If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to: David Kissane, School of Medicine, University of Notre Dame Australia, 160 Oxford Street, Darlinghurst, NSW, 2010, Australia.
Affiliations
School of Medicine (D.W.K, J.A., J.L., N.M., K.N., I.B.), University of Notre Dame AustraliaSt. Vincent's Hospital (D.W.K., J.A., J.L., R.C., K.N.), Sydney, NSWCabrini Health (D.W.K., N.M., I.B.), Melbourne, VictoriaSchool of Clinical Sciences (D.W.K., N.M., L.W., P.P., I.B.), Monash University, VictoriaMonash Health (D.W.K., P.P.), Melbourne, Victoria
School of Medicine (D.W.K, J.A., J.L., N.M., K.N., I.B.), University of Notre Dame AustraliaSt. Vincent's Hospital (D.W.K., J.A., J.L., R.C., K.N.), Sydney, NSW
School of Medicine (D.W.K, J.A., J.L., N.M., K.N., I.B.), University of Notre Dame AustraliaSt. Vincent's Hospital (D.W.K., J.A., J.L., R.C., K.N.), Sydney, NSW
School of Medicine (D.W.K, J.A., J.L., N.M., K.N., I.B.), University of Notre Dame AustraliaCabrini Health (D.W.K., N.M., I.B.), Melbourne, VictoriaSchool of Clinical Sciences (D.W.K., N.M., L.W., P.P., I.B.), Monash University, Victoria
School of Medicine (D.W.K, J.A., J.L., N.M., K.N., I.B.), University of Notre Dame AustraliaSt. Vincent's Hospital (D.W.K., J.A., J.L., R.C., K.N.), Sydney, NSW
School of Medicine (D.W.K, J.A., J.L., N.M., K.N., I.B.), University of Notre Dame AustraliaCabrini Health (D.W.K., N.M., I.B.), Melbourne, VictoriaSchool of Clinical Sciences (D.W.K., N.M., L.W., P.P., I.B.), Monash University, Victoria
Psycho-existential symptoms are common yet often missed or neglected in palliative care. Screening can be an effective way to recognize and respond to this need.
Objectives
We aimed to implement routine use of the Psycho-existential Symptom Assessment Scale (PeSAS) as a screening tool in Australian palliative care services and discern the symptom prevalence identified.
Methods
In a multi-site rolling design, we established implementation site committees and embarked on experiential workshops to train clinicians in the tool's efficient use. Patient symptom prevalence data were collected to compare uptake across sites. Descriptive statistics were applied.
Results
Over one year, we trained 216 clinicians across six palliative care services in the use of the PeSAS as a screening tool and collected data from 1405 patients. Clinicians reported significant growth in their sense of efficacy in assessing psycho-existential wellness. Services using electronic records implemented most easily. Psycho-existential symptoms with clinically significant prevalence (scores ≥ 4/10) included anxiety 41.1%, discouragement 37.6%, hopelessness 35.8%, pointlessness 26.9%, depression 30.3%, and the wish to die 17%. The precision of measurement within 3% was found for severe ratings (score ≥ 8/10) including anxiety 10.6%, depression 10.2%, the wish to die 7.6%, and confusion 3.6%.
Conclusion
Clinicians can be trained to screen with the Psycho-existential Symptom Assessment Scale, which serves as a valuable measure to better recognize symptoms of psycho-existential distress among palliative care patients. Implementation barriers included the prior ethos of the service, confidence in talking about these themes, electronic data entry, and perceived time pressures.
Because psycho-existential states including anxiety, depression, demoralization and the wish to die can remain unaddressed by palliative care services, implementation of a standard screening tool for psycho-existential symptoms is recommended. Over one third of palliative care patients report significant prevalences of these symptoms.
Introduction
Unidentified and unaddressed psycho-existential symptoms cause patient suffering, poor quality of life, vulnerability, suicidal thinking, and lead to acute admissions.
These symptoms generate comparable hospitalizations to those needed to manage physical symptoms, a substantial health cost. They also create a societal burden, contributing to caregiver burnout, poor coping, and increased community services.
An old wisdom states, “More mistakes are made in medicine by not looking than not knowing.” This adage is especially true for psychological and existential states, as research across 20 years has shown how depression has inadvertently passed untreated.
Prevalence, associations, and adequacy of treatment of major depression in patients with cancer: a cross-sectional analysis of routinely collected clinical data.
Through use of a visual analogue scale (0–10), Bruera et al. promoted both assessment and subsequent monitoring of these common symptoms. Across 25 years, use of the ESAS has helped our understanding of symptom trajectories, clusters and modulating factors.
which excluded symptoms such as anxiety and depression. Thus, clinicians bypassed screening for these, perhaps because of management complexity and lack of confidence about how to respond. Moreover, a systematic review of 152 assessment tools used in palliative care revealed the relative paucity of tools assessing spiritual and existential domains.
highlighted the prominence of existential distress. Four systematic reviews of studies assessing demoralization in palliative care have revealed prevalences that include 13%–18% in 2295 patients using the validated Demoralization Scale (DS),
Mental state of demoralisation across diverse clinical settings: A systematic review, meta-analysis and proposal for its use as a ‘specifier’ in mental illness.
Assessing demoralization in medically ill patients: Factor structure of the Italian version of the demoralization scale and development of short versions with the item response theory framework.
underpinned the development of the Psycho-existential Symptom Assessment Scale (PeSAS) as a screening tool to better identify psycho-existential distress. In this study, following the model used for the ESAS, we aimed to implement the routine use of the PeSAS among our palliative care services, in the process observing the prevalence of these symptoms and training multidisciplinary teams to respond appropriately.
Development of the Psycho-existential Symptom Assessment Scale (PeSAS)
Here, we briefly review the prior development of the PeSAS. For inclusion in a screening tool, symptoms need to be important, burdensome, worrying, and disabling. Furthermore, for routine assessment, symptoms must occur commonly. Three statistical approaches were used to identify the most helpful psycho-existential symptoms to screen for.
Firstly, using the Demoralization Scale alongside the Demoralization Interview
as a gold standard for the presence of demoralization, we found comparable areas under the curve (AUC) and sensitivity and/or specificity between 24 items (AUC 0.821, sensitivity 76.7%, specificity 72.8%) and 6 items (AUC 0.810, sensitivity 75.7%, specificity 71.7%).
Assessing demoralization in medically ill patients: Factor structure of the Italian version of the demoralization scale and development of short versions with the item response theory framework.
Latent class analysis differentiation of adjustment disorder and demoralization, more severe depressive-anxiety disorders, and somatic symptoms in a cohort of patients with cancer.
For those with severe psychopathology, the probability of endorsing pointlessness was 80%, feeling trapped 77%, suicidal 76%, loss of control 75%, anxiousness 74%, hopelessness 68%, and lowered mood 52%. For those with moderate psychopathology, the probability of symptoms being endorsed was 69% for feeling trapped, 66% pointlessness, 50% hopelessness, and 42% loss of control. Restated simply, these probabilities make these symptoms worthwhile screening targets.
Thirdly, we used the strength of associations between symptoms in a network analysis to discern which symptoms form communities of related symptoms, and which were most central to each community.
This approach revealed four symptom clusters: firstly, that feeling trapped and becoming discouraged/distressed were strongly correlated; secondly, a cluster of pointlessness, hopelessness and a desire to die; thirdly, a cluster of the symptoms of anhedonic depression; and finally, a cluster involving loss of control and self-worth, where roles were lost and non-specific emotional symptoms predominated. This network analysis was replicated in a separate dataset,
and confirmed that discouragement, entrapment, hopelessness, pointlessness, loss of control, loss of roles and desire to die were key symptoms to screen for.
These different methodological approaches demonstrated convergence towards the same brief group of symptoms as screening targets. Furthermore, the inclusion of anxiety and depression as in the ESAS has a long body of scholastic support, while confusion as a marker of cognitive impairment or delirium/dementia syndromes seemed essential. The resultant screening tool, the Psycho-existential Symptom Assessment Scale (PeSAS), is illustrated in Appendix 1; the optimal ordering of symptoms was tested for patient comfort and comprehension through several iterations. Rather than offer definitions of any symptom, we confirmed that patient understanding was best achieved through the introduction of a couple of synonyms that clarified the meaning of a particular symptom. Illustrative synonyms are shown in Table 1.
Table 1Synonyms Used by Clinicians to Promote Understanding of Screening Items
PeSAS Symptom
Common Symptom Synonyms
Anxiety
Worry, Nervous, Restless
Discouragement
Low morale, disheartened, low confidence
Trapped by illness
Stuck, feeling blocked
Hopelessness
No future, pessimistic, loss of hope
Pointlessness
No meaning, no purpose, no value to life
Loss of control
Not in charge, helpless, can't plan
Loss of roles
Partner, parent, job, loss of identity or self-worth
Depression
Sad, low mood, no interest, joy or pleasure
Wish to die
Not go on, want to end it, opposite of wish to live, suicidal thoughts
Confusion
Poor memory, disoriented, delirious, confused about things
graduated roll out across successive services was planned. The project was introduced to a site implementation committee, following which experiential training workshops were conducted using role play with simulated patients to train staff to conduct PeSAS screening. An electronic version of PeSAS was available through an electronic medical record or web-based app. Patient data were uploaded by sites to a central data manager for comparative evaluation. Local implementation committees monitored barriers. The implementation model is illustrated in Fig. 1.
Fig. 1Model of implementation for Psycho-existential Symptom Assessment scale (PeSAS): a dynamic interaction occurs between national and local site teams, training and supervision, and outcome measurement.
Approval was received from the Human Research and Ethics Committees of all participating institutions.
Sample and Settings
As part of a national implementation program, this study reports on year one of the introduction of PeSAS screening to six palliative care services in Victoria, New South Wales and Australian Capital Territory during 2021. These services represent a blend of inpatient, consultation and community programs and differ in size and staffing numbers. Numbers of clinicians trained by discipline and sociodemographic characteristics are in Table 2; services are deidentified.
Table 2Clinicians’ Socio-Demographic Characteristics by Discipline and Across Contributory Services, Showing Distribution of Disciplines and Eventual Patient Numbers Screened by Services
Characteristics
Consultants Physicians
Trainee Physicians
Nurses
Psychosocial Staff
Pastoral Care
Total Trained
Patient numbers screened
Age ranges % 20–29 years % 30–49 years % >50 years
Each site leader convened a multidisciplinary implementation committee and identified local site champions and a research data nurse. The rationale for psycho-existential screening was introduced with the implementation plan. Site committees selected the components of a service to initially engage with (inpatient, consult or community program).
Training
Training workshops (3 hour, groups of 8–12 staff) were conducted, training between 30 and 50 staff per site. Workshops comprised presentation of introductory principles, demonstration videos, fishbowl role play exercises, feedback discussion and review of management principles.
Clinicians were trained to obtain symptom scores in two to three minutes, to summarize the 2–3 most elevated scores and empathize with elicited distress. They were instructed to explore the “four P's” of assessment: what precipitated, predisposed to, perpetuated, or protected against these symptoms. Participants were advised on how to seek permission to discuss scores with the multidisciplinary team and effect referral for additional psychosocial support.
Training included an overview of symptom management, including treatment of anxiety and depression. Postures of existential vulnerability were contrasted with postures of resilience, so that clinicians could appreciate how to promote optimal adaptation. Risk assessment for suicidal patients and development of management plans were covered. Participants were given a booklet summarizing all that was taught.
Introduction of Screening
Screening was launched at components of each service (e.g., inpatient, community) as patients were admitted. It was introduced as part of routine clinical activity and embedded into the performa used for history taking and examination of each assessed patient. This was reflective of our concern for psychological coping as part of our interest in the whole person. Completion of PeSAS using available software was encouraged. Clinicians incorporated PeSAS scores into discussions of goals of care held by multidisciplinary teams. Follow-up scoring was repeated as clinically indicated. Patients were ineligible if confused, had an altered conscious state, had days to live, or when language was a barrier.
Data Collection
Clinician data: Pre-workshop data included the discipline of the clinician, age band and years of experience, gender, country of birth, and religious affiliation. Pre and post workshop data included ratings of perceived confidence (0–10) in assessing physical, psycho-existential and suicidal symptoms, empathic responsiveness, making a psychiatric diagnosis, managing distress and effecting psychosocial referral. Post-workshop data also included perceptions of new skill development, review of communication skills and likelihood of enhanced quality of care provision.
Service data: Services were deidentified through the use of letters A to F. Basic data included implementation committee minutes, number of workshops, total clinicians trained, and follow-up supervision and case discussions.
Patient data: A minimal patient dataset included phase of illness as described by the Palliative Care Outcomes Collaboration,
Deidentified aggregate clinician data by service compared their pre-and post-workshop efficacy ratings in the use of the PeSAS screening tool. Aggregate patient data enabled symptom prevalence computation as frequencies, while the precision of measurement was represented by the difference in 95% confidence intervals. T-tests were conducted for mean differences and Chi-Square analysis for frequencies. A two-tailed significance level was set at 0.05.
A network structure representing the PeSAS items was estimated through exploratory graph analysis (EGA),
using the EGA function of the EGAnet package, accessed through R 4.0.0. EGA uses the Gaussian Graphical Model, based on regularized partial correlations which represent the strength of association between each pair of items after associations with all other items have been controlled for. The network contains nodes (symptoms) and edges (associations between symptoms, represented by connecting lines for the partial correlations between symptoms). The network was graphically represented based on the Fruchterman–Reingold algorithm.
Clinicians totaling 216 were trained across six services in 2021. They were experienced, mostly nurses, with their disciplines and socio-demographics displayed in Table 2. Significant gains in self-perceived efficacy in screening are shown in Table 3. This growth in confidence achieved an overall Cohen's d = 0.86, a large effect size.
Table 3Clinician Efficacy in Conducting Psycho-existential Symptom Assessment Scale (PeSAS) Screening Comparing Pre and Post training Confidence Scores Out of 10 for 204 Clinicians Who were Among Attendees at 21 Workshops
Skills Confidence
Pre-training Mean (SD)
Post-training Mean (SD)
Change in Means
Cohen's D
Assess physical symptoms
7.22 (1.98)
7.92 (1.53)
0.70
0.42
Assess psychological symptoms
6.39 (1.71)
7.81 (1.11)
1.42
1.01
Assess existential symptoms
5.67 (1.83)
7.80 (1.10)
2.13
1.45
Assess suicidal symptoms
5.67 (2.15)
7.55 (1.40)
1.88
1.06
Empathic responsiveness
7.16 (1.88)
8.23 (1.25)
1.07
0.67
Diagnose psychiatric disorder
4.46 (2.16)
6.40 (1.87)
1.94
0.96
Manage overall distress
6.71 (1.71)
7.88 (1.23)
1.17
0.80
Effect referrals when needed
7.46 (1.65)
8.45 (1.22)
0.99
0.69
Cohen's D: 0.2 = small effect size, 0.5 = medium effect size, 0.8 = large effect size
Clinicians rated the usefulness of training (scored of 10) with mean (SD) values for confidence in using the communication skills as 8.7 (1.2), likelihood of better patient care 8.8 (1.2), personal review of their communication 8.6 (1.5), and usefulness of role play 8.5 (1.5).
Services
From a total of 1405 patients screened, service A (n = 323 patients screened) had a medical-nursing model; service B (n = 226) had a broader mix of clinicians; service C (n = 49) depended on a nurse champion; service D (n = 625) had a strong psychosocial model; service E (n = 70) used a nurse champion; and service F (n = 112) had a medical-nursing care model. Services A, B and D used software to record PeSAS data; services C, E and F employed pencil and paper.
Each service's implementation model was flexible in allowing clinicians to gradually develop their skill and confidence, and employ clinical judgement about the incorporation of PeSAS screening into their practice. As a clinician's confidence grew, most clinicians incorporated screening into their initial patient assessments; some deferred it until later patient encounters.
Patients
Prevalences of PeSAS scores are grouped into severe (≥8) and moderate (≥4 ≤7), alongside clinically significant (≥4) total scores as displayed in Table 4. Precision of measurement (the difference in 95% confidence intervals) was good, with greater accuracy for severe symptom ratings (precision strong, generally <3%) than moderate or total symptom ratings (precision <5%).
Table 4Prevalence of Psycho-existential Symptoms Among 1405 Palliative Care Patients in Clinically Significant (≥4), Moderate 4-7 and Severe (≥8) Categories
To confirm the discriminatory ability of PeSAS, patients with high PESAS scores (≥75th percentile = 41) had significantly lower AKPS scores (M = 49.3, SD = 13.7) than patients with lower PESAS scores (M = 53.2; SD = 14.5) (t-test P < 0.001, d = 0.3). Patients with high PESAS scores were also significantly more likely to be in unstable, deteriorating, or terminal PCOC phases of illness (68.7%) than patients with lower PESAS scores (57.2%) (Chi-Square P = 0.001; adjusted standardized residual = 3.2).
Patients can choose to omit items if they wish. From 1405 patients, 128 (9%) returned incomplete data. Reasons included that patients can be unsure, hesitant to ascribe a rating value, want to remain private, or prefer to offer a more detailed commentary about their experience than can be portrayed numerically.
The network of symptoms is depicted graphically in Fig. 2 using exploratory graph analysis. Hopelessness was the central symptom (node), pointlessness was most strongly associated with the wish to die, while feeling trapped (coping) was associated with loss of roles and loss of control. A median network produced from 10,000 bootstrap iterations was highly similar to the original network, providing credence to the network's stability.
Fig. 2Exploratory graph analysis of psycho-existential symptoms assessed in 1277 palliative care patients on admission to a clinical service (patients with complete data). The thickness and darkness of interconnecting lines between symptoms reflect the regularized partial correlations, a measure of the strength of association between each pair of symptoms after associations with all other items have been controlled for.
Standardized centrality values of nodes are displayed in Supplementary Figure 3. Higher levels of centrality suggest greater involvement in the network. In the estimated network, the most central nodes were hopelessness, depression and pointlessness.
Discussion
Clinicians unfamiliar with screening for psycho-existential symptoms were trained to implement this in palliative care. As a result of experiential roleplay exercises, they grew significantly in confidence in assessment. Many were surprised by the efficiency possible. They valued the model of screen first, then assess in detail, and plan for management later. Clinicians also appreciated the structured approach in the PeSAS, helping gain a greater understanding of each person.
The data are noteworthy for the precision involved in prevalence estimates of these symptoms, especially in the severe category. These prevalences are consistent with formal studies of psycho-existential distress.
Mental state of demoralisation across diverse clinical settings: A systematic review, meta-analysis and proposal for its use as a ‘specifier’ in mental illness.
Around one half of patients felt trapped by their illness, a clear signifier of distress. Loss of control over their life and loss of roles (career, partner, parent) were also prominent. Demoralization occurred in one third of patients, with symptoms of discouragement, hopelessness and pointlessness of life proving prominent. States of anxiety and depression also featured in many, while the wish to die was severe in 7.6% and moderate in 9.4%, consistent with clinical practice.
Patients are free to decline an item response and 9% of our participants returned incomplete screening data. Openness is dependent on a trusting clinician-patient relationship, which can be established over time. Masking of distress can also result from a somatizing or alexithymic style, defensive coping mechanisms, shame or fear of stigma over revealing psychological distress or a lack of trust in the clinician. However, the majority of patients express relief when they feel well understood by a compassionate clinician who has inquired about dimensions of their suffering that others have avoided.
The network map based on the strength of associations between these symptoms confirmed the centrality of hopelessness,
the way pointlessness is the key driver of the wish to die, the links between poor coping (feeling trapped) and loss of control and roles, and the quadratic relationship between anxiety, discouragement, depression and hopelessness. Therapeutic strategies to restore hope and meaning are central alongside management of anxiety and depression.
Implementation research was adapted to each site and the local problems that arose,
inclusive of Covid-19 lockdowns. Use of an implementation committee, engagement of site leadership, electronic recording of PeSAS and local champions proved helpful. Services employing real-time electronic PeSAS entry did best. When a site's physician leader was less involved (e.g., services C and E), trainee physicians bypassed screening. When a biomedical and pharmacological orientation prevailed, introduction of PeSAS was slower (e.g., service F). One service committed their psychosocial staff to take responsibility (service D) with their uptake most successful. Other services had fewer psychosocial staff, depending instead on nurses. When nurse specialists/practitioners worked independently in the community (e.g., services A and B), their appreciation of the importance of psycho-existential distress reflected their wisdom as clinicians.
By comparing services, we could gradually “wake up” institutions to appreciate any historical neglect of psycho-existential issues. This dialectical process in implementation work has been well described,
necessitated much nurturance, and succeeded better in services that were more psychosocially attuned.
There are strengths and limitations to this study. Limitations include the cross-sectional nature of this report; its focus on service issues and clinical processes rather than patient outcomes over time; the complexity of system change with its action research model; lack of generalizability across diverse services; challenges with non-electronic data collection; and the lack of a controlled trial. Implementation research has a practical, real-world orientation, which can be derailed by extensive data collection rather than a focus on services and clinical processes.
Future work will focus on program maintenance with longitudinal data published next. Nevertheless, the prevalence of psycho-existential distress found in these patients was noteworthy. This work extends the broad distress screening programs
that many cancer services have introduced with a symptom assessment tool that is specific for palliative care. The study's value is its description of an implementation program for palliative care, where studies of a systematic approach to quality care improvement for psycho-existential distress have been few.
Conclusion
This implementation of a routine psycho-existential screening tool has identified distress in over one third of the patients screened. Clinicians can be successfully trained to screen in this manner. Real-time electronic entry of symptom data is most valuable. Attention to psycho-existential suffering in palliative care is clearly warranted.
Disclosures and Acknowledgments
The authors have nothing to declare. Australian Commonwealth Department of Health Palliative Care National Program grant number 4-E1QGPW9 awarded to David W Kissane
The authors thank the staff who helped with this project, especially Antonio Claridad, Danielle Carboon, Rachel Eade, Nicholas Glasgow, Nicola Lowrie, Jessica Mather-Hillon, Cynthia Masters, Graham Moss, Nicole Tait, and Grace Walpole.
Prevalence, associations, and adequacy of treatment of major depression in patients with cancer: a cross-sectional analysis of routinely collected clinical data.
Mental state of demoralisation across diverse clinical settings: A systematic review, meta-analysis and proposal for its use as a ‘specifier’ in mental illness.
Assessing demoralization in medically ill patients: Factor structure of the Italian version of the demoralization scale and development of short versions with the item response theory framework.
Latent class analysis differentiation of adjustment disorder and demoralization, more severe depressive-anxiety disorders, and somatic symptoms in a cohort of patients with cancer.
00855-7&id=gr1.jpg){kind=link}
00855-7&id=gr2.jpg){kind=link}