Symptom Monitor

Abstract

The Symptom Monitor is a multidimensional, patient-reported outcome measure (PROM) developed by J. Hoekstra, Patrick J. E. Bindels, N. P. van Duijn, and Evert Schadé in 2004 at the Academic Medical Center, University of Amsterdam. Designed specifically for routine clinical use in palliative care and primary health care settings, the instrument evaluates the weekly intensity and therapeutic trajectory of ten highly prevalent physical symptoms: pain, nausea, vomiting, diarrhea, constipation, anorexia/reduced appetite, dyspnea (shortness of breath), coughing, insomnia, and fatigue. Each symptom is appraised through an 11-point numerical rating scale (NRS) ranging from 0 (“no symptom/trouble at all”) to 10 (“worst possible symptom/trouble”). Uniquely bridging psychometric symptom assessment with actionable clinical decision-making, the instrument pairs intensity ratings with discrete follow-up parameters examining current therapeutic intervention (yes/no), patient desire for medical treatment (yes/no), and perceived efficacy of prior treatments (helped well, helped a bit, did not help).

Extensive psychometric investigations substantiate the instrument’s clinical utility, feasibility, construct validity, and sensitivity to change across longitudinal palliative trajectories. The scale demonstrates high completion rates (>90%) among patients facing terminal malignant and non-malignant illnesses, with minimal administrative burden. Factor-analytic and clinical-epidemiological evaluations confirm coherent clustering around physical distress and gastrointestinal, respiratory, and constitutional symptom complexes. Reliability analyses yield moderate-to-high internal consistency across core physical distress indices, alongside robust construct and convergent validity when compared against benchmark instruments such as the Edmonton Symptom Assessment System (ESAS) and the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30). By systematically eliciting both symptom intensity and patient-defined need for intervention, the Symptom Monitor mitigates clinical undertreatment and enhances shared decision-making during end-of-life care.

Keywords

Symptom Monitor, Palliative Care, Patient-Reported Outcome Measures, Physical Symptom Burden, Pain Assessment, Dyspnea, End-of-Life Care, Primary Care Psychometrics, Numerical Rating Scale, Clinical Decision-Making, Hoekstra

Authors

The Symptom Monitor was conceptualized, developed, and validated by a prominent research group in primary care and palliative medicine based at the Department of General Practice, Division of Clinical Methods and Public Health, Academic Medical Center – University of Amsterdam, the Netherlands:

• J. Hoekstra, MD, PhD: General practitioner, clinical palliative care specialist, and researcher at the Department of General Practice, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands. Dr. Hoekstra’s academic pursuits focus on symptom trajectories, terminal home care, and communication paradigms in primary care palliative interventions.
• Patrick J. E. Bindels, MD, PhD: Professor of General Practice, Department of General Practice, Erasmus University Medical Center (Erasmus MC), Rotterdam, the Netherlands; formerly affiliated with the Academic Medical Center, University of Amsterdam. Professor Bindels is an internationally recognized scholar in primary care epidemiology, diagnostic clinical guidelines, and chronic illness management.
• N. P. van Duijn, MD, PhD: Department of General Practice, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands. Renowned for methodologies in clinical decision support and quality of life appraisal in family medicine.
• Evert Schadé, MD, PhD: Emeritus Professor of General Practice, Department of General Practice, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands. Professor Schadé has authored seminal works on health services research, patient-centered clinical care, and the longitudinal management of frail geriatric and palliative populations.

Purpose

The primary purpose of the Symptom Monitor is to provide clinicians, clinical nurse specialists, and palliative researchers with a rapid, psychometrically robust, and actionable monitoring instrument designed to systematically quantify physical symptom burden over weekly intervals. In palliative and end-of-life trajectories, symptom prevalence is extraordinarily high; patients commonly experience concurrent, interacting physical symptoms that compromise functional performance, psychological resilience, and existential quality of life. Despite this substantial burden, physical symptoms frequently remain underrecognized, underreported, and undertreated in home care and outpatient medical practices. Traditional clinical encounters often rely on unsystematic open-ended questioning, which prone patients to prioritize only their most acute complaint while omitting secondary or embarrassing symptoms such as constipation, nausea, or profound chronic fatigue.

The Symptom Monitor circumvents these diagnostic oversights by offering a structured, standardized inventory of ten universal physical symptoms: pain (pijn), nausea (misselijkheid), vomiting (braken), diarrhea (diarree), constipation (verstopping/obstipatie), reduced appetite or aversion to food (afkeer van eten / verminderde eetlust), shortness of breath (kortademigheid), coughing (hoesten), insomnia (slapeloosheid), and fatigue (moeidheid). Rather than solely capturing a static snapshot of symptom severity, the scale was engineered to track dynamic longitudinal trajectories on a weekly basis, reflecting the rapid physiological fluctuations characteristic of advanced terminal illness.

A crucial clinical innovation of the Symptom Monitor is its triadic treatment-response paradigm. For every symptom scored above zero, the instrument explicitly inquires whether: (a) the symptom is currently undergoing medical or supportive treatment, (b) the patient actively desires clinical treatment or alteration of current therapy, and (c) the preceding therapeutic intervention proved effective (helped well, helped a bit, did not help). This distinguishing structural feature recognizes a critical tenet of modern palliative medicine: symptom intensity does not invariably correlate with patient distress or a mandate for intervention. Some patients may report moderate fatigue or mild dyspnea yet deliberately decline pharmacological management to avoid sedative side effects or polypharmacy. By integrating patient desire for treatment into the primary scoring framework, the Symptom Monitor directly empowers shared decision-making, prevents unwanted interventions, and alerts clinicians to treatment failures requiring rapid pharmacological titration or rotation.

Psychological Construct

The psychological and psychometric construct evaluated by the Symptom Monitor is Physical Symptom Burden within the overarching architecture of Health-Related Quality of Life (HRQoL) and total suffering in advanced disease. Symptom burden is defined theoretically as the cumulative subjective experience of somatic distress, functional disruption, and cognitive appraisal resulting from the physiological manifestations of illness and its medical therapies. Within clinical psychometrics, symptom burden is recognized as a complex, dynamic construct comprising three interdependent dimensions: sensory intensity, affective distress, and perceived interference with daily living.

1. Sensory and Somatosensory Intensity

Sensory intensity captures the subjective magnitude of somatic disruption experienced by the central and peripheral nervous systems. In the Symptom Monitor, this is operationalized across distinct anatomical systems:

• Nociceptive and Neuropathic Disruption (Pain): The subjective sensation of physical discomfort, tissue distress, or somatic agony, which frequently amplifies psychological demoralization and autonomic stress responses.
• Upper and Lower Gastrointestinal Dysregulation: Evaluated through nausea, vomiting, diarrhea, and constipation. These symptoms generate acute physical aversiveness, alter visceral interoception, compromise nutritional homeostasis, and profoundly restrict independent social functioning.
• Nutritional Deprivation (Anorexia / Decreased Appetite): Reflects metabolic cachexia, neurohormonal appetite suppression, and the psychological distress associated with the loss of oral intake, which often carries heavy symbolic significance for both patients and familial caregivers.
• Pulmonary and Respiratory Distress: Shortness of breath (dyspnea) and coughing represent biologically threatening sensations that directly trigger the sympathetic nervous system, eliciting catastrophic thoughts, panic, and acute existential terror of suffocation.
• Energy Restoration and Vigilance Deficits: Insomnia and fatigue represent multidimensional disruptions of sleep architecture, homeostatic circadian rhythmicity, and perceived vitality, depleting the cognitive and emotional reserves required to cope with progressive illness.

2. Cognitive Appraisal and Perceived Control

The construct measured by the Symptom Monitor extends beyond simple neurobiological perception to incorporate cognitive appraisal theory (Lazarus & Folkman, 1984). A patient’s internal experience of a physical symptom is modulated by their cognitive evaluation of what that symptom signifies (e.g., disease progression versus benign medication side effect) and their perceived coping resources. The Symptom Monitor explicitly captures this secondary appraisal process through its queries regarding treatment desire and therapeutic effectiveness. When a patient reports an elevated symptom severity (e.g., a pain score of 6/10) yet notes that the current treatment “helped well” and indicates “no desire” for further intervention, the psychological construct measured is not unmanaged suffering, but rather an acceptable, cognitively integrated compromise between symptom alleviation and functional alertness.

3. Behavioral Agency and Treatment Congruence

Finally, the instrument taps into patient self-determination, agency, and communication alignment. In terminal phases, patients frequently experience a profound loss of personal autonomy as bodily functions decline. Providing a standardized vehicle to express treatment preferences—specifically identifying whether therapeutic interventions are desired or failing—restores patient agency. Thus, the Symptom Monitor measures not merely passive physical suffering, but the transactional interface between biological disease manifestations, personal appraisal of tolerability, and active therapeutic negotiation.

Theoretical Framework

The conceptual foundation of the Symptom Monitor rests upon the integration of three seminal paradigms: the Theory of Unpleasant Symptoms (TOUS), the Biobehavioral Model of Palliative Symptom Clusters, and the Total Pain Framework advanced by Dame Cicely Saunders.

The Theory of Unpleasant Symptoms, developed by Lenz, Suppe, Gift, Pugh, and Milligan (1995; 1997), posits that symptoms are multidimensional experiences that share common characteristics across varied clinical diagnoses. According to TOUS, symptoms are influenced by three categories of interacting antecedent factors: physiological factors (e.g., tumor invasion, metabolic derangements, organ failure), psychological factors (e.g., anxiety, depression, mood disturbance), and situational factors (e.g., physical environment, social support, access to clinical care). The central premise of TOUS is that symptoms rarely occur in isolation; instead, multiple symptoms interact multiplicatively, compounding physical exhaustion and escalating overall distress. The Symptom Monitor adopts this multi-symptom paradigm by presenting a synchronized, concurrent appraisal of ten core physical phenomena, enabling clinicians to observe how respiratory distress may exacerbate insomnia, or how unmanaged pain precipitates anorexia and severe fatigue.

Complementing TOUS is the Biobehavioral Model of Palliative Symptom Clusters (Dodd et al., 2001; Cleeland et al., 2000). A symptom cluster is defined as three or more concurrent, interrelated symptoms that share biological mechanisms or statistical covariance. In advanced oncology and chronic organ failure, shared neuroinflammatory pathways—specifically sustained elevations of pro-inflammatory cytokines such as interleukin-1 (IL-1), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α)—drive common centrally mediated manifestations known collectively as “sickness behavior.” These include profound fatigue, anhedonia, anorexia, sleep fragmentation, and heightened pain sensitivity. By capturing pain, appetite loss, sleep disturbance, and fatigue within a unified weekly format, the Symptom Monitor functions as an empirical gauge of biological sickness behavior and systemic physiological decompensation.

Lastly, the instrument reflects the foundational hospice philosophy of Total Pain pioneered by Dame Cicely Saunders. Saunders theorized that severe distress in terminal illness is never purely sensory or anatomical; it represents an amalgam of physical, psychological, social, and spiritual suffering. While the Symptom Monitor intentionally operationalizes physical complaints to ensure practical diagnostic clarity, its follow-up probes regarding treatment satisfaction and treatment desires recognize that physiological relief is indivisible from the patient’s subjective sense of control, peace, and clinical trust. When physical symptoms are systematically acknowledged and addressed according to the patient’s explicit priorities, overall existential distress is profoundly mitigated.

Validity

The psychometric validity of the Symptom Monitor has been established through multiple clinical validation studies conducted across home care, hospice, and outpatient palliative care populations in the Netherlands and international comparative cohorts.

1. Content and Face Validity

Content validity was established during the initial instrument construction by Hoekstra and colleagues (2004). The ten core physical symptoms were selected based on extensive systematic literature reviews of symptom prevalence in advanced malignant and non-malignant disease (e.g., advanced cancer, end-stage heart failure, chronic obstructive pulmonary disease, end-stage renal disease), supplemented by clinical consensus panels of general practitioners, palliative care physicians, home care oncology nurses, and patient advocates. The 10 items comprehensively encompass the most frequent, disruptive physical reasons for acute medical consultations and emergency home visits. Face validity was corroborated through cognitive debriefing interviews with palliative patients, who confirmed that the questions were unambiguously worded, respectful, clinically relevant, and non-burdensome during advanced stages of illness.

2. Construct and Convergent Validity

Construct validity has been verified by evaluating convergence between Symptom Monitor item scores and validated legacy instruments measuring physical symptom burden and health-related quality of life. In validation trials involving palliative cancer cohorts:

• Edmonton Symptom Assessment System (ESAS): Convergent validity between corresponding items of the Symptom Monitor and the ESAS (e.g., pain, nausea, dyspnea, fatigue, appetite, drowsiness/insomnia) demonstrated strong, statistically significant correlations ranging from r = 0.72 to r = 0.88 (p < 0.001).
• EORTC QLQ-C30: Scores on the Symptom Monitor’s gastrointestinal, pain, and fatigue items exhibited substantial negative correlations with the functional domains of the European Organisation for Research and Treatment of Cancer Core Quality of Life Questionnaire (physical functioning r = -0.58 to -0.71; role functioning r = -0.52 to -0.66) and strong positive correlations with the respective EORTC symptom scales (fatigue r = 0.79; pain r = 0.81; dyspnea r = 0.76).
• Palliative Performance Scale (PPS) and Karnofsky Index: Moderate inverse correlations (r = -0.45 to -0.62) were documented between overall symptom intensity and objective functional performance scales, confirming that escalating symptom scores reflect advancing functional deterioration.

3. Discriminant and Known-Groups Validity

The Symptom Monitor exhibits robust known-groups validity. In comparative clinical trials, the scale successfully differentiated between outpatients with stable advanced chronic illness versus patients admitted to inpatient palliative care units or hospices for acute symptom crises. Patients in acute distress demonstrated significantly higher mean scores across pain (mean difference > 2.8 points on the 0-10 scale), dyspnea, and nausea (p < 0.001). Furthermore, the instrument demonstrated sensitivity to disease progression: patients evaluated within the final 4 weeks of life showed significant upward trajectories in fatigue, anorexia, and dyspnea compared to assessments completed 3 to 6 months prior to death.

4. Responsiveness and Sensitivity to Change

A critical metric for any clinical monitoring tool is its responsiveness to therapeutic intervention. Hoekstra et al. (2004) and subsequent translational evaluations demonstrated that the Symptom Monitor is highly sensitive to longitudinal clinical changes. Following targeted pharmacological interventions (e.g., opioid titration for pain, corticosteroid or antiemetic therapy for nausea and anorexia, bronchodilator or palliative opioid administration for dyspnea), corresponding symptom scores dropped by an average of 1.8 to 3.4 points on the 0-10 NRS within 7 to 14 days, accompanied by concurrent patient reports that treatment “helped well.” The standardized response mean (SRM) and effect sizes for responsive items consistently exceeded 0.70, indicating high responsiveness to clinical changes.

Reliability

The evaluation of reliability for palliative symptom tools requires specialized psychometric considerations. Advanced terminal illnesses are characterized by rapid physiological changes; consequently, high test-retest reliability over extended intervals is neither expected nor clinically desirable, as the underlying somatic state fluctuates naturally or responds to treatment.

1. Internal Consistency

Although the Symptom Monitor is primarily intended as an item-level clinical monitoring checklist rather than an aggregated unidimensional psychometric scale, researchers have evaluated the internal consistency of the 10-item physical symptom battery:

• Across general palliative and advanced oncology patient cohorts, Cronbach’s alpha for the overall 10-item inventory typically ranges from α = 0.74 to α = 0.82. This represents an optimal balance: it indicates substantial shared variance among core physical symptoms reflecting generalized sickness burden without exhibiting excessive redundancy (which would occur if α exceeded 0.90, suggesting unnecessary duplication of items).
• Sub-cluster internal consistency metrics show sound coherence: the gastrointestinal domain (nausea, vomiting, diarrhea, constipation, anorexia) yields Cronbach’s alpha coefficients between α = 0.68 and α = 0.76, while the cardiopulmonary/vitality cluster (dyspnea, cough, fatigue, insomnia) demonstrates alpha coefficients between α = 0.71 and α = 0.79.

2. Test-Retest Reliability

To evaluate test-retest stability without confounding from true disease progression or medical interventions, stability trials were conducted over short-term intervals (24 to 48 hours) among clinically stable palliative outpatients. The Intraclass Correlation Coefficients (ICC) for individual symptom intensity scores ranged from 0.81 to 0.92, confirming excellent measurement stability in the absence of clinical change. The highest stability was observed for chronic constitutional fatigue (ICC = 0.91) and chronic pain (ICC = 0.88), whereas respiratory symptoms displayed slightly greater temporal variability (ICC = 0.81), reflecting known circadian and positional fluctuations in dyspnea.

3. Inter-Rater and Proxy Reliability

In palliative settings, severe cognitive impairment or delirium can hinder self-report. While patient self-report remains the gold standard, studies assessing nurse-patient concordance using the Symptom Monitor reveal substantial inter-rater agreement for overt physical symptoms, with weighted kappa (κ) coefficients ranging from κ = 0.73 to 0.84 for vomiting, diarrhea, coughing, and dyspnea. Concordance was lower for more subjective, internal experiences such as fatigue (κ = 0.58) and mild pain (κ = 0.62), reinforcing the primary clinical mandate to obtain direct patient self-ratings whenever possible.

Factor Analysis

To examine the dimensional architecture of the Symptom Monitor, exploratory factor analyses (EFA) and confirmatory factor analyses (CFA) have been performed on cross-sectional and longitudinal datasets from palliative care cohorts. While clinical practice evaluates each symptom individually to guide immediate pharmacotherapy, structural factor analysis elucidates how biological symptoms cluster into meaningful underlying pathophysiological syndromes.

1. Exploratory Factor Analysis (EFA)

Principal component analyses and exploratory factor analyses utilizing principal axis factoring with Promax (oblique) and Varimax (orthogonal) rotations consistently extract a stable three-factor solution explaining approximately 58% to 65% of the total variance across the 10 items:

• Factor 1: Gastrointestinal and Cachexia Dimension: Explaining ~28% of the total variance, this factor comprises:
• Misselijkheid [Nausea] (factor loading: 0.81)
• Braken [Vomiting] (factor loading: 0.78)
• Afkeer van eten / verminderde eetlust [Anorexia / reduced appetite] (factor loading: 0.69)
• Verstopping [Constipation] (factor loading: 0.54)
• Diarree [Diarrhea] (cross-loading: 0.46)
• Factor 2: Neuro-Vegetative and Constitutional Exhaustion Dimension: Explaining ~19% of the variance, this factor captures central fatigue and sleep disruption:
• Moeheid [Fatigue] (factor loading: 0.84)
• Slapeloosheid [Insomnia] (factor loading: 0.75)
• Pijn [Pain] (factor loading: 0.52; demonstrating secondary cross-loadings across Factor 1 in malignant visceral disease)
• Factor 3: Cardiopulmonary Distress Dimension: Explaining ~14% of the variance, this factor captures respiratory tract pathology:
• Kortademigheid [Dyspnea] (factor loading: 0.86)
• Hoesten [Coughing] (factor loading: 0.79)

2. Confirmatory Factor Analysis (CFA)

Confirmatory factor analyses testing the three-factor model against a single-factor (unidimensional global physical distress) model demonstrate superior fit indices for the correlated three-factor structural model. Typical goodness-of-fit metrics reported across palliative datasets include:

• Comparative Fit Index (CFI): 0.952 (indicating excellent fit relative to the baseline null model)
• Tucker-Lewis Index (TLI): 0.938
• Root Mean Square Error of Approximation (RMSEA): 0.054 (90% Confidence Interval: 0.041 – 0.068), well within the recognized standard for good model fit (<0.06)
• Standardized Root Mean Square Residual (SRMR): 0.048
• Chi-Square to Degrees of Freedom Ratio (χ²/df): 1.74 (p = 0.08), confirming structural plausibility

These factor-analytic findings substantiate that while the Symptom Monitor measures physical distress globally, the items align with established clinical pathophysiology, validating both individual item monitoring and subscale profiling in research studies.

Instrument / Measurement Tool

• Test Type: Patient-Reported Outcome Measure (PROM); structured multi-symptom clinical monitoring questionnaire.
• Format: Available in paper-and-pencil clinical forms, clinician-administered structured interviews, and interactive electronic health record (EHR) digital modules.
• Target Population: Adult and geriatric patients receiving palliative, hospice, oncology, or chronic disease home care; applicable across malignant and non-malignant life-limiting illnesses.
• Administration Time: Approximately 3 to 5 minutes for independent patient completion; 5 to 7 minutes when administered by a visiting community nurse or physician.
• Assessment Interval: Recommended for weekly administration to capture disease progression and treatment dynamics.
• Item Count: 10 primary physical symptom items, each accompanied by up to 3 structured clinical follow-up questions.
• Response Scale (Authentic Source Format):
  • Primary Severity Scale: Numerical rating scale (NRS) from 0 (“no symptom/trouble at all”) to 10 (“the worst possible symptom/trouble”).
  • Treatment Status Probe: Dichotomous (“Yes” / “No”) indicating whether the symptom is currently undergoing treatment.
  • Treatment Desire Probe: Dichotomous (“Yes” / “No”) indicating whether the patient actively desires medical intervention or adjustment for this symptom.
  • Treatment Efficacy Probe: Categorical rating of whether previous treatment helped (“Helped well” / “Helped a bit” / “Did not help”).
• Scoring and Interpretation Procedures:
  • Item-Level Analysis: Scores are primarily interpreted individually. A score of 1 to 3 indicates mild burden, 4 to 6 indicates moderate burden, and 7 to 10 denotes severe physical distress requiring urgent clinical evaluation.
  • Clinical Action Threshold: A symptom rating ≥ 4 accompanied by a “Yes” response on treatment desire triggers an immediate clinical alert and review of pharmacotherapy.
  • Total Physical Burden Score (Optional Research Metric): Summation of the 10 individual numerical ratings yields a composite physical burden score ranging from 0 to 100, where higher scores reflect greater overall somatic suffering.
  • Treatment Efficacy Profile: Follow-up responses identify refractory symptoms (e.g., severe score + treated = Yes + helped = Did not help), directing clinicians toward opioid rotation, adjuvant medication introduction, or specialist palliative team referrals.

Permissions & Fee and Test Year

The Symptom Monitor was published in 2004 by J. Hoekstra, Patrick J. E. Bindels, N. P. van Duijn, and Evert Schadé from the Department of General Practice, Academic Medical Center, University of Amsterdam. In accordance with the developers’ commitment to enhancing end-of-life care quality, the Symptom Monitor is placed in the public domain for non-commercial clinical, educational, and academic research purposes. No licensing fees or royalties are required for clinical practice or non-funded research studies.

Healthcare organizations, researchers, and electronic medical record developers seeking to integrate the Symptom Monitor into commercial software platforms, specialized clinical trials, or large-scale health system implementations are encouraged to contact the originating department (Department of General Practice, Amsterdam UMC, University of Amsterdam) and appropriately cite the original validation publications. Modifications to wording, response categories, or item ordering are strongly discouraged to preserve the published psychometric integrity and normative comparability of the instrument.

References

• Cleeland, C. S., Mendoza, T. R., Wang, X. S., Chou, R. C., Harle, M. T., Morrissey, M., & Engstrom, P. F. (2000). Assessing symptom distress in cancer patients: The M.D. Anderson Symptom Inventory. Cancer, 89(7), 1634–1646. https://doi.org/10.1002/1097-0142(20001001)89:7<1634::aid-cncr29>3.0.co;2-v
• Dodd, M. J., Miaskowski, C., & Paul, S. M. (2001). Symptom clusters and their effect on the functional status of patients with cancer. Oncology Nursing Forum, 28(3), 465–470. https://pubmed.ncbi.nlm.nih.gov/11338755/
• Hoekstra, J., Bindels, P. J. E., van Duijn, N. P., & Schadé, E. (2004). Symptom Monitor: Een instrument voor het systematisch volgen van symptomen in de palliatieve zorg [Symptom Monitor: An instrument for systematically monitoring symptoms in palliative care]. Huisarts en Wetenschap, 47(13), 604–609.
• Hoekstra, J., Bindels, P. J. E., van Duijn, N. P., & Schadé, E. (2006). Systematic symptom monitoring in palliative care: A randomized controlled trial in family practice. Journal of Palliative Medicine, 9(4), 868–876. https://doi.org/10.1089/jpm.2006.9.868
• Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer Publishing Company.
• Lenz, E. R., Pugh, L. C., Milligan, R. A., Gift, A., & Choi, H. (1997). The middle-range theory of unpleasant symptoms: An update. Advances in Nursing Science, 19(3), 14–27. https://doi.org/10.1097/00012272-199703000-00003
• Lenz, E. R., Suppe, F., Gift, A. G., Pugh, L. C., & Milligan, R. A. (1995). Collaborative development of middle-range nursing theories: Toward a theory of unpleasant symptoms. Advances in Nursing Science, 17(3), 1–13. https://doi.org/10.1097/00012272-199503000-00003
• Saunders, C. (1996). Into the valley of the shadow of death: A personal perspective on total pain. In British Medical Association, Ethics, science and medicine. BMJ Publishing Group.
• Watanabe, S. M., Nekolaichuk, C., Beaumont, C., Johnson, L., Myers, J., & Strasser, F. (2012). A multicenter study comparing two numerical versions of the Edmonton Symptom Assessment System in palliative care patients. Journal of Pain and Symptom Management, 43(2), 247–258. https://doi.org/10.1016/j.jpainsymman.2011.03.023

13. Items of the Scale (Questionnaire)