Economic evaluation of treatment for acute lymphoblastic leukaemia in childhood.

Berlin-Frankfurt-Munster (BFM) and Dana-Farber Cancer Institute (DFCI) consortia's treatment strategies for acute lymphoblastic leukaemia (ALL) in children are widely used. We compared the health effects and monetary costs of hospital treatments for these two strategies. Parents of children treated at seven centres in Canada, Italy and the USA completed health-related quality of life (HRQL) assessments during four active treatment phases and at 2 years after treatment. Mean HRQL scores were used to calculate quality-adjusted life years (QALYs) for a period of 5 years following diagnosis. Total costs of treatment were determined from variables in administrative databases in a universally accessible and publicly funded healthcare system. Valid HRQL assessments (n = 1200) were collected for 307 BFM and 317 DFCI patients, with costs measured for 66 BFM and 28 DFCI patients. QALYs per patient were <1.0% greater for BFM than DFCI. Median HRQL scores revealed no difference in QALYs. The difference in mean total costs for BFM (US$88 480) and DFCI (US$93 026) was not significant (P = 0.600). This study provides no evidence of superiority for one treatment strategy over the other. Current BFM or DFCI strategies should represent conventional management for the next economic evaluation of treatments for ALL in childhood.

Norms-based assessment of patient-reported outcomes associated with adalimumab monotherapy in patients with ankylosing spondylitis.

OBJECTIVES: To compare the impact of ankylosing spondylitis (AS) on health-related quality of life (HRQL) and of adalimumab on initial and sustained improvement in HRQL for patients with active AS versus the general US population. METHODS: Data from the 5-year ATLAS trial were analysed. HRQL burden of AS and treatment impact on HRQL were assessed by comparing health status and utility scores from ATLAS (Short Form 36 Health Survey [SF-36] and Health Utilities Index Mark 3 [HUI3]) with population norms. RESULTS: Baseline scores for all measures were comparable between adalimumab and placebo. All scores for both groups were significantly worse than general population norms (all p<0.0001). Within- and between-group improvements in SF-36 Physical Component Summary and SF-6D scores from baseline to Weeks 12 and 24 were clinically relevant for patients receiving adalimumab. For patients initially randomised to adalimumab, HRQL scores improved from Weeks 25 to 52 and remained relatively stable through 3 years but remained lower than for the general US population at all time points. CONCLUSIONS: Findings demonstrate a significant burden of AS on HRQL. Treatment with adalimumab significantly improved physical functioning and other measures of HRQL compared with placebo. Clinically relevant improvements in HRQL outcomes over 3 years represent a significant benefit of adalimumab. Because of the advanced AS disease, patient health status remained below that of the general population. Treatment earlier in the course of AS may be needed to restore HRQL to the level of the general population.

Computerized administration of health-related quality of life instruments compared to interviewer administration may reduce sample size requirements in clinical research: a pilot randomized controlled trial among rheumatology patients.

OBJECTIVES: Computerized health-related quality of life (HRQoL) administration may facilitate clinical trials incorporating HRQoL assessment in rheumatology patients by reducing sample size requirements. We tested this hypothesis in a pilot randomized controlled trial. METHODS: Chinese-speaking adult rheumatology outpatients were randomized to computerized (PC) or interviewer (IA) administration of the EQ-5D (utility & VAS), Health Utilities Index (HUI2 & HUI3) and Family Functioning Measure (FFM). We compared measurement variability (i.e., variance) between PC and IA for each instrument before (Levene's test) and after adjusting for the effects of age, gender and education (multivariable modeling) and computed the variance ratio (VR) for PC over IA. RESULTS: In 138 patients (mean age: 48), the mean (SD) time for administration was similar for PC (n = 67) and IA (n = 71) at 17.7 (7.94) versus 17.3 minutes (7.49), respectively. More subjects expressed a preference for PC (n = 21) over IA (n = 13). Mean HRQoL scores were not significantly different for PC versus IA except for higher VAS scores with IA (difference -7.7, 95% CI -14.0 to 1.3, p = 0.018). Variances and adjusted VR were smaller with PC for the EQ-5D (adjusted VR 0.34, 95% CI 0.18 to 0.65), HUI3 (0.49, 0.27 to 0.89) and FFM (0.95, 0.61 to 1.46), but larger for the HUI2 (1.30, 0.67 to 2.55) and VAS (1.05, 0.55 to 2.00). CONCLUSION: The reduced variability in 3 of 5 instruments and good acceptance of computerized HRQoL assessment, if confirmed in larger studies, may lead to smaller sample size requirements, with potential reductions in cost and recruitment time for clinical trials and cohort studies.

A comprehensive multiattribute system for classifying the health status of survivors of childhood cancer.

PURPOSE: A multiattribute health status classification system was devised to describe comprehensively the health status of survivors of childhood cancer. METHODS: The system consists of seven attributes: sensation, mobility, emotion, cognition, self-care, pain, and fertility. Three to five levels of functioning are defined for each attribute. Any specific combination of seven attribute levels constitutes a health state. In the first survey, the system was used to classify the health status of 20 children currently undergoing therapy for high-risk acute lymphoblastic leukemia (ALL), Wilms' tumor, or neuroblastoma, and eight who had completed treatment. A second survey consisted of 13 children with brain tumors on active treatment. RESULTS: In general, independent ratings by clinicians were in agreement, and consensus was readily achieved in 1 to 2 minutes per patient. Children on therapy experienced a higher burden of morbidity than those off treatment. Brain tumor patients experienced more morbidity than patients in the first survey. CONCLUSION: The multiattribute system provides a compact but comprehensive tool for long-term follow-up of survivors of childhood cancer. It captures both multiple sequelae and varying levels of severity. By using a mathematical utility function, a single summary score of health-related quality of life may be assigned to each health state. Additional studies to establish reproducibility, validity, responsiveness, and generalizability are indicated.

Health-related quality of life in survivors of Wilms' tumor and advanced neuroblastoma: aA cross-sectional study.

PURPOSE: In pediatric oncology, Wilms' tumor and advanced neuroblastoma represent opposite ends of the spectra of survival probability and therapeutic intensity. Consequently, it was envisaged that survivors of Wilms' tumor would enjoy better health status and health-related quality of life (HRQL) than survivors of advanced neuroblastoma. PATIENTS AND METHODS: Health status questionnaires were sent to the parents of all eligible children and to the children themselves if they were > or = 8 years of age. Responses were received from 84% of 93 eligible families. Responses were converted by established algorithms into levels of two multiattribute health status classification systems known as Health Utilities Index Mark 2 and Mark 3. These systems are linked to measures of preference, in the form of multiattribute utility functions, which provide scores of morbidity for single-attribute levels and of global HRQL for comprehensive health states. RESULTS: A greater burden of morbidity was identified in the survivors of advanced neuroblastoma than in survivors of Wilms' tumor based on the assessments of the parents of these children. In particular, survivors of advanced neuroblastoma exhibited deficits in hearing and speech. It is possible that this morbidity burden reflects the prevalent use of platinum compounds (causing ototoxicity) in this group. Within parent-child dyads there was a high level of percentage agreement on responses in all attributes except cognition. CONCLUSION: Extension of this study to a larger sample size of patients will provide clarification of these observations.

Economic evaluation of allogeneic bone marrow transplantation: a rudimentary model to generate estimates for the timely formulation of clinical policy.

PURPOSE: To provide an evidence-based approach to the formulation of clinical policy with respect to allogeneic bone marrow transplantation (BMT) that involves perceived trade offs between two major factors: costs and consequences. The report also highlights key informational deficiencies. PATIENTS AND METHODS: Adults with acute myeloid leukemia (AML) in second complete remission (2CR) and those with acute lymphoblastic leukemia (ALL) in first complete remission (1CR) were assigned to BMT or control groups solely on the availability of a suitable donor. All hospital-borne costs were estimated, based on services used according to manual chart review, in four categories: diagnostic and therapeutic costs, professional fees, drug costs, and ward costs. Incremental costs and incremental life-years were calculated, and the quotient determined a cost per life-year gained by BMT for AML (2CR) and ALL (1CR). RESULTS: The incremental cost (in 1992 Canadian dollars) per life-year gained by BMT (cost-effectiveness) for AML (2CR) was $29,200; and for ALL (1CR) it was minus $29,200. CONCLUSION: For AML (2CR), allogeneic BMT creates better outcomes than standard treatment, but is more costly. For ALL (1CR), both the costs and outcomes are similar for BMT and standard therapy. Quality adjustments made to life-years gained did not change these conclusions.

Health-related quality of life in survivors of tumours of the central nervous system in childhood--a preference-based approach to measurement in a cross-sectional study.

There is an evident need to measure the comprehensive burden of morbidity experienced by survivors of brain tumours in childhood. To this end, a questionnaire based on the Health Utilities Index mark 2 (HUI2) and mark 3 (HUI3) systems was completed independently for a cohort of such children by their parents, by a nurse, by physicians and by a selected group of the children themselves. Each of the HUI2 and HUI3 systems consists of a multi-attribute health status classification scheme linked to a preference function which provides utility scores for levels within single attributes (domains of health) and for global health states. All eligible families (n = 44) participated. Even cognitively impaired children of at least 9.5 years of age could complete the questionnaire. The greatest burden of morbidity, occurring in two-thirds of children, was in the attribute of cognition. Surprisingly, almost one-third of children experienced pain. Global health status was lowest in children who underwent radiotherapy before the age of 5 years and the corresponding utility scores were related inversely to the volume irradiated. Children with demonstrable disease had lower scores than those in whom disease was not evident. There was a high level of agreement (intraclass correlation coefficients > 0.5) on formal assessment of inter-rater reliability for global health-related quality of life utility scores. The usefulness of measures of health status and health-related quality of life, in children surviving brain tumours, has been demonstrated by this study.

Applicability of the Health Utilities Index to a population of childhood survivors of central nervous system tumours in the U.K.

This paper describes the application of a multi-attribute, preference-linked health status and health-related quality of life measurement system--the Health Utilities Index (developed in Canada)--to a group of subjects in the U.K. Children who had survived tumours of the central nervous system (n = 30, age 6-16 years) formed the study group. Respondents (children, parents, physicians and physiotherapists) found the activity (completion of a 15-item questionnaire) to be acceptable and not burdensome (it was accomplished easily by all children > or = 10 years of age). Instrumental reliability was established by acceptable intra- and interobserver agreement and construct validity was supported by strong similarities between the results obtained in this study and those reported from a similar group of children in Canada. The greatest burden of morbidity was reported for the attributes of emotion and cognition (each affected in > 50% of the children). Pain was surprisingly prevalent (affected in approximately one-third of children). The finding of a large number of unique health states emphasises the complex morbidity burden experienced by these children who self-reported poorer overall health (as reflected in utility scores) than did the proxy respondents. The information obtained from this study is readily interpretable and clinically useful. The results of this study also illustrate that extreme caution must be exercised in undertaking linguistic modifications to established instruments for, in this instance, these resulted in a loss of the ability to detect the most severe emotional morbidity and reduced the comparability of results between studies. With this provision, the Health Utilities Index is evidently applicable in the U.K. and the original version has been recommended for use in brain tumour studies by the U.K. CCSG (the U.K. Children's Cancer Study Group).

Parental perspectives of the health status and health-related quality of life of teen-aged children who were extremely low birth weight and term controls.

OBJECTIVES: To compare the health status and health-related quality of life of teen-aged children who were extremely low birth weight (ELBW) with matched controls from the perspective of their parents. STUDY DESIGN: Geographically defined cohort; longitudinal follow-up; cross-sectional interviews. PARTICIPANTS: parents of 149/169 (88%) ELBW children between 12 and 16 years of age (including 41 children with neurosensory impairments) and 126/145 (87%) parents of term controls. Health status of the teenagers was classified according to the 6 attributes of the Health Utilities Index Mark 2, based on information obtained during parent interviews. Parents were asked to imagine themselves living in their own child's health state and 4 preselected hypothetical health states when providing directly measured standard gamble utility scores. RESULTS: Parents of ELBW children reported a higher frequency and more complex functional limitations than parents of controls for their own children's health status. Also, the mean utilities were lower (ELBW =.91 vs controls =. 97) and the variability in their scores was greater. There were no differences in the valuation of the hypothetical health states provided by parents of ELBW and control children. CONCLUSIONS: ELBW children were reported to have a greater burden of disability than were control children based on parental descriptions. Nonetheless, parents of ELBW children, on average, rated the health-related quality of life of their children fairly high. Thus, differences in reported functional status are not necessarily associated with lower utility scores.

Economic evaluation of cardiac rehabilitation soon after acute myocardial infarction.

Although there are extensive clinical evaluations of cardiac rehabilitation after acute myocardial infarction (AMI), no full economic evaluation is available. Patients with AMI and mild to moderate anxiety or depression, or both, while still in hospital were randomized to either an 8-week rehabilitation intervention (n = 99) or usual care (n = 102). Comprehensive costs and health-related quality of life, measured with the time trade-off preference score, were obtained in a 12-month trial, and together with survival data derived from published meta-analyses, cost-utility and cost-effectiveness of early cardiac rehabilitation were estimated. The best estimate of the incremental net direct 12-month costs for patients randomized to rehabilitation was $480 (United States, 1991)/patient. During 1-year follow-up, rehabilitation patients had fewer "other rehabilitation visits" (p < 0.0001) and gained 0.052 quality-adjusted life-year more than did the group with usual care. The cost-utility ratio was $9,200/quality-adjusted life-year gained with cardiac rehabilitation during the year of follow-up. This economic evaluation of cardiac rehabilitation does not consider the important distinctions between affordability and worth of alternative health-care services. The data provide evidence that brief cardiac rehabilitation initiated soon after AMI for patients with mild to moderate anxiety or depression, or both, is an efficient use of health-care resources and may be economically justified.

Effects on quality of life with comprehensive rehabilitation after acute myocardial infarction.

This investigation was designed to determine the impact of a brief period of cardiac rehabilitation, initiated within 6 weeks of acute myocardial infarction (AMI), on both disease-specific and generic health-related quality of life, exercise tolerance and return to work after AMI. With a stratified, parallel group design, 201 low-risk patients with evidence of depression or anxiety, or both, after AMI, were randomized to either an 8-week program of exercise conditioning and behavioral counseling or to conventional care. Although the differences were small, significantly greater improvement was seen in rehabilitation group patients at 8 weeks in the emotions dimension of a new disease-specific, health-related Quality of Life Questionnaire, in their state of anxiety and in exercise tolerance. All measures of health-related quality of life in both groups improved significantly over the 12-month follow-up period. However, the 95% confidence intervals around differences between groups at the 12-month follow-up effectively excluded sustained, clinically important benefits of rehabilitation in disease-specific (limitations, -2.70, 1.40; emotions, -4.86, 1.10, where negative values favor conventional care and positive values favor rehabilitation) and generic health-related quality of life (time trade-off, -0.062, 0.052; quality of well-being, -0.042, 0.035) or in exercise tolerance (-38.5, 52.1 kpm/min); also, return to work was similar in the 2 groups (relative risk, 0.93; confidence interval, 0.71, 1.64).(ABSTRACT TRUNCATED AT 250 WORDS)

Economics, health-related quality of life, and cost-effectiveness methods for the TACTICS (Treat Angina With Aggrastat [tirofiban]] and Determine Cost of Therapy with Invasive or Conservative Strategy)-TIMI 18 trial.

Concern over escalating health care costs has led to increasing focus on economics and assessment of outcome measures for expensive forms of therapy. This is being investigated in the Treat Angina With Aggrastat [tirofiban] and Determine Cost of Therapy with Invasive or Conservative Strategy (TACTICS)-TIMI 18 trial, a randomized trial comparing outcome of patients with unstable angina or non-Q-wave myocardial infarction treated with tirofiban and then randomized to an invasive versus a conservative strategy. Hospital and professional costs initially and over 6 months, including outpatient costs, will be assessed. Hospital costs will be determined for patients in the United States from the UB92 formulation of the hospital bill, with costs derived from charges using departmental cost to charge ratios. Professional costs will be determined by accounting for professional services and then converted to resource units using the Resource Based Relative Value Scale and then to costs using the Medicare conversion factor. Follow-up resource consumption, including medications, testing and office visits, will be carefully measured with a Patient Economic Form, and converted to costs from the Medicare fee schedule. Health-related quality of life will be assessed with a specific instrument, the Seattle Angina Questionnaire, and a general instrument, the Health Utilities Index at baseline, 1, and 6 months. The Health Utilities Index will also be used to construct a utility. By knowing utility and survival, quality-adjusted life years will be determined. These measures will permit the performance of a cost-effectiveness analysis, with the cost-effectiveness of the invasive strategy defined and the difference in cost between the invasive and conservative strategies divided by the difference in quality-adjusted life years. The economic and health-related quality of life aspects of TACTICS-TIMI 18 are an integral part of the study design and will provide a comprehensive understanding of the impact of invasive versus conservative management strategies on a broad range of outcomes after hospitalization for unstable angina or non-Q-wave myocardial infarction.

Health-related quality of life during post-induction chemotherapy in children with acute lymphoblastic leukemia in remission.

Using a multi-attribute approach, we and others have reported on the comprehensive health status of survivors of cancer in childhood, after completion of therapy. However, there is a dearth of information about health status and associated health-related quality of life (HRQL) in children and adults with cancer, during the treatment process; a deficit which this study was intended to redress. All children (n=18) receiving 'maintenance' chemotherapy for acute lymphoblastic leukemia (ALL) at a single institution were assessed prospectively through a single cycle of therapy. Patients, family members and health care professionals used global ratings, the health utilities index mark 2 (HUI2) and mark 3 (HUI3) classification systems, and detailed descriptions of temporary states of health, to provide independent assessments of health status at weekly intervals. Utility scores were derived for each comprehensive health state and for single attribute levels of the HUI2 system, and for the temporary health state descriptions. The classification of the subjects into the most appropriate temporary health states was challenging even for older children and some of the parents. The HUI instruments were used much more easily and produced highly comparable information. The most frequently affected attributes were pain, emotion and mobility/ambulation; in that order. The global (p=0.005) and specific morbidity burdens were, as predicted, greatest in the middle of the cycle of chemotherapy, reflecting the toxicity of steroid use. HUI2 global utility scores demonstrated moderate responsiveness with an intra-class correlation coefficient of 0.43. Cyclical administration of therapy, in particular corticosteroids, produces important changes in health status and HRQL in children with ALL. The HUI are valid and responsive systems for quantifying the burden of morbidity during the treatment of cancer in childhood. These are sensitive tools which can be used in a wide array of other circumstances to quantify changes in health status.

A multiattribute approach to health-status measurement and clinical management illustrated by an application to brain-tumors in childhood.

A comprehensive multi-attribute system was devised to assess global morbidity burden. The system was used to obtain qualitative and quantitative measures of health-status for a case series of survivors of brain tumors in childhood. Health-related quality-of-life scores for each global health-state were calculated from a utility function derived from measures of preferences collected during surveys of a random sample of parents of school age children in the general public (n=194). Global multi-attribute health status (MAHS) utility scores revealed a wide range of severity in morbidity burden. We suggest that the assessment of health-related quality-of-life should become an integral part of clinical management.

The monetary costs of childhood cancer to the families of patients.

The medical costs of cancer treatment are well described, but there are few reports of expenses incurred by families of children with malignant disease. The objective of this study was to describe the monetary costs borne by families of patients with childhood cancer and to determine whether these costs represent an important component of the burden of illness. Families completed a prospective diary survey about daily expenses incurred during each sample week of therapy. We also undertook a retrospective, cross-sectional questionnaire survey about expenses, incurred during the entire duration of treatment, associated with major or one-time cost items. Seventy families of children treated for high risk acute lymphoblastic leukemia, 19 families of children treated for Wilms' tumor stages 2-5, and 16 families of children treated for neuroblastoma stages 3 and 4 completed diaries or questionnaires detailing the costs resulting from the diseases and their treatment. The mean total expenses (in 1986 Canadian dollars) incurred by families of childhood cancer patients over the entire course of therapy are $26,070 for acute lymphoblastic leukemia? $20,074 for Wilms' tumour, and $10,376 for neuroblastoma. On-going weekly costs rather than major one-time purchases account for the largest share of expenses. Overall, in spite of universal first dollar coverage for medical care in Canada, family-borne costs during the course of these illnesses are at least one-third of the average family's after-tax income.

Does it matter whom and how you ask? inter- and intra-rater agreement in the Ontario Health Survey.

A large amount of information in the 1990 Ontario Health Survey (OHS) was collected from proxy respondents using questions administered in face-to-face interviews. Can this type of information represent candid self-reported measures of health status? Inter-rater agreement was assessed using Cohen's kappa statistic for responses to questions that were answered both by individuals about themselves and by proxies on their behalf. Intra-rater agreement, assessing the effect of mode of survey administration (in-person interviews versus self-completed written questionnaires) on the responses, was also investigated using the kappa statistic. We conclude that: (1) proxy responses in the OHS for impairments of emotion and pain are not reliable indicators of self-response (kappa < 0.32) because proxy respondents consistently under-report the burden of morbidity; (2) levels of morbidity reported by subjects to interviewer-administered questionnaires may underestimate morbidity, relative to morbidity reported by subjects using self-administered questionnaires completed in privacy. We also hypothesize that the relative magnitudes of inaccuracy introduced by interviewer administration relative to proxy reporting depends on the phenomenon being measured. When assessing pain, mode of administration is quantitatively a more important source of disagreement than type of respondent.

Quality of life after myocardial infarction.

The objective of this work was to develop and test a questionnaire to measure health-related quality of life for patients after myocardial infarction (MI). In a cross-sectional survey, 63 patients identified the most frequent and important problems following acute myocardial infarction. The Quality of Life after Myocardial Infarction (QLMI) instrument was developed on the basis of these most frequent and important problems. The QLMI was administered, along with instruments measuring health utilities, social function, and emotional function, in a randomized trial of rehabilitation versus conventional care. The most frequent and important problems fell into areas of symptoms, restriction, confidence, self-esteem, and emotions, each of which is represented in the 26-item QLMI. Effect sizes of the overall QLMI in differentiating between rehabilitation and control groups (0.35), and in detecting improvement over 12 months (1.22) were comparable or larger than any other instrument. The Pearson's correlation coefficient between QLMI administered at 8 and 12 months following AMI varied between 0.75 and 0.87 for the five domains and the overall score. We found substantial correlations of the QLMI with other measures with moderate concordance with predictions about how the instrument should behave if it is a valid measure of health-related quality of life. The QLMI demonstrates a high degree of reliability, and is more responsive than other questionnaires. Relations between the QLMI and other measures provide moderate to strong evidence of its validity in discriminating between patients following AMI according to their health-related quality of life, and in measuring changes in health-related quality of life over time.

Generic and specific measurement of health-related quality of life in a clinical trial of respiratory rehabilitation.

The purpose of this study was to compare the performance of measures of health-related quality of life in a randomized controlled trial of respiratory rehabilitation versus conventional community care for patients with chronic airflow limitation. The study included 89 stable patients with moderate to severe chronic airflow limitation with measurement of health status at 12, 18, and 24 weeks. Outcomes included two disease-specific (the Oxygen Cost Diagram and the Chronic Respiratory Questionnaire [CRQ]) measures, a generic health profile (the Sickness Impact Profile [SIP]), and two utility measures (the Standard Gamble and the Quality of Well-Being index [QWB]). Of the measures, only the four domains of the CRQ (dyspnea, fatigue, mastery, and emotional function) showed statistically significant differences (P < or = 0.05) between treatment and control groups. Correlation between change in the CRQ and change in other relevant measures, including the 6-minute walk test and global ratings of change in dyspnea, fatigue, and emotional function were generally weak to moderate (from 0.19 to 0.51). All correlations between change in the QWB, SIP, and Standard Gamble and other measures were very weak or weak (up to 0.30). Correlation between change in the three generic measures were all very weak (<0.15). The results suggest that unless investigators include responsive and valid disease-specific measures of health-related quality of life in controlled trials in chronic diseases, they risk misleading conclusions about the effect of treatments on health status.

Economic analysis of respiratory rehabilitation.

STUDY OBJECTIVE: We report on the incremental costs associated with improvements in health-related quality of life (HRQL) following 6 months of respiratory rehabilitation compared with conventional community care. DESIGN: Prospective randomized controlled trial of rehabilitation. SETTING: A respiratory rehabilitation unit. PARTICIPANTS: Eighty-four subjects who completed the rehabilitation trial. INTERVENTION: Two months of inpatient rehabilitation followed by 4 months of outpatient supervision. MEASUREMENTS AND RESULTS: All costs (hospitalization, medical care, medications, home care, assistive devices, transportation) were included. Simultaneous allocation was used to determine capital and direct and indirect hospitalization costs. The incremental cost of achieving improvements beyond the minimal clinically important difference in dyspnea, emotional function, and mastery was $11,597 (Canadian). More than 90% of this cost was attributable to the inpatient phase of the program. Of the nonphysician health-care professionals, nursing was identified as the largest cost center, followed by physical therapy and occupational therapy. The number of subjects needed to be treated (NNT) to improve one subject was 4.1 for dyspnea, 4.4 for fatigue, 3.3 for emotion, and 2.5 for mastery. CONCLUSION: Cost estimates of various approaches to rehabilitation should be combined with valid, reliable, and responsive measures of outcome to enable cost-effectiveness measures to be reported. Comparison studies with the same method are necessary to determine whether the improvements in HRQL that follow inpatient rehabilitation are cheap or expensive. Such information will be important in identifying the extent to which alternative approaches to rehabilitation can influence resource allocation. A consideration of cost-effectiveness from the perspective of NNT may be useful in the evaluation of health-care programs.