncology meetings and referred patients to palliative care physician as needed. Control group received standard oncology care which included psychosocial care (meetings with psychologists, social workers, dieticians).	Perception of curability, goals of therapy, and terminal illness status	No difference between groups in perception of curability or terminal illness status at any of the assessment points. Intervention group had more accurate understanding of goals of therapy at 12 weeks, but not at 18 and 24 weeks.	No difference in aggressive end-of-life care and the use of healthcare resources in the last 30 days of life.31
Enzinger et al., 2020, 202132,33	186	Patients with advanced colorectal or pancreatic cancer starting palliative chemotherapy at cancer centers in America	Educational tool (booklet and video) Patients received booklet and video describing information on the specific chemotherapy regimen recommended to them by their physician. Information included data on treatment logistics, purpose, potential benefits, alternatives, and sample patient experiences. Noncurative nature of palliative chemotherapy was conveyed, along with optional lifeexpectancy estimates. Control group received usual care	Perceived likelihood of curability; perception of treatment intent as cure; life expectancy estimate	No difference between groups in prognostic understanding	No difference in end-of-life care preferences and decisional outcomes

Intervention Features

Most interventions were directed towards patients, but two included caregivers,²³ one of which also included oncologists.¹⁶ The modality of the interventions varied from provision of decision aid²⁹ to discussions with health-care professional.²³ Clinicians administering the interventions varied and included nurses, social workers, oncologists, and palliative care physicians. Intervention intensity ranged from onetime discussion²⁸ to recurring weekly or monthly meetings until death.^23,30

The interventions featured a variety of components. Table 2 shows these components and their frequency across the studies. Provision of information regarding prognosis or available treatment options was a part of all the study interventions. Most interventions (n = 6) also involved assisting patients with planning their endof-life care and providing them emotional and decisional support during the discussion. Gauging patients’ level of prognostic understanding and intervening accordingly was a component of more than half of the interventions (n = 5). Involvement of specialty palliative care was a feature of three of the interventions. Finally, communication training was a key part of just one of the studies, which involved training patients and clinicians on how to have conversations about prognosis and treatment choices.¹⁶

Table 2.

Intervention Components and Their Frequency Across Interventions

Intervention components	Chen 2019	Leighl 2011	Epstein 2017	Temel 2011	Step 2019	O’Donnell 2018	AlSagheir 2020	Vanbutsele 2018	Enzinger 2020	Total
Provision of prognostic or treatment information	x	x	x	x	x	x	x	x	x	9
End-of-life care/advance care planning	x		x	x	x	x		x		6
Provision of emotional support (e.g., cultivating patient readiness)	x		x	x	x	x		x		6
Provision of decision support (e.g., eliciting values)	x	x	x	x		x		x		6
Assessment of prognostic understanding and/or tailoring of information	x		x		x	x			x	5
Use of information aid (e.g., booklet; audio recording)	x	x					x		x	4
Palliative care team involvement				x		x		x		3
Documentation of end-of-life care discussions	x				x	x				3
Communication training for patients/clinicians			x							1

The nine interventions appeared to fall into four categories:

1. Decision/Information aids accompanying medical consults (n = 3): In these interventions — by Leighl and colleagues,¹⁴ AlSagheir and colleagues,²⁹ and Enzinger and colleagues³² — decision or information aids were used in conjunction with routine oncology consults to enhance patients understanding of their disease course, the role of treatment, and future treatment decisions.

2. Specialty palliative care (n = 2): These were palliative care interventions, where prognostic understanding was one of several palliative care needs addressed by the palliative care team. In studies by Temel and colleagues¹⁵ and Van Butsele and colleagues,³⁰ palliative care clinicians met weekly or monthly with patients until death, with the meetings focused in addition to prognostic understanding, on issues such as symptom management, coping, development of care plans, and referrals to other providers.

3. Communication intervention (n = 1): Epstein and colleagues provided communication coaching sessions to patients, caregivers, and oncologists, prior to oncology visits, to enhance prognostic communication.¹⁶ The coaching focused on enhancing ability to engage in conversations about prognosis and treatment choices.

4. Targeted discussions regarding prognosis, treatment decision making and advance care planning (n = 3): These interventions involved discussion(s) with a clinician focused specifically on informed decision making and care planning. Chen and colleagues’ intervention involved repeated weekly or monthly meetings until death with a masters level nurse, that focused on prognostic and endof-life care discussions and care planning.²³ Similarly, the intervention by O’Donnell and colleagues involved a structured conversation with a palliative care social worker focused on prognostic understanding, information preferences, values, goals of care, hopes and fears, unacceptable function, tradeoffs, and family prognostic understanding.²² Finally, the intervention by Step and colleagues also took the form of a meeting focused on prognostic and treatment goal discussion, but featured a multidisciplinary approach.²⁸ They held a 60-minute structured meeting with patients, consisting of consecutive 15-minute meetings each with oncologist, oncology nurse, and oncology social worker (final 15 minutes spent on questions).

Intervention Efficacy

The efficacy of interventions in improving prognostic understanding was mixed (see Table 1). Three of the nine studies showed a statistically significant improvement in prognostic understanding,^22,23,28 whereas three did not.^{16,29 32,33} The remaining three studies showed mixed evidence,^14,15,30 showing improvement in one aspect of prognostic understanding but not another, or improvement at only one of the post-intervention assessment timepoints. For example, the palliative care intervention study by Vanbutsele and colleagues reported no differences in perception of curability or terminal illness status between study arms, but showed a change in understanding of goals of therapy, but only at 12 weeks, and not at 18 and 24 weeks post-intervention.³⁰ The two interventions showing no apparent change in prognostic understanding were the communication training intervention by Epstein and colleagues and a video decision aid intervention by AlSagheir and colleagues.^16,29

Studies with repeated post-intervention assessments (n = 5) suggested that temporal dynamics may need to be considered in evaluating intervention effects. In Lehigh and colleagues decision aid intervention, intervention effects were only apparent a week later, and not immediately after receipt of information, leading the authors to suggest that it may take some time for patients to “absorb” information.¹⁴ Another important temporal pattern is reported by Chen and colleagues, Temel and colleagues, and Vanbutsele and colleagues. In these studies, the gap in prognostic understanding between the intervention and control groups seemed to disappear over time as patients approached death, with patients in the control group “catching up” and showing improved prognostic understanding.^15,23,30 Authors’ suggest that as disease worsens, information indicating a poor prognosis may become more apparent (e.g., more physical symptoms), leading to increased prognostic understanding in both groups. Therefore, the value of interventions may be in increasing prognostic understanding earlier in the disease course.²³

Prevalence of Inaccurate Prognostic Understanding Post Intervention

Given the often-described persistence of inaccurate prognostic understanding even in the face of prognostic information,^12,34 we looked at the reported prevalence of inaccurate prognostic understanding post-intervention. In studies where prognostic understanding was operationalized as more than one facet and/or assessed multiple times post intervention, prevalence rates were averaged across facets and/or assessment points. Five of the eight studies reported prevalence rates at post-intervention (see Table 3). Even after intervention, inaccurate prognostic understanding appeared common, with between 31 and 95 percent of intervention group participants reporting inaccurate perceptions of their prognosis.

Table 3.

Prevalence of Inaccurate Prognostic Understanding After Intervention

Study	Intervention Group	Control Group
Chen et al., 201923	31%	39%
Epstein et al., 201716	49%	53%
Temel et al., 201115	39%	51%
AlSagheir et al., 202029	95%	92%
Enzinger et al., 2020, 202132,33	42%	42%

Notes: Included in table are five of the intervention references that reported prevalence estimates post intervention. In studies where prognostic understanding was operationalized as more than one facet and/or assessed multiple times post intervention, prevalence values were averaged across facets and/or assessment points. Sample size of studies as follows: Chen et al., 2019, N = 460; Epstein et al., 2017, N = 265; Temel et al., 2011, N = 151; AlSagheir et al., 2020, N = 92; Enzinger et al., 2020, 2021, N = 186.

Impact on Other Decisional and Care-Received Outcomes

For most of the identi^fied interventions, their impact on decisional and care-received outcomes had also been reported. These outcomes included aggressive end-of-life care, hospice utilization, documentation of advance care preferences, and decision to undergo chemotherapy and satisfaction regarding treatment decision (Table 1, last column). Chen and colleagues’ advance care planning intervention showed that intervention group participants with accurate prognostic understanding had lower likelihood of receiving CPR in the last month of life, relative to control group participants without accurate prognostic understanding.²³ Similarly, Temel and colleagues palliative care intervention was associated with a lower likelihood of aggressive end-of-life care and a higher likelihood of documented advanced care preferences.^15,27 Finally, the structured goals of care intervention by Step and colleagues was associated with a higher likelihood of documented advanced care preferences.²⁸

Discussion

This systematic review examined interventions that have been evaluated within a randomized controlled design for their impact on prognostic understanding. The twelve identified references, evaluating nine unique interventions, indicate that prognostic understanding can be improved through interventions. However, there are limitations and gaps to be addressed in future research. Specifically, results highlight a need to further examine what intervention components are most potent, the needed dose or intensity of intervention, reasons for continued inaccurate perceptions post intervention among many patients, and the developmental trajectory of prognostic understanding at the end of life.

Not all studies showed an intervention effect (n = 3),^16,29 and some showed difference in only one of the assessed aspects of prognostic understanding (n = 3).^14,15,30 In the present review, there were no clear patterns indicating one intervention component or type as more efficacious. Future research needs to establish, of the numerous intervention components identified here, which components are most important and effective? Is provision of information, provision of decisional and emotional support, assistance with advance care planning, use of information aids, or training in communication, most potent in altering prognostic understanding? Similar questions pertain to intervention intensity and intervention modality. Is a single time-point intervention enough, or are repeated exposures needed? Could general palliative care or oncologic consults suffice,^14,15 or are targeted conversations dedicated solely for prognostic and care planning discussions, necessary?^22,28 Answers to these questions are essential in more widespread dissemination and implementation of these interventions into routine practice.^12,13

Regarding intervention intensity, Chen and colleagues’ intervention reflects the most comprehensive and intensive of the approaches.²³ Masters level-nurses met weekly or monthly with patients until death for conversations focused on advance care planning, and during this process, utilized decision aids such as educational booklets and videos, and facilitated prognostic communication with patients’ physicians. Such a comprehensive approach while potentially more effective, involves more resources, posing implementation challenges in low resource settings and limits on scalability. Future research should therefore address whether less resource intensive interventions (e.g., information aids alone)^14,29 could be as effective. Alternatively, cost-benefit analyses could highlight the return on investment of the resource-intensive interventions, demonstrating their necessity.³⁵

Provision of prognostic information was a component of all the identified interventions. This reflects a key assumption in the literature, that more prognostic information translates to better prognostic understanding.⁷ However, the present results suggest limits to this assumption. Even after intervention, inaccurate prognostic understanding was common among patients, with between 31 and 95 percent of patients exhibiting inaccurate understanding. For example, Chen and colleagues intervention, where repeated discussions were held with patients, showed that across assessment points after starting the intervention, approximately 24% to 38% of patients inaccurately reported their cancer as curable.²³ This is particularly striking considering that patient perception of curability is often an “easier” facet of prognostic understanding, relative to other facets such as life expectancy perception or patient terminal illness acknowledgement. A previous descriptive study (N = 334) examined four different facets, or operationalizations, of prognostic understanding: patient perception of curability, terminal status of illness, advanced stage of cancer, and life expectancy.³⁶ Study results showed that prevalence rates of inaccurate perception of curability was the lowest at 33.6%, whereas inaccurate perception of terminal status was 52.1%, the late stage nature of cancer was 58.8% and of shortened life expectancy was 91.8%. So in the Chen and colleagues study, other operationalizations of prognostic understanding would have likely shown much higher rates of inaccurate prognostic understanding at post-intervention. Therefore, future research needs to examine what contributes to inaccurate prognostic understanding even in the face of prognostic information, and augment interventions to address such factors.^12,13,37

There are several possibilities as to where the breakdown in intervention efficacy occurs: perhaps the information was not given to patients; or even if given, the information was not understood by patients; or perhaps, information was received and understood but psychological barriers buffered the impact of the information.^13,38,39 Future research needs to examine to what extent patients received and understood the prognostic information. It would also be important for interventions to incorporate elements that make it emotionally easier to hear difficult prognostic information and facilitate absorption of information.^12,34 In the reviewed interventions, provision of emotional support was a common element among the interventions, reflecting attempts to mitigate the psychological difficulties associated with hearing aversive prognostic information.

It is important to acknowledge that while we have homogenously referred to the identified interventions as prognostic understanding interventions, the interventions varied in the extent to which enhancing prognostic understanding was a primary goal of the intervention. In many cases, the interventions were multifaceted with multiple goals, and it is unclear if a more central focus on enhancing prognostic understanding may have demonstrated more potent improvements in understanding. Nevertheless, provision of prognostic information was at least a small component in all of the identified interventions, and the prevalence of inaccurate understanding despite such information warrants further research.

The review results suggest that in intervention planning the developmental trajectory of prognostic understanding should be considered. The value of interventions appears to be in facilitating prognostic understanding earlier in the disease trajectory. Closer to death, prognostic understanding appears to improve regardless of intervention receipt (as seen among control group participants in multiple studies).^15,23,30 This raises several questions to be further addressed in future research. How far in advance should prognostic understanding be facilitated? Does better prognostic understanding earlier in the disease course, pose any difficulties for patients, such as interfering with the ability to maintain persistence with further disease-directed treatments and positive mood?⁴⁰ Overly optimistic expectations of disease control may be a coping strategy by which patients are managing their worry and distress related to the illness. Earlier recognition of the potential for illness progression may therefore warrant helping patients develop alternate coping strategies to manage mood and feelings of despondency.¹²

It is notable that despite having a broad eligibility criteria of any study sample with advanced life-limiting illness, almost all of our identified studies were conducted among patients with advanced cancer. The literature on interventions for prognostic understanding thus appear to be dominated by samples of patients with cancer, and it is unclear to what extent the findings from these studies may translate to other patient populations, such as those with advanced renal disease or heart failure. Challenges associated with prognostic understanding have been shown to exist across multiple illness types,^17–19 and the tendency to maintain overly optimistic expectations is an inherent characteristic of human cognition.⁴¹ So, it may be the case that prognostic understanding challenges may be somewhat similar across illness types. Nevertheless, there may be specific aspects to the cancer context (e.g., the prospect of cure with the advent of novel cancer therapies) that may not translate to other illness contexts, limiting the generalizability of the existing literature. Therefore, future research examining non-cancer populations are needed.

Summary

This systematic review evaluating interventions for prognostic understanding suggests that interventions hold the promise to improve prognostic understanding. Future research needs to further examine what intervention components are most efficacious, the needed dose or intensity of intervention, reasons for continued inaccurate perceptions post intervention, and the developmental trajectory of prognostic understanding at the end of life. Additionally, research should examine interventions for prognostic understanding in non-cancer populations.

Appendix 1:

PubMed/MEDLINE Search Strategy

1	(“Prognostic understanding” OR “prognostic awareness” OR “prognostic perception” OR “prognostic communication” OR “prognostic disclosure” OR ((“Prognosis”[Mesh] OR prognos* OR “realistic”) AND (“Awareness”[Mesh] OR “awareness” OR “perception” OR “perceive” OR “perceived” OR “knowledge” OR “understanding” OR “understand” OR “comprehend” OR “comprehension” OR expecta*)))
2	((“Terminal Care”[Mesh] OR “terminal care” OR “Terminally Ill”[Mesh] OR “terminally ill” OR “terminal illness”) OR ((terminal* OR “advanced” OR “life limiting” OR “incurable” OR “end-of-life”) AND (“illness” OR “illnesses” OR “disease” OR “diseases” OR “cancer” OR “cancers” OR “care”)))
3	(intervention[tiab] OR randomized controlled trial[pt] OR controlled clinical trial[pt] OR randomized[tiab] OR placebo[tiab] OR “clinical trials as topic”[MeSH Terms:noexp] OR randomly[tiab] OR trial[ti])
4	(English[lang])
5	1 AND 2 AND 3 AND 4
6	(“Animals”[Mesh] NOT “Humans”[Mesh])
7	5 NOT 6

Mesh = Medical Subject Heading, tiab = Title and Abstract Fields, ti = Title Field, pt = Publication Type Field, MeSH Terms:noexp = Do not include MeSH terms found below this term in the MeSH hierarchy.