Fax: (301) 435-3710
Version of Record online: 4 SEP 2009
Copyright © 2009 American Cancer Society
Supplement: Cancer Survivorship Research: Mapping the New Challenges Atlanta, Georgia, Supplement to Cancer
Volume 115, Issue Supplement 18, pages 4362–4373, 15 September 2009
How to Cite
Yabroff, K. R. and Kim, Y. (2009), Time costs associated with informal caregiving for cancer survivors. Cancer, 115: 4362–4373. doi: 10.1002/cncr.24588
This research is dedicated to the memory of Dorothy Yabroff and Heekyoung Kim.
Presented at the Fourth Biennial Cancer Survivorship Research Conference entitled “Cancer Survivorship Research: Mapping the New Challenges,” Atlanta, Georgia, June 18-20, 2008.
- Issue online: 4 SEP 2009
- Version of Record online: 4 SEP 2009
- Manuscript Accepted: 5 MAY 2009
- Manuscript Revised: 8 APR 2009
- Manuscript Received: 15 JAN 2009
- American Cancer Society National Home Office, Intramural Research
- informal caregiving;
- burden of illness;
- time cost;
To the authors' knowledge, little work has been done to estimate the time costs associated with informal caregiving for cancer survivors.
Data from a national survey of caregivers of cancer patients in 2003 to 2006 were used to estimate the time associated with informal caregiving in the 2 years after patient diagnosis with bladder, breast, colorectal, kidney, lung, melanoma of the skin, ovarian, prostate, or uterine cancer, or non-Hodgkins lymphoma (NHL). Caregivers reported the duration and daily intensity of caregiving as well as the types and frequency of support provided. The median wage rate in 2006 ($16.28) was used to value caregiver time, and other methods to value time were evaluated with sensitivity analyses.
On average, caregivers provided care for 8.3 hours per day for 13.7 months. The number of months and daily hours spent caregiving were the highest for cancer survivors diagnosed with distant disease compared with survivors with regional or localized disease (P < .05). Approximately half of caregivers provided emotional, instrumental, tangible, or medical support, although the frequency varied dramatically. Informal caregiver time costs over the 2-year period after diagnosis were the highest for caregivers of patients diagnosed with lung ($72,702; 95% confidence interval [95% CI], $56,814-$88,590) and ovarian ($66,210; 95% CI, $40,750-$91,670) cancers and NHL ($59,613; 95% CI, $43,423-$75,803) and the lowest for caregivers of patients with breast cancer ($38,334; 95% CI, $31,442-$45,226).
Time spent by informal caregivers was substantial and was an important component of the burden of cancer care. Incorporation of the value of informal caregiver time will be important when evaluating the costs and benefits of cancer control interventions. Cancer 2009;115(18 suppl):4362–73. © 2009 American Cancer Society.
In 2005, approximately 10.7 million individuals in the United States were alive with a history of cancer.1 An additional 1.4 million cancers were diagnosed in 2006,2 and the number of new cancer patients is expected to increase over time due to population changes in the United States. As a result, a larger prevalence of cancer survivorship is expected in the future.1, 3 These cancer survivors will receive medical care throughout the trajectory of their cancer experience, including the initial period after diagnosis, long-term survivorship, recurrence, and at the end of life. Many cancer survivors will also experience lasting effects of disease or its treatment leading to health and employment limitations.4 Throughout their cancer care experiences, informal caregivers play an important role in providing patients with emotional, financial, and personal care support.5 Informal caregivers also provide transportation and accompany patients to medical appointments, coordinate between multiple care settings, and provide home care.5
Much work has been devoted to describing the medical costs of cancer care in the United States6, 7; however, less attention has been devoted to describing the economic burden of informal caregiving for cancer patients. Both patient and caregiver time are recommended for inclusion in cost-effectiveness analyses,8 yet these data are rarely considered, potentially resulting in a bias toward underestimating the costs of interventions that place a higher burden on the cancer patient and informal caregiver. Several studies have assessed components of the burden of informal caregiving for cancer in the general population9 or in convenience samples of caregivers of newly diagnosed or seriously ill cancer patients for limited periods of time.10-12 To our knowledge, no studies have reported informal caregiver time costs by tumor site of the cancer patient or by stage of disease at diagnosis. Both are important for estimating the cost-effectiveness of any cancer control intervention. In this study, we used data from a large national study of cancer patient caregivers to estimate the time costs associated with informal caregiving for cancer survivors with 10 of the most prevalent cancers.
MATERIALS AND METHODS
The American Cancer Society's Quality of Life Survey for Caregivers was designed to assess the impact of cancer on the quality of life of family members and close friends who are caring for cancer survivors. Caregivers were nominated by cancer survivors who completed the American Cancer Society's Study of Cancer Survivors-I (SCS-I).13 For the SCS-I, survivors were identified from 11 state cancer registries, including Alabama, Connecticut, Iowa, Idaho, Minnesota, New Jersey, Ohio, Pennsylvania, Rhode Island, South Carolina, and Wyoming. Eligible survivors were at least 18 years old and identified during the 12-month period after their diagnosis between January 2000 and September 2003 (with the exception of New Jersey, where the eligibility period lasted for 15 months after diagnosis). Patients were diagnosed with 1 of the 10 most common cancers, including bladder, breast, colorectal, kidney, lung, melanoma of the skin, ovarian, prostate, or uterine or non-Hodgkins lymphoma (NHL), according to the registry. SEER (Surveillance, Epidemiology and End Results) summary stage of localized, regional, or distant disease at diagnosis (with an exception of bladder cancer including in situ) was also obtained from the registries. Patients were required to be fluent in either English or Spanish and be a resident of the United States. From this group of potentially eligible survivors, state-wide probability samples were stratified by cancer type, and younger patients (ie, aged <55 years) were oversampled to ensure sufficient representation of this age group.
Each cancer survivor who participated in the SCS-I was then asked to nominate 1 individual in a family-like relationship who consistently provided help, including emotional support, financial support, symptom management, personal care, or transportation support during their cancer experience.14 These individuals were defined as informal caregivers in our study. On average, cancer survivors were approximately 14 months postdiagnosis when they nominated their caregivers for the caregiver survey. Approximately 0.4% of survivors reported that they did not receive any consistent help and did not nominate a caregiver. Eligible caregivers were aged ≥18 years, able to speak/read English or Spanish, and resided in the United States. A total of 2650 caregivers were identified; 1635 returned the baseline caregiver survey (61.7% response rate) in 2 separate survey periods. Because information regarding caregiving hours and types of caregiving were only available from the second baseline survey period (August 2003 to February 2006), we restricted our sample to the 896 caregivers who completed the second baseline data collection. At the time caregivers completed this baseline survey, cancer patients had been diagnosed an average of 25.1 months (standard deviation [SD], 5.5 months).
Each eligible caregiver was mailed a packet containing an introductory letter, survey, self-addressed stamped envelope, frequently asked-questions brochure, and a 60-minute telephone card as an incentive. Returning the completed survey served as consent to participate. If the caregivers had not returned the survey by 3 weeks after the initial mailing, telephone follow-up calls were made. A second packet, which included the same materials except the telephone card, was mailed 5 weeks after the initial mailing, and a second follow-up call was made 8 weeks after the initial mailing. This study was conducted in compliance with the regulations of the Emory University Institutional Review Board.
Caregiver demographic characteristics included self-reported age (<24 years, 25-34 years, 35-44 years, 45-54 years, 55-64 years, and ≥65 years), gender, race/ethnicity (white non-Hispanic, black non-Hispanic, Hispanic), marital status (married or partnered, other), educational attainment (<high school graduate, high school graduate, college graduate, more than college), family income (<$20,000, $20,000-$39,999, $40,000-$74,999, and ≥$75,000), and employment at the time of the survey (yes, no) and at the time they were providing care to the patient (yes, no). The relationship with the cancer patient (spouse or partner, child or child-in-law, parent, sibling, friend, other) was also measured.
Medical characteristics of cancer survivors included information concerning the type of cancer (bladder, breast, colorectal, kidney, lung, melanoma of the skin, ovarian, prostate, or uterine cancer, or NHL), date of cancer diagnosis, and stage of cancer at diagnosis (localized, regional, distant, unstaged/unknown/missing) was obtained from the state cancer registry. As an objective indicator of the severity of the care-recipient's cancer, a severity of cancer index was created for each survivor based on the mortality rate for the type and stage of cancer and the time since diagnosis.1, 15 Cancer patient status at the time of the survey (remission, metastasis or recurrence, new cancer, or deceased) was assessed by the caregiver's self-report.
Self-reported duration of providing help for cancer-related concerns of the patient, including emotional support, financial support, symptom management, personal care, or transportation was based on responses to the question, “How long were you (or have you been) helping this person through his/her experience with cancer?” and was reported in months. Hours of care per day were based on responses to the question, ‘On average, about how many hours a day did you help this person through his/her experience with cancer?’. Frequencies of providing different types of care were also measured using a 17-item scale.16 Individual items from this scale are listed in Table 1, and assessed traditional aspects of social support17-19 (emotional, instrumental, and tangible) and also assistance for medical needs. Caregivers rated frequency for each item using a 4-point Likert-style format (monthly, weekly, 2-3 days a week, daily). Cronbach α ranged from .56 to .66 for the 4 frequency scales, indicating that similar items for corresponding types of care were related (internally consistent), but each item contributed some unique information.
|Emotional support||Talk to about things that worry him/her|
|Boost his/her spirits when he/she felt low|
|Show him/her trustworthiness|
|Help him/her relax|
|Instrumental support||Give him/her information, suggestions, and guidance about cancer, treatments, and survivorship|
|Talk to medical staff for him/her|
|Tell about hotlines or services|
|Help communicate with distant family members|
|Tangible support||Help with household chores|
|Daily living needs (eg, grocery shopping, preparing meals)|
|Daily instrumental needs (eg, dressing, bathing)|
|Provide financial help|
|Medical support||Accompany to treatments|
|Help arrange medical care|
|Help with managing cancer-related symptoms|
|Help administer medications|
Time Cost Estimates for Informal Caregiving in 2 Years After Patient Diagnosis
Time costs were derived from a combination of the total number of hours spent caregiving and a value for each hour. To estimate the total number of hours spent caregiving, the number of months spent caregiving was multiplied by 30 days per month and the number of hours per day. The reported duration of caregiving was limited to the 2-year period after patient diagnosis to standardize estimates across caregivers. Some caregivers (18.4%) reported providing care 24 hours per day, and for the purposes of estimating the economic value of informal caregiving, we restricted the maximum amount of care in a single day to an approximation of waking hours, 16 hours per day. This approach to truncating informal caregiving hours has also been used elsewhere.20
Time spent by caregivers supporting cancer survivors represents time not spent pursuing usual activities, including work and leisure. In economics, the value of these activities is referred to as the opportunity cost of providing informal care. A variety of approaches have been used to estimate the opportunity cost of patient or informal caregiver time. Because the hourly wage is the value of an hour of work, it can also be viewed as the opportunity cost of not working to engage in another activity. Human capital approaches value time based on wages, or apply average national age-specific and sex-specific wages to individuals by age and sex. This approach explicitly values time differently for populations with higher earnings (eg, middle-aged men with graduate education) than for populations with lower earnings (eg, young women without college education). This approach also requires assumptions regarding the value of household productivity for retired populations and homemakers, who may not be earning wages for their work. Other wage-based approaches to valuing time apply a single average wage to all individuals or identify the wage for the profession providing the specific service, such as home health aid. Willingness-to-pay approaches, in contrast, ask individuals to state how much they would be willing to pay for an additional year of life or estimate the amount from their behavior in labor or other markets, and incorporate both lost productivity and the intrinsic value of life.
In this study, we used the national median wage rate in 2006 ($16.78)21 as our base-case method for valuing time, because it does not require assumptions regarding different values of productivity in different population groups, and it has been used previously to value cancer patient time.22 We also evaluated a variety of other approaches to valuing informal caregiver time in a sensitivity analysis, including applying 1) age- and gender-specific median wages in 2006 (age categories: <24 years, 25-34 years, 35-44 years, 45-54 years, 55-64 years, and ≥65 years),21 2) the median wage for a home health aid ($9.70 in 2006),21 3) age- and educational attainment level-specific median wages in 2006,21 and 4) a previously published value of 1 year of life based on a willingness-to-pay approach of $150,00023 and converted to an hourly rate ($17.12). This value of a year of life is consistent with other approaches and has been used elsewhere to value patient time.22
Demographic characteristics of caregivers and medical characteristics of cancer patients were assessed with descriptive statistics. Mean months of caregiving and hours per day spent caregiving were stratified by the caregiver and patient characteristics, and compared with F statistics. Scales were evaluated for internal consistency, using Cronbach α. We restricted the sample to caregivers with complete information on gender, relationship to the patient, and caregiving duration (hours per day and months of caregiving) within 2 years of patient diagnosis (N = 688). In all group comparisons, caregivers were included only when all the variables involved in a certain analysis had complete data (pairwise deletion).
The caregiver sample was predominantly female, aged <65 years, married, white non-Hispanic with at least a college education (Table 2). Most caregivers were spouses of cancer patients. Most caregivers were employed when their patient was diagnosed and at the time of the survey. Because a larger proportion of the patients aged ≥65 years were not employed, employment in the younger group is even higher. Caregivers provided support for breast cancer patients (29.1%), followed by prostate (18.9%), colorectal (13.8%), lung (10.5%), NHL (7.8 %,), ovarian (5.1%), and kidney cancer patients (4.2%). The remaining 11.2% of caregivers provided support for patients with bladder cancer, melanoma of the skin, or uterine cancer. Approximately half of the cancer patients were originally diagnosed with in situ or localized disease, with approximately 25% diagnosed with regional disease. The remainder had distant or unstaged disease. Similarly, the cancer severity index score was low. At the time of the survey, approximately half of the patients were in remission or free of disease. In general, caregivers with incomplete information were similar to caregivers with complete information, except they were more likely to be older, not employed, and providing care to patients with distant compared with earlier stage disease (P < .05).
|Married or partnered||572||83.1%|
|<High school graduate||20||2.9 %|
|High school graduate||146||21.2%|
|More than college||140||20.3%|
|Employed/doing volunteer work when patient was diagnosed|
|Relation to cancer patient|
|Spouse or partner||451||65.6%|
|Other (bladder, skin, and uterine)||73||10.6%|
|Stage of disease at diagnosis|
|In situ (only for bladder cancer cases)||15||2.2%|
|Status of cancer patient at time of survey|
|Cancer spread or recurred||49||7.1%|
|Cancer Severity Index (0 indicates least severe; ∼1, most severe)||Mean = .18||SD = .25|
On average, caregivers spent 13.7 months providing care (Table 3). In stratified analysis, the number of months spent caregiving varied by the patients' medical characteristics. The number of months for caregiving was more similar by caregivers' demographic characteristics. Caregivers spent the longest time for caring for patients with ovarian (16.7 months) or lung (16.1 months) cancers and the shortest time caring for patient with kidney (11.4 months) cancer. Caregivers also spent more months caring for patients with distant disease (17.9 months), followed by regional (14.5 months) and localized disease at diagnosis (12.2 months). At the time of the survey, approximately half of caregivers were no longer providing informal care.
|Cancer Patient Characteristics||Time Spent Caregiving|
|Mean No. of Months (SD)||F statistic||Mean Hours per Day (SD)||F statistic||Mean Total in Hours (SD)||F statistic|
|Overall||13.7 (7.8)||—||8.3 (8.3)||—||2932.3 (3414.0)|
|Age, y||F(5,663) = 1.09;||F(5,632) = 0.41||F(5,632) = 0.60|
|<24||11.4 (7.6)||P =.196||8.3 (7.5)||P =.844||2661.0 (2452.0)||P =.703|
|25-34||12.3 (7.3)||7.3 (6.9)||2349.0 (2438.4)|
|35-44||15.1 (7.8)||7.7 (8.4)||2982.2 (3628.2)|
|45-54||13.8 (7.6)||8.8 (8.5)||3165.1 (3504.7)|
|55-64||13.2 (8.0)||8.6 (8.7)||2710.6 (3351.4)|
|≥65||13.7 (8.0)||8.0 (8.2)||2979.3 (3530.4)|
|Gender||F(1,686) = 3.08;||F(1,652) = 1.20;||F(1,652) = 0.01|
|Male||14.4 (7.6)||P =.080||7.8 (8.0)||P =.273||2925.2 (3319.9)||P =.968|
|Female||13.3 (7.9)||8.6 (8.5)||2936.2 (3469.9)|
|Race/ethnicity||F(2,666) = 0.82;||F(2,6834 = 0.91;||F(2,634) = 1.75|
|White, non-Hispanic||13.6 (7.8)||P =.440||8.3 (8.4)||P =.403||2867.4 (3422.7)||P =.175|
|Black, non-Hispanic||14.9 (7.4)||9.3 (6.8)||3757.7 (3151.8)|
|Hispanic||11.6 (8.2)||4.9 (8.3)||1548.8 (1620.0)|
|Marital status||F(1,678) = 0.10;||F(1,645) = 0.00;||F(1,645) = 0.00|
|Married or partnered||13.6 (7.8)||P =.752||8.3 (8.3)||P =.975||2899.8(3397.5)||P =.980|
|Other||13.4 (7.4)||8.3 (8.4)||2909.1 (3382.5)|
|Educational attainment||F(3,673) = 0.89;||F(3,641) = 7.49;||F(3,641) = 7.41|
|<High school graduate||16.3 (6.3)||P =.446||11.2 (9.2)||P =.000||4746.7 (4125.1)||P =.000|
|High school graduate||13.8 (8.2)||10.6 (9.1)||3651.3 (3798.4)|
|College graduate||13.5 (7.8)||8.1 (8.1)||2877.6 (3350.2)|
|More than college||13.3 (7.5)||6.1 (7.5)||1998.8 (2707.8)|
|Family income||F(3,567) = 0.60;||F(3,542) = 3.35;||F(3,542) = 2.01|
|<$20,000||12.9 (7.9)||P =.617||12.0 (9.3)||P =.019||3553.8 (3433.7)||P =.111|
|$20,000-$39,999||14.7 (7.8)||9.2 (8.1)||3578.0 (3755.5)|
|$40,000-$74,999||13.7 (7.9)||8.6 (8.6)||2974.7 (3433.6)|
|≥$75,000||13.6 (7.6)||7.2 (7.9)||2592.8 (3253.5)|
|Current employment||F(1,567) = 0.31;||F(1,542) = 6.72;||F(1,647) = 0.23|
|Employed||13.9 (7.5)||P =.575||10.0 (8.9)||P =.010||2880.6 (3400.1)||P =.631|
|Not Employed||13.5 (8.4)||7.8 (8.1)||3031.0 (3451.0)|
|Employed/doing volunteer work when patient was diagnosed||F(1,680) = 2.15;||F(1,647) = 5.17;||F(1,647) = 1.61|
|Yes||14.0 (7.6)||P =.143||7.8 (8.2)||P =.023||2800.8 (3355.9)||P =.205|
|No||13.0 (8.0)||9.3 (8.4)||3158.2 (3504.9)|
|Relation to cancer patient||F(5,682) = 0.41;||F(5,648) = 8.72;||F(5,648) = 8.30|
|Spouse or partner||13.9 (7.9)||P =.846||9.7 (8.8)||P =.000||3504.3 (3697.8)||P =.000|
|Child (child-in-law)||13.4 (7.8)||5.9 (6.7)||2009.1 (2555.0)|
|Parent||12.3 (7.9)||8.3 (7.6)||2498.3 (3160.5)|
|Sibling||13.4 (7.4)||4.1 (5.3)||1542.5 (2230.5)|
|Friend||13.5 (7.5)||5.4 (7.6)||1017.4 (823.4)|
|Other||12.2 (6.5)||4.2 (4.3)||1290.0 (1671.6)|
|Cancer Patient Characteristics|
|Cancer type||F(7,680) = 3.45;||F(7,646) = 3.89;||F(7,646) = 4.99|
|Breast||13.6 (7.3)||P =.001||6.4 (7.4)||P =.000||2175.9 (2658.4)||P =.000|
|Colorectal||13.5 (7.5)||8.2 (8.0)||2818.0 (3130.5)|
|Prostate||12.5 (8.0)||9.1 (8.8)||2970.0 (3482.8)|
|Lung||16.1 (7.7)||10.8 (8.9)||4610.9 (4199.0)|
|Ovarian||16.7 (7.8)||10.3 (9.5)||3774.2 (3797.0)|
|NHL||15.2 (6.5)||10.7 (9.5)||3744.5 (3651.9)|
|Kidney||11.4 (9.8)||9.8 (8.0)||3352.2 (4339.4)|
|Other (bladder, skin, and uterine)||12.1 (8.0)||6.8 (7.3)||2345.8 (3246.1)|
|Stage of disease at diagnosis||F(3,667) = 14.78;||F(3,633) = 3.04;||F(3,633) = 6.65|
|Localized||12.2 (7.9)||P =.000||7.8 (8.1)||P =.028||2510.4 (3221.2)||P =.000|
|Regional||14.5 (7.5)||8.3 (8.0)||3151.3 (3451.2)|
|Distant||17.9 (6.1)||10.9 (9.7)||4362.4 (3942.9)|
|Unstaged/unknown/missing data||12.8 (6.6)||9.8 (9.4)||3211.5 (3039.0)|
|Status of cancer patient at time of survey||F(3,467) = 15.28||F(3,444) = 8.75;||F(3,444) = 16.18|
|Remission/cancer free||12.0 (7.6)||P =.000||7.9 (7.9)||P =.000||2454.4 (3117.1)||P =.000|
|Cancer spread or recurred||18.0 (6.9)||7.7 (8.0)||3672.7 (3581.7)|
|New cancer||13.3 (9.0)||9.9 (10.5)||3148.0 (3921.9)|
|Deceased||18.1 (5.2)||15.3 (9.1)||6494.1 (3818.1)|
Caregivers spent, on average, 8.3 hours a day providing care. Approximately half reported <8 hours per day, and approximately 25% reported >16 hours per day. The number of hours per day spent caregiving was inversely related to caregivers' socioeconomic status, with the longest hours a day reported by caregivers with less than a high school education or with family income of <$20,000 per year, and the shortest by individuals with more than college education or making >$75,000 (Table 3). Caregivers who were employed at the time of their patient's diagnosis spent fewer hours caregiving than caregivers who were not employed. However, employed caregivers at the time of survey spent longer hours than caregivers who were not employed. Spousal or parental caregivers spent the most hours per day caregiving.
The number of hours a day spent providing care also varied by the patients' medical characteristics. Caregivers of patients with lung cancer, ovarian cancer, or NHL spent the most hours per day caregiving (>10 hours) and caregivers of patients with breast and bladder cancer, melanoma of the skin, or uterine cancer spent the fewest hours per day (<7 hours). Stage of disease at patient diagnosis was positively associated with average number of hours spent caregiving. Caregivers of patients who had died at the time of the survey reported the longest hours per day for caregiving during the period they were providing care (15.3 hours).
The average total number of hours spent caregiving in the 2 years after patient diagnosis was 2932. As with the number of months and hours per day spent caregiving, the total number of hours spent caregiving was highest for patients diagnosed with more advanced stage of disease, and for patients who were deceased at the time of the survey. The total number of hours also varied by type of cancer, and was the highest for NHL and lung and ovarian cancer patients and the lowest for breast cancer patients and patients with bladder cancer, melanoma of the skin, or uterine cancer (Fig. 1).
Approximately half of the caregivers provided assistance with medical support, including accompanying patients to treatment, talking with medical staff, managing cancer-related symptoms, and administering medications (Table 4). Less than 20% of caregivers provided this support daily or monthly, with more caregivers providing medical support 2 to 3 days per week (34%) or weekly (29%). Caregivers also reported providing emotional (52%), tangible (46%), and instrumental (44%) support.
|No.||Percentage Providing Care*||Caregiving Frequency|
|Daily||2-3 Days a Week||Weekly||Monthly|
Dollar Value Estimates of Informal Caregiving
At the 2006 median wage, the average value of informal caregiver time over the 2 years after patient diagnosis was $47,710 (95% confidence interval [95% CI], $43,423-$51,997) for the 10 cancer sites we evaluated (Table 5). Estimates from our sensitivity analysis ranged from $28,363, using a home health aide wage rate, to $50,817 using gender- and age-specific wages. The willingness-to-pay value of life estimate over the 2 years ($50,060) fell within the 95% CI for the base-case estimate. Estimates using gender- and educational attainment-specific wages were very similar to gender- and age-specific wages and are not reported separately. In the 2 years after diagnosis, the value of informal caregiver time varied by cancer site, with the highest time cost for lung ($72,702; 95% CI, $56,814-$88,590) and ovarian ($66,210; 95% CI, $40,750-$91,670) cancer patients and NHL ($59,613; 95% CI, $43,423-$75,803) patients, and the lowest for breast ($38,334; 95% CI, $31,442-$45,226) cancer patients. The value of caregiver time also varied by cancer stage at diagnosis from $40,973 (95% CI, $35,326-$46,620) for localized disease to $71,278 (95% CI, $56,303-$86,253) for distant disease at diagnosis.
|Sensitivity Analysis of Hourly Value of Time|
|Base Case Median Hourly Wage ($16.78)||Home Health Aide ($9.70)||Gender-Specific and Age-Specific Wage Rates||Willingness to Pay Value of Life ($17.12)|
|Estimate, $||95% CI, $||Estimate, $||Estimate, $||Estimate, $|
|Relation to cancer patient|
|Tumor site of cancer patient|
|Cancer patient stage of disease at diagnosis|
In this study, we used data from a large national survey of patient-nominated caregivers to estimate the time costs associated with informal caregiving in the 2 years after cancer diagnosis. Approximately 99% of the cancer survivors had an informal caregiver. To our knowledge, it is the first study to estimate time costs separately for informal caregivers of patients diagnosed with cancers of the bladder, breast, colorectal, kidney, or lung; melanoma of the skin; ovarian, prostate, or uterine cancer; or NHL, or by patient stage of disease at diagnosis. Caregivers reported spending a substantial amount of time providing medical, emotional, instrumental, and other tangible support for cancer patients in the 2 years after diagnosis, with an average time cost ranging from $38,334 for caregivers of breast cancer patients to $72,702 for caregivers of lung cancer patients. These estimates are of a similar magnitude to the direct medical costs associated with cancer care.24-26 The number of cancer survivors is expected to increase in future years, because the cancer incidence increases with age,1 life expectancy is increasing, and the US population is aging and growing. In addition, ongoing efforts to control healthcare costs in the United States and shorter hospitalizations suggest even greater reliance on informal caregiving in the future. Thus, quantifying the informal caregiver time and time costs as presented here, may inform healthcare policy and practice for families supporting a member with cancer.
Informal caregiver time costs will also be important inputs in cost-effectiveness analyses (CEAs), as recommended by The Panel on Cost-Effectiveness in Health and Medicine,8 particularly when comparing healthcare strategies or interventions that differentially rely on informal caregivers. Few studies have explicitly assessed the impact of including time costs in CEAs,27-29 and in some situations, the optimal strategy can vary depending on whether these costs are included or excluded.27, 28 For example, investigators reanalyzed the cost-effectiveness of self-monitoring of blood glucose by patients with type 2 diabetes taking oral antidiabetic medicine compared with no self-monitoring, and found that inclusion of patient time changed the cost-effectiveness ratio from $23,380 per quality‒adjusted life-year (QALY) to $76,340 per QALY over a 5-year horizon.27 To our knowledge, no studies have explicitly evaluated the impact of including and excluding caregiver time costs. Exclusion of this important component of healthcare cost may have a substantial impact on conclusions regarding cost-effectiveness and potentially bias decisions toward healthcare interventions that places greater burden on patients and caregivers.
In addition to time costs associated with informal caregiving for cancer patients, families may also experience other financial burdens related to out of pocket costs of medical care30, 31 and reduced participation in the workforce.32 Because health insurance is commonly tied to employment in the United States, changes in employment may introduce additional challenges to the families of cancer patients. Informal caregiving may also present costs to employers, including retraining and other replacement costs, absenteeism, lower levels of productivity while at work (ie, presenteeism), and work interruptions. Efforts to minimize potential productivity losses associated with informal caregiving will be important from the perspective of patients and their families as well as employers, especially as the prevalence of informal caregivers in the workforce increases.
Our informal caregiver time cost estimates are generally consistent with the few other time cost and time estimates reported elsewhere in the published literature.20, 22 In our study, caregiver time costs were positively associated with patient stage at diagnosis. Others have reported a positive association between caregiver time costs and severity of disease in dementia, with higher time costs in informal caregivers of patients with severe dementia compared with mild dementia.20 In addition, the relative caregiver time cost across tumor sites, with higher estimates for caregivers of lung and ovarian cancer patients and lower estimates for caregivers of breast cancer patients, is consistent with studies that report direct medical and patient time costs associated with cancer care.22, 24-26 Exploring and understanding the clinical, patient, and caregiver factors associated with the content and intensity of caregiving will be an important area for additional research.
Studies that quantify the amount of time spent providing informal care generally use either population-based samples or convenience samples of patient-nominated caregivers.9, 12, 33, 34 In a recent national survey conducted in the general population by the National Alliance for Caregiving and the American Association for Retired Persons, estimates of time spent providing informal care ranged from 13.3 hours per week for frail elderly to 31.8 hours per week for cancer patients, and 29.9 hours per week for individuals with dementia.33 Because the majority of cancer survivors are neither newly diagnosed nor at the end of life,1 the periods requiring the most intensive care, a general population sample will also identify more caregivers of patients in the long-term survivorship phase, the period requiring the least intensive care. Other studies, based on patient-nominated caregivers are expected to yield much higher estimates of time spent providing informal care, in part, because patients are typically identified at a time when they are actively receiving medical care, at diagnosis, treatment for recurrence, or at the end of life. For example, male and female spousal caregivers of hospice patients with lung cancer reported an average of 105 hours and 127 hours per week (approximately 15 hours and 18 hours per day) spent caregiving, respectively.12 Given the intensity of patient care needs at the end of life, these estimates from a hospice patient-based sample would be expected to be higher than those from a general population-based sample. Our estimates are based on a sample nominated by patients approximately 14 months after diagnosis, and it is neither a general population sample nor an end of life sample. Nonetheless, our caregiver time cost estimates are generally consistent with these reported time estimates. For example, using an estimated 31.8 hours per week of informal caregiving for cancer patients33 over a 2-year period (104 weeks) results in approximately 3307 total hours, a finding that is similar to our total hour estimate of 2932 hours over an average of 13.7 months. Applying the median wage rate in 2006 yields a time cost estimate of $55,495, also similar to our base-case estimate of $47,710.
Applying a value to informal caregiver time is complicated, particularly for the elderly who are more likely to be retired. Because time spent providing informal care to cancer patients represents time that cannot be used for usual activities, including both work and leisure, we used the median wage rate as our base case and valued time in the same manner for all caregivers. We also evaluated several approaches to value the time of informal caregiving for cancer patients as a sensitivity analysis and found that, regardless of the method used to value time, caregiver time costs are substantial. Because these different approaches rely on different assumptions regarding the value of time, it will be important for future work estimating time costs associated with informal caregiving to be explicit concerning assumptions and approach used.
There were several limitations with our study. Eligibility for the caregiver survey was based on patient completion of SCS-I, and the response rate to the patient survey was relatively low (34%). Respondents to the SCS-I were more likely to be female, younger, white, and diagnosed with localized or regional disease compared with the overall population identified that was eligible for the survey.13 In addition, the portion of our caregiver sample with complete data was more likely to younger, employed, and providing care to patients diagnosed with earlier stage disease. Cancer survivors were approximately 14 months after diagnosis when they nominated their caregivers for the caregiver survey. Only cancer survivors who survived to this point after diagnosis were able to participate in the survey and nominate their caregivers. Finally, we did not include the value of time for family members other than the nominated caregiver, nor did the measure of time address multitasking (eg, providing emotional support while driving to a physician's appointment). In one of the few studies to report labor costs of family members in addition to those of the primary caregiver, approximately 50% of families had additional family members providing informal care.10 To the extent that caregivers of cancer patients with rapidly fatal cancers or distant disease at diagnosis experience even greater burden and patients receive care and support from multiple individuals, our estimates are likely an undercount of the time costs of informal caregiving for cancer patients.
Caregivers reported the number of months and the average daily amount of time spent caregiving at a single point for the entire 2 years after patient diagnosis. Because self-report is thought to be most accurate for more recent reporting,35 caregivers' estimates may better reflect their experiences at the time of the survey, rather than an average of the entire trajectory of the cancer caregiving experience. Studies of healthcare use and costs of care among cancer patients report the largest burden in the initial period after diagnosis and in the last year of life.24 Because the trajectory of caregiving demands are likely to follow a similar pattern, future work in estimating the burden of informal caregiving might focus specifically on different aspects of the cancer care trajectory, rather than a single point in time. A variety of methods for measuring time use exist, including retrospective questionnaires, prospective questionnaires or diaries, and direct observation.36 Evaluation of the validity, reliability, respondent burden, and feasibility of these methods for measuring patient and caregiver time will be an important area for future research.
Despite these limitations, to our knowledge, this study is one of the first to use a cancer registry-based sample of patient-nominated caregivers in the United States to estimate caregiver time costs for multiple cancer sites. Regardless of the method used to value time, we found that caregiver time costs are substantial. Our time cost estimates may be useful in cost-effectiveness analyses of cancer control interventions, particularly those that differentially rely on informal caregivers. Quantifying the economic burden of informal caregiving may inform healthcare policy and practice for the family dealing with cancer.
Conflict of Interest Disclosures
Cosponsored by the National Cancer Institute's Office of Cancer Survivorship, the Office of Cancer Survivorship of the Centers for Disease Control and Prevention, and the American Cancer Society's Behavioral Research Center.
The American Cancer Society's Quality of Life Survey for Caregivers was funded by the American Cancer Society National Home Office, Intramural Research.
- 1http://seer.cancer.gov/csr/1975_2004/ Accessed on July 2, 2009., , , et al. SEER Cancer Statistic Review, 1975-2005. Bethesda, Md: National Cancer Institute; Year. Available at:
- 2American Cancer Society. Cancer Facts & Figures 2005. Atlanta: American Cancer Society; 2005.
- 6Economic impact of cancer in the United States. In: SchottenfeldD, FaumeniJF, eds. Cancer Epidemiology and Prevention. 3rd ed. New York, NY: Oxford University Press; 2006., .
- 8Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996., , , .
- 14Sampling in population-based cancer caregivers research. Qual Life Res. In press., , , , , .
- 19Enabling and Empowering Families. Cambridge, Mass: Brookline Books; 1988., , .
- 21U.S. Department of Labor. Bureau of Labor Statistics Data. Available at: http://www.bls.gov/ data Accessed on July 2, 2009.
- 23http://ssrn.com/abstract=235716. Accessed on July 2, 2009., , , , , . The Economic Impacts of the Tobacco Settlement. Available at:
- 36Time-Use Measurement and Research. Washington, DC: National Academy Press; 2000., , , , , .