Study inclusion
A total of 693 titles and abstracts were identified. After duplicates were removed, 681 articles remained and were screened by title and abstract. Of these, 397 met the inclusion criteria, with full texts reviewed for relevance. One hundred and four studies met the inclusion criteria and were included for narrative analysis. Six additional studies were added through a separate grey literature search using ProQuest, CADTH, Google, and SUMSearch; six studies were added after reference list searching of included studies (N = 116) (Fig. 1).
Characteristics of included studies
Included studies were grouped based on year of publication, country (based on lead author), design, aims, diseases included (i.e. type of cancer, multiple diagnoses and non-cancer illnesses) and definition category (see Appendix I). Studies that examined EPC in more than one life limiting illness were considered ‘multiple diagnosis’ (e.g., Cancer and/or Dementia and/or COPD and/or Sepsis).15
The operationalized criteria for when EPC was initiated were categorized under five main definitions: 1) time-based; 2) prognosis-based; 3) location-based; 4) treatment-based; 5) symptom-based. See Appendix II for breakdown of definitions by type and subtype of life limiting illness and operational definition category. Although some definitions had elements from multiple definition categories, a primary category was selected for each definition based on which criteria was stated first or had more prominence; for example, where EPC was defined as specialist palliative intervention occurring within three days of being admitted to acute care with advanced stage of disease15,16, this could be considered time and location-based but was sorted into time-based category first.
Seventy-eight of the included studies examined EPC in cancer populations. Table 1 provides frequencies of EPC definitions used in primary and multiple primary cancers (n = 78). Table 2 provides frequencies of EPC definitions used in multiple diseases such as cancer, heart failure, respiratory diseases, sepsis, frailty, organ failure and neurodegenerative diseases using a single EPC intervention (n = 38).
Table 1
Frequencies of EPC Definitions used in Cancer Populations
| Cancer | Cancer Total |
Definition Category | GIa | Hematologic | Lung | Multiple primary cancersb | Otherc | |
Time-Based | | | | | | |
Time from advanced cancer diagnosis to EPCd intervention | 4 | 1 | 8 | 19 | 4 | 36 |
Timing of EPC intervention before death | | | 1 | 5 | | 6 |
Prognosis-Based | | | | | | |
Based on disease stage | | | 3 | 3 | 1 | 7 |
Based on prognosis | | 1 | | 4 | | 5 |
Using the surprise question | | | | 3 | | 3 |
Location-Based | | | | | | |
Based on access point within the healthcare system | | | 1 | 4 | 2 | 7 |
Treatment-Based | | | | | | |
Based on physician’s judgement | | | | 2 | | 2 |
Prior to rescue therapies | | 2 | | 2 | | 4 |
Symptom-Based | | | | | | |
Based on Symptoms | | 1 | 1 | 1 | | 3 |
Using a trigger-based tool | | 2 | 1 | 2 | | 5 |
Total | 4 | 7 | 15 | 45 | 7 | 78 |
a Gastrointestinal |
b Studies describing multiple primary origins (such as breast, and/or gastric, and/or oral, and/or pancreatic) were captured under ‘multiple primary cancers’. |
c Studies describing single origin primary cancers with fewer than three articles were captured in the ‘other’ column. |
d Early palliative care |
Table 2
Frequencies of EPC Definitions used in Multiple Diagnoses and Non-cancer Diseases
Definition Category | Multiplea | Dementia | Liverb | Heartb | HIVc | Respiratory | Trauma | Trans plant | Total |
Time-Based | | | | | | | | | | |
Time from advanced diagnosis to EPCd intervention | 1 | | | | 1 | 2 | | | | 4 |
Timing of EPC intervention before death | 2 | | 1 | | | | | | | 3 |
Prognosis-Based | | | | | | | | | | |
Based on disease stage | 2 | | | 2 | 2 | 2 | | | | 8 |
Based on prognosis | 1 | | | | | | | | | 1 |
Using the surprise question | | | 1 | | | | | | | 1 |
Location-Based | | | | | | | | | | |
Based on access point within the healthcare system | 3 | | | | 1 | | | 1 | | 5 |
Treatment-Based | | | | | | | | | | |
Based on physician’s judgement | | | | | | | | | | |
Prior to rescue therapies | | | 1 | 1 | | | | | 1 | 3 |
Symptom-Based | | | | | | | | | | |
Based on Symptoms | 3 | | | 2 | | 1 | | | | 6 |
Using a trigger-based tool | 3 | 1 | | | | 3 | | | | 7 |
Total | 15 | 1 | 3 | 5 | 4 | 8 | | 1 | 1 | 38 |
a Multiple diagnoses encompassed studies that discussed EPC among patients who had more than one life-limiting illness including cancer, heart failure, respiratory diseases, sepsis, frailty, organ failure and neurodegenerative diseases. |
b Disease |
c Human Immunodeficiency Virus. |
d Early palliative ca |
EPC definitions in Cancer:
Time-Based Definitions (Cancer).
Most EPC operational definitions described among patients with cancer were time-based (time from advanced cancer diagnosis to EPC initiation, n = 36 or timing of EPC initiation before death, n = 6). Most of these definitions align with and often cite the seminal article published by Temel et al (2010). In these studies (n = 15) patients were enrolled if diagnosed with advanced cancer within the previous 6–8 weeks; some required an Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1, or 2.3,17−30 The remaining studies and guidelines either broadly stated that EPC should be initiated at the time of diagnosis of an advanced or incurable cancer or recommended that it be initiated within specific timeframes: from 2 weeks to one year after diagnosis (n = 21).31–51 EPC interventions varied between unstructured needs-based care to standardized EPC consultations .33,35 Retrospective studies often described EPC as specialist palliative care initiated more than three months before death.52–57
Prognosis-Based Definitions (Cancer).
In 15 studies, prognosis-based indicators were used to distinguish EPC initiation among cancer patients. Seven studies used staging criteria (stages III/IV) to identify patients eligible to receive EPC services.58–65 Four authors used prognosis of 6–24 months, as determined by the physician, as the operational definition of appropriateness for EPC initiation.66–69 Tanzi and colleagues (2020) initiated EPC in patients with a prognosis of greater than one month once patients were on their last active treatment. In three articles, the ‘surprise question’ (would the physician be surprised if this patient dies in the next year?) was used to determine appropriateness for EPC intervention.70–72
Location-Based Definitions (Cancer).
Seven studies considered EPC as care delivered in outpatient or homecare settings, the rationale for this was that outpatient care is generally provided earlier in the disease process, before the onset of overly burdensome symptoms.73–75 In four articles, authors examined specialist palliative care consultation within the hospital setting for advanced cancer patients, where initiation of palliative services with 2–3 days of admission to an acute care hospital was considered EPC.76–79
Treatment-Based Definitions (Cancer).
Treatment-based criteria to define the initiation of EPC was less common. In two studies, authors used physician’s judgment to determine when to initiate EPC after providing education to enhance awareness and optimize their referral-based practices. Greater awareness led to increased consults.80,81 Four authors defined EPC as care initiated prior to a definitive therapy such as a hematopoietic stem cell transplant or curative-intent.72,82−84
Symptom-Based Definitions (Cancer).
In eight articles, authors used symptom-based indicators to define initiation of EPC; three of which described EPC interventions after patients exhibited disease or treatment related symptoms or felt distress85–87; symptoms were assessed using routinely collected likert-type scales to assess severity of symptoms. (e.g. The Edmonton Symptom Assessment Scale (ESAS).85 In the remaining five articles, authors described trigger-based criteria for EPC initiation which combined an increased symptom burden, measured by specific questionnaires (e.g. The European Organization for Research and Treatment of Cancer quality of life questionnaire [EORTC QLQ-C30]) and/or increased tertiary services such as frequent emergency department visits combined with advanced disease stage (such as stage III/IV cancer).88–92
EPC Definitions used for Studying Multiple Diagnoses and Non-cancer Diseases.
Fifteen of the included studies examined EPC interventions among patients with multiple diagnoses. Multiple diagnoses include studies that discuss the initiation of EPC among those with one or more of the following diseases: cancer, heart failure, respiratory diseases, sepsis, frailty, organ failure and neurodegenerative diseases. Some authors used specific criteria for each disease such as advanced disease stage (e.g. heart failure with a score of 2–4 using the New York Heart Association (NYHA) Functional Classification, or a score of 1–4 on the Global Initiative for Chronic Obstructive Lung Disease Scale (GOLD)).8 In other cases, authors did not differentiate between diseases but used other criteria to define initiation of EPC such as palliative care delivered more than 60 days before death.93 Table 2 describes frequencies of EPC definitions used in multiple (n = 15) and non-cancer diseases (n = 23) and are categorized by type of operational definition.
Time/Location-Based Definitions (Multiple Diagnoses).
In two retrospective studies, authors defined EPC as initiation within a specific timeframe before death; for example, EPC was considered specialist consultation more than three months before death.93,94 Two authors defined EPC as specialist palliative intervention initiated within three days of being admitted to an acute care hospital with advanced stage of cancer, dementia, cardiovascular, COPD, sepsis or other diseases15 and frailty16, and one study chose with 24 hours of admission for heart failure, cirrhosis, COPD, cardiovascular, frailty, dementia, renal, HIV, and cancer patients.95
Prognosis-Based Definitions (Multiple Diagnoses).
Beernaert et al. (2016) explored EPC needs in cancer, COPD, heart failure and dementia patients separately using standardized tools to identify disease stage and need for EPC.8 For example, authors used the NYHA Functional Classification for heart failure and considered patients with a score of 2-4.8 Pesut and colleagues (2017) used a prognosis of less than a year among cancer, heart failure, COPD, neurodegenerative and other life-limiting illnesses to initiate EPC.96,97
Symptom-Based Definitions (Multiple Diagnoses).
Three articles in this category used trigger-based criteria; two of which considered EPC in ICU patients using advanced disease stage (stage IV cancer, multi-organ failure, class III or IV heart failure) and tertiary service utilization for symptom management (> 1 hospital or ICU admission within three months).98,99 One article developed a tool for primary care physicians to initiate EPC in heart failure, COPD, and cancer patients using advanced disease stage, tertiary service use, and worsening symptoms.100 Gomez-Batiste et al. (2017) created a standardized tool to identify patients with multiple chronic conditions in need of palliative care services using a tool called the NECesidades Paliativas (NECPAL). It uses prognosis (i.e. the surprise question), symptoms (refractory symptoms using the ESAS), and disease stage e.g. NYHA stage III or IV heart failure, or renal failure with a GFR < 15 to determine initiation of EPC for.101 Johnston et al. (2018) completed a mixed methods study looking to evaluate a home care palliative care model with an early palliative referral before burdening symptoms for cancer, cardiac, respiratory, dementia/frailty, neurologic and other conditions.102 Similarly, Chidiac et al. (2018) completed a review highlighting the various definitions and defined EPC as, “specialist palliative care interventions delivered earlier in the course of illness and before the onset of burdening symptoms, using an integrated model of care”.103 p. 231
EPC in other non-cancer diseases:
In the remaining 23 studies, authors examined EPC initiation in single, non-cancer diseases including: heart failure (n = 4), HIV (n = 4), COPD (n = 5), respiratory disease (n = 3), liver disease (n = 3), organ transplantation (n = 2), dementia (n = 1), and trauma patients (n = 1).
Time/Location-Based Definitions (Non-cancer Diseases).
Time-based definitions were less common in non-cancer diseases. Barnes et al. (2019) defined EPC as specialist palliative care services initiated 30 days prior to death among patients with end-stage liver disease.104 Lindell et al. (2018) recommended EPC at the time of diagnosis of interstitial pulmonary fibrosis while Iyer et al (2019) suggested a broader definition of EPC which integrated planning, emotional, spiritual and social support along with chronic disease management at the time of diagnosis of COPD.105,106 One author used a combination of location and time as their definition of EPC in trauma patients, with EPC initiation occurring within 24 hours of admission into a trauma ICU.107 For those with HIV, time-based initiation of EPC was discussed in two articles7,108, and one definition was solely location-based, where EPC was defined as palliative care initiated to HIV patients in the outpatient setting.109
Prognosis-Based Definitions (Non-cancer Diseases).
A large portion (n = 7) of definitions in non-cancer diseases are prognosis-based and use advanced disease stage as their operational definition of EPC. Within the HIV population, one author suggested that EPC should be defined by clinical criteria, specifically a CD4 T-cell in the 300–400 cells/mL.110 Webel and colleagues (2019) defined EPC as care initiated early in the disease trajectory for HIV patients. When examining the role of EPC among patients with COPD, authors based their definition on advanced disease stage (stage III or IV based on the GOLD criteria).4,111 Similarly, among people with heart failure, disease stage was the operational definition for when EPC should be initiated (AHA Stage C/D or NYHA Class III/IV).112,113 In the remaining article, EPC was based on prognosis using the surprise question among patients with end-stage liver disease.114
Treatment-Based Definitions (Non-cancer Diseases).
All treatment-based definitions were related to rescue therapies, specifically EPC being delivered pre-organ transplant (for liver, lung, and heart transplant patients).5,115,116
Symptom-Based Definitions (Non-cancer Diseases).
Most studies examining EPC in non-cancer populations used symptom-based definitions compared to cancer populations (35% vs 10%) In one study, authors examined EPC in patients with dementia using a trigger-based tool called the Gold Standard Framework Proactive Identification Guidance tool which combines the surprise question, client choice and need, with clinical indicators such as advanced disease stage and decreased function.117 Similarly in the COPD population, Scheerens et al. used a trigger-based method which included advanced disease stage (GOLD III or IV) plus one or more of the following clinical indicators: oxygen dependence, frequent hospitalization, refractory dyspnea, recent intubation, and/or recent weight loss.118
Other articles used advanced disease with the presence of disease or treatment related symptoms which aligns with the American Thoracic society guidelines for palliative care for patients with respiratory diseases as early as diagnosis if disease or treatment related symptoms are present.119–121 This definition was also found within the heart failure population.122,123