- © 2008 by American Society of Clinical Oncology
Economic Evaluation Alongside a Clinical Trial of Psycho-Educational Interventions to Improve Adjustment to Survivorship Among Patients With Breast Cancer
- Jeanne S. Mandelblatt,
- Jennifer Cullen,
- William F. Lawrence,
- Annette L. Stanton,
- Bin Yi,
- Lorna Kwan and
- Patricia A. Ganz
- From the Department of Oncology, Georgetown University Medical Center, and Lombardi Comprehensive Cancer Center Cancer Control Program, Washington, DC; and the Division of Cancer Prevention and Control Research, Jonsson Comprehensive Cancer Center, and Departments of Psychology and Psychiatry/Biobehavioral Sciences, University of California at Los Angeles, Los Angeles, CA
- Corresponding author: Jeanne Mandelblatt, MD, MPH, Lombardi Comprehensive Cancer Center, 3300 Whitehaven Ave, Suite 4400, Washington, DC 20007; e-mail: mandelbj{at}georgetown.edu
Abstract
Purpose There is little economic research on psychosocial interventions. We aimed to collect data alongside a randomized trial to compare the costs and benefits of three psycho-educational strategies to improve transition to cancer survivorship.
Methods We evaluated the incremental delivery costs per unit increase in energy (using the Medical Outcomes Study vitality scale) or decrease in distress (from the Revised Impact of Events Scale) in the 6 months postintervention. We also evaluated 1-year post-treatment health care costs.
Results The costs of the control, video, and video plus counseling arms were $11.30, $25.85, and $134.47 per person, respectively. The video costs were $2.22 per unit increase in energy compared with control; among women who were the least prepared for transition, the video was more effective, resulting in even lower costs. The video cost $7,275 per unit change in distress versus control, but costs were lower in the subgroup least prepared for transition ($355). The counseling arm was more expensive and less effective than the video for virtually all end points. However, in one group, women more prepared for transition, counseling cost $1,066 per unit decrease in distress compared with the video. Health care costs tended to increase as intervention intensity increased.
Conclusion There are no standards for evaluating cost-effectiveness of trials with psychosocial end points. In this trial, the educational video was the most cost-effective way to improve transition to survivorship. It will be important to confirm whether there is an increased use of services after such interventions and if this represents appropriate use of rehabilitative and supportive care or over-use.
INTRODUCTION
Breast cancer is the second leading cancer in United States, women and as a result of screening and treatment advances, there are an estimated 2.3 million breast cancer survivors alive today.1-4 Given this large and growing population, increasing attention is now being focused on the quality of life of survivors.4 In our prior work, we conducted a clinical trial of psycho-educational approaches for improving transition from active treatment to survivorship. The trial evaluated three strategies of increasing intensity (and costs).5 In this article, we evaluate the costs relative to the benefits of each approach. We hypothesized that a video plus counseling and a booklet (the most expensive intervention) would be more effective than a video and booklet (the next most expensive) and that both would be more effective than a booklet alone (the least expensive), so that the costs of the most intensive approach would be offset by its additional benefits. We also explored whether the most intensive strategy would lower short-term health care costs. The results are intended to inform decisions about dissemination of the interventions.
METHODS
This economic analysis was conducted alongside a clinical trial of psycho-educational interventions for women completing breast cancer treatment. Using a societal perspective, we compared the incremental costs and effects of three different strategies for improving the transition to survivorship in the 6-month period after active treatment completion. The primary measures of effect are changes (improvements) in psychological status. We did not use life years saved because adaptation to survivorship affects quality but not length of life. Costs include intervention delivery (exclusive of research development) and patient time as per current standards for economic studies.6 We also examined health care costs in the 1 year after treatment by arm.
Patient Recruitment
Participants were recruited from three geographical sites: Los Angeles, CA, Washington, DC, and Lawrence/Kansas City, KS. The study is described in detail elsewhere.5 Briefly, women were eligible if they had received surgery for invasive breast cancer of any size or nodal status. Exclusion criteria included use of neoadjuvant chemotherapy, high-dose chemotherapy with bone marrow or stem-cell rescue or protracted reconstructive surgery, and inability to read and write in English. Institutional review board–approved informed consent was obtained for all participants.
Data Collection
Women completed a baseline survey 4 to 6 weeks after primary treatment; the survey was repeated at 6 and 12 months after the intervention. We focused on the 6-month results because this is the period of immediate transition and by 12 months, most women have adjusted to survivorship. The research staff used weekly logs to record the time and resources used to deliver the interventions. Finally, participants were mailed calendars every 3 months to document all health services used; those not returning calendars were provided mail and telephone reminders.7 We have found more than 90% agreement between self-report calendars and medical record data.8
Exactly 418 women (74.9%) completed the original trial; 396 women (94.7%) returned calendar data. Seven women were missing treatment or other information, leaving 389 women (69.7%) for economic analyses. There were no significant differences between those completing economic data and those not doing so.
Intervention Arms
We compared three strategies for improving survivorship transitions. The least expensive approach (a booklet control condition) is compared with the next most expensive (an educational video plus the control booklet); the video is compared incrementally to the most expensive option (counseling plus the video plus the booklet). The interventions are described in detail elsewhere.5 In brief, women randomly assigned to the control arm were mailed a copy of the 1994 National Cancer Institute publication Facing Forward.5a Women assigned to the educational videotape arm also received the booklet and a videotape entitled, “Moving beyond Cancer.” This 23-minute film was designed to address re-entry challenges in physical health, emotional well-being, interpersonal relations, and life perspectives. The video includes peer modeling of active coping approaches to fatigue and other survivorship concerns. The last arm included psycho-educational counseling (one individually conducted in-person session and one telephone session) plus the video and booklet.
Measures of Effect
The outcomes for this analysis were changes in distress and energy 6 months postintervention, as measured by the Revised Impact of Events Scale (IES-R) and the Medical Outcomes Study (MOS) Short-Form (SF) 36 vitality scale, respectively. The vitality scale captures energy level and fatigue. Because of skewness in the IES-R score distribution, a logarithmic transformation was performed to produce a relatively normal distribution of IES scores; the scores were retransformed for cost-effectiveness analysis. Higher change scores for the MOS represent more energy and a higher IES change score represents less distress pre- to postintervention. There were some baseline imbalances in the study arms,5 so we controlled for baseline depression in calculating outcomes. We also evaluated costs stratified by baseline level of preparedness for transition because the original trial showed differences in outcomes by this variable.5 Preparedness was defined by responses to two items (“Overall, I feel very well prepared about what to expect during recovery; Overall, I feel the medical team has done a great deal to prepared me for what to expect during my recovery.”).
Finally, because we hypothesized that the more intensive strategies would lower health care use as a result of better adaptation, we also evaluated the overall 12-month health care costs by study arm.
Assignment of Costs for Intervention Arms
Treatment intervention costs were treated as fixed values based on the costs of delivering the intervention, including reproduction of study materials, distribution/mailing of study materials, staff/counselor time, and patient time costs (Table 1). Research development costs and their associated overhead were excluded because we are interested in dissemination. The value of patient-time costs were approximated using United States Bureau of Labor Statistics data for the mean hourly wage for all occupations in 2002 when the study was conducted.9 The counselor time costs were estimated from the counselors and data from a similar study and valued using wage rates.10
Intervention Delivery Costs per Participant by Study Arm in US Dollars
Assignment of Costs for Health Care Services
Details of the costs are summarized in a previous publication.7 Briefly, all episodes of health care utilization in the 12 months after active treatment ended were coded using Current Procedural Terminology, Fourth Edition.11 Coding was reviewed by two coders (W.F.L., J.C.); any disagreements were resolved by consensus. Coders were blinded to the study arm.
Costs of outpatient and laboratory services were calculated using the average Medicare reimbursement. Costs for inpatient hospitalizations for specific procedures (eg, reconstruction surgery) were determined by average Medicare reimbursement; costs for hospitalizations for specific conditions (eg, pneumonia) were estimated using an average daily cost of hospitalization12 multiplied by the number of days the participant was hospitalized. Costs of phone calls to providers were estimated as taking 5 minutes of time and multiplied by an average hourly wage. We did not include the costs of hormonal therapy; use of this modality did not differ by study arm.5
Cost and Effects Calculation
To assess costs per unit change in adaptation to survivorship, we examined the incremental difference between intervention costs per unit change in our two primary health outcomes (mean change in MOS Vitality and IES-R scores between baseline to 6 months’ follow-up), controlling for baseline depression. We then calculated the incremental cost divided by the incremental effect of the more expensive compared with the next least expensive (and intensive) intervention. Results are presented overall and stratified by level of baseline preparedness. Discounting was not performed because of the abbreviated follow-up time (≤ 1 year). Health care costs by arm were calculated using regressions controlling for baseline depression and site (to capture any geographic differences in care not addressed by site-blocked random assignment). Because values were skewed, values were log transformed to meet regression assumptions and then retransformed to dollars for presentation. All statistical analyses were conducted using SAS version 9.1 (SAS Institute, Cary, NC). Differences in health care costs by arm were tested using Mantel-Haenszel tests for trend. Uncertainty in estimates was tested using one-way sensitivity analysis (eg, one standard deviation difference or one half or double the value).
RESULTS
Women in the study ranged from 27 to 87 years (mean, 58.2 years) and were largely well educated and had a high income; 13% were non-white (Table 2).
Demographic Characteristics of Breast Cancer Survivors by Intervention Arm (n = 389)
Costs per Unit Effect
The incremental costs for the most intensive arm (counseling plus video plus booklet) were fairly high but, contrary to expectation, this arm was not more effective in increasing energy or decreasing distress than the other arms. Therefore, this approach was dominated by the other arms. There was one exception to this: the counseling arm was more effective in decreasing distress than the video among women who perceived themselves to be more prepared for transition to survivorship, resulting in an incremental cost of $46,604 for each unit decrease in distress (Table 3).
Incremental Costs and Effects by Intervention Arm for 6-Month IESR Change in Breast Cancer Survivors (n = 389)
The arm that used the video plus the booklet improved vitality compared with the booklet alone and had low costs ($2.22 per unit improvement in energy). The video arm was also more effective in improving energy than the control among the sub-group of women who felt less prepared for the transition to survivorship than for women who perceived themselves as more prepared (the opposite direction of interaction as distress). Therefore, this intervention was even less costly per unit benefit in the less prepared compared with the more prepared group (Table 4).
Incremental Costs and Effects by Intervention Arm for 6-Month MOS SF 36 Energy Scale Change in Breast Cancer Survivors (n = 389)
Contrary to expectation, there were no significant differences in health care costs over the 12 months’ postintervention by study arm (Table 5). However, there was a trend for women in the most intensive arm to have higher costs for all categories of services (P ∼ .10).
Cost (US dollars) of Health Care Utilization* Among Early-Stage Breast Cancer Survivors in the 1 Year After Completion of Active Treatment by Type of Service and Intervention Arm (n = 388)†
In sensitivity analyses, we evaluated how effective an intervention arm would have to be to be as cost-effective as the next least intensive arm. First, we considered the situation where counseling might be less expensive than in the trial. However, because counseling was less effective than the video for nearly all comparisons, varying its costs did not change conclusions. Next, we estimated how much more effective the counseling arm would need to be to be economically equivalent to the video arm. For increasing energy, the counseling arm would have had to have been six times more effective than the video arm for the results to be economically equivalent (or 7.6 and 4.0 times more effective in the low and high preparedness subgroups, respectively). In terms of distress, the counseling arm would only have needed to be 20% more effective than the video to be economically equivalent. This result is due to the fact that the video was less effective (and more costly) in impacting distress relative to energy and the fact that counseling dominated the video for women who are most prepared for transition (Table 3).
DISCUSSION
Although a number of studies have addressed the economic benefit of interventions to increase screening participation13-17 and cancer treatment,18-22 there are no studies that we are aware of that have addressed costs and effects of behavioral interventions targeting adjustment to breast cancer survivorship. We found that an approach with moderate costs, such as distribution of an educational video that portrays peer modeling of coping, might be considered to be a reasonable investment for improving adaptation, especially increasing energy levels, compared with distributing print material. However, contrary to expectation, combining the video with individual counseling sessions was not sufficiently more effective to warrant the additional expenditure. Our results also underscore the fact that the most cost-efficient approach to improving adaptation will be to match the intervention with need based on individual factors, such as level of preparedness for transition.
We have made some progress in developing spending thresholds for health interventions based on costs per life years saved,6,23 and guidelines for conduct of economic analyses now recommend incorporating quality of life into outcome measures.6 However, there are no standards for acceptable costs of achieving more intermediate outcomes such as adaptation to survivorship or quality of life per se. Willingness to pay (WTP)is one method that might be used to develop such standards.24,25 However, one study has shown that quality of life ratings are not a significant predictor of WTP,26 and other work has suggested that WTP results are sensitive to race/ethnicity and culture,27,28 indicating that the work in this area will need to consider equity issues and conceptual underpinnings of the approach.26
An economic approach to intermediate psychosocial outcomes will also have to consider the magnitude of unit changes that would be considered personally and clinically meaningful. For instance, is a one-unit increase on an energy scale meaningful in day to day activities? Although we have no direct evidence to answer this question, the magnitude of effect of the interventions on energy and distress was comparable to those seen in other commonly applied psycho-educational interventions for cancer patients29,30 and are likely to be clinically relevant.5 Fatigue and low energy are the most commonly reported side effects of cancer treatment and cause significant impairment in emotional, social, and employment function.31,32 Therefore, investments in improving these intermediate outcomes can have important collateral benefits in individual and family productivity.33 Even if greater change was considered important (eg, 8 to 10 unit changes), given its low costs, the video might still be considered a reasonable investment. In addition, the costs of the video intervention were much lower than the costs of commonly used stress management approaches for improving quality of life during chemotherapy29 or well-accepted interventions for increasing breast cancer screening.34 Moreover, the majority of the costs of the video were patient time costs and most survivors would be likely to be willing to spend this time. Therefore, it seems reasonable to say that the costs of the video are within accepted ranges of health care expenditures for comparable outcomes. Development of standards for cost thresholds and units of effects will be needed to facilitate comparisons across psychosocial interventions targeting shorter- and intermediate-term quality of life.
We expected that if the interventions were effective in meeting patient needs, health care costs might be lower. Unfortunately, over the short-term period of observation, we did not find differences in overall health care use or costs because the study was not powered for this outcome in the original design. There were interesting trends, however, with health care costs increasing as the intensity of the intervention increased. Our results are in contrast to those of an earlier study where Stanton et al8 found that participants in a study of expressive writing during the transition to survivorship had fewer appointments for cancer-related morbidity than women in the control group, although there was no difference in use of other services. Our results cannot rule out a combination of lower cancer morbidity-related visits combined with higher use of appropriate cancer supportive care over the 12 months of observation. It will be important to confirm whether there is, in fact, an increased use of services after use of video and counseling interventions, and if so, whether this represents appropriate use of rehabilitative and supportive care (that was being otherwise under-used) or induced over-use.
This study also demonstrates that it is feasible to collect economic data alongside psychosocial interventions.35 Other researchers have conducted economic evaluations of behavioral interventions to improve cancer screening,15,16 and one study has been done to test a stress management intervention during chemotherapy,29 but this is the first study that we are aware of that focuses on cancer survivorship. With the current priority given to survivorship research by the National Cancer Institute and the Institute of Medicine,4,36 more interventions can be expected to be developed and tested.8,37 Our approach should be broadly portable to assessing the economic consequences of these psychosocial interventions.
There are several limitations that should be considered when evaluating our results. First, the patients included in this trial had high education and income. It is possible that less advantaged women might be more vulnerable to poor adaptation and might benefit more from intervention or require different or more costly types of interventions. This will be an important area for future research. Second, this study was limited to volunteers for a randomized controlled trial in three geographic regions and may not be generalizable to all breast cancer survivors in the United States. Third, the study arms were unbalanced for baseline depression. Although we controlled for depression, it is possible that there are residual unmeasured differences between the groups. This would only bias conclusions about relative costs and effects if any unmeasured differences by arm were differential with respect to outcome effectiveness; we have no data to assess this possibility. Fourth, although the video produced benefits at the lowest costs, this benefit was for a defined, limited period of time: the 6 months after active treatment ended. The video tended to produce continued positive effects at 12 months, but this trend was not statistically significant.5 This finding is consistent with other research that shows that most women return to baseline status by 12 months.5 Next, we were limited by self-report of health care use, but this should not have affected internal comparisons by arm. Finally, we could not separate cancer-related services from care for other noncancer conditions or care for medical, psychological, and rehabilitative services and did not collect pharmaceutical costs. Collection of these data should be considered in future economic companions to clinical trials of psychosocial interventions.
Despite these limitations, we have found that the most labor-intense intervention that includes one-on-one counseling does not seem to be warranted in terms of costs and effects. However, distribution of the educational video would seem to be a reasonable investment relative to return on improved adaptation. As more interventions are developed to improve the quality of life for survivors, it will be important to develop standard approaches to evaluating the cost-effectiveness of psycho-social interventions and discuss what patients and payers would consider a reasonable monetary investment per unit change in meaningful patient-reported psychosocial outcomes.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The author(s) indicated no potential conflicts of interest.
AUTHOR CONTRIBUTIONS
Conception and design: Jeanne S. Mandelblatt, Jennifer Cullen, William F. Lawrence, Annette L. Stanton, Patricia A. Ganz
Financial support: Patricia A. Ganz
Administrative support: Patricia A. Ganz
Provision of study materials or patients: Annette L. Stanton, Lorna Kwan, Patricia A. Ganz
Collection and assembly of data: Jeanne S. Mandelblatt, Jennifer Cullen, William F. Lawrence, Annette L. Stanton, Bin Yi, Lorna Kwan, Patricia A. Ganz
Data analysis and interpretation: Jeanne S. Mandelblatt, Jennifer Cullen, William F. Lawrence, Bin Yi, Lorna Kwan
Manuscript writing: Jeanne S. Mandelblatt, Jennifer Cullen
Final approval of manuscript: Jeanne S. Mandelblatt, Jennifer Cullen, William F. Lawrence, Annette L. Stanton, Bin Yi, Lorna Kwan, Patricia A. Ganz
Footnotes
-
Supported by Grants No. R01- CA63028 and K05 CA 96940 (J.M.) from the National Cancer Institute.
-
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received September 2, 2007.
- Accepted December 7, 2007.
