| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author 2007. Published by Oxford University Press.
ARTICLE |
Patient Time Costs Associated With Cancer Care
Affiliations of authors: Division of Cancer Control and Population Sciences, National Cancer Institute, Bethesda, MD (KRY, WWD, MLB, JLW); Massachusetts General Hospital Cancer Center and Harvard Medical School, Boston, MA (EBL); Information Management Systems, Inc, Rockville, MD (AF, MT)
Correspondence to: K. Robin Yabroff, PhD, MBA, Health Services and Economics Branch, Applied Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Executive Plaza North, Rm. 4005, 6130 Executive Blvd., MSC 7344, Bethesda, MD 20892-7344 (e-mail: yabroffr{at}mail.nih.gov).
 |
ABSTRACT |
|---|
 TopNotes Abstract Context and CaveatsMethodsResultsDiscussionReferences |
|---|
BACKGROUND: Although costs of medical care for cancer have been investigated extensively, patient time costs associated with cancer care have rarely been estimated systematically. In this study, we estimated patient time costs associated with cancer care in patients aged 65 years and older in the United States.
METHODS: We identified 763 527 patients with breast, colorectal, corpus uteri, gastric, head and neck, lung, melanoma of the skin, ovary, prostate, renal, and urinary bladder cancers from linked Surveillance, Epidemiology, and End Results–Medicare files and 1 145 159 noncancer control subjects among Medicare enrollees who were matched by sex, age-group, and geographic location. Frequency of service use was calculated by category for patients and control subjects using Medicare claims data from 1995 to 2001. For each service category, time estimates were combined with service frequency and an hourly value of patient time. Net patient time costs were summed in the initial, continuing, and last-year-of-life phases of care for each tumor site. Net time cost estimates for the initial phase of care were applied to national estimates of numbers of new cancers in 2005 to obtain national time costs for the initial phase of care.
RESULTS: Net patient time costs during the initial phase of care ranged from $271 (95% confidence interval [CI] = $213 to $329) and $842 (95% CI = $806 to $878) for melanoma of the skin and prostate cancer, respectively, to $5348 (95% CI = $4978 to $5718) and $5605 (95% CI = $5273 to $5937) for gastric and ovarian cancers, respectively. Net patient time costs for care during the last year of life ranged from $1509 (95% CI = $1343 to $1675) for melanoma of the skin to $7799 (95% CI = $7433 to $8165), $7435 (95% CI = $7207 to $7663), and $7388 (95% CI = $7018 to $7758) for gastric, lung, and ovarian cancers, respectively. In 2005, patient time costs for the initial phase of care were $2.3 billion.
CONCLUSIONS: Patient time costs for cancer care in the United States are substantial and vary by tumor site and phase of care, likely reflecting differences in stage at diagnosis and availability and intensity of treatment.
Prior knowledge Most estimates for costs of cancer care are based on costs of medical treatments. The costs associated with patient time spent travelling to and from care, waiting for, and receiving care are rarely reported. Study design Estimates of medical service use in US cancer patients and similar non-cancer patients aged 65 years and older based on Medicare claims data were combined with service-specific time estimates from national surveys and an hourly value of time. Contribution Patient time costs vary widely by cancer type and phase of care. From the estimated number of patients diagnosed with cancer in 2005 in the US, the estimated patient time cost during the first year of treatment for all 11 types of cancer examined was $2.3 billion. Implications Costs of time associated with medical care for cancer patients are substantial. Differences observed among cancer type and phase of care are likely due to differences in stage at diagnosis and availability and intensity of treatment. Limitations The authors used data from Medicare claims and national surveys to calculate time estimates rather than direct measurement for most services. Although Medicare claims accurately record types of treatment received and duration of hospitalization, they do not provide hourly estimates of treatment duration or travel and waiting time. Survey data based on national averages may underestimate time spent by cancer patients travelling, waiting, and receiving care.
|
The majority of studies of the cost of cancer care have focused on the direct monetary medical costs of care (1). Direct, nonmedical costs of care, including patient time associated with medical care, have rarely been measured systematically (1), in part, because these data are not collected routinely. Time spent traveling to and from care, waiting for appointments, receiving consultation from providers, and receiving services or procedures represents time not spent pursuing usual activities, including work and leisure. Patient time costs therefore represent an important component of the burden of illness from the perspective of society, employers, and patients. Organizations such as the Panel on Cost-Effectiveness in Health have recommended including patient time costs in cost-effectiveness evaluations of interventions (2), yet few studies have estimated time costs empirically (1).
Recently, investigators used data from Surveillance, Epidemiology, and End Results linked to Medicare claims (SEER–Medicare) to estimate service frequencies for colorectal cancer patients and similar noncancer control subjects, combined these frequencies with estimates of time for specific services, and calculated net patient time costs associated with colorectal cancer care (3). This approach systematically measured medical care received by cancer patients and similar patients without cancer from diagnosis to death and yielded population-based estimates of net patient time costs. Colorectal cancer patient time costs were reported to be 19% of direct medical costs in the first year following diagnosis, 37% of direct medical costs in the last year of life, and 16% of direct medical costs in the continuing phase of care (3). In this study, we applied this method to estimate time costs of patients with cancer at 11 of the most prevalent tumor sites in the United States: breast, colorectal, corpus uteri, gastric, head and neck, lung, melanoma of the skin, ovarian, prostate, renal, and urinary bladder cancers.
|
Methods |
|---|
|
TopNotesAbstractContext and CaveatsMethodsResultsDiscussionReferences |
|---|
Overview
Briefly, we applied the SEER-Medicare–based method to patients with cancers of 11 anatomic sites and estimated frequencies of relevant medical services, including physician office visits, emergency room visits, chemotherapy, radiation therapy, hospitalizations, and ambulatory surgeries (3). In addition, we estimated frequencies of imaging procedures (e.g., magnetic resonance imaging [MRI] and computed tomography [CT] scans) and blood transfusions for all tumor sites and injectable hormone therapy for prostate cancer patients. We then estimated patient time spent traveling to, waiting for, and receiving care for each service category using data from national surveys and other sources. Average service frequencies were then combined with estimates of patient time for each category of service, and the value of patient time was assigned. Net patient time costs associated with cancer care for each tumor site were calculated by subtracting mean values for control subjects from mean values for patients by service category in the initial, continuing, and last-year-of-life phases of care. Net patient time costs were then summed by phase of care. To estimate the magnitude of patient time costs in the initial phase of care nationwide, we multiplied the sum of the net costs in the initial phase by the estimated number of new cancer patients in the United States in 2005.
Data Sources
The SEER tumor registry program collects information about all incident cases among geographically defined areas that included 14% of the US population during the time of our study (4). Among persons aged 65 and older who have been diagnosed with cancer according to the SEER database, 95% have been linked with Medicare enrollment data (4). Tumor stage and histology information was obtained from the SEER data. Patient demographic characteristics were obtained from Medicare enrollment data, and service frequencies for patients and noncancer control subjects were obtained from Medicare claims. The SEER–Medicare database was created with the permission of the National Cancer Institute, the Centers for Medicare and Medicaid Services (CMS), and the Institutional Review Boards of each of the participating SEER registries. The final analytic dataset used for this analysis was stripped of identifiers. A more detailed description of SEER–Medicare is available at http://healthservices.cancer.gov/seermedicare/.
Selection of Cancer Patients and Control Subjects
A total of 916 105 patients with breast, colorectal, corpus uteri, gastric, head and neck, lung, melanoma of the skin, ovarian, prostate, renal, and urinary bladder cancers who were initially diagnosed between 1973 and 1999 and were aged 65 years and older during the study observation period, 1995–2001, were selected from the SEER–Medicare database to be included in the cancer cohort. Cancer patients with a prior cancer diagnosis (n = 109 011) or who were identified as having had cancer through a death certificate or autopsy (n = 4647) were excluded. Histology codes for all tumor sites were reviewed by a practicing oncologist to exclude unusual tumor types (n = 35 437; details available from authors by request). The remaining 767 010 cancer patients were eligible for inclusion in the cohort. Stage of disease at initial diagnosis is reported using the SEER historic staging system to allow comparison of stage distribution across tumor sites and years of diagnosis.
Potential control subjects were individuals without any cancer diagnosis recorded by SEER and aged 65 years and older during the observation period, 1995–2001. A total of 1 145 159 control subjects were selected from a 5% random sample of Medicare enrollees and frequency matched to patients by sex, 5-year age strata (65–69, 70–74, 75–79, 
80 years), and SEER registry area. Because prior studies have shown differences in service use in urban and rural areas (5), patients and control subjects were classified as living in either Metropolitan Statistical Areas (MSAs) or non-MSAs using the US Department of Agriculture urban–rural continuum coding system (6) by linking county of residence to the Area Resource File (7).
Estimating Service Frequencies for Each Medical Service Category
As in the prior study of patient time costs associated with colorectal cancer care (3), we used Medicare claims to identify six categories of medical services: physician office visits, emergency room visits, chemotherapy, radiation therapy, hospitalizations, and ambulatory surgery. We added two service categories, imaging and blood transfusions, for all tumor sites. We also included a category of injectable hormone therapy for prostate cancer patients only. Claims data during the observation period, 1995–2001, were used for both patients and control subjects.
Historically, managed care organizations have not been required to submit claims or encounter data for specific services received by their Medicare enrollees. For months in which patients received coverage through managed care or were without both Medicare part A and B, their service utilization and follow-up time were excluded because these data would not completely capture the care received during this period. An additional 3483 cancer patients were excluded because they were enrolled in managed care or did not have both Medicare part A and B during the observation period, and the remaining 763 527 cancer patients were included in the cohort. We were not able to include home health and hospice services in our analyses because service units were not identifiable from the Home Health and Hospice claims files.
Phase of Care Definitions
Phase of care definitions were adapted from standard methods used to estimate direct medical costs (8). For cancer patients, months of observation and service frequencies between 1995 and 2001 were divided into three clinically relevant phases of care: initial, last year of life, and continuing care, based on the month of service on the Medicare claim. Date of death (or its absence) in the Medicare enrollment file through 2002 was used to determine vital status. The initial phase was defined as the first 12 months following diagnosis, the last-year-of-life phase was defined as the final 12 months of life, and the continuing phase was defined as all months between the initial and last-year-of-life phases of care. Not all cancer patients contributed to all phases of care, however. For patients who survived less than 24 months after diagnosis, the final 12 months of observation and services were allocated to the last-year-of-life phase because the content of care for patients with short survival is more similar to the last-year-of-life phase than to the initial phase (9). The remaining months of observation and services were allocated to the initial phase, with no contribution to the continuing phase. Patients who were diagnosed before 1994 and survived beyond 2002 contributed months and services only to the continuing phase. Within each phase of care, average monthly estimates of service frequency for each service category were calculated.
Because control subjects did not have a date of cancer diagnosis, they were randomly assigned a "pseudodiagnosis date" that corresponded to the date of diagnosis of one of the cancer patient cohorts. The pseudodiagnosis date and the date of death (or its absence) were used to define phase of care for control subjects. Months of observation and services were assigned to phases of care in the same manner used with patients. In addition to frequency matching by sex, 5-year age-group, and SEER area strata, control subjects were also matched to patients by phase of care in up to a 1 : 5 patient : control subject ratio based on the number of control subjects available for each stratum. To reflect service use associated with cancer care in the last year of life, cancer patients in the last-year-of-life phase of care who died of cancer were matched to control subjects in the continuing phase, and cancer patients in the last year of life who died of other causes (e.g., accident) were matched to last-year-of-life control subjects. As with patients, average monthly estimates of service frequency were calculated for each phase of care for control subjects.
Identification of Services From Claims Data
The Medicare files, Common Procedure Terminology codes (CPT-4), International Classification of Diseases, 9th Revision (ICD-9), codes, and other unique CMS codes used to identify medical service categories (i.e., physician office visits, emergency room visits, chemotherapy, radiation therapy, hospitalizations, ambulatory surgeries, imaging procedures, i.e., MRI, CT scans, and blood transfusions for all tumor sites and injectable hormone therapy for prostate cancer patients) are listed in Supplementary Table 1 (available online). To avoid double counting patient time associated with medical care for the same services identified in more than one file, we developed a hierarchy based on the dates and categories of services and the Medicare file that was the source of the claim. During hospitalization (duration was defined by the number of days elapsed between admission and discharge), no other physician supplier or outpatient services were counted because patient time was already identified for the duration of the hospitalization. Similarly, no other physician services were counted on the same day as an outpatient surgery. Chemotherapy, radiation therapy, or outpatient surgeries that were identified from both the Outpatient and Physician Supplier files on the same day (or within 3 days) were counted only once.
Time Estimates for Medical Service Categories
Estimates of patient time associated with travel to and from care, waiting for appointments, and receiving care were calculated separately for each service category as in the prior study (3) and are listed in the Supplementary Table 1 (available online). We used secondary data from national sources to estimate time, when possible. The average time spent with a physician during an office visit was calculated from the 2001 National Ambulatory Medical Care Survey (10). Patient time for emergency room visits was calculated from the difference between arrival time and discharge time from the 2002 National Hospital Ambulatory Medical Care Survey Emergency Department (11). Chemotherapy duration was estimated using a weighted average of CPT-4 chemotherapy infusion codes with time descriptions from cancer patients separately for each tumor site cohort. Time for imaging studies (i.e., CT and MRI) and blood transfusions were estimated from guidelines for these procedures (12,13). Times for radiation therapy and hormone therapy injection were assumed to be 30 and 15 minutes, respectively, based on input from practicing oncologists. Patient time in the hospital was calculated from Medicare claims as the difference between admission and discharge dates and multiplied by 16 hours, an estimate of waking hours that could alternatively be spent pursuing usual activities, including work and leisure. Patient time spent in ambulatory surgical and recovery suites was calculated from the 2001 Medicare Current Beneficiary Survey (14) as the difference between admission time and discharge time for outpatient surgeries.
Estimates of round-trip travel times and waiting times for medical care were estimated from related questions from the 1992 National Health Interview Survey (15). Travel time was added to all service time estimates. Chemotherapy and radiation therapy occurring on the same day as a physician visit were categorized separately, but because we assumed that patients would make only a single round-trip for multiple services occurring on the same day, we assigned a single travel time (rather than a travel time for each service) to estimates of time for these services. Waiting times were added to estimates for office visit, chemotherapy, radiation therapy, imaging, blood transfusion, and hormone therapy. Time estimates for emergency room visits, hospitalizations, and outpatient surgeries were based on the difference between admission and discharge time, so waiting time was not added to these estimates. Patient time estimates within phase of care were calculated separately by MSA and non-MSA status.
Because time estimates for physician visits, emergency room visits, outpatient surgeries, round-trip travel, and waiting time were calculated from secondary data, means and standard errors were estimated directly. On average, the standard error was 1.5% of the mean time estimate for these services. To estimate a standard error for delivery of services not calculated from primary data (i.e., chemotherapy, radiation therapy, imaging, blood transfusion, and hormone therapy), we multiplied the mean time estimate for each service by 1.5%.
The prior study of patient time costs used the median US wage rate in 2002 from the Bureau of Labor Statistics (16), $15.23 per hour, as the value of patient time (3). We used the same value in this study to allow for comparisons between studies.
Statistical Analysis
Calculating patient time costs associated with cancer care. For each tumor site, service counts were calculated by category of service (i), phase of care (j), patient/control status of subject (k), and MSA status (m). The number of visits in each category of service in each phase of care was summed and divided by the number of patient months of observation to develop an average monthly service count labeled nijkm. We report service frequencies for hospitalizations and physician visits by phase of care.
To estimate patient time, PTijk, for each service category, phase of care, and case status, the following formula was used:
 | [1] |
in which Tim is the average patient time by service category and MSA status and pjm is the proportion of events by phase of care and MSA status. We calculated standard errors and 95% confidence intervals (CIs) for all patient times and time costs using the formula for the variance of a product of two independent variables (17). Net patient time costs of cancer care were obtained by subtracting the monthly average service frequencies in control subjects from patients in each phase of care, applying the formula described above to estimate patient time, and multiplying patient time by the value of $15.23 per hour.
Sensitivity analyses. We conducted sensitivity analyses on key assumptions related to time costs identified in the prior study (3)—hospitalization time and the method used to value patient time. We also selected alternate approaches for measuring hospitalization time and methods for valuing time from the sensitivity analysis conducted in the prior study (3). We assessed the impact of using the entire time spent in the hospital, 24 hours, rather than the estimated 16 waking hours, to value patient time during hospitalization. We also assessed the impact of other methods for valuing patient time, including the median wage rate among individuals aged 65 and older ($14.10 per hour) (16) and a hedonic value of life of $150 000 per year ($17.12 per hour) (18), rather than the median wage rate for individuals of all ages ($15.23 per hour) to value patient time.
National estimates of patient time costs by phase of care. Finally, we multiplied estimated net patient time costs of cancer care for the initial phase of care by overall US estimates of new cancer patients in 2005 (19) by tumor site to develop national estimates of patient time costs for this phase of care. We also multiplied monthly net patient time costs for colorectal cancer (the only site for which US prevalence estimates are currently available by phase of care) in the continuing care and last-year-of-life phases by estimates of colorectal cancer prevalence in the continuing and last-year-of-life phases of care in 2005 [10 718 556 months and 1 047 523 months, respectively (20)].
|
Results |
|---|
|
TopNotesAbstractContext and CaveatsMethodsResultsDiscussionReferences |
|---|
Sample Characteristics
At least 2000 cancer patients and 10 000 matched control subjects were included in the sample for each of the 11 tumor sites and phases of care (Table 1). The match ratio for each tumor site and phase of care was based on the available number of potential control subjects for each stratum used for matching (i.e., age-group, sex, SEER registry area, and phase of care) up to a 5 : 1 ratio, to ensure similar distributions of patients and control subjects across strata. The distribution of MSA and non-MSA residence among control subjects was similar to that for patients for each tumor site and phase of care.
|
Stage distribution at initial diagnosis varied considerably across the 11 tumor sites (Table 2). More than half of the patients with breast, corpus uteri, melanoma of the skin, prostate, and urinary bladder cancers were initially diagnosed with localized disease; a smaller percentage of patients with tumors at other sites were diagnosed with localized disease. More than 10% of patients with gastric, lung, ovarian, and renal cancers were initially diagnosed with distant disease.
|
Service Counts for Hospitalizations and Physician Visits by Phase of Care and Tumor Site
For hospitalizations and physician visits, cancer patients had statistically significantly higher service use than matched control subjects in almost every tumor site and phase of care (Table 3). In the initial phase of care, the mean hospitalization length of stay was highest for patients with gastric and ovarian cancers (21.1 and 20.8 days, respectively), and the mean number of physician visits was highest for patients with lung and ovarian cancers (20.3 and 23.9 visits, respectively). Hospitalization length of stay was shortest for patients with melanoma of the skin (2.2 days), prostate cancer (3.8 days), and breast cancer (4.0 days). In the last-year-of-life phase of care, the mean hospitalization length of stay was highest for patients with gastric, lung, and ovarian cancers (35.4, 32.4, and 31.9 days, respectively). The mean number of physician visits ranged from 8.5 visits for patients with uterine cancer to 16.4 visits for patients with ovarian cancer. Differences in service use per month between patients and control subjects in the continuing phase of care were small (data not shown).
|
Patient Time by Phase of Care
Patient time estimates for the initial phase of care varied widely among patients but were very consistent among control subjects (Table 4). Net patient time estimates associated with medical care ranged from 17.8 hours for melanoma of the skin to 351.3 hours for gastric cancer and 368.1 hours for ovarian cancer. Net time estimates in the last year of life were lowest for melanoma of the skin (99.1 hours) and highest for ovarian (485.3 hours), lung (488.3 hours), and gastric (512.2 hours) cancers. Among control subjects in the last year of life, time estimates varied from 70.0 hours for prostate cancer control subjects to 163.6 hours for melanoma of the skin control subjects, potentially reflecting differences in the distributions of cancer-specific and other-cause deaths across tumor sites.
|
Patient Time Costs by Phase of Care
Net patient time costs associated with medical care were lowest for melanoma of the skin and prostate cancers, $271 (95% CI = $213 to $329) and $842 (95% CI = $806 to $878), respectively, and highest for gastric and ovarian cancers, $5348 (95% CI = $4978 to $5718) and $5605 (95% CI = $5273 to $5937), respectively (Table 5). For most of the tumor sites, hospitalizations were the primary component of net patient time costs. Among control subjects only, total patient time cost estimates for the initial phase of care were very similar across control site and ranged from $621 (95% CI = $599 to $643) to $674 (95% CI = $656 to $692) (data not shown). Net patient time costs were less than $60 per month in the continuing care phase for all tumor sites (data not shown). Net patient time costs for care during the last-year-of-life phase ranged from $1509 (95% CI = $1343 to $1675) for melanoma of the skin to $7799 (95% CI = $7433 to $8165), $7435 (95% CI = $7207 to $7663), and $7388 (95% CI = $7018 to $7758) for gastric, lung, and ovarian cancers, respectively. As with the initial phase of care, hospitalizations were the largest component of net patient time costs for most tumor sites in the last year of life.
|
Sensitivity Analysis of Assumptions Regarding Hospitalization Time and Value of Time
Using 24 hours rather than 16 hours to value hospitalization time led to increases in the net time cost for the initial, continuing, and last-year-of-life phases of care for all tumor sites, with the largest increases occurring among patients with tumors at sites requiring the longest hospitalization length of stay (i.e., gastric, lung, ovary) (Table 6). The different methods of valuing patient time led to point estimates that were similar to or only slightly outside of the 95% confidence intervals of the base case estimates.
|
National Estimates of Patient Time Costs in the United States in 2005
The national estimate of patient time costs associated with the initial phase of cancer for these 11 sites was approximately $2.3 billion in 2005 (Table 7). Patient time costs for colorectal cancer care, the only tumor site for which prevalence has been estimated by phase of care (20), were estimated to be $268 million in the continuing and $376 million in the last-year-of-life phases of care in 2005.
|
|
Discussion |
|---|
|
TopNotesAbstractContext and CaveatsMethodsResultsDiscussionReferences |
|---|
This study applies an approach for estimating net patient time costs associated with cancer care (3) for 11 of the most prevalent tumor sites in the United States. To our knowledge, this study is the first to estimate net patient time costs over the course of cancer care for breast, colorectal, corpus uteri, gastric, head and neck, lung, melanoma of the skin, ovarian, prostate, renal, and urinary bladder cancers in a large, population-based cohort. These time costs were substantial; for 2005, the estimated cost for the initial phase of care alone was approximately $2.3 billion. Time costs associated only with colorectal cancer care in the continuing and last-year-of-life phases of care in 2005 were $268 million and $376 million, respectively. Our phase-specific estimates of patient time costs can be used as inputs for developing estimates of the cost-effectiveness of novel treatments or technologies for early detection for the most prevalent tumor sites. The national estimates of patient time costs could also be combined with other components of direct and indirect medical costs not measured in this study, leading to greater understanding of the overall burden of cancer in the United States.
We observed large differences across tumor sites in net patient time cost estimates for the initial and last-year-of-life phases of care. In general, tumor sites among which the majority of patients in our study were initially diagnosed with early-stage disease, such as breast, corpus uteri, melanoma of the skin, prostate and urinary bladder, had the lowest net patient time costs in both the initial and last-year-of-life phases of care. Tumor sites among which a sizable portion of patients in our study were initially diagnosed with regional and distant disease, and that are reported to have relative 5-year survival of less than 50%, such as gastric, lung, and ovary (4), had the highest net patient time costs in both the initial and last-year-of-life phases of care. These relative relationships by tumor site in patient time cost estimates are generally consistent with prior estimates of the direct medical costs of care across multiple tumor sites (8,21–23). National estimated time costs in the initial phase of care varied tremendously, reflecting differences in incidence across the tumor sites as well as differences in net time costs.
Time costs associated with colorectal cancer care in this study are largely consistent with those reported in a previous study (3). In this study, we refined the methods developed in that study (3) and were able to include a longer observation period, substantially more patients and control subjects, and additional services. We also modified the inclusion criteria and the calculation of time costs in the last year of life.
Our estimates of patient time costs are likely to be broadly generalizable to the US population. With few exceptions, the majority of cancer patients in the United States with tumors in the 11 sites included in this study are aged 65 years and older (4) and thus, based on age, are eligible for Medicare. In addition, 5-year relative survival rates for individuals younger than 65 years and those 65 years and older are similar for most of the tumor sites included in this study (4). Exceptions include patients with melanoma of the skin, the majority of whom are younger than 65, and patients with corpus uteri, ovarian, and renal cancers, among whom 5-year relative survival rates are higher in younger (<65 years of age) patients than in older patients (4).
Net patient time cost estimates may be understated for younger cancer patients, who tend to undergo more aggressive surgical care (24–26), undergo more adjuvant treatment (27), and travel farther for cancer care (27,28) than do older cancer patients. In addition, the young spend less on health care than the elderly (29), so service use and resulting time costs in younger control subjects might be lower than in older control subjects. Lower time costs in younger control subjects might also contribute to higher net patient time costs in younger cancer patients.
Our net estimates are likely to understate the amount of time spent receiving care from the perspective of the patients and their families. We did not include preparation time before medical care (e.g., preparation for colonoscopy), posttreatment recovery time spent at home, or time spent addressing health insurance issues in our patient time cost estimates. Family or caregiver time spent traveling and waiting with patients for care was not included, nor were hospice and home health care services. Because patients may require assistance before or after treatment, caregivers frequently accompany patients to care (30). Finally, hospice and home health services are reported to be used more frequently by cancer patients than patients with other conditions (31).
Applying a value to patient time spent seeking medical care is complicated, particularly for the elderly, the majority of whom may be retired. As in the previous study of time costs associated with colorectal cancer care, several methods were used to value patient time, including a median wage rate for individuals of all ages, a median wage rate for individuals aged 65 and above, and a hedonic assumption about the value of life (3). These all yielded very similar net patient time cost estimates. Another method of valuing patient time, the human capital approach, requires assuming that the value of patient time for retired individuals without household productivity is zero (32). Because time spent seeking medical care represents a lost opportunity for usual activities, including both work and leisure, we chose not to make the assumptions required for the human capital approach and valued patient time consistently across tumor sites with a median wage rate as our base case.
There are some limitations associated with using claims data to identify services. Recent validation studies indicate that codes are reliable for identifying receipt of chemotherapy and radiation therapy services (33–36), although there is some variability across tumor sites. In addition, claims data do not provide information on hourly duration of most services or travel and waiting time. Although we used data on adults from national sources for many time estimates, these estimates may underestimate time for cancer patients, who tend to have more complex care needs than the average adult patient. Hospitalizations accounted for the majority of net patient time costs across the phases of care, however, and hospital length of stay was directly observed from claims data. Other time estimates for service duration and travel and waiting time were based on data from large, nationally representative surveys and may not reflect the additional time required for cancer care.
Net patient time costs for cancer care are substantial and vary by tumor site and phase of care. Differences in net patient time cost estimates by tumor site are likely to reflect differences in severity of disease at diagnosis and availability and intensity of treatment during the initial and last-year-of-life phases of care.
|
NOTES |
|---|
|
TopNotesAbstractContext and CaveatsMethodsResultsDiscussionReferences |
|---|
We are grateful to Kevin Knopf of the Annapolis Oncology Center/DeCesaris Cancer Institute for comments on earlier versions of this work. We are grateful to Susan Schappert for producing estimates from the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Medical Care Survey Emergency Department, Daniel Waldo and Jessica Herrera Cancel of the CMS for producing estimates from the Medicare Current Beneficiary Survey, and Timothy McNeel of Information Management Services, Inc, for producing estimates from the National Health Interview Survey. The authors take sole responsibility for the study design, data collection and analysis, interpretation of the data, and the preparation of the manuscript.
Presented in part at the 2005 International Health Economics Association meeting, Barcelona, Spain.
|
REFERENCES |
|---|
|
TopNotesAbstractContext and CaveatsMethodsResultsDiscussionReferences |
|---|
(1) Brown ML, Yabroff KR. Economic impact of cancer in the United States. In: David Schottenfeld, Joseph F. Faumeni J, editors. Cancer epidemiology and prevention. New York (NY): Oxford University Press; 2004.
(2) Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Cost-effectiveness in health and medicine. New York (NY): Oxford University Press; 1996.
(3) Yabroff KR, Warren JL, Knopf K, Davis WW, Brown ML. Estimating patient time costs associated with colorectal cancer care. Med Care 2005;43:640–8.[CrossRef][ISI][Medline]
(4) Ries LAG, Kosary CL, Hankey BF, Edwards BK. SEER Cancer Statistics Review, 1973–1996. Bethesda (MD): National Cancer Institute; 1999.
(5) Kleinman JC, Makuc D. Travel for ambulatory medical care. Med Care 1983;21:543–57.[ISI][Medline]
(6) Butler MA, Beale CL. Rural-urban continuim codes for metro and nonmetro counties. Beltsville (MD): Agriculture and Rural Economy Division, Economic Research Service, U.S. Department of Agriculture; 1994 Staff Report No. 9425.
(7) Stambler HV. The Area Resource File—a brief look. Public Health Rep 1988;103:184–8.
(8) Brown ML, Riley GF, Schussler N, Etzioni RD. Estimating health care costs related to cancer treatment from SEER-Medicare data. Med Care 2002;40(Suppl):104–17.
(9) Brown ML, Riley GF, Potosky AL, Etzioni RD. Obtaining long-term disease specific costs of care: application to Medicare enrollees diagnosed with colorectal cancer. Med Care 1999;37:1249–59.[CrossRef][ISI][Medline]
(10) National Center for Health Statistics. NAMCS description. 2004. Available at http://www.cdc.gov/nchs/about/major/ahcd/namcsdes.htm. [Last accessed: November 27, 2006.]
(11) National Center for Health Statistics. NHAMCS description. 2004. Available at http://www.cdc.gov/nchs/about/major/ahcd/nhamcsds.htm. [Last accessed: November 27, 2006.]
(12) Preparing for transfusion therapy. Bethesda (MD): Patient Information Publications, Warren Grant Magnuson Clinical Center, National Institutes of Health; 2006.
(13) Radiological Society of North America. RadiologyInfo: the radiology information resource for patients. Available at http://www.radiologyinfo.org/index.cfm?bhcp=1. [Last accessed: November 27, 2006.]
(14) Centers for Medicare & Medicaid Services. MCBS survey overview. 2004. Available at http://www.cms.hhs.gov/MCBS. [Last accessed: November 27, 2006.]
(15) National Health Interview Survey (1992). NCHS CD-ROM Series 10, No. 6. Hyattsville (MD): National Center for Health Statistics; 1995.
(16) U.S. Department of Labor. Bureau of Labor Statistics Data. 2003. Available at http://www.bls.gov/cps/#news. [Last accessed: November 27, 2006.]
(17) Raj D. Sampling theory. New York (NY): McGraw-Hill; 1968.
(18) Nordhaus W. The health of nations: the contribution of improved health to living standards. Measuring gains from medical research: an economic approach. Chicago (IL): University of Chicago Press; 2003.
(19) American Cancer Society. Cancer facts & figures 2005. Atlanta (GA): American Cancer Society; 2005. p. 1–32.
(20) Mariotto AB, Yabroff KR, Feuer EJ, De Angelis R, Brown M. Projecting the number of patients with colorectal carcinoma by phases of care in the US: 2000–2020. Cancer Causes Control. In press 2006.
(21) Taplin SH, Barlow W, Urban N, Mandelson MT, Timlin DJ, Ichikawa L, et al. Stage, age, comorbidity, and direct costs of colon, prostate, and breast cancer care. J Natl Cancer Inst 1995;87:417–26.
(22) Riley GF, Potosky AL, Lubitz JD, Kessler LG. Medicare payments from diagnosis to death for elderly cancer patients by stage and diagnosis. Med Care 1995;33:828–41.[CrossRef][ISI][Medline]
(23) Fireman BH, Quesenberry CP, Somkin CP, Jacobson AS, Baer D, West D, et al. Cost of care for cancer in a health maintenance organization. Health Care Financ Rev 2006;18:51–76.
(24) Smith TJ, Penberthy L, Desch CE, et al. Differences in initial treatment patterns and outcomes of lung cancer in the elderly. Lung Cancer 1995;13:235–52.[CrossRef][ISI][Medline]
(25) Farrow DC, Hunt WC, Samet JM. Geographic variation in the treatment of localized breast cancer. N Engl J Med 1992;326:1097–101.[Abstract]
(26) Satariano ER, Swanson GM, Moll PP. Nonclinical factors associated with surgery received for early stage breast cancer. Am J Public Health 1992;82:195–8.
(27) Athas WF, Adams-Cameron M, Hunt WC, Amir-Fazil A, Key CR. Travel distance to radiation therapy and receipt of radiotherapy following breast-conserving surgery. J Natl Cancer Inst 2000;92:269–71.
(28) Gross CP, Herrin J, Wong N, Krumholz HM. Enrolling older persons in cancer trials: the effect of sociodemographic, protocol, and recruitment center characteristics. J Clin Oncol 2005;23:4755–63.
(29) Meara E, White C, Cutler DM. Trends in medical spending by age, 1963–2000. Health Aff 2004;23:176–83.
(30) Guidry JJ, Aday LA, Zhang D, Winn RJ. Transportation as a barrier to cancer treatment. Cancer Pract 1997;5:361–6.[ISI][Medline]
(31) Christakis NA, Escarce JJ. Survival of Medicare patients after enrollment in hospice programs. N Engl J Med 1996;335:172–8.
(32) Grosse SD. Productivity loss tables. In: Haddix AC, Teutsch SM, Corso PS, editors. Prevention effectiveness: a guide to decision analysis and economic evaluation. New York (NY): Oxford University Press; 2003. p. 245–57.
(33) Virnig BA, Warren JL, Cooper GS, Klabunde CN, Schussler N, Freeman J. Studying radiation therapy using SEER-Medicare-linked data. Med Care 2002;40(Suppl):IV-49–54.
(34) Warren JL, Harlan LC, Fahey A, Virnig BA, Freeman JL, Klabunde CN, et al. Utility of the SEER-Medicare data to identify chemotherapy use. Med Care 2002;40(Suppl):IV-55–61.
(35) Lamont EB, Herndon JE, Weeks JC, Henderson IC, Lilenbaum R, Schilsky RL, et al. Criterion validity of Medicare chemotherapy claims in Cancer and Leukemia Group B breast and lung cancer trial participants. J Natl Cancer Inst 2005;97:1080–3.
(36) Cooper GS, Yuan Z, Stange KC, Dennis LK, Amini SB, Rimm AA. Agreement of Medicare claims and tumor registry data for assessment of cancer-related treatment. Med Care 2000;38:411–21.[CrossRef][ISI][Medline]
Manuscript received April 6, 2006; revised October 18, 2006; accepted November 16, 2006.
Editorial about this Article
Related Article in JNCI
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2007 99: 2-3.
J Natl Cancer Inst 2007 99: 1.
This article has been cited by other articles:
|
|
 |
|
  K. R. Yabroff, E. B. Lamont, A. Mariotto, J. L. Warren, M. Topor, A. Meekins, and M. L. Brown Cost of Care for Elderly Cancer Patients in the United States J Natl Cancer Inst, May 7, 2008; 100(9): 630 - 641. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Lipscomb Estimating the Cost of Cancer Care in the United States: A Work Very Much in Progress J Natl Cancer Inst, May 7, 2008; 100(9): 607 - 610. [Full Text] [PDF]
|
|
|
|
|
|
S. Lauzier, E. Maunsell, M. Drolet, D. Coyle, N. Hebert-Croteau, J. Brisson, B. Masse, B. Abdous, A. Robidoux, and J. Robert Wage Losses in the Year After Breast Cancer: Extent and Determinants Among Canadian Women J Natl Cancer Inst, March 5, 2008; 100(5): 321 - 332. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Bennett and E. A. Calhoun Evaluating the Total Costs of Chemotherapy-Induced Febrile Neutropenia: Results from a Pilot Study with Community Oncology Cancer Patients Oncologist, April 1, 2007; 12(4): 478 - 483. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. G. Kessler and S. D. Ramsey The Forest and the Trees: the Human Costs of Cancer J Natl Cancer Inst, January 3, 2007; 99(1): 2 - 3. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||