 |
 |
|
|
|
Assessment in palliative care is a complex process requiring not only a multidimensional perspective that covers a range of physical symptoms but also careful consideration of individuals’ existential and social concerns and the state of their mental health. While remaining broadly based, however, palliative care assessments must respect the limits imposed by advanced medical illness on patients’ endurance and concentration. Although several questionnaires have been developed to assess symptoms (1–3) or health-related quality of life (4,5), self-administered scales may have limited applicability for some older adults who are terminally ill. In the clinical setting, it is more common for assessments to be structured around simple measures such as visual analog scales (VAS) (6–8). VAS measures can be administered quickly, repeated as required, and tend to have good statistical properties. Nevertheless, this approach also has disadvantages. One is that some patients have difficulty “grasping the metaphor of the continuum represented by the VAS” (9). In addition, VAS measures and other assessment methods, such as numerical rating scales, can be criticized for a lack of intuitive meaning. For example, it is difficult to know what it means to have “5 cm of pain,” and the effort to identify ranges of scores that have more direct intuitive or functional significance remains an active area of research (10,11). These measures may also provide too global an assessment for some important clinical purposes. In the case of depression, VAS scores are comparatively nonspecific for screening terminally ill patients who have clinical depressive disorders identified on the basis of conventional diagnostic criteria (12). Given these constraints, we considered that an alternative approach based on the methodology of clinician-administered structured interviews could be a worthwhile addition to research in palliative care. Structured interviews are associated most closely with psychiatric research, where they have become the criterion standard for reviewing the defining symptoms of mental disorders (13,14). In this context, structured interviews enhance the reliability of clinical assessments and serve a research function by standardizing diagnostic practices across studies. As with mental health problems, assessments in palliative care rely largely on patient reports of subjective symptoms and concerns, which suggests that the rationale for structured interview assessment is similarly applicable. Moreover, structured interview assessments have been adopted as an approach to single-item screening in studies of such constructs as depression, dignity, and desire for death among people receiving palliative care services (12,15,16), despite the fact that little is known about their reliability, temporal stability, or correlation with other assessment methods. In this study, we undertook a preliminary validation of a structured interview addressing physical and psychosocial issues common in palliative care. MethodsItem Selection Initially, we reviewed studies that reported the prevalence of physical symptoms within populations having advanced illness (18–20). We considered it important to include highly frequent discomforting physical symptoms, and we selected the specific problems of pain, dyspnea, drowsiness, nausea, and weakness. We drew psychosocial issues from 2 sources. First, as noted, we reviewed the growing literature regarding end-of-life concerns associated with patient requests for euthanasia or physician-assisted suicide. This review included studies of medical patients who had been surveyed about their attitudes toward these practices (21,22), as well as studies of physicians in the Netherlands who had participated in them (23). The specific items included sense of personal control (defined as being able to tolerate one’s situation and cope adequately), sense of dignity, perception of being a burden to others, hopelessness, and desire for death (15). Second, we drew information from studies and reviews of mental disorders in primary care (24) or palliative care populations (15,25–27). It was evident from this literature that anxiety and depressive disorders are among the most prevalent mental health problems in the palliative care setting. We believed that including items addressing subjective anxiety, depressed mood, and loss of interest or pleasure in activities would allow us to use the interview to screen for the central criterion symptoms of several common anxiety and depressive disorders described in the DSM-IV (28). Item Construction Finally, we constructed the severity thresholds for the mental health screening questions so that they could be related directly to the DSM-IV criteria for the diagnosis of anxiety and depressive disorders. This process resulted in a 13-item Structured Interview for Symptoms and Concerns (SISC). The next step in refining the measure was to circulate the items in draft form to palliative care physicians and nurses and to solicit their comments about item content, wording, and scaling. An early version was also pilot-tested with a group of 10 participants who were inpatients at a palliative care unit. After revision based on this feedback, we considered the interview ready for use. Table 1 shows examples of representative items.
Participants and Procedure The study recruited 70 participants, all of whom were receiving palliative care for advanced cancer. However, 1 participant did not complete all aspects of the protocol, including the SISC, which resulted in a group of 69 participants for the validation study. Forty-six participants (66.7%) were inpatients at a regional palliative care unit, and 23 (33.3%) were patients in the oncology services of an acute care hospital who received palliative care consultations during their admissions. Whenever possible, participants completed 2 interviews separated by 1 to 3 days. The interviews were administered by a clinical psychologist, by doctoral students in psychology, or by research associates in palliative care. The first interview was attended by both a primary interviewer and an observer; these individuals made independent ratings that were used to determine interrater reliability. The initial interview involved an extended protocol that took an average of 63.8 minutes to complete (range 25 to 145 minutes); it was broken into shorter sessions when necessary. It comprised an introduction to the study, the signing of the consent forms, a review of demographic information, an attitude survey, administration of the SISC, and a debriefing. Along with each SISC item, participants also completed a 10-cm VAS addressing the same symptom or concern. When responding to specific SISC and VAS items, participants were asked to answer with respect to their experience “over the past day.” To conform to DSM-IV standards, longer time frames were also reviewed to assess specific core criterion symptoms of anxiety and depressive disorders, that is, anxiety, depression, and loss of interest or pleasure in activities. Whenever a participant reported one of these symptoms at a severity level and duration sufficient to indicate the presence of a specific anxiety or depressive disorder, we used the evaluation guide of the Primary Care Evaluation of Mental Disorders (PRIME- MD) (24) to probe additional criteria. The PRIME-MD is a brief diagnostic assessment tool that has been validated for use in primary care and is well-suited to the palliative care setting. An independent interviewer with no knowledge of the participant scores from the first session conducted the second assessment. To determine the test–retest reliability of the items, only the SISC items and corresponding VAS measures were administered at this interview. A relatively short retest interval was used because patient situations can quickly change in palliative care. The second interview took an average of 28.6 minutes (range 10 to 60 minutes). Statistical Analysis Although most SISC items were positively skewed (indicating a higher frequency of low scores), the results of the statistical tests were comparable when either parametric or nonparametric analyses were used. Therefore, we report the parametric tests. ResultsParticipant Characteristics VAS measures were available for only 68 participants at the first interview because 1 individual had a visual impairment that precluded completing the scales. In addition, second interviews were available for only 46 participants: for the rest, an interview could not be scheduled within the prescribed time frame (n = 12), the participant declined to take part in a follow-up (n = 7), or the participant died or deteriorated medically (n = 4). Of the participants who took part in the second interview, 3 completed the SISC but declined or were unable to fill out the VAS assessment. Prevalence of Symptoms and Concerns
Reliability and Correlation With Visual Analog Scales
Both the SISC items and the VAS measures showed moderate temporal stability over the 1- to 3-day follow-up interval. For the SISC items, retest reliability coefficients ranged from r = 0.50 to r = 0.90, whereas the VAS coefficients ranged from r = 0.42 to r= 0.91. All the test–retest coefficients were statistically significant for both methods (P < 0.006). Further, comparisons of the reliability coefficients across methods based on Fisher’s z transformations showed that the SISC items and VAS measures had equivalent stability over time. Prevalence of Anxiety and Depressive Disorders
To probe the SISC items’ sensitivity to individual differences, we conducted a series of exploratory analyses that compared subgroups of participants who either did (n = 16) or did not (n = 53) meet DSM-IV diagnostic criteria for any anxiety or depressive disorder. Table 5 summarizes the results. These analyses revealed that participants diagnosed with a mental disorder differed from the other respondents in acknowledging more severe problems on 10 of the 13 interview items, including the physical symptoms of drowsiness (P = 0.01) and weakness (P = 0.02), as well as all psychosocial and mental health concerns (all P < 0.03). The VAS measures showed the same pattern, with the exception that the specific symptom of weakness did not differ significantly between the groups (P > 0.10).
Participant Preference DiscussionThe SISC was developed as a first-generation structured interview for assessing common symptoms and concerns of patients approaching the end of life. The interview is similar to a bedside clinical assessment and does not require that the patient be able to read or follow complex instructions. However, people who are interviewed must be mentally competent and able to reflect on their circumstances, which does restrict its applicability in palliative care. In addition, the typical, unhurried administration of an interview covering all relevant areas takes at least 10 minutes and can range up to an hour. We selected the 13 items comprising this initial version of the SISC because they represent physical symptoms that are highly prevalent in the palliative care setting or because they reflect important end-of-life concerns related to the maintenance of dignity and emotional and social well-being. These are central considerations in the practice of palliative care. Nevertheless, it must be acknowledged that many different problems can arise in terminally ill individuals, with implications for their quality of life. In this study, we strove to balance a comprehensive but long interview with one that was shorter but restricted in content. Therefore, the items do not completely cover all possible symptoms and concerns; however, the assessment format can easily be extended to cover other problems. In ongoing research (31), we have expanded the interview to be compatible with other conceptualizations of quality of life in the context of palliative care (32). The psychometric characteristics of the structured interview items indicate that they have excellent interrater reliability, show good concordance with respondent VAS measures of the same constructs, are sensitive to individual differences, and are regarded by participants as being at least as acceptable as the VAS method. For the most part, the SISC items exhibit moderate temporal stability that is similar to VAS measures. However, it is noteworthy that, whereas the patient completed the VAS measures on both occasions, a different individual, who was unaware of the previous interview results, completed the second SISC assessment. This suggests that little of the variability in SISC scores is attributable to interrater differences and that the moderate temporal stability shown by both the SISC and VAS items reflects variability inherent in the underlying constructs or other sources of error (33,34). The SISC assessment format offers several advantages over the VAS measures. First, the scale ratings were constructed to be clinically relevant: the verbal labels and functional descriptors offer an intuitively meaningful way of understanding and communicating information about symptoms. In this way, patients who may have a symptom or concern at a minimal and nondistressing level can be distinguished from those for whom a symptom is a more pervasive issue requiring further review and intervention, if possible. A second advantage of the SISC format is that the the mental health item ratings can serve as initial screens for specific anxiety and depressive disorders. We found that 23% of the participants fulfilled diagnostic criteria for at least 1 of these disorders, which is within the range of prevalence estimates reported by other investigators (15,25–27). We also found a high degree of comorbidity between anxiety and depressive disorders. The diagnoses were of varying severity, but the most common problem was major depression. In the primary care setting, mental disorders are associated with a substantial decline in health-related quality of life (33). This appears to be true of palliative care as well. In the present study, participants diagnosed with a mental disorder also experienced more severe physical symptoms in some areas. Further, they perceived themselves to be less in control of their circumstances, more burdensome to their families, and engaged in a greater struggle for the preservation of dignity. They also expressed a greater desire for death. Clearly, the emergence of mental disorders introduces a significant source of suffering at the end of life, and recommendations for routine screening with simple measures are well founded (12,35). The structured interview approach is particularly well suited to this purpose, although the core screening items must be supplemented with additional measures to review all relevant criteria for the disorders in question. In summary, the structured interview approach provides a reliable and valid assessment that incorporates clinically relevant and intuitively meaningful rating scales. It offers a structured format for conducting an interview into some of the most common symptoms and concerns arising at the end of life and may be more generally applicable for palliative care research. Funding and SupportThis study was supported by funding from the National Health Research and Development Program of Health Canada and by a Career Scientist Award to Ian Graham from the Ontario Ministry of Health. AcknowledgementsWe thank John Scott, who was involved in the development of the survey questionnaire, and Dorothyann Curran, who helped with the data analysis. We also acknowledge the palliative care physicians and nurses at the participating institutions. Finally, we thank those individuals who participated in the research for sharing their time so generously. References1. Noyes R, Kathol RG, Debelius-Enemark P, Williams J, Mutgi A, Suelzer MT, and others. Distress associated with cancer as measured by the Illness Distress Scale. Psychosomatics 1990;31:321–30. 2. Hollen PJ, Gralla RJ, Kris MG, Cox C, Belani CP, Grunberg SM, and others. Measurement of quality of life in patients with lung cancer in multicenter trials of new therapies. Psychometric assessment of the Lung Cancer Symptom Scale. Cancer 1994;73:2087–98. 3. Portenoy RK, Thaler HT, Kornblith AB, Lepore JM, Friedlander-Klar H, Kiyasu E, and others. The Memorial Symptom Assessment Scale:an instrument for the evaluation of symptom prevalence, characteristics and distress. Eur J Cancer 1994;30A:1326–36. 4. Hollen PJ, Gralla RJ. Comparison of instruments for measuring quality of life in patients with lung cancer. Semin Oncol 1996;23:31–40. 5. Cohen SR, Mount BM, Bruera E, Provost M, Rowe J, Tong K. Validity of the McGill Quality of Life Questionnaire in the palliative care setting: a multi-centre Canadian study demonstrating the importance of the existential domain. Palliat Med 1997;11:3–20. 6. Ahles TA, Ruckdeschel JC, Blanchard EB. Cancer-related pain. II. Assessment with visual analog scales. J Psychosom Res 1984;28:121–4. 7. Bruera E, Kuehn N, Miller MJ, Selmser P, Macmillan K. The Edmonton Symptom Assessment System (ESAS): a simple method for the assessment of palliative care patients. J Palliat Care 1991;7:6–9. 8. Chochinov HM, Tataryn D, Clinch JJ, Dudgeon D. Will to live in the terminally ill. Lancet 1999;354:816–9. 9. McDowell I, Newell C. Measuring health: guide to rating scales and questionnaires. 2nd ed. New York: Oxford University Press; 1996. 10. Serlin RC, Mendoza TR, Nakamura Y, Edwards KR, Cleeland CS. When is cancer pain mild, moderate or severe? Grading pain severity by its interference with function. Pain 1995;61:277–84. 11. Collins SL, Moore RA, McQuay HJ. The visual analogue pain intensity scale: what is moderate pain in millimetres? Pain 1997;72:95–7. 12. Chochinov HM, Wilson KG, Enns M, Lander S. Are you depressed? Screening for depression in the terminally ill. Am J Psychiatry 1997;154:674–6. 13. Endicott J, Spitzer RL. A diagnostic interview: the Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry 1978;35:837–44. 14. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for DSM-IV Axis I Disorders–Patient Edition. (SCID-I/P, Version 2.0, 8/98 revision). New York: Biometrics Research Department, New York State Psychiatric Institute; 1998. 15. Chochinov HM, Wilson KG, Enns M, Mowchun N, Lander S, Levitt M, and others. Desire for death in the terminally ill. Am J Psychiatry 1995;152:1185–91. 16. Chochinov HM, Hack T, Hassard T, Kristjanson LJ, McClement S, Harlos M. Dignity in the terminally ill: a cross-sectional cohort study. Lancet 2002;360:2026–30. 17. Wilson KG, Scott JF, Graham ID, Kozak JF, Chater S, Viola RA, and others. Attitudes of terminally ill patients toward euthanasia and physician-assisted suicide. Arch Intern Med 2000;160:2454–60. 18. Coyle N, Adelhardt J, Foley KM, Portenoy RK. Character of terminal illness in the advanced cancer patient: pain and other symptoms in the last four weeks of life. J Pain Symptom Manage 1990;5:83–93. 19. Ventafridda Y, Ripamonti C, DeConno F, Tamburini M, Cassileth BR. Symptom prevalence and control during cancer patients’ last days of life. J Palliat Care 1990;6:7–11. 20. Curtis EB, Krech R, Walsh TD. Common symptoms in patients with advanced cancer. J Palliat Care 1991;7:25–9. 21. Emanuel EJ, Fairclough DL, Daniels ER, Clarridge BR. Euthanasia and physician-assisted suicide: attitudes and experiences of oncology patients, oncologists, and the public. Lancet 1996;347:1805–10. 22. Ganzini L, Johnston WS, McFarland BH, Tolle SW, Lee MA. Attitudes of patients with amyotrophic lateral sclerosis and their care givers toward assisted suicide. N Engl J Med 1998;339:967–73. 23. Van der Maas PJ, van Delden JJM, Pijnenborg L, Looman CWN. Euthanasia and other medical decisions concerning the end of life. L | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||