|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology Depression in Patients With Lung Cancer: Prevalence and Risk Factors Derived From Quality-of-Life DataFrom The Cancer Research Campaign Psychological Medicine Group, Christie Hospital National Health Service Trust, Withington, Manchester; Cancer Division, Medical Research Council Clinical Trials Unit, London, United Kingdom. Address reprint requests to P. Hopwood, MD, Cancer Research Campaign Psychological Medicine Group, Stanley House, Christie Hospital NHS Trust, Withington, Manchester, M20 4BX, UK; email penelope.hopwood@ man.ac.uk.
PURPOSE: To evaluate self-reported depression rates in patients with inoperable lung cancer and to explore demographic, clinical, and quality-of-life (QOL) factors associated with depression and thus identify patients at risk. PATIENTS AND METHODS: Nine hundred eighty-seven patients from three palliative treatment trials conducted by the Medical Research Council Lung Cancer Working Party formed the study sample. 526 patients (53%) had poor prognosis small-cell lung cancer (SCLC) and 461 patients (47%) had good prognosis non–small-cell lung cancer (NSCLC). Hospital Anxiety and Depression Scale data and QOL items from the Rotterdam Symptom Checklist were analyzed, together with relevant demographic and clinical factors.
RESULTS: Depression was self-rated in 322 patients (33%) before treatment and persisted in more than 50% of patients. SCLC patients had a three-fold greater prevalence of case depression than those with NSCLC (25% v 9%; P < .0001). An increased rate for women was found for good performance status (PS) patients (PS of 0 or 1) but the sex difference reduced for poor PS patients (PS of 3 or 4) because of increased depression rates for men ( CONCLUSION: Depression is common and persistent in lung cancer patients, especially those with more severe symptoms or functional limitations. Psychologic screening and appropriate intervention is an essential part of palliative care.
IMPORTANT ASPECTS of a patient’s well-being may be adversely affected by the diagnosis of cancer or its treatment, and so a fundamental aim of palliative care is to relieve distressing symptoms in a broad range of domains.1,2 Yet in the field of cancer, reports of palliative clinical trials are often limited to the impact of treatment on survival, toxicity, or physical symptoms, rather than psychologic problems such as depression. Mood disorder may obviously be part of a reaction to the news of diagnosis,3 but in many patients it will persist, causing an added burden during treatment and leading to more difficulty with general management and symptom control,2,4 increased length of hospital stay,5 decreased compliance with treatment,6 and, possibly, reduced survival.7 Depressed mood is also an important predictor of quality of life (QOL), showing a stronger relationship than fatigue.8 Moreover, the presence of depression or anxiety in the patient significantly increases the risk of emotional problems in the partner,9,10 thus compounding distress in the family. Although there has been consistent emphasis on the importance of emotional functioning as an essential QOL domain, it is well documented that doctors and nurses fail to detect patients’ emotional distress and patients themselves do not disclose it unless asked.11-14 In a recent study of more than 1,000 cancer patients, physicians and patients were more likely to agree on the absence of depression, and there was a marked underestimation of existing depression, particularly when it was severe. Doctors were primarily influenced by patients’ outward signs of distress or anxiety, which may be misleading, and seriously underestimated depression in patients with pain or poor performance status.13 A direct consequence of the underrecognition of depression is its undertreatment, particularly when mood disorder can be deemed understandable.2 Studies of psychologic distress in patients with advanced disease have reported a prevalence of affective disorder varying from 23%15 to 47%,16 but research has usually been carried out in heterogeneous patient samples and the effect of disease type and symptom pattern has not been evident. The implications for patients with a short expected survival time (such as with lung cancer) as compared with those who may live several years (such as patients with breast cancer) are not known. The highest frequency of depression has been found in patients with more advanced stages of malignancy, more severe illness, and poorer physical status.17,18 In a study of advanced cancer patients who were about to receive palliative radiotherapy, Kaasa et al18 reported a high level of psychologic distress: 69% patients scored above the threshold value (for the normal range) on the 20-item General Health Questionnaire, a frequently used measure for psychologic morbidity. In a more recent evaluation using the Hospital Anxiety and Depression Scale19 (HADS), Aass et al20 found a prevalence of 20% for borderline and case depression and 31% for borderline and case anxiety in a survey of 716 patients at the Norwegian Radium Institute; depression and anxiety rates in a small subgroup of lung cancer patients were 26% and 23%, respectively. There is therefore good general evidence that depression is a significant problem in advanced cancer patients. Despite the large numbers of patients with lung cancer who receive palliative treatment, there is little published data concerning the prevalence of psychologic distress in this group, and hence little attention is paid to the need to provide psychologic support. QOL data contain important information about emotional distress, but such data are usually collected to make comparisons between treatment regimens. Using data from the Rotterdam Symptom Checklist21 (RSCL), the authors reported the number and diversity of symptoms in patients at presentation for treatment for lung cancer22 and highlighted the fact that psychologic symptoms were among the most prevalent items reported by patients. These large data sets also provide an invaluable opportunity to explore specific domains, and we therefore decided to investigate the prevalence of depression in this disease group. Psychologic scales or subscales lend themselves to this kind of analysis, but they are not sufficiently accurate to generate a diagnosis in the clinical setting. Used alone, they could create an unacceptable workload of additional clinical assessments for routine staff. Other risk factors for depression are therefore needed in this patient group to help identify those patients who may be most in need of such an assessment and subsequent intervention. A number of factors associated with depression have been found in other settings and warrant exploration in lung cancer, to better identify at-risk patients in this disease group. Factors include female sex (which is widely reported in population studies), functional status,23 and symptom burden.23 Thus patients attending general practitioners’ surgeries had a two-fold increased risk of psychiatric comorbidity if they presented with five or more somatic symptoms and a five-fold increased risk of affective disorder if they presented with impaired function.23 In the field of cancer, Dugan et al24 reported that advanced stage at diagnosis, lung cancer as a primary tumor, and higher Eastern Cooperative Oncology Group rating were the variables most associated with clinically significant depressive symptoms in a study of more than 1,000 patients. More specifically, Kurtz et al10 reported that patients’ physical symptoms (and, to a lesser degree, patient immobility) were strong predictors of depression in a heterogeneous group of 150 cancer patients. In advanced breast cancer, a relationship was found between fatigue and depression, but the expected association of depression with pain was not confirmed.25 In contrast, Kaasa et al18 found associations between intense pain and poor performance status with patients’ psychologic distress. The authors did not confirm the expected sex difference but reported that female patients had significantly more intrusive thoughts about their disease. From a study at the same hospital, Aass et al20 again found a lack of association between sex and depression and confirmed also an association between fatigue and depression in cancer patients. However, a causal relationship between these parameters in cancer patients has not been confirmed.8 The association of any of these factors with depression in a population of lung cancer patients has not been explored. We used a large QOL data set to estimate the prevalence and persistence of depression in patients with lung cancer and explored the association with histologic cell type, age, sex, functional status, symptom burden, and individual symptoms as suggested by previous work to better identify patients at risk. The prevalence of anxiety was estimated for comparative purposes.
Sources of Data Data from three multicenter randomized trials conducted by the Medical Research Council Lung Cancer Working Party were used. The trials were as follows:
2. LU16 was a comparison of oral etoposide versus standard intravenous multidrug chemotherapy for SCLC.27 Three hundred seventy patients with a median age of 67 years (range, 35 to 83 years) were entered from 38 centers in the United Kingdom. Sixty-three percent of patients were male and on entry, 57% had extensive disease and 97% patients had a WHO PS of 2 or 3. 3. LU13 was a comparison of palliative two-fraction radiotherapy versus more intensive 13-fraction radiotherapy for patients with inoperable non-small-cell lung cancer (NSCLC) and good performance status.28 A total of 509 patients with a median age of 66 years (range, 33 to 89 years) were entered from 11 centers in the United Kingdom. Seventy-nine percent were male, 79% had squamous histology, and on entry, 78% had a WHO PS of 0 or 1. Each trial involved integral QOL assessments as part of the trial protocol. Symptoms of depression and anxiety were assessed using the HADS.19 This is a 14-item self-report measurement tool designed for use in medical outpatient settings to assess depression and anxiety. Each subscale is scored from 0 to 21, with higher scores indicating greater distress. QOL parameters were assessed using the RSCL,21 which comprises subscales for physical and psychologic functioning and activities of daily living (ADL). Additional items were added to reflect key symptoms of lung cancer and its treatment. Details of QOL administration and further QOL analyses for these trials are presented in the trial reports.26-28
Methods of Analysis Change in depression scores over time. Persistence of depression was determined by comparing patients with complete HADS data at baseline and at first follow-up. New incidence of depression at first follow-up in patients with normal pretreatment scores was also determined. The change in depression over time was tested using Wilcoxon’s signed ranks test. Analyses were carried out for the whole sample and by disease subgroups. Factors associated with depression. The relationship between pretreatment depression scores and concurrent assessments of anxiety, cell-type, age, sex, functional status (WHO PS and patient-rated ADL), selected symptoms, and physical symptom burden were investigated using univariate analyses. ADL items were transformed into a five-point impairment scale (devised by the authors) that gave a more clinically relevant indication of limitation of activity than a simple summary score of this subscale, which was uninterpretable.30 Symptom burden was the number of physical items reported on the RSCL. If responses to more than two items were missing, then the data were not used. If one or two items were not completed, then the number of reported symptoms was calculated as a proportion of those scored. Independent factors associated with depression. Pretreatment factors considered in the univariate analyses above were entered into a stepwise logistic regression analysis, using SPSS31 to identify factors or variables that exhibit statistically significant, independent associations with the dependent variable of interest (in this case, depression). In view of the large number of analyses carried out, only those factors with associations at a significance level of less than .01 are reported.
Compliance with the completion of pretreatment HADS and RSCL forms was defined as any form completed on or before the first day of any anticancer treatment. Sixty-eight percent of forms were completed on the day of start of treatment, 29% within 1 week, and only 3% more than 1 week from the start of treatment (maximum, 28 days). The proportion of assessable patients with complete pretreatment HADS depression items and those with both baseline and follow-up forms are listed in Table 1. The number of HADS forms with missing items (20, or 2%) was trivial compared with missing data due to missing forms, and so no missing items were imputed.
Compliance with completion of RSCL forms was similar to that of the HADS, as the forms were handed out together. However, of the 987 patients with baseline HADS, 72 patients (7%) who were from trial LU16 completed the European Organization for Research and Treatment of Cancer QLQ-C3032 as part of a randomized comparison of QOL instruments within the trial. An additional 32 patients (3%) had more than two missing items (the maximum for imputation being two) leaving 883 RSCL forms (90%) available. Missing items were a source of missing data on 230 (26%) of the RSCL forms. Missing items were therefore imputed in 198 forms (22%) for which either one (163 forms) or two (35 forms) items were missing, in order to maximize the number of patients with assessable data for the analysis. Thirty-two forms (4%) had more than two missing items and were not considered assessable.
Prevalence of Depression and Anxiety
Thirty-four percent of patients (334 of 974) self-reported anxiety, of whom 18% (171) were at the level of probable case and 17% (163) were at the level of borderline severity, as listed in Table 2. A significantly higher prevalence in patients with SCLC was again observed (P < .0001). Concurrent depression and anxiety (of case or borderline severity) occurred in 21% of patients (203 of 974) overall: 28% of patients (146 of 517) with SCLC and 12% (57 of 457) of those with NSCLC. The depression and anxiety subscales of the HADS were highly correlated (r = .57) and may mask other associated factors for depression. Therefore, for the purposes of further analysis, only nonpsychologic parameters associated with depression will be explored.
Change in Depression Rates From Pretreatment to First Follow-Up The assessable sample was made up of those patients with complete HADS depression items on both occasions, as listed in the lower section of Table 1. Two hundred sixty-nine patients (27%) had missing data at first follow-up, leaving 718 patients available for an analysis of change. One hundred seventy-four patients with SCLC (33%) and 95 patients with NSCLC (21%) had missing data (P < .0001). The difference was due largely to an excess of poor PS patients in the SCLC group who did not complete follow-up forms. Logistic regression analysis confirmed that poor PS was the most important predictor of noncompliance (odds ratio, 1.35; 95% confidence interval, 1.10 to 1.66; P = .0043). It should be noted that the prevalence of pretreatment depression in this sample of 718 patients was 29% (206 of 718 patients). Patients with only baseline depression scores available had a pretreatment prevalence of 43% (116 of 269 patients). In considering change over time, more than one half of patients with depression at baseline (121 of 206) self-rated again as depressed at first follow-up. An additional 86 of 512 patients who had normal scores at baseline (17%) scored in the depressed or borderline depressed range at first follow-up, giving an overall prevalence of 29% (207 of 718) on this occasion, as listed in Table 3. There was no statistically significant change in depression rates (P = .46, Wilcoxon signed ranks test).
Change in Depression Rates in SCLC and NSCLC Subgroups A very similar pattern of change was found for both the SCLC and NSCLC subgroups, as listed in Tables 4 and 5, with a reported prevalence of depression at first follow-up of 34% and 24%, respectively. This was again made up of more than one half of the pretreatment patients with depression and a similar proportion of patients with normal scores pretreatment who became new cases at follow-up (18% and 16%, respectively) in each disease subgroup.
Equal proportions of SCLC and NSCLC patients self-reported scores indicating worsening of depression (14% of the total patients for each subgroup, but improvement was reported in 21% of all SCLC patients and only 11% of all NSCLC patients.
Factors Associated With Pretreatment Depression Sex. We expected to find an increased prevalence in female patients, in keeping with reported differences in the general population. This was observed for the sample overall (41% prevalence in female v 29% in male patients, P = .0004) and in the NSCLC subgroup (31% female v 19% male patients, P = .007) but not in the SCLC subgroup (46% in female v 41% male patients, P = .26). Clinician-rated WHO PS. The overall proportions of subjects with depression for PS of 0 through 4 were 3%, 19%, 36%, 57%, and 55%, respectively, indicating a very significant trend (P < .0001), as shown in Fig 1. This trend was consistent across disease subgroups. In NSCLC, the results were 4%, 18%, and 41% for PS categories 0, 1, and 2, respectively (patients with PS of 3 or 4 were ineligible for this trial). In SCLC the results were 0%, 30%, 33%, 57%, and 55% for the range of PS categories 0, 1, 2, 3, and 4, respectively.
This is likely to explain the observed difference in prevalence between the SCLC and NSCLC patient groups, as there were more patients with poor PS status in the SCLC subgroup. This difference in PS range resulted from the requirements of the individual trial designs. The results are explored in a secondary analysis (see below, Secondary Analyses).
Patient-reported functional impairment.
Depression rates were compared for patients ranked according to their impairment score derived from the ADL subscale of the RSCL, as shown in Fig 1. There was a significant association between higher depression rates (increased from 12% to 64%) and increased functional impairment ( Symptom burden. In the baseline sample, 883 patients with assessable data reported on average 11.3 physical symptoms pretreatment, as reflected by their ratings on the RSCL. There was a small expected difference between the SCLC (average, 12.0 symptoms) and NSCLC (average, 10.6 symptoms) subgroups, which was probably explained by differences in PS (mean numbers of symptoms according to PS grades 0, 1, 2, 3, and 4 were 8.8, 10.6, 11.6, 12.8, and 14.9, respectively, P < .0001).
It was expected that increasing symptom burden would be associated with an increased prevalence of depression, and this was confirmed. Figure 1 shows a dramatic increase in prevalence with increasing number of symptoms at presentation (P < .0001, Fig 1 shows a strong relationship between increasing prevalence of depression and severity of tiredness (P < .0001), breathlessness (P < .0001), cough (P = .0002), general pain (P < .0001), and chest pain (P = .0006).
Summary of the Univariate Analyses The analyses were repeated by investigating pretreatment factors in 718 patients with depression data at first follow-up, and the results were generally comparable. The association of depression with disease cell type and pain were less significant, but the most robust factors associated at baseline remained highly significant (PS, functional status, symptom burden, tiredness, and breathlessness, P < .0001 for all).
Secondary Analyses
As PS worsened, the prevalence of depression increased and the sex difference almost disappeared. PS therefore had a greater influence on depression ratings than sex. It is worth noting, however, that numerically, there were always more affected male patients than females with depression, this difference being two-fold in the better PS groups.
Factors Independently Associated With Pretreatment Depression The results of the logistic regression analysis are listed in Table 7. The most important factor was patient-rated functional impairment: the risk of depression increased by 41% for each point on the impairment scale. Physical symptom burden was the next most important factor. Tiredness and PS were also independently associated with depression.
The results overall show that one third of patients presenting for treatment for lung cancer report levels of depression that warrant further assessment. Patients who report functional impairment and a high symptom burden (indicated by the overall number of symptoms) or severe breathlessness, cough, pain, or tiredness are at increased risk of depression. Thus 63 of 99patients (64%) with the worst level of functional impairment were depressed, as compared with 35 of 284 (12%) of those with the best reported functional level. When coexisting anxiety was excluded, this was the clearest indicator of the risk of depression. More than one half of patients presenting with self-rated depression on the HADS scale continued to report depressive symptoms at first follow-up, suggesting that this morbidity is not simply a transient reaction to the diagnosis. A further one in six patients who were not depressed at the outset became so within the space of one follow-up visit, and these patients may warrant continued monitoring. Overall, there is no significant improvement in depression rates as treatment gets underway. These results suggest that assessment of psychologic status should be carried out at an early stage of treatment in lung cancer patients.
Depression is not routinely assessed in patients who receive palliative cancer treatment and the majority of published results of palliative treatment trials in lung cancer fail to address patients’ psychologic distress, despite the fact that this has a major impact in determining QOL. We found a concerning level of depressive symptoms, especially in patients with poor PS, which indicates that the psychologic status of lung cancer patients needs to be systematically addressed in their overall management. The HADS is widely used in the cancer field, but it is acknowledged that the prevalence of mood disorder based on self-report questionnaires must be interpreted within the limits of the accuracy of such scales. In a validation study in advanced breast cancer, one of the authors reported a positive predictive power (ie, case finding rate) for the HADS of 50%, meaning that only half of the patients with a high score may suffer clinical depression. Other depressed patients will be missed however, because of false-negative (ie, inappropriately low) questionnaire ratings. Even so, it can be argued that the prevalence of clinically significant rates of depression may be overestimated using self-report scales. Other methodological problems that may affect the prevalence rate reported include the fact that trial patients may not be typical of the population of lung cancer patients and the high attrition rate. We know of no published evidence to suggest that depression rates in lung cancer patients treated outside of clinical trials are different from those of trial patients, but our analysis cannot confirm this. With respect to missing data due to attrition, our data suggest that the prevalence rate is a conservative one, because depression rates increased with poor PS, which itself accounted for missing data. In contrast, however, other groups33 have reported lower prevalence rates using standardized psychiatric diagnostic criteria (Diagnostic and Statistical Manual of Mental Disorders, third edition, revised34) generated by structured interviews and considered these criteria to be too strict. Minagawa et al33 suggested that this resulted in the classification of many patients with considerable emotional distress as not being depressed. These patients would likely remain untreated. The HADS has proved to be an acceptable first-screening tool for depression, because it does not include somatic items that may cause diagnostic confusion, it is brief, and it is easily completed in the hospital setting.35 We have suggested that the greater prevalence of depression in SCLC was largely accounted for by PS grade rather than cell type but we acknowledge that there is a limitation in using trials data, because PS is predefined for inclusion criteria within any trial. However, the effect of cell type was not significantly related to depression in the regression analysis, so we feel it is appropriate to identify the impact of PS as a risk factor. Although depressed mood may be deemed understandable at the start of treatment for cancer, there was little reassurance from our analyses that most of the reported depression was a transient reaction to the diagnosis. There was persistence of high scores in both SCLC and NSCLC groups. Although we did not having information on patients’ emotional status before diagnosis, a study in which patients were observed from their first presentation with chest symptoms reported a doubling of depression, as measured by the HADS, in those with a cancer diagnosis 3 months later.3 In this Scottish study, depression rates were in the normal population range at baseline but increased significantly and were maintained beyond the start of treatment, whereas anxiety levels dropped spontaneously. Commonly held assumptions about older age and female sex being likely to increase the risk of depression were not supported, which is in accordance with other recent studies of the physically ill.3,18,36 It is also important to stress that male patients are numerically more important in this regard, although they are less likely to reveal their emotional distress. The strong relationship between depression scores and impaired functional status was expected in light of the findings of other studies,37 but the cause-and-effect nature of the association remains unclear from the literature. Until this is clarified by further research, it is in the interest of patients to consider both possibilities, as reversible causes of impaired activity may otherwise be overlooked. Thus clinical depression is known to cause agitation or anergia and fatigue, which will add to patients’ functional limitations but may be improved with antidepressant treatment. This would help maintain optimal activity within the limits of patients’ health status. Research in patients with chronic symptoms and airflow limitation has demonstrated a relationship between depression and daily wheeze. Moreover, depression scores had an independent significant effect on QOL as measured by the Sickness Impact Profile.36 Thus lung cancer patients with a history of chronic obstructive airway disease may already be emotionally vulnerable. In the advanced cancer setting, pain is frequently associated with the risk of depression at the expense of considering other physical symptoms. Our data show that other common symptoms such as breathlessness, cough, and tiredness are equally important, and clinicians should be more aware of this. Doctors frequently underestimate such symptoms, making the routine use of patient self-assessment mandatory. Because our data were based on assessments within clinical trials, we did not have information on past psychiatric history or social support, which may also be important predictors of depression.19,37,38 Such information can be readily obtained in the context of a brief psychologic assessment. Other risk factors for a poor adjustment that have been identified by psychosocial research also warrant evaluation in this setting, for example, the number of unresolved concerns expressed by patients.39 This would help counteract the limitations of self-report scales in case detection. Alternatively, a higher threshold score can be used, increasing diagnostic certainty, when resources for assessment are limited. The focus of these analyses was on pretreatment indicators of depression; given the short survival time of advanced lung cancer patients, it is important that the well-being of the patient is assessed globally from the start of treatment and that assumptions are not made that the patient’s mood is an understandable reaction that will improve as soon as treatment gets underway. The assessment process itself may be valuable to the patient, in feeling that relevant concerns are being explored and interest is being shown in the patient and not just in the disease. The relationship between depression and global QOL was not explored in this study and warrants further research: it is not known whether identification and treatment of depression would result in changes in global QOL scores. Future studies to explore this need to be based on a clinical diagnosis of depression, given the limited accuracy of self-report measures. The assessment of depression in such a physically ill group is acknowledged to be difficult, but good guidelines now exist.40 We support the advice for clinical purposes to be inclusive rather than exclusive of symptoms that may have multiple causes, such as anorexia, to best protect the patient from risk of underdiagnosis of depression. It is often claimed that the poor prognosis of lung cancer is a disincentive to treat depression, because patients may not survive long enough to benefit from antidepressant therapy. Such an attitude is antagonistic to the ethos of managing the whole patient instead of just the disease and is to be discouraged. The value to the patient of having mood disorder identified, discussed, and of receiving appropriate support and advice about intervention is considerable. Treatment of depression can improve the patient’s sleep, appetite, concentration, interest, enjoyment, and even tolerance of treatment. Moreover, there is a very real possibility of preventing comorbidity in partners and spouses if active intervention is used. The treatment of depression in cancer patients is now well described in the literature.14,40-42 We have shown that clinicians can use information that is readily available to them to alert them to depression and thereby aid the provision of comprehensive palliative care.
Supported by the Cancer Research Campaign (P.H.) and Medical Research Council (R.J.S.). We thank members of the Medical Research Council Lung Cancer Working Party (A.J. Bailey, N.M. Bleehen, J.J. Bolger, P.I. Clark, C.K. Connolly, D.J. Girling, P.S. Hasleton, P. Hopwood, F.R. Macbeth, D. Machin, K. Moghissi, M.I. Saunders, R.J. Stephens, N. Thatcher [chairman], and R.J. White), in addition to all of the clinicians who entered patients onto the three trials whose data were used for these analyses. We also thank the Cancer Trials Office trial coordinators: J. Cartmell, G. Lallemand, J. Sandercock, B. Slomka, and S. Twiddy.
1. Richards MA: Quality of life: The main outcome measure of palliative care. Palliat Med 11:89-92, 1997 2. Tiernan E: Depression in terminally ill cancer patients. J Ir Med Assoc 91:47-48, 1998 3. Montazeri A, Milroy R, Hole D, et al: Anxiety and depression in patients with lung cancer before and after diagnosis: Findings from a population in Glasgow, Scotland. Epidemiol Community Health 52:203-204, 1998 4. Koenig HG, Shelp F, Goli V, et al: Survival and healthcare utilization in elderly medical inpatients with major depression. J Am Geriatr Soc 37:599-606, 1989[Medline] 5. Koenig HG, Cohen HJ, Blazer DG, et al: A brief depression scale for use in the medically ill. Int J Psychiatry Med 22:183-195, 1992[Medline] 6. Stoudemire A, Thompson TL: Medication non-compliance: Systematic approaches to evaluation and intervention. Gen Hosp Psychiatry 5:233-239, 1983[Medline] 7. Spiegel D, Bloom JR, Kraemer HC, et al: Effect of psychosocial treatment on survival of patients with metastatic breast cancer. Lancet 2:888-889, 1989[Medline] 8. Vissor MRM, Smets EMA: Fatigue, depression and quality of life in cancer patients: How are they related? Support Care Cancer 6:101-108, 1998[Medline] 9. Harrison J, Haddad P, Maguire P: The impact of cancer on key relatives: A comparison of relative and patient concerns. Eur J Cancer 31A:1736-1740, 1995 10. Kurtz ME, Kurtz JC, Given CW, et al: Relationship of caregiver reactions and depression to cancer patients’ symptoms, functional status and depression: A longitudinal view. Soc Sci Med 40:837-846, 1995 11. Ford S, Fallowfield L, Lewis S: Can oncologists detect distress in their out-patients and how satisfied are they with their performance during bad news consultations? Cancer 70:767-770, 1994 12. Hopwood P: The quality of life: Clinical judgement versus self-report measures. Cancer Topics 8:1122-1124, 1992 13. Passik SD, Dugan W, McDonald MV, et al: Oncologists’ recognition of depression in their patients with cancer. Clin Oncol 16:1594-1600, 1998 14. Valente SM, Saunders JM: Evaluating depression among patients with cancer. Cancer Pract 2:65-71, 1994[Medline]
15.
Plumb M, Holland JC: Comparative studies of psychological function in patients with advanced cancer: Self-reported depressive symptoms. Psychosom Med 39:264-276, 1977 16. Derogatis LR, Morrow GR, Fetting J, et al: The prevalence of psychiatric disorder among cancer patients. JAMA 249:751-757, 1983[Abstract] 17. Massie MJ, Holland JC: Depression and the cancer patient. J Clin Psychiatry 75:12-17, 1990 18. Kaasa S, Malt U, Hagen S, et al: Psychological distress in cancer patients with advanced disease. Radiother Oncol 27:193-197, 1993[Medline] 19. Zigmond AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 67:361-370, 1983[Medline] 20. Aass N, Fossa SD, Dahl AA, et al: Prevalence of anxiety and depression in cancer patients seen at the Norwegian Radium Hospital. Eur J Cancer 33:1597-1604, 1997 21. de Haes JCJM, van Knippenberg FCE, Neijt JP: Measuring psychological and physical distress in cancer patients: Structure and application of the Rotterdam Symptom Checklist. Br J Cancer 62:1034-1038, 1990[Medline] 22. Hopwood P, Stephens RJ: Symptoms at presentation for treatment in patients with lung cancer: Implications for the evaluation of palliative treatment. Br J Cancer 71:633-636, 1995[Medline] 23. Kisely SR, Goldberg DP: Physical and psychiatric co-morbidity in general practice. Psychiatry 169:236-242, 1996 24. Dugan W, McDonald MV, Passik SD, et al: Use of the Zung self-rating scale in cancer patients: Feasibility as a screening tool. Psychooncology 7:483-493, 1998[Medline] 25. Hopwood P, Howell A, Maguire GP: Psychiatric morbidity in patients with advanced cancer of the breast: Prevalence measured by two self-rating questionnaires. Br J Cancer 64:349-352, 1991[Medline] 26. Medical Research Council Lung Cancer Working Party: Randomised trial of four-drug vs less intensive two-drug chemotherapy in the palliative treatment of patients with small-cell lung cancer (SCLC) and poor prognosis. Cancer 73:406-413, 1996 27. Girling DJ: Comparison of oral etoposide and standard intravenous multidrug chemotherapy for small cell lung cancer: A stopped multicentre randomised trial—Medical Research Council Lung Cancer Working Party. Lancet 348:563-566, 1996[Medline] 28. Macbeth FR, Bolger JJ, Hopwood P, et al: Randomised trial of palliative two-fraction versus more intensive 13-fraction radiotherapy for patients with inoperable non-small cell lung cancer and good performance status: Medical Research Council Lung Cancer Working Party. Clin Oncol 8:167-175, 1996 29. Hopwood P, Howell A, Maguire GP: Screening for psychiatric morbidity in patients with advanced breast cancer: Validation of two self-report questionnaires. Br J Cancer 64:353-356, 1991[Medline] 30. Stephens RJ, Hopwood P: Constructing a clinically useful functional impairment scale from the Rotterdam Symptom Checklist (RSCL). Lung Cancer 18:208, 1997 (suppl 1) (abstr 805 31. SPSS Base 7.0 for Windows: User’s Guide. Chicago, IL, SPSS Inc, 1996
32.
Aaronson N, Ahmedzai S, Bergman B, et al: The European Organization For Research and Treatment of Cancer QLQ-C30: A quality of life instrument for use in international clinical trials in oncology. J. Natl Cancer Inst 85:365-376, 1993 33. Minagawa H, Uchitomi Y, Yamawaki S, et al: Psychiatric morbidity in terminally ill cancer patients. Cancer 78:1131-1137, 1996[Medline] 34. American Psychiatric Association:Diagnostic and Statistical Manual of Mental Disorders (ed 3, revised). Washington, DC,American Psychiatric Association, 1987 35. Carroll BT, Kathol RG, Noyes ,et al: Screening for depression and anxiety in cancer patients using the Hospital Anxiety and Depression Scale. Gen Hosp Psychiatry 15:69-74, 1993[Medline] 36. Jones PW, Baveystock CM, Littlejohns P: Relationships between general health measured with the Sickness Impact Profile and respiratory symptoms, physiological measures and mood in patients with chronic airflow limitation. Am Rev Respir Dis 140:1538-1543, 1989[Medline] 37. Godding PR, McNulty RD, Wittrock DA, et al: Predictors of depression among male cancer patients. J Nerv Ment Dis 183:95-98, 1995[Medline]
38.
Harrison J, Maguire P: Predictors of psychiatric morbidity in cancer patients. Br J Psychiatry 165:593-598, 1994 39. Heaven C M, Maguire P: The relationship between patients’ concerns and psychological distress in a hospice setting. Psychooncology 7:502-507, 1998[Medline] 40. McDaniel JS, Musselman DL, Porter MR, et al: Depression in patients with cancer. Arch Gen Psychiatry 52:89-99, 1995[Abstract] 41. Holland JC: Managing depression in the patient with cancer. Cancer 37:366-371, 1987 42. Schwenk TL: Cancer and depression. Oncology 25:505-513, 1998 Submitted April 30, 1999; accepted October 19, 1999. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
2 for trend, P < .0001). Multivariate analysis showed that functional impairment was the most important risk factor; depression increased by 41% for each increment on the impairment scale. Pretreatment physical symptom burden, fatigue, and clinician-rated PS were also independent predictors, but cell type was not. 
