Quality of life of 125 adults surviving 6–18 years after bone marrow transplantation

Sot'. Sci. Med. Vol. 40, No. 4, pp. 479-490, 1995 Pergamon

0277-9536(94)00153-7

Copyright ~3 1995 ElsevierScience Ltd Printed in Great Britain. All rights reserved 0277-9536/95 $9.50 + 0.00

QUALITY OF LIFE OF 125 ADULTS SURVIVING 6-18 YEARS AFTER BONE MARROW TRANSPLANTATION NIGEL E. BUSH, MEL HABERMAN,GARY DONALDSON and KEITH M. SULLIVAN Fred Hutchinson Cancer Research Center, Pain and Toxicity Research Program, 1124 Columbia Street, Seattle, WA 98104-2092, U.S.A.

Abstract--Background: Recent studies examining the medical and psychosocial sequelae of bone marrow transplantation have reported most survivors do relatively well while a smaller group continues to experience less than optimal quality of life (QOL). Many of these studies are limited by small sample sizes, limited scope, and focus on a narrow (1-4 year) window of survival. Methods: The descriptive survey examined the QOL, late medical complications, psychological distress, demands of long-term recovery, and health perceptions of 125 adults surviving (~18 (mean 10) years after marrow transplantation. Seven wide-ranging tests covering 271 items were completed on average in 90 min. Two tests were developed by the authors specifically for assessing QOL in this population. Results: 74% of long-term survivors of bone marrow transplanation reported their current QOL was the same or better than before transplantation, 80% rated their current health status and QOL as good to excellent, and 88% said the benefits of transplantation outweighed the side effects. Ten years or more post-transplantation, long-term survivors continued to experience a moderate incidence of lingering complications and demands, including emotional and sexual dysfunction, fatigue, eye problems, sleep disturbance, general pain and cognitive dysfunction. However, the severity or degree of distress attributed to those complications was, for most survivors, consistently low. Nearly all were back to work or school. Only 5% rated both their QOL and health status as poor. Long-term survivors demonstrated good mood and low psychological distress compared to cancer and population norms, and had the same perceptions as the general population of their current health and expectation of future health. Demands attributed to long-term survival appeared to impose little hardship. The most frequently cited demand of recovery was the perceived lack of social support as time went on. Conclusions: Almost all long-term survivors were leading full and meaningful lives. Persistent complications were, on the whole, dismissed as relatively trivial and the overwhelming majority viewed themselves as cured and well.

Key words~uality of life, bone marrow transplantation, long-term survivors, cancer, oncology

[11, 14-22] that Q O L is a multidimensional construct composed minimally of four domains:

Traditional cancer outcome variables, such as relapse and disease-free survival, tell us little about the nature or quality of survival after bone marrow transplantation (BMT). It is no longer sufficient to measure the effectiveness of B M T without regard to the greater experience of the patient. Despite improved prospects for survival following marrow transplantation, little is known about the demands associated with longterm recovery or the problems related to physical, psychological and social functioning, that is, the quality of life following BMT.

(a) Physical function (daily activities, self-care, mobility, role activities); (b) Psychological function (emotional/mental state, distress, mood, perceptions of wellbeing, satisfaction, happiness); (c) Social role function (ability to engage in normative social interactions and occupations, interpersonal relations, sexual functioning, family dynamics); and (d) Disease~treatment symptoms (somatic discomfort, pain, nausea, etc.).

REVIEW OF TIlE LITERATURE

Quality of life

Late medical complications of BMT

The importance of measuring quality of life (QOL) in health care research is well documented [1-12]. M o r [13], summarizing definitions of Q O L found in the cancer literature, defined Q O L as a composite of physical, emotional, and eccaomic considerations which is frequently dominated by the physical sequelae of the disease. A more operational consensus is emerging among current health researchers

The general well-being of B M T survivors may depend largely upon the severity of medical complications that arise late after B M T [23]. Late complications of B M T are well documented in the literature [23-27]. Many long-range medical complications are caused by high-dose chemotherapy and/or irradiation therapy, chronic graft-vs-host-disease (GVHD), and recurrent leukemia [23]. Rates of leukemia relapse 479

480

NIGEL E. BUSHet al.

range from 20% in persons with chronic mylogenous leukemia (CML) in chronic phase of acute nonlymphocytic leukemia in first remission to 75% in persons with CML in blast crisis or acute leukemia in relapse [24, 28]. Chronic G V H D develops in 35% of HLAidentical recipients between 70-400 days post-BMT [23]. Symptoms include cutaneous hypo/hyperpigmentation and maculo-papular or puritic rashes, mucositis, cataracts, joint contractures, esophageal and vaginal strictures, chronic liver disease, diarrhea, anorexia, a wasting syndrome and persistent severe immunodeficiencies [28]. Chronic G V H D related restrictive pulmonary disease peaks l year post-BMT and normalizes within 3~, years while rates of obstructive disease may increase with the passage of time [23]. Additional complications include gonadal failure, growth abnormalities, and early menopause [23]. Sullivan and colleagues [29] documented causes of death at 3.9-7.4 years post-BMT (median 5.1 years) to include cancer relapse, interstitial pneumonia, infection with/without acute GVHD, hepatic failure, chronic G V H D with/without infection, hemorrhage, leukoencephalopathy and secondary malignancy. The existing research on the later complications of BMT clearly indicates some longterm survivors continue to have significant, even life-threatening, health problems. Quality o f life after B M T

Few studies to date have examined QOL following BMT. In two related projects, Andrykowski and colleagues [30,31] conducted longitudinal assessments of 16 and 23 adult allogeneic marrow recipients. Initial assessments were conducted a mean of 28 months post-BMT with final assessment at a mean of 52 months. Results from both studies indicated that, although many patients experienced long-term physical complications there appeared to be little change in these complications over time; rather there was a stable-ceiling effect of functioning within several years of BMT with further recovery being minimal. Furthermore, the older the patient at time of BMT, the poorer was the long-term functioning. Wolcott et al. [32] examined 26 adults who were an average of 42 months post-BMT. Although threequarters of the sample appeared to be doing well, the remaining 25% reported ongoing medical problems and 15-25% reported significant emotional distress, low self-esteem and a less than optimal life satisfaction. Hengeveld et aL [33] assessed 17 survivors who were from l to 5 years post-transplant and found that about half reported problems related to physical complications, sexual problems, infertility and inability to work. However, most survivors showed surprisingly good affective adjustment and were overwhelmingly positive in their outlook. Similarly, Belec [34] showed that, of 24 adults between 1-3 years post-BMT, fully 92% perceived the quality of their

lives to be acceptable. Chao et al. [35] found that 51 of 58 surviving adult patients reported an above-average to excellent QOL 1 year following autologous BMT, while Wingard et al. [36] found similarly good to excellent health and functional ability 4 years after transplant. In a survey of 135 BMT recipients who were 6 months to 12 years (mean-4 years) post BMT, Baker et al. [37] found that the retention of important social roles was significantly related to QOL and inversely related to negative mood. In a recent prospective study with pre-transplant baseline assessments, Syrjala et al. [38] found that the large majority of 67 long-term BMT survivors returned to full-time employment with normal physical and psychosocial functioning, although recovery took longer than 1 year for 40%. Finally, Ferrell et al. [39, 40] conducted a detailed qualitative study of 119 adult BMT recipients which described the personal meaning of QOL from the perspective of the BMT survivor and explored interventions related to improving QOL. In summary, the studies to date show a relatively encouraging view of recovery, but do indicate that a sub-population of survivors continues to experience physical and psychological complications. However, the majority of these studies have not extended their assessments beyond 4 years post-BMT and typically involve small sample sizes and limited study design. With few exceptions, they have generally relied on generic or relatively global measures of QOL with little regard to disease or treatment-specific factors [9]. The current study is part of an ongoing effort to assess quality of life in adults who have survived from 1 year post-transplant to well beyond 5 years. In contrast to previous, shorter-term studies in the literature, the current project is unique in examining a large sample of patients who have survived an average of 10 years beyond their bone marrow transplants, a point at which they might be deemed 'cured' of their primary disease and assumed to be functionally and psychologically stable. This project is also distinguished from most previous studies by giving particular attention to the unique and specific demands of recovery from BMT. A future extension of this study will make longitudinal measures of patients between 1 and 4 years post-transplant. The current project was designed as a descriptive survey with the consensus definition of QOL [11, 14-22] as being composed minimally of four dimensions: physical, psychological and social functioning, and disease/treatment symptoms.

METHODS

Setting and sample

The study was conducted at the Fred Hutchinson Cancer Research Center, Seattle, WA, and was derived from a population of 1541 individuals

Quality of life after marrow transplantation who had marrow transplants in Seattle prior to 30 June 1983. Of 423 recipients who were alive at the time of analysis, 195 met criteria for inclusion in the study: (a) the transplant had been performed a minimum of 6 years previously; (b) patients were not in primary disease relapse; (c) subjects were i 8 or more years of age at time of transplant, and resided in the U.S.A. or Canada.

Instruments Seven self-report instruments were selected or developed to provide as comprehensive an assessment as possible, minimally encompassing the four domains of QOL defined above, and allowing respondents to provide extensive quantitative and qualitative information. Instruments included: Biodemographics. This instrument includes 26 items on income, employment, health insurance and expenses as well as standard questions about age, religion, marital status, etc.

European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 [41-43]. This is a 30 item core or generic QOL tool for use with cancer patients, to which disease specific modules, such as the BMT-specific instrument described below, must be appended. It includes multiple items which encompass each of the four QOL domains. The QLQ-C30 indexes six multi-item scales of functioning: Physical Functioning, Role Functioning, Emotional Functioning, Cognitive Functioning, Social Functioning, and Gobal Health Status/QOL. It further measures nine multi-item scales or single items of symptomatology/problems: Fatigue, Nausea and Vomiting, Pain, Dyspnea, Constipation, Diarrhea, Sleep Disturbance, Appetite Loss, and Financial Impact. Items are scaled on a Yes or No basis or on 4 point Likert-type scales where 1 = not at all, 2 = a little bit, 3 = quite a bit, and 4 = very much. Global Health Status/QOL is assessed by two 7-point Likert scales. All multi-item scales and single item measures subsequently are linearly transformed to standard 0-100 scales. For the functioning scales, a higher score represents a higher level of functioning while for the symptom scales/items a higher score represents a higher level of symptomatology. Late complications of BMT module. This BMTspecific module/addendum to the EORTC QLQ-C30 was developed by the author (NEB). Content validity for the 50 items was derived from inservice discussions with Hutchinson Center staff and from the literature [23-29, 35, 36]. The module includes items which primarily index the disease/treatment symptoms domain of QOL, although the remaining 3 domains are also represented. The items, scaled identically to the EORTC QLQ-C30 with 4-point Likert scales linearly transformed to 0-100 scales, are rated for occurrence/severity of symptoms. Multiple

481

items are categorized by: Skin, Eyes, Mouth/Throat, Joints/Muscles involvement, Pulmonary problems, Sex/Warmth/Intimacy, Cognitive Dysfunction, Infections, and Fear of Relapse/Dying. When calculating composite scores, items from two categories, Pulmonary problems and Cognitive Dysfunction, were combined with items from the EORTC QLQ-C30 to form more complete BMT-specific subscales. Single items index Physical Appearance, Hair and Nail loss, Teeth Problems, Abnormal Sense of Taste, Heartburn and Abdominal Pain, Sinusitis and Runny Nose, chronic GVHD, and Minor Symptoms/Ailments.

Demands of BMT Recovery Inventory (DBMT) [44]. The 59 item DBMT is an adaptation of the Demands of Illness Inventory. The instrument quantifies the number and intensity of demands or hardships associated with long-term recovery across five dimensions: personal meaning; social relationships; self-image; monitoring symptoms; and treatment issues. Although the social, psychological and symptomatology domains of QOL are indirectly measured by the DBMT, the instrument largely goes beyond the minimal domains to assess more qualitative aspects of recovery. Items are scored on a 5-point Likert-type scale (0 = 'Not at All,' 1 = 'A Little Bit,' 2 = 'Moderately,' 3 = 'Quite a Bit,' 4 = 'Extremely'). The total scores range from a 0 to 59 for the number of demands experienced and from 0 to 236 for the intensity of demands. Higher scores signify a greater number of demands experienced or a greater intensity of demands. Profile of Mood States (POMS) [45]. The 65 item POMS measures mood and the psychological distress domain of QOL along six dimensions: TensionAnxiety, Depression-Dejection, Anger-Hostility, Vigor-Activity, Fatigue-Inertia, ConfusionBewilderment, and an overall Mood rating. The total score ranges from 0 to 260 with a higher score indicating greater mood disturbance or psychological distress. The POMS has been widely employed in health research and demonstrates excellent reliability and validity.

Ware Health Perceptions Questionnaire (I,VHPQ) [46, 47]. the WHPQ contains 32 items measuring six dimensions: Current Health--the extent to which the respondent presently sees him/herself as being healthy or ill; Prior Health--whether the respondent perceives a favorable (healthy) or unfavorable (unhealthy) prior health history; Health Outlook-the respondent's prediction of things to come; Health Worry~Concern--the extent to which the person is worried or concerned about his/her state of health; Resistance-Susceptibility--perception of the ability to resist illness; and Rejection of Sick Role-characteristic reaction to illness. Items are scored on a 5-point Likert-type scale. Higher scores indicate a more positive perception of health. Reliability and validity of the WHPQ have been previously well established by Ware [46]. The WHPQ is difficult to categorize in terms of the four QOL domains. It was

482

NIGEL E. Bush et al.

selected for the study to tap the more intimate, qualitative thoughts of survivors regarding their past, present and future health. Long-term B M T recovery questionnaire. This 9 item open-ended questionnaire, devised by the authors, elicits information on reestablishing one's daily life after BMT, demands of recovery, coping strategies, limitations imposed by BMT, current health problems, quality of life, and concerns about the future. This tool provides an abundance of rich qualitative data, most of which is described in a separate publication [48]. Procedures

The instrument packet explained the survey instruments, the purpose of the study and the promise of confidentiality for respondents. Respondents were instructed to fill out the questionnaires based on their experiences within the last two weeks; in other words, based on their recent quality of life. Eighty-nine percent of the sample reported their QOL over the previous two weeks was completely typical of the prior six months. Informed consent, as specified in the cover letter, was implied by a return of the questionnaire packet. All procedures were approved by the Institutional Review Board. Of the 195 questionnaire packets mailed, 125 completed packets were returned; 14 uncompleted packets were returned by the post office because of an incorrect address; 2 individuals refused participation (by postcard reply); and 54 failed to respond. The valid response rate of 70% (excluding incorrect addresses) represents an excellent compliance for an unsolicited survey, especially given the size of the instrument packet. Response rates of 50% or less are more typical [49]. Respondents' comments provided some insight into the excellent response in the present survey: "It is important to me that you've given me the opportunity to tell you the problems I've overcome". "I felt good giving you this information because it helped me put things in perspective". The mean time for completion of the instrument packet was 90 min. Statistical analyses

The primary data summarized are patients' reports by QOL category. An incidence rate was calculated that describes the percent of items in a scale marked as problematic. This incidence rate is insensitive to gradations in the intensity of individual items, treating 'a little bit-, 'quite a bit', and 'very much' of a problem identically. Scale means, which are sensitive to the intensity of individual items, are also provided as another descriptive index of severity. Traditional medical outcomes, such as survival and relapse, have self-evident validity. In contrast, the validity of QOL outcomes obviously depends on measurement properties of the scales chosen to assess this complex construct. A necessary condition for validity is that scales demonstrate adequate

reliability and precision of measurement. Reliability is documented with Cronbach's alpha coefficient, a conventional psychometric measure of the internal consistency of a scale--the extent to which an overall scale score reflects agreement among individual items composing the scale. The theoretical range of Cronbach's alpha is from 0, denoting no internal consistency, to 1.00, indicating perfect internal consistency. In practice, internal consistency reliabilities limit the maximum correlations one can detect, and should therefore be in the range of 0.7~).9. Conversely, two scales that correlate to the extent of their reliabilities provide no independent measurement at all. Ideally, reliabilities should be high and should exceed the intercorrelations among scales. The t-statistic was used to test that two means were equal or that a coefficient was zero; when there were more than two groups or classifications, the analysis of variance F-statistic was provided. These inferential statistics are accompanied by effect size statistics that describe magnitudes of effects. Significance tests quantify the probability of obtaining a nonzero difference; with fairly large samples, significant findings may correspond to effects that are clinically trivial or uninteresting. The effect size, expressed here as proportion of variance accounted for, provides a more pragmatic assessment of the importance of the observed result. Higher effect sizes suggest important differences; a proportion of variance (PV) of 10%, for example, would indicate that the tested effect explains 10% of the variance of the outcome. For tests of group differences, PVs < 5 % are generally considered small; PVs between 5% and 10% are moderate; and PVs > 10% are large and likely to be of clinical importance. PVs for correlations are simply the square of the correlation coefficients; in this context, the above guidelines for interpreting effect sizes should be roughly doubled. Probability tests unadjusted for multiple comparisons were reported, even though the numerous tests conducted increase the risk of Type I errors. This risk is acceptable because the authors do not generally argue that significant results denote clinical importance. This is consistent with the descriptive emphasis on the magnitudes of effects. To provide a single multivariate test that the set of outcomes varied with multiple predictors (age at BMT, time since BMT, and sex), a canonical correlation analysis was conducted. No significant linear pattern emerged between the two sets of variables. To check for nonlinear time dependence, the Lowess procedures of Cleveland [50] was used to plot a smooth nonparametric nonlinear trend for the QOL outcomes with time since BMT. This objective algorithm describes local trends in the data that may be missed by visual inspection of parametric fits. Finally, intercorrelations among different questionnaires were substantially less than reliabilities, ranging from effectively 0 to around 0.5 (most generally 0.1 and

Quality of life after marrow transplantation 0,3), indicating the relative independence of major measures and good discriminant validity. RESULTS

Patient characteristics. The cohort of 61 women and 64 men was predominately Caucasian (93%), married (72%) and young with a mean age of 38 (range 26-62, SD 7.6) years. Mean age at time of transplant was 28 (range 19-51, SD 7.6) years. The mean time since BMT was 10.1 (range 6-18.4, SD 2.5) years. Patients were well educated and relatively affluent with 56% having 2 or more years of college and 63% reporting a total family income > $30,000/year. Ninety percent had health insurance and the leading source of this insurance was from present employers. Seventy-four percent were employed and of those not employed, only 3 were actively seeking work. Of the nine disabled survivors, seven were permanently and two temporarily disabled. The original diagnoses included acute leukemia (40%), aplastic anemia (34%), chronic leukemia (20%), and other conditions (6% lymphoma, myeloma or other neoplasm). Eighty-seven percent were HLA-identical sibling transplants; 11% syngeneic; and 2% autologous transplants. Over 20 different chemotherapy and total body irradiation (TBI) preparatory regimens were used. The majority (45%) received chemotherapy plus fi'actionated TBI (1220cGy in 6 sessions or 1575cGy in 7 sessions), 22% received chemotherapy plus a single dose of 920-1000cGy TBI and 34% received only chemotherapy. Forty-three percent of survivors were in remission at the time of transplant, 26% in relapse, 22% in chronic phase, and the remaining 9% were in the accelerated phase or blast crisis. Following BMT, 59% of allogeneic recipients experienced acute GVHD, and 65% developed chronic GVHD. Fred Hutchinson Cancer Research Center medical records on the 22 females and 48 males who did not complete and return packets or whos' whereabouts were unknown were examined for biases in respondent characteristics. Demographic, disease and treatment characteristics of non-respondents did not differ statistically from those of respondents. Mean age at time of transplant was 27 (SD 6.9) years. The mean time since BMT was 9.9 (SD 2.7) years. Original diagnoses of non-respondents included acute leukemia (52%), aplastic anemia (23%), chronic leukemia (13%), and other conditions (10% lymphoma, myeloma or other neoplasm). As with respondents, 87% of non-respondents were HLAidentical sibling transplants; 11% syngeneic; and 2% autologous transplants. Sixty-nine percent of non-respondents were in remission at the time of transplant, 14% in relapse, 12% in chronic phase, and the remaining 5% were in the accelerated phase of blast crisis. Finally, following BMT, 45% of allogeneic

483

recipients experienced acute GVHD. Given that a greater proportion of non-respondents were transplanted in remission and fewer experienced acute GVHD, it is unlikely that non-respondents comprised a disproportionately sick cohort. EORTC QLQ-C30 and B M T specific module. Although the EORTC QLQ-C30 is multidimensional, the dimensions are not strictly independent: the overall score of the QLQ-C30 in the present study had high internal consistency (Cronbach's alpha = 0.92). Six of the nine multi-items subscales were reliable, with alphas ranging from 0.7-0.91 with the three remaining subscales indicating questionable reliabilities (Physical Functioning = 0.57, Social Functioning = 0.38, Nausea/Vomiting = 0.48). The overall Cronbach's alpha for the BMT Module was 0.87 with category alphas ranging from 0.71-0.89 with four exceptions (Skin = 0.55, Mouth/ Throat = 0.69, Infections = 0.32, Fear of Relapse/ Dying = 0.66). The EORTC QLQ-C30 correlated well with the BMT Module (r = 0.77, P < 0.001 ) and Global Health Status/QOL (r = 0.66, P < 0.001). Since the QLQ-C30 and the BMT Module were highly correlated, they were also combined in a single factor score having greater breadth than either component alone. A mean overall (~100 score was calculated for the combined EORTC QLQ-C30 and BMT Module to derive a potential single summary of QOL. Functioning scale scores were reversed when calculating this measure so that a higher score represented a lower level of functioning. Thus, a higher composite score on the combined EORTC QLQ-C30 and BMT Module represented poorer QOL or greater problems overall. The Cronbach's alpha for the combined EORTC QLQ-C30 and BMT Module was 0.93. The combined QLQ-C30 and Module correlated well with the Global Health Status/QOL (r =0.65, P < 0.001). Although the internal consistency reliability coefficients of a few subscales on the EORTC QLQ-C30 and BMT Module were below criterion for new scales of 0.70 [51], the data from these subscales were cited for normative comparisons and retained for analysis until the scales can be tested on future samples of short-term survivors. The mean score for the combined EORTC QLQ-C30 and BMT Module was only 16.66 (range 1.3 52.7, SD 12.4) on the 0 100 scale, signifying a high QOL. Table 1 shows items and categories ranked by incidence of occurrence. Among the highest reported incidence rates were emotional dysfunctioning, fatigue, eye problems, sleep disturbance, pain, cognitive dysfunction, fear of relapse/dying and joint and muscle discomfort. Table 1 also shows the mean scores and standard deviations for the symptom and functioning scales, again with functioning scores reversed so that a higher score represents poorer functioning. The mean values for the individual subscales were all low, signifying that the adverse effects of BMT across all subscales of QOL were mild in intensity. The only

NtGEL E. BUSH et al.

484

Table I. EORTC QLQ-C30 and BMT Module categories rank ordered by incidence ratings (N = 125)

Rank/category I. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Emotional dysfunction Fatigue Eye problems Sleep disturbance General pain Cognitive dysfunction Fear relapse & dying Joint/muscle pain Financial impact Social dysfunction Skin problems Constipation Mouth/throat problems Diarrhea Pulmonary problems Appetite loss Role dysfunction Nausea/vomiting Physical dysfunction Infections

Number of items

Incidence rates (%)*

Mean (0-100)

SD

4 3 6 I 2 4 2 3 1 2 5 1 3 I 5 1 2 2 5 7

63.3 55.9 48.9 42.7 41.0 38.9 38.7 37.6 33.8 32.3 26.9 26.6 25.3 24.2 21.9 17.7 13.3 12.9 10.5 3.6

29.2 23.1 27.6 19.6 19.0 18.1 18.5 18.9 16.9 15.7 14.0 10.2 13.8 10.7 10.3 7.52 13.3 4.6 10.5 2.1

24.3 19.2 24.1 26.8 24.2 22.2 23.5 22.6 28.0 24.0 13.9 18.1 20.3 21.0 15.1 17.4 25.5 9.8 17.5 5.5

*Incidence rates are defined as the percentage of items reported as indicating a problem within each category; i.e. the number of items within a category scored as a 2 or greater, divided by the number of items in the category, x 100, where 1 = Not at All, 2 = A little Bit, 3 = Quite a Bit, and 4 = Very Much.

QLQ-C30 data available for comparison came from multinational trials on 305 patients with lung cancer [43], a sample one might assume to be considerably sicker than equivalent BMT candidates and certainly more so than the present long-term survivors. Compared by independent t-tests to lung cancer patients in pre-treatment phase, transplant survivors reported better Physical Functioning, Role Functioning, Global Quality of Life (all P < 0 . 0 0 1 , mean P V = 1 3 . 6 % , PV r a n g e = 12.4-15.9%), and Social Functioning ( P < 0 . 0 2 , P V = 1.4%), and Lower Fatigue, General Pain, Dyspnea, Sleep Disturbance, Appetite Loss, and Constipation (all P < 0.001, mean PV = 6.8%, PV range = 2.5-16.6%). No differences were detected in Cognitive or Emotional Functioning, Nausea and Vomiting, or Financial Impact. Similar differences were evident between transplant survivors and lung cancer patients in treatment phase [although BMT survivors showed less nausea and vomiting (P < 0.001, PV = 8.0%)1. Males and females transplant survivors did not differ by independent t-test on any QLQ-C30 or module items with three exceptions: males reported greater mouth/throat problems (P < 0.016, PV = 4.7%) and greater pulmonary problems (P < 0.024, P V = 4 . 1 % ) than females. Not surprisingly, as in the general population, males also experienced more hair loss than females (P <0.027, PV = 4%). In response to a series of questions concerning sexual functioning, 27% of survivors reported that they were dissatisfied with their appearance compared to before BMT. Between 28 and 36% of respondents were dissatisfied with their Sexual Appeal, Ability to share Warmth and Intimacy, or

Interest in Sexual Thoughts or Feelings. Over 26% (23% of males and 32% of females) revealed that they had physical problems which reduced their satisfaction with sex and intimacy. Despite the high incidence rate of some symptoms, 80% of survivors rated their 'Global QOL' as good to excellent (mean 77.15, SD 21.8), and 77% of survivors assessed their 'Global Physical Health Status' as good to excellent (mean 73.2, SD 20.1). This finding was cross-validated by a question in the demographic instrument in which 80% rated their current health as good to excellent. Only 5% rated both their gobal QOL and Health Status as poor. Further validation was provided by the Long-Term BMT Recovery Questionnaire. When asked to compare their current QOL with that prior to BMT, 74% of the survivors said their current QOL was the same or better than before, and 88% said the benefits of transplantation outweighed the side-effects. Demands of BMT recovery. The overall Cronbach's alpha for the DBMT was 0.93. Subscale were reliable, ranging from 0.84-0.87 with two exceptions (selfimage=0.48, social relations=0.68). The eight most distressing demands of recovery are ranked by frequency of occurrence in Table 2. Ninety-six percent of the recipients indicated people were less supportive over time and that this demand was the single most distressing hardship of long-term survival. Ninety percent of the recipients indicated they continued to think about the value of their life as well as their priorities of life (83%) and that both demands were moderately stressful. Self-image appeared to be a concern for long-term survivors with 90% reporting some unhappiness with their physical appearance and 77% indicating they thought about their sexual appeal. Similar to the

Quality of life after marrow transplantation

485

Table 2. Demands of BMT Recovery items rank ordered by frequency of occurrence (N = 125) Rank

Item description

Percent

People are less supportive over time Thinking about the value of my life Not satisfied with the way 1 look Cannot always rely on my body Reordering the priorities of my life Thinking about my sexual appeal Monitoring closely any new symptoms Need information about long-term side effects

96 90 90 87 83 77 71 66

(78)* (60) (52) (45) (40) (26) (30) (27)

Mean

SD

2,73 2.52 1,71 1.79 1.94 1.56 1.55 1.90

1.55 1.25 1.19 1.25 1.23 1.21 1.33 1.62

*The first percentage represents the number of respondents choosing a 1-4 rating while the second, in parentheses, represents a more conservative percentage, i.e. respondents selecting either a 3 or 4 rating (0 = Not at All, I = A little Bit. 2 = Moderately, 3 = Quite a Bit, 4 = Extremely).

low mean intensity ratings for the majority of demands items, the mean intensity ratings for all five categories of demands fell between a 0.41 and 1.54 on the 5 point scale. This finding signifies that the demands attributed to long-term survival appear to impose little hardship an average of 10 years after BMT. A consistent picture emerged when the DBMT was correlated with the major QOL outcomes. An inverse relationship existed between demands and Global Health Status/QOL (r = -0.55, P < 0.001), signifying the higher the intensity of demands the lower the perceived QOL/health status. Similarly, a positive relationship existed between demands of survival and the mean overall score for the combined EORTC QLQ-C30/BMT Module (r =0.76, P < 0.001), implying the higher the intensity of demands the poorer the perceived Q O L Lastly, demands correlated positively with the POMS (r = 0.68, P < 0.001), implying the higher the intensity of demands the poorer the survivors' mood adjustment. Profile o f mood states. The overall Cronbach's alpha for the POMS was 0.93 with subscales ranging from 0.87-0.94. Table 3 lists the subscale means for the POMS. Mood and level of psychological distress were assessed and compared to published norms using independent means t-tests. Compared to samples of untreated adult cancer patients [52] and college students or psychiatric outpatients [45], BMT recipients showed less tension, fatigue, confusion and total distress (all P <0.01, mean P V = 5.7%, PV range = 0.4-14.2%) and more vigor (all P <0.05, mean PV = 6.3%, PV range= 0.9-12.1%). Interestingly, BMT survivors reported higher anger levels (closer to published norms) than untreated cancer patients (P < 0.05, PV = 1.7%). Finally, the current Table 3. Subscale means and standard deviations for the Profile of Mood States (N = 123) Subscale Tension-anxiety Depression-dejection Anger-hostility Vigor-activity Fatigue-inertia Confusion-bewilderment Total mood score

cohort of 125 adults with a mean of 10 years survival was compared with published data from a sample of 16 adults who had survived a mean of 4.3 years after BMT [30]. The 10 year survivors reported less tension (P < 0.01, PV = 5.2%), fatigue (P < 0.02, PV = 4.0%), confusion (P < 0.01, PV=7.1%) and total psychological distress (P < 0.05, PV = 3.4%) than survivors living roughly one-half as long. Although not significantly different, the 10 year survivors also demonstrated less depression and anger and more vigor than the shorterterm survivors. Ware health perceptions questionnaire. The overall Cronbach's alpha for the WHPQ was 0.89 with subscales ranging from 0.764).92 with two exceptions (prior health = 0.59, rejection of sick role = 0.46). Table 4 lists the subscale mean scores for the WHPQ. Male and female survivors had essentially the same perceptions about their health with one exception. Female survivors perceived their current health as better than male survivors (P < 0.007, PV = 5.9%). When compared to normative general population samples [46,47] using a population t-test, BMT survivors did not significantly differ in assessing their current health and health outlook. Not surprisingly, transplant survivors perceived their prior health as worse (P < 0.001, PV = 55.8%), felt more susceptible to illness (P < 0.001, PV = 11.0%), and were more worried about their health (P <0.001, PV = 23.9%) than the general population. However, BMT survivors were also more inclined to reject a sick role than the general population (P <0.001, PV = 13.58%). Patients with poor outcomes. Only 6 of the 125 respondents rated their global QOL and Health Status as poor, too small a subgroup for useful Table 4. Means and standard deviations for the Ware Health Perceptions Questionnaire (N = 123) Males (n = 63)

Mean

SD

Subscale

9.1 9.6 9.0 17.9 7.6 5.4 54.7

6.5 10.7 8.4 6.3 6.3 5.2 37.1

Current health Prior health Health outlook Health worry Resistance to illness Reject sick role *t = 2.76,

dr=

Females (n = 60)

Mean

SD

Mean

SD

3.47 2.29 3.44 2.37 3.02 3.16

1.06 1. I I 1.04 0.83 1.06 0.48

3,95* 2.26 3.72 2.67 3.10 3.25

0.79 I. 13 1,67 0.98 1.01 0.52

121, P <0.007.

486

NIGELE. BUSHet al.

statistical comparisons. Of the 5 males and 1 female in the subsample, 3 had been diagnosed with aplastic anemia, 2 with chronic leukemia, and 1 with acute leukemia. Mean age (38 years), age at transplant (30 years) and time since transplant (10.2 years) were similar to the rest of the sample. A visual examination of the data suggested that this small minority experienced a wide spectrum of relatively severe long-term symptoms and difficulties. The six survivors showed very much higher incidence rates and mean levels of symptom/complication severity and poorer functioning on all QLQ-C30, BMT Module and POMS measures, and on the majority of DBMT and WHPQ measures compared with the remaining 119 subjects in the sample. Most notably, incidence rates of between 80 and 100% were reported for general pain, social dysfunction, emotional dysfunction, cognitive dysfunction and eye problems. Five of the six survivors were receiving financial support for disability and were unemployed at the time of survey. The sources of some of those problems were uncovered by a post-hoc examination of archival records. A littany of long-term medical complications was revealed, including prolonged chronic G V H D toxicities such as osteoporosis, contractures, muscle pain, neuropathy and fatigue. Also evident were instances of cataracts, hearing loss, sexual dysfunction, and decreased lung and kidney function. One patient complained of being unable to hold down a job because of BMTrelated complications while another related his divorce to his debilitation. A third had recently undergone a second bilateral hip replacement. Disease, treatment and biodemographic predictors

Unless otherwise stated, none of the outcome measures varied significantly as a function of standard biodemographic, disease or treatment variables; in particular no differences were found for type of disease (e.g. malignant vs non-malignant), type of transplant, or conditioning regimen. The only biodemographic variables that were significantly correlated with QOL outcomes were age at transplant and time since transplant. The greater the receipient's age at the time of the transplant, the greater the current fatigue (r = 0.16, P < 0.037, PV = 2.6%), the poorer the memory and concentration (r = 0.15, P < 0.04, PV = 2.3%), and the poorer the overall cognitive functioning (r = 0.15, P < 0.04, PV = 2.3%). Moreover, the greater the age at the time of BMT the lower the survivors current vigor/activity (r = 0.17, P < 0.027, PV = 2.7%), the poorer the self-image (r = - 0 . 3 2 , P <0.0001, PV = 10.2%), the lower the health worry/concerns (r =0.233, P <0.005, P V = 5.4%) and the more inclined survivors were to reject a sick role (r = 0.26, P < 0.002, PV = 6.8%). Time since BMT. Because the present study was restricted to a single time-point, the time since BMT as a covariate of major QOL outcomes was of particular interest. As with age at time of transplant,

a number of outcomes had significant but trivial (r =0.154).21) correlations with time since BMT, which generally indicated a lessening of problems with more recent transplants. Canonical correlation analyses between major summary outcomes (combined EORTC QLQ-C30 and BMT Module, Total POMS Mood Distress, Overall Demands of BMT, and Global Health Status/QO1) and multiple predictors (Age at BMT, Current Age, and Time Since BMT, and Gender) revealed no significant multivariate linear relationships. Finally, nonlinear time trends were fitted, using lowess smoothing, for key outcome variables versus time since BMT. As shown in Fig. 1, a sequence of peaks and troughs was noted for a number of different outcomes over the period between approx. 100 months and 140 months post-BMT. The decline in QOL approaching 10 years post-BMT followed by a return to normality is noteworthy, approx. 2/3 of a standard deviation which is a magnitude considered a 'moderate-to-large' effect in an intervention context. The apparent trend is reinforced by its occurrence across several different outcome measures. Chronic graft-vs-host disease. A one-way ANOVA comparison of means revealed that patients with a history of chronic GVHD (9.8 years mean survival) had received transplants longer ago than survivors without chronic GVHD (8 years mean survival) IF = 7.07, df = 3.77, P < 0.0003, PV = 27.6%]. This finding could be interpreted in two ways. The likelihood of experiencing clinical chronic GVHD may increase the longer one survives after BMT. Alternately, prophylaxis and treatment for GVHD was less effective in the early days of BMT resulting in a higher late incidence of the condition. Surprisingly, survivors with chronic GVHD reported higher satisfaction with their current treatment and the progress of their recovery and were more confident that their health problems could be correctly managed (combined mean of three items = 2.49) than survivors free of chronic G V H D (combined m e a n = 1.17) [P < 0.006, PV = 9.0%]. DISCUSSION In a climate of cost containment and reform for health care, the high price of bone marrow transplantation is of considerable importance. While prior studies have shown cost-effectiveness compared to conventional treatment [53], critics suggest that the quality of life of cancer survivors is restricted. The present study addresses those concerns in a population of 125 adults surviving 6-18 years after marrow transplantation. Seven wide ranging and relatively independent tests encompassing the four major domains of QOL showed remarkable concordance. Far from being an unproductive or dysfunctional burden on society, > 90% of respondents were leading full and meaningful lives an average of 10 years after transplantation.

Quality of life after marrow transplantation

Quality of Life by Time

Mean

487

Since BMT

CombinedEORTCQLQ-C30&BMTModule 100 90 8O

~o 60 •

era

•

•

40

3o 2o "'

10

~

o

,,,

60

•

.... 0

i•

18o

1~o

•

1~o

•

•

]~

TIME SINCE BMT - MONTHS

300 [

,o :i io

240

TotalIOMSMiod Distressby TimeSinceBMT ~

-

-

•

ePo

200

•

o

6o

oj

•

.

.

~'0

~o

--

•

18o

I~o

==

l;)o

•

~

•

-

i~o

•

I~o

28o

•

2~o

240

TIME SINCE B M T - MONTHS

Total BMT Demandsby Time Since BMT 140 120 1oo IXI t,

0

80 •

•

$ 60

•• I

•

• eI

6

6 6

• •

•

•

•

•

W r.~ ..i

40 O•

ell

20 0 60

i•

•

_ ~

•

•

1~o

• •O

]}o

•

o °° oo

•

•

• •

l;)o

1~o

•

17o

2bo

2?0

240

TIME SINCE BMT • MONTHS

Fig. I. Mean Quality of Life (upper panel), total Profile of Mood States Distress (middle panel) and total Demands of BMT Recovery Inventory (lower panel) displayed by time after bone marrow transplantation (BMT). Higher scores on the Y-axes indicate diminished quality of life, increased mood distress or greater demands, respectively. ;SM ~ / 4 ~ F

488

NIGELE. BUSHet al.

Nearly all were back to work or school. Seventy-four percent perceived their current quality of life as the same or better than before transplant and 88% reported the benefits outweighed the side effects of transplantation. Respondents demonstrated good mood and low psychological distress in comparison to population and cancer norms. Demands attributed to long-term survival appeared to impose little hardship. Respondents had the same perceptions of current health status and expectations of future health as the general population. Eighty-six percent reported little or no fear of disease relapse. In fact, although more worried about health and susceptibility to illness, they were more inclined than the general population to reject a sick role. Although the design in the present study necessarily confounds time since BMT with historical changes in treatment protocols and other cohort effects, it is also important to know that the overall time trend was essentially stable. In particular, there was no evidence of deterioration of QOL as the time from transplant increased. Life at 16 years posttransplant was evidently just as rich and satisfying as life after 8 years. The abnormality at 10 years shown in Fig. 1, found consistently across several variables, suggests a temporary decline in QOL followed by a return to the previous level. Former patients remember clearly the date of transplant and may well invest the tenth anniversary with particular significance. Anecdotal reports suggest that anxiety about survival past a bench mark may induce a temporary decline in QOL that reverses once the milestone is past. Although the overall impression from participants in this study was compellingly positive, survivors were not without disease, treatment and readjustment related complications. For example, 26 to 36% of respondents reported physical problems or dissatisfaction with sex and intimacy. Schmidt et al. [54], found that 40% of allogeneic survivors indicated problems with their sexual lives, while Wingard et al. [55] reported that 22% of transplant survivors indicated some degree of sexual dissatisfaction. As a whole, the data from the current study show the incidence rates of some QOL deficits, especially physical symptoms, psychological distress and demands of recovery, were still relatively high an average of 10 years after transplant. However, the data also show that the intensity or degree of distress attributed to those complications was, for most survivors, consistently low. Recipients seemed to regard long-term complications more as a habitual irritation of daily life than as a burden or debilitation. The descriptive data point to some additional implications. Health providers can help survivors examine their expectations of recovery, of themselves and of others. For example, survivors appear to perceive others as less supportive as time goes on. Health providers must not only continue to support those doing well but monitor risk for those reporting

problems. Six of the 125 patients in the current study rated both their global health and QOL as poor and have special needs for care. Participants were chosen, by design, to include the longest survivors of transplantation available. Subjects in the sample were transplanted during the pioneering days of BMT. As such, the current sample is not representative of the entire population of transplant survivors. For example, there were 16 syngeneic transplant recipients, a group unlikely to develop chronic GVHD. Furthermore, diagnosis and treatment of chronic GVHD and regimen-related toxicities is now significantly more effective than in the early days of BMT. While one might assume that with earlier diagnosis and treatment long-term complications are now better controlled, investigators are now using more toxic conditioning regimens and routinely transplanting patients of greater age and H L A disparity. Thus, despite improvements in the treatment of regimen-related toxicities and chronic GVHD, one cannot assume that contemporary recipients will have fewer long-term complications in the future. Monitoring of late complications and QOL remains key to long-term follow-up and the BMTspecific assessment tools developed for this study appear reliable measures of life late after marrow transplantation. The BMT-specific module developed for the EORTC QLQ-C30 and the Demands of BMT Recovery questionnaire are currently undergoing further testing in patients surviving 1-5 years after transplantation. A word of caution must be added regarding the 57 survivors who received questionnaire packets but who did not complete and return them. Although non-respondents did not differ from respondents in terms of demographics, disease or treatment charcteristics at time of transplantation, there are no available data on their current quality of life. It is possible that non-respondents enjoyed a similarly high quality of life. However it is also possible that they represented a subgroup with poorer outcomes than respondents; that some non-respondents may have been less inclined to fill out the questionnaires if they were sicker. Renormalizing one's life after a life-threatening treatment is a relative notion with respect to a designated baseline or frame of reference. Although comparative QLQ-C30 data were available for older lung cancer patients prior to treatment, the current survey lacked a true baseline for bone marrow transplant recipients. One simply does not know what life was actually like for the sample before transplant or during the many years of survival prior to this study. de Haes and van Knippenburg [56] offered Helson and Bevan's adaptation level theory ( c f Ref. [57]) as an explanation for the relatively high QOL of some cancer survivors. They postulate that cancer patients adapt to disease and treatment demands with the passage of time by adjusting their perceptions of normality accordingly. Similarly, although many

Quality of life after marrow transplantation transplant survivors in the current sample reported a high incidence of problems, the vast majority perceived the quality of their lives and their physical health as good to excellent. Although life is obviously not the same as prior to transplantation, survivors appear to readily adjust their perceptions o f normality to fit current circumstances or expectations. Thus, the overwhelming impression gained from transplant survivors in this study is that they are productive, stable and well-adjusted, despite everpresent but relatively innocuous physical complications. In their eyes are are 'cured'. Narrative data strongly suggest that the actual process of renormalizing and adapting to life after transplantation is not all that important in the bigger scheme of things. One 8 year survivor summarized the general sentiment of the sample as a whole: " W h a t matters most", she wrote, "is life itself... Any life after BMT, must be considered a bonus, a miracle".

Acknowledgements--This investigation was supported by NIH Clinical Pain Reserch Grant No. NS07210; by the Seattle Marrow Transplant Nursing Consortium, by the Nursing Support Program, NCI Grant No. CA15704-18, and by NIH Grants CA38552, CA18221, CA18029 and HL36444. The authors wish to acknowledge the assistance of C. Richard Chapman, Ph.D., Joleen Kelleher, RN, MN, Neil K. Aaronson, Ph.D., Janet Nims, RN, BSN and Betty Ferrell, RN, Ph.D., FAAN.

REFERENCES

1. Moinpour C. M., Feigi P., Metch B., Hayden K. A., Meyskens F. L. Jr and Crowley J. Quality of life endpoints in cancer clinical trials: Review and Recommendations. J. Natl. Cancer Inst. 81, 485, 1989. 2. Aaronson N. K. Methodological issues in psychosocial oncology with special reference to clinical trials. In: Assessment o f Quality o f Life and Cancer Treatment (Edited by Ventafridda V., van Dam F. S. A. M., Yancik R. and Tamburini M.), pp. 29-41. Excepta Medica, Amsterdam, 1986. 3. Flanagan J. C. Measurement of quality of life: current state of the art. Arch. Phys. Med. Rahabil. 63, 56, 1982. 4. Frank-Stromberg M. Selecting an instrument to measure quality of life. Onc. Nurs. Forum. 11, 88, 1984. 5. de Haes C. J. M. and van Knippenberg F. C. E. The quality of life of cancer patients: a review of the literature. Soc. Sci. Med. 20, 809, 1985. 6. Clark A. and Fallowfield L. J. Quality of life measurement in patients with malignant disease. R. Soc. Med. 79, 165, 1986. 7. Celia D. F. and Cherin E. A. Measuring quality of life in patients with cancer. Proceedings of the Fifth National Conference on Human Values and Cancer, pp. 23-31. American Cancer Society, New York, 1987. 8. Fayers P. M. and Jones D. R. Measuring and analysing quality of life in cancer clinical trials: a review. Stat. Med. 2, 429, 1983. 9. Guyatt G. H., Feeny D. H. and Patrick D. L. Measuring health-related quality of life. Ann. Intern. Med. 118, 622, 1993. 10. Hollandsworth J. G. Jr. Evaluating the impact of medical care and technologies on the quality of life: A 5-year update. Soc. Sci. Med. 26, 425, 1988.

489

11. Aaronson N. K. Quality of life: What is it? How should it be measured? Oneology 2, 69, 1988. 12. Testa M. A., Anderson R. B., Nackley J. F., Hollenberg N. K. et al. Quality of life and antihypertensive therapy in men. N. Engl. J. Med. 328, 907, 1993. 13. Mor V. Cancer patients' quality of life over the disease course: lessons from the real world. J. Chronic. Dis. 40, 535, 1987. 14. Ware J. E. Jr. Standards for validating health measures: Definition and content. J. Chronic. Dis. 40, 473, 1987. 15. Caiman K. C. Definitions and dimensions of quality of life. In: The Quality of Life of Cancer Patients (Edited by Aaronson N. K. and Beckman J.), pp. 1-9. Raven Press, New York, 1987. 16. Spitzer W. O., Dobson A. J., Hall J., Chesterman E., Levi J., Shepherd R., Battista R. N. and Catchlove B. B. Measuring the quality of life of cancer patients: a concise QL-index for use by physicians. J. Chron. Dis. 34, 585, 1981. 17. Tchekmedyian N. S. and Celia D. S. Quality of Life in Oncology Practice and Research. Dominus Publishing Co., New York, 1991. 18. Smart C. R. and Yates J. W. Quality of life. Cancer 60, 620, 1987. 19. Schipper H., Clinch J., McMurray A. and Levitt M. Measuring the quality of life of cancer patients: the functional living index-cancer: development and validation. J. Clin. Oncol. 2, 472, 1984. 20. Sehraub S., Bransfield D. D., Monpetit E. and Fournier J. Comparison of three methods to measure quality of life. In: The Quality of Life of Cancer Patients (Edited by Aaronson N. K. and Beckman J.), pp. 83-92. Raven Press, New York, 1987. 21. Nayfield S. G., Gantz P. A., Moinpour C. M., Celia D. F. and Hailey B. J. Report from a National Cancer Institute (USA) Workshop on Quality of Life Assessment in Cancer Clinical Trials. Quality Life Res. 1,203, 1992. 22. Bullinger M. Quality of life assessment in palliative care. J. Palliat. Care. 8, 34, 1992. 23. Sullivan K. M., Agura E., Appelbaum F., Badger C., Bearman S., Erickson K., Flowers M., Hansen J., Loughran T., Martin P., Matthews D., Petersdorf E., Radich J., Riddell S., Rovira D., Sanders J., Schuening F., Siadak M., Storb R. and Witherspoon R. P. Chronic graft-versus-host disease and other late complications of bone marrow transplantation. Semin. Hematol. 28, 250, 1991. 24. Applebaum F. R. and Thomas E. D. Treatment of acute leukemia in adults with chemoradiotherapy and bone marrow transplantation. Cancer 55, 2202, 1985. 25. Buchsel P. C. Long-term complications of allogeneic bone marrow transplantation: nursing implications. Onc. Nuts. Forum. 13, 61, 1986. 26. Nims J. W. and Strom S. Late complications of bone marrow transplant recipients: nursing care issues. Semin. Oncol. Nurs. 4, 47, 1988. 27. Press O. W., Schaller R. T. and Thomas E. D. Bone marrow transplant complications. In Complications of Organ Transplantation (Edited by Toledo-Pereyra L. H.), pp. 399-424. Marcel Dekker, New York, 1987. 28. Sullivan K. M. and Storb R. Bone marrow transplantation and graft-versus-host disease. In Organ Transplantation: Current Clinical and Immunological Concepts (Edited by Brent L. and Sells R. A.), pp. 91-118. Bailliere Tindall, London, 1989. 29. Sullivan K. M., Storb R., Buchner C. D., Fefer A., Fisher L., Weiden P. L. Appelbaum F. R., Banaji M., Hansen J., Martin P., Sanders J. E., Singer J., Witherspoon R. P. and Thomas E. D. Graft-versus-host disease as adoptive immunotherapy in patients with advanced hematologic neoplasms. N. Engl. J. Med. 320, 828, 1989.

490

NIGEL E. BUSH et al.

30. Andrykowski M. A., Henslee P. J. and Barnett R. L. Longitudinal assessment of psychosocial functioning of adult survivors of allogeneic bone marrow transplantation. Bone Marrow Transplant 4, 505, 1989. 31. Andrykowski M. A., Henslee P. J. and Farrall M. G. Physical and Psychosocial Functioning of adult survivors of aUogeneic bone marrow transplantation. Bone Marrow Transplant 4, 75, 1989. 32. Wolcott D. L., Wellisch D. K., Fawzy F. I. and Landsverk J. Adaptation of adult bone marrow transplant recipient long-term survivors. Transplantation 41, 478, 1986. 33. Hengeveld M. W., Houtman R. B. and Zwaan F. E. Psychological aspects of bone marrow transplantation: a retrospective study of 17 long term survivors. Bone Marrow Transplant 3, 69, 1988. 34. Betec R. Quality of life: perceptions of long-term survivors of bone marrow transplantation. Onc. Nurs. Forum. 19, 31, 1992. 35. Chao N. J., Tierney D. K., Bloom J. R., Long G. D., Barr T. A., Stallbaum B. A., Wong R. M., Negrin R. S., Horning S. J. and Blume K. G. Dynamic assessment of quality of life after autologous bone marrow transplantation. Blood 80, 825, 1992. 36. Wingard J. R., Curbow B., Baker F. and Piantadosi S. Health, functional status, and employment of adult survivors of bone marrow transplantation. Ann. Intern. Med. 114, 113, 1991. 37. Baker F., Curbow B. and Wingard J. R. Role retention and quality of life of bone marrow transplant survivors. Soc. Sci. Med. 32, 697, 1991. 38. Syrjala K. L., Chapko M. K., Vitaliano P. P., Cummings C. and Sullivan K. M. Recovery after allogeneic marrow transplantation: prospective study of predictors of long-term physical and psychosocial functioning. Bone Marrow Transplant. 11, 319, 1993. 39. Ferrell B., Grant M., Schmidt G. M., Whitehead C., Fonbuena P. and Forman S. J. The meaning of quality of life for bone marrow transplant survivors. Part I. The impact of bone marrow transplant on quality of life. Cancer Nuts. 15, 153, 1992. 40. Ferrell B., Grant M., Schmidt G. M., Whitehead C. Fonbuena P. and Forman S. J. The meaning of quality of life of bone marrow transplant survivors. Part 2. Improving quality of life for bone marrow transplant survivors. Cancer Nurs. 15, 247, 1992. 41. Aaronson N, K., Bullinger M. and Ahmedzai S. A modular approach to quality-of-life assessment in cancer clinical trials. Recent Results Cancer Res. 111, 231, 1988. 42. Aaronson N. K., Bakker W., Stewart A. L., van Dam F. S. A. M., van Zandwijk N., Yarnold J. R. and Kirkpatrick A. Multidimensional approach to the measurement of quality of life in lung cancer clinical trials. In: The Quality of Life of Cancer Patients (Edited by Aaronson N. K. and Beckman J.), pp. 63-82. Raven Press, New York, 1987.

43. Aaronson N. K., Ahmedzai S., Bergman B., Bullinger M., Cull A., Duez N. J., Filiberti A., Flechtner H., Fleishman S. B., de Haes J. C. J. M., Kaasa S., Klee M., Osoba D., Razavi D., Rofe P. B., Schraub S., Sneeuw K., Sullivan M. and Takeda F. 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, 356, 1993. 44. Haberman M., Wood N. and Packard N. Demands of chronic illness: reliability and validity assessment of a demands-of-illness inventory. Holist. Nuts. Pract. 5, 25, 1990. 45. McNair D. M., Loor M. and Droppelman L, F. EDITS manual for the Profile o f Mood States. Educational and Industrial Testing Service, San Diego, 1971. 46. Ware J. E. Scales for measuring general health perception. Hlth. Serv. Res. 11, 396, 1976. 47. Ware J. E., Snyder M. K. and Wright W. R. Development and Validation Scales to Measure Patient Satisfaction with Health Care Services. Southern Illinois University, Carbondale, IL, 1976. 48. Haberman M., Bush N. E., Young K. and Sullivan K. M. Quality of life of adult long-term survivors of bone marrow transplantation: a qualitative analysis of narrative data. Oncol. Nurs. Forum. 20, 1545, 1993.

49. Miller D. C. Handbook of Research Design and Social Measurement, 5th Edn. Sage Publications, Newbury Park, 1991. 50. Cleveland W. S. Robust Locally-weighted Regression and Smoothing Scatterplots. J. Am. Stat. Assoc. 74, 829, 1979. 51. Nunnally J, C. Psychometric Theory, 2nd Edn. McGraw-Hill, New York, 1978. 52. Celia D., Tross S., Orav E., Holland J., Silberfarb P. and Rafla S. Mood states of patients after the diagnosis of cancer. J. Psychosoc. Onc. 7, 45, 1989. 53. Welch H. G. and Larson E. B. Cost effectiveness of bone marrow transplantation in acute nonlymphocytic leukemia. N. Engl, J. Med. 321, 807, 1989. 54. Schmidt G., Fonbuena P., Chao N. et al. Quality of life in survivors of allogeneic BMT for malignant disease. Proc. Am. Soc. Clin. Oncol. 18, 311 (abstract), 1989. 55. Wingard J. R., Curbow B., Baker F., Zabora J. and Piantadosi S. Sexual satisfaction in survivors of bone marrow transplantation. Bone Marrow Transplant. 9, 185, 1992. 56. de Haes C. J. M. and van Knippenberg F. C. E. Quality of life of cancer patients: review of the literature. In The Quality of Life o f Cancer Patients (Edited by Aaronson N. K. and Beckman J.), pp. 167-182. Raven Press, New York, 1987. 57. Helson H. and Bevan W. Contemporary Approaches to Psychology. Van Nostrand, Princeton, 1967.