Emotional Symptoms

For the purpose of this conference, we have divided physical and emotional symptoms into two separate domains. This was purely an administrative decision. It should be noted that research has shown that physical and emotional symptoms are indeed strongly related. Physical symptoms in a terminal illness are associated with emotional distress, depression, and anxiety. An important question is whether we recommend a separate instrument for emotional symptoms such as anxiety and depression, or if we recommend an instrument that measures pain, depression, anxiety, and other symptoms. Based on a review of the literature, we have listed both sets of instruments. Note that the instruments that focus on emotional symptoms are listed first, followed by those for general symptoms.

A literature review of the Medline database was conducted with the search terms: emotional symptoms, emotional, depression, anxiety, fear, measurement, instrument and (terminal care or palliative care). We also searched the Cancerlit database. We found a total of 118 articles published between 1983 and 1997. In addition, we selected references from the methods sections of those articles pulled, and the George Washington University online catalogue was searched for relevant books.

Based on a review of these 118 articles, we found 18 potential instruments for consideration which are listed below:

a. CES-D scale (Radloff 1977) - a 20-item, self-report depression scale used in the Kane and colleagues' randomized controlled trial of hospice.

b. Mental Health Inventory (Ware 1979) - used in Kane and colleagues' randomized controlled trial of hospice. This instrument has been shortened to five questions (the MHI-5, Berwick 1991).

c. Profile of Mood States (Cella 1987) - a 65-item, adjective rating scale used in the National Hospice Study and The Study to Understand Prognoses and Preferences for Outcomes, and Risks of Treatments.

d. Beck Depression Index (Beck 1961) - a widely used scale that measures the severity of depression by evaluating 21 symptoms.

e. Hospital Anxiety and Depression Scale (Zigmond 1983) - a self-assessment scale with depression and anxiety subscales.

g. Geriatric Depression Scale (Yesavage 1983)- simple scale developed to diagnose depression in older persons.

h. General Health Questionnaire (Goldberg 1972) - a self-administered instrument designed to detect depression, anxiety, social impairment, and hypochondriasis

i. The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments Afterdeath interview (Lynn 1996) - It should be noted that there is an important debate on how to handle physical/somatic symptoms in depression scales. I have predominately listed scales which focus on the measure of psychological symptoms, given that the assessment of physical symptoms will be an important part of the "toolkit".

j. Support Team Assessment Schedule (Butters 1992) - instrument that requires staff and patient to evaluate symptoms and needs.

k. Dartmouth COOP Functional Assessment Charts (Nelson 1996) - quick, easily administered survey that relies on visual images to assess multiple domains.

l. Symptom Distress Scale (McCorkle 1978) - a measure of ten symptoms, including mood and concentration, that is interviewer administered.

m. Rotterdam Symptom Checklist (de Haes 1990) - a self-administered scale that includes an eight item subcale for psychological distress.

n. Edmonton Symptom Assessment System (Bruera 1991) - a scale comprised of eight visual analogue scales indicating various physical and emotional symptoms.

o. The Quality of Life Index (Spitzer 1981) - a scale designed to measure the general well-being of patients with terminal illness or chronic disease.

p. EORTC QLQ (Anderson 1993) - an instrument created for cancer patients incorporating nine multi-item scales.

q. Memorial Symptom Assessment Scale (Portenoy 1994) - a patient rated instrument to provide information on symptom prevalence, characteristics and distress.

r. Chronic Respiratory Disease Questionnaire (Guyatt 1987) - a 19 item scale that measures dyspnea, fatigue, emotional function, and feeling of control over disease.

Based on a review of these 18 instruments, I have identified the following instruments-Profile of Moods State, Rand Mental Health Inventory (the 5 question version), CES-D, and the Memorial Symptom Assessment Scale, which is the instrument that I recommend. I will review the other instruments first, and then state why I have recommended the Memorial Symptom Assessment Scale.

i. Conceptual and Measurement model (Does the scale represent a single domain or do model scales measure distinct domains? Is the variability of the scale reported? If so, please document it. What is the intended level of measurement i.e., ordinal, interval, ratio or category?)

This is a self-administered questionnaire (also can be interviewer administered) which emphasizes measures of affective components (see attached instrument). Factor analyses suggests four constructs - depressed affect, positive affect, somatic symptoms and retarded activities, and interpersonal problems. The scale has been used extensively in general populations, persons with AIDS or symptomatic HIV disease, and frail older persons. As noted previously, the CES-D was used in one randomized trial of hospice care. In this study, they found that the mean score was 13 and 7 over 8 weeks of follow-up. The intended level of measurement is ordinal (although it has been analyzed as an interval level variable).

ii. Reliability (Did they address internal consistency? Did they address reproducibility?)

Internal consistency has been assessed among multiple samples with Crohnbach alpha between .76-.91. The test-retest has shown only a moderate amount of agreement (most correlations are between .50-.60).

iii. Validity - (How did they address content related validity? Any information on construct related validity? Any information on criterion validity?)

The scale was created based on a review of existing instruments. The scale has been shown not only to differentiate those with and without a clinically diagnosed depression, but it also differentiates those who reported needing help with an emotional problem. As a screening tool, it has been tested in outpatient settings and among patients with severe depression. However, it has not been applied as a screening tool in patients with a terminal illnesses.

iv. Responsiveness (Any information? Has the scale ever been used as an outcome measure? If so, in what populations?)

It has been used as an outcome measure in Kane and colleagues' (1985) study of hospice. Repeated measures analyses of covariance did not show an impact of the intervention (there was a trend to suggest that intervention patients were less depressed).

v. Interpretability (What populations has it been applied too? Is the score translated into a clinically relevant event? Does the score predict outcome events?)

There are suggested cut points of 16 as well as national norms published which can be stratified by age, race, income, gender, occupation, and marital status.

vi. Burden (Any information on cost or time to administer? Does the instrument impact on the respondent? How long does a survey take to complete? Response rates? Any problems with missing data?)

Short 20 item instrument that can be self-administered or be interviewer administered.

vii. Alternative forms (What are the modes of administration? Alternatives? If alternative exist, provide what is known for each of the above categories)

There is a 10 item form (Burnham 1988) which has shown good reliability and correlation with the 20 item instrument.

b. Mental Health Inventory 5 item version (adapted from Rand Mental Health Inventory, Ware 1979, by Berwick 1991)

This 5 item instrument was adapted from the 20 item Rand Mental Health Inventory. Kane and colleagues used the anxiety intervention.

ii. Reliability (Did they address internal consistency? Did they address reproducibility?)

Internal consistency has been found to be high (.86-.90) and test-retest agreement was good.

iii. Validity - (how did they address content related validity? Any information on construct related validity? Any information on criterion validity?)

The scale was created based on review of existing instruments. Berwick and colleagues compared the MHI5 to General Health Questionnaire, a longer version of the Mental Health Inventory, Somatic Symptom Schedule, and Diagnostic Interview Schedule (DIS). They found that MHI5 had an ROC of .79 for detecting any mental disorders.

iv. Responsiveness (Any information? Has the scale ever been used as an outcome measure? If so, what populations?)

A version of the MHI (Anxiety subscale) has been used as an outcome measure in Kane and colleagues' (1985) study of hospice. Repeated measures analyses of covariance did not show an impact of the intervention (there was a trend to suggest intervention patients were less anxious).

v. Interpretability (What populations has it been applied to? Is the score translated into a clinically relevant event? Does the score predict outcome events?)

The questions have been widely used with populations norms for SF-36 (McHorney). It has not, to the best of our knowledge, been used in a terminally ill population, with the exception of the anxiety components of the MHI in Kane and colleagues' study. In the Medical Outcome Study, the SF-36 was used among chronically ill patients.

vi. Burden (Any information on cost or time to administer? Does the instrument impact on the respondent? How long does a survey take to complete? Response rates? Any problems with missing data?)

Short 5 item questionnaire that is meant to be self-administered in less than 5 minutes.

vii. Alternative forms (What are the modes of administration? Alternatives? If alternative exist, provide what is known for each of the above categories)

The SF-36 has varied the time interval from the past month to the past week (called the acute version). It has been self-administered as well as interviewer administered.

C Profile of Mood State short form and the subscales of Anxiety and Depression (Cella 1987, Cella 1993, Lynn 1996)

The POMS was originally developed from the Total Mood Disturbance Score. This version consists of 11 items from the 65 item POMS. The intended measurement is ordinal, but it has been analyzed using parametric statistics. In doing this adaptation, the authors put forward the argument that psychological distress is heirical with the majority of the information coming from one construct (which they call a mood disturbance score). In a population of patients with cancer, they found a mean score of 10.2 (SD 8.72).

Two subscales (anxiety and depression) were utilized in the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment. This was a 14 item scale for which factor analysis confirmed two separate constructs of anxiety and depression, as suggested by Cella and colleagues (1993). This instrument was utilized in interviews with patients with a serious illnesses and with family members to report on patient emotions. This instrument was administered about one month after the patients' deaths (Lynn 1996).

ii. Reliability (Did they address internal consistency? Did they address reproducibility?)

Internal consistency has been found to be high (.91) among a population of cancer patients and test-retest agreement was good. The use of the 14 item scale in SUPPORT showed similarly high Crohnbach's alpha reliability (.91-.95). In this study, we examined the correlation between patient and surrogate response, which was .43 for the entire scale, .39 for the anxiety scale and .44 for the depression scale. Missing data was not a problem.

iii. Validity - (How did they address content related validity? Any information on construct related validity? Any information on criterion validity?)

Use of the 11 item has been shown to be associated with the extent of disease and pain. The full POMS has been found to be sensitive to psychotherapy and drug effects.

iv. Responsiveness (Any information? Has the scale ever been used as an outcome measure? If so, what populations?)

The full POMS was utilized in the National Hospice Study, and missing data was a problem. Only 60% of patients were able to respond for themselves at the time of study entry and this number decreased to 44% at the seventh interview conducted 14 weeks later (Mor 1986).

v. Interpretability (What populations has it been applied too? Is the score translated into a clinically relevant event? Does the score predict outcome events?)

The POMS has been utilized in the seriously ill (Lynn 1996), patients with cancer (Cella 1987), and terminally ill patients (Mor 1986 ).

vi. Burden (Any information on cost or time to administer? Does the instrument impact on the respondent? How long does a survey take to complete? Response rates? Any problems with missing data?)

Short 10 item questionnaire that is interviewer administered in less than 10 minutes.

vii. Alternative forms (What are the modes of administration? Alternatives? If alternative exist, provide what is known for each of the above categories)

This is a multidimensional scale which assesses multiple symptoms common in patients with cancer. The MSAS measures the prevalence, severity, and distress of 23 symptoms. A factor analysis yielded two constructs that distinguished 3 major symptoms groups: the psychological symptoms, high prevalence physical symptoms and low prevalence physical symptoms. The intended measurement is ordinal.

ii. Reliability (Did they address internal consistency? Did they address reproducibility?)

Internal consistency has been found to be high in the first two major symptom groups (psychological symptoms and high prevalence symptoms, with an alpha of .88 and .83). It was low in the low prevalence symptoms. There are no test-retest reports on this self-administered scale.

iii. Validity - (How did they address content related validity? Any information on construct related validity? Any information on criterion validity?)

The scale was developed by team members of the Pain Service at Memorial Sloan Kettering Hospital. They tested the construct validity with a number of measures, including the Symptom Distress Scale, Rand Mental Health Inventory, Functional Living Index-Cancer, and Karnofsky Performance Scale. Canonical correlation analyses suggested that both the distress measure and the frequency or severity measure added important information to the measure of quality of life. Correlations are shown in the enclosed article with each of these measures.

iii. Responsiveness (Any information? Has the scale ever been used as an outcome measure? If so, what populations?)

iv. Interpretability (What populations has it been applied to? Is the score translated into a clinically relevant event? Does the score predict outcome events?)

From the clinical standpoint, the reports of frequency, severity, and distress have a firm basis in clinical practice. It has strong face validity. Measurements of frequency and severity of symptoms have been widely used in the measurement of pain and other symptoms. As noted above, the score is associated with multiple measures of quality of life.

v. Burden (Any information on cost or time to administer? Does the instrument impact on the respondent? How long does a survey take to complete? Response rates? Any problems with missing data?)

A self-administered questionnaire that asks about symptoms in the past week and takes less than 15 minutes to complete.

vi. Alternative forms (What are the modes of administration? Alternatives? If alternative exist, provide what is known for each of the above categories)

Authors have created a short form which they call the Global Distress Index. Additionally, they are developing measures specific for certain diagnoses (e.g. prostrate cancer), and a pediatric population.

The four highlighted instruments are each worthy of consideration. Each has important strengths that would justify its recommendation. An important reason for recommending the MSAS is that the instrument assesses both physical and emotional symptoms. Brevity is an important guiding principle for this recommendation, but not at the cost of reliability and validity. Of the instruments reviewed, the MSAS measures the frequency, severity, and distress of each symptom, which has strong face validity from a clinical standpoint. The emotional symptoms mainly focus on psychological symptoms. One concern that I had with the MHI 5 was how well the question "How many times, in the past month, have you been a happy person" would be accepted by a terminally ill population. The MSAS shows adequate internal consistency and construct validity with a number of measures of health related quality of life.

The SUPPORT study administered a multidimensional instrument which examined the frequency, severity, and satisfaction with treatment of specific symptoms, and found that similarly designed questions can easily be asked among patients with nine serious illnesses (about one in two died in six months of follow-up). Additionally, this instrument was completed by both patients and surrogates with a moderate correlation in responses (about .40). This instrument was not recommended because of concerns over the validity of measuring satisfaction with symptom control (i.e., patient reported severe pain, yet they were satisfied with pain control).

There are several drawbacks to this recommendation. First, the scale was designed for measuring symptoms in patients with cancer and thus has only been tested in patients with cancer (i.e., some symptoms may have such a low prevalence that it is not worth the time burden for other diseases). However, the measure of frequency and severity of symptoms has been done on varying populations. Second, the proposed measure has not been utilized as an outcome measure. The only three scales which have been used as outcome measures in dying persons are the POMS, CECS-D, and Mental Health Inventory. Portenoy and colleagues note that one of the main purposes of the MSAS is for documenting the frequency and severity of symptoms. My vision for the "Toolkit" is that it acts as a screening instrument that provides an overall picture of the dying experience. If a health care institution identifies a particular problem, I would strongly recommend that they consider a longer instrument to focus on particular symptoms. There are plans to assess the MSAS as an outcome variable and examine serial measurements over time.

It is my recommendation that a patient instrument incorporate the MSAS and that consideration being given to modifying the instrument (with permission) for administration to a close family member after the patient's death. Despite these concerns, my recommendation is that this instrument be considered for prospective study of patients and interview conducted after the patients death. I have chosen not to recommend a chart based instrument.

The two instruments that I would recommend for studying emotional distress would be the POMS or the Mental Health Inventory. Each have been used in evaluation of intervention of palliative care interventions.

i. We need a better understanding of the use of proxies. This is important in that each of the studies of dying persons have found that about one in three persons can not be interviewed in the last week of life. A prospective study of patient, proxy, and caregiver assessment of symptoms is needed to assess the validity of proxy reports. Geriatric and palliative care medicine aim not only to treat a patient, but also to provide compassionate care for the patient and their family (WHO 1989).

iii. A better understanding is needed regarding what is the "normal" amount of emotional distress. Consideration should be given to adaptation of instruments to the degree in which emotional symptoms impact on social role function (similar to the SF-36).

iv. The relationship of scale scores to institutional quality problems needs to be assessed.

Anderson NK, Ahmedzai S, Bullinger M et al.The European Organization for Research and Treatment of Cancer QOL-C30: a quality of life instrument for use in international trials in oncology. J Natl Cancer Inst 1993;85:365-376.

Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561-571.

Bedford A, Foulds GA, Sheffield BF. A new personal disturbance scale (DSSI/sAD). Br J Soc Clin Psychol 1976;15:387-394.

Berwick DM, Murphy JM, Goldman PA, et al. Performance of a five-item mental health screening test. Med Care 1991;29:169-176.

Bruera E, Kuehn N, Miller M, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J of Palliative Care 1991;7:6-9.

Butters E, Higginson I, George R, Smits A, et al. Assessing the symptoms, anxiety and practical needs of HIV/AIDS patients receiving palliative care Qual Life Research 1992;1:47-51.

Cella DF, Jacobson PB, Orlav EJ, Holland JC, et al. A brief POMS measure of distress for cancer patients. J Chron Dis 1987; 40:939-942.

Cella DF, Tulsky DS, Gray, G et al. The functional assessment of cancer therapy scale: development and validation of the general measure. J Clin Oncology 1993;11:570-579.

de Haes JC, van Kippenberg FC, Neijt JP. Measuring psychological and physical symptom distress in cancer patients: structure and application of the Rotterdam Symptom Checklist. Br J Cancer 1990;62:1034-1038.

Goldberg DP. The detection of psychiatric illness by questionnaire. Oxford U Press: London 1972.

Guyatt GH, Berman LB, Townsend M, Pugsley SO, et al. A measure for the quality of life for clinical trials in chronic lung disease. Thorax 1987;42:773-778.

Lynn J, Teno JM, Philips RS, Wu AW. Dying experience of older and seriously ill patients: findings from the SUPPORT and HELP projects. Provisionally accepted, Annals of Internal Medicine 1996.

McCorkle R. Development of a symptom distress scale. Cancer Nursing 1978:373-378.

Mor V. Assessing patient outcomes in hospice: what to measure? In: Psychosocial Assessment in Terminal Care, Dush DM, Cassileth BR, Turk DC, eds. Haworth: New York, 1986.

Nelson EC, Wasson JH, Johnson DJ, Hays RD. Dartmouth COOP functional health assessment charts: brief measures for clinical practice. In: Spilker, B, ed. Quality of Life and Pharmacoeconomics in Clinical Trials, second edition. Lippincon-Raven Publishers: Philadelphia, 1996.

Portenoy RK, Thaler HT, Kornblith AB, Lepore JM et al. The memorial symptom assessment scale: an instrument for the evaluation of symptom prevalence, characteristics and distress. Eur J of Cancer 1994;30A:1326-1336.

Radloff, SF. The CES-D Scale: A self-report depression scale for research in the general population. Appl Psychol Measurement 1977;1:385-401.

Spitzer WO, Dobson AJ, Hall J, et al. Measuring the quality of life of cancer patients: a concise QL-index for use by physicians. J Chron Disease 1981;34:585-597.

Ware JE Jr, Johnston SA, Davies-Avery A, at al. Conceptualization and measurement of health for adults in the Health Insurance Study. Vol. III. Mental Health. Santa Monica, CA: Rand Corporation, 1979.

World Health Organization: Cancer Pain Relief and Palliative Care. Geneva: WHO, 1989.

Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression scale: a preliminary report. J Psychiatr Res 1983;17:37-43.

Zigmond AS, Snaith, RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983;67:361-370.

Back to Toolkit Table of Contents

Instruments to Assess Functional Status

EMOTIONAL SYMPTOMS