2936
Assessing the Burden of HPV-Associated Cancers in the United States Supplement to Cancer
Years of Potential Life Lost and Productivity Costs Because of Cancer Mortality and for Specific Cancer Sites Where Human Papillomavirus May Be a Risk Factor for Carcinogenesis United States, 2003 Donatus U. Ekwueme, PhD1 Harrell W. Chesson, PhD2 Kevin B. Zhang, PhD3 Appathurai Balamurugan, MD, MPH4 1
Division of Cancer Prevention and Control, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia. 2 Division of STD Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia. 3
Macro International, Calverton, Maryland.
4
Arkansas Central Cancer Registry/Epidemiology Branch, Arkansas Department of Health, Little Rock, Arkansas.
BACKGROUND. Although years of potential life lost (YPLL) and mortality-related productivity costs comprise a substantial portion of the burden of cancers where human papillomavirus (HPV) may be a risk factor for carcinogenesis (called HPV-associated cancers in this report), estimates of these costs are limited. The authors estimated the mortality-related burden (in terms of YPLL and productivity costs) of HPV-associated cancers (without regard to the percentage of each of these cancers that could be attributed to HPV) and all malignant cancers in the United States in 2003.
METHODS. The authors used 2003 national mortality data and US life tables to estimate YPLL for HPV-associated cancers and all malignant cancers. YPLL was estimated by using the life expectancy method. The human capital approach was used to estimate the value of the expected future lifetime productivity losses caused by premature deaths from HPV-associated cancers and all malignant cancers. Indirect mortality costs were estimated as the product of the number of deaths and the expected value of individuals’ future earnings, including an imputed value of housekeeping services.
RESULTS. In 2003, HPV-associated cancers accounted for 181,026 YPLL, which represented 2.4% of the estimated 7.5 million YPLL attributable to all malignant cancers in the United States. The average number of YPLL was 21.8 per HPV-associated cancer death and 16.3 per death from overall malignant cancers. Overall, HPV-associated cancers had the largest relative contribution to YPLL in women ages 30 to 34
This supplement to Cancer was supported by Cooperative Agreement Number U50 DP424071-04 from the Centers for Disease Control and Prevention (CDC). The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. We sincerely thank Drs. Scott D. Grosse and Mercy Mvundura for providing us with the updated present value of future lifetime earnings data from 2000 to 2003. In addition, we are indebted to the anonymous reviewers for many helpful comments and suggestions.
years. The lifetime productivity cost from mortality in 2003 was $3.7 billion for HPVassociated cancer mortality and $133.5 billion for overall malignant cancer mortality.
CONCLUSIONS. HPV-associated cancers impose a considerable burden in terms of premature deaths and productivity losses. Cancer 2008;113(10 suppl):2936–45. Published 2008 by the American Cancer Society.*
KEYWORDS: human papillomavirus, years of potential life lost, productivity costs, cancer mortality.
Received April 14, 2008; revision received June 2, 2008
Buford Highway, MS K-55, Atlanta, GA 30341; Fax: (770) 488-4639; E-mail: dce3@cdc.gov
Address for reprints: Donatus U. Ekwueme, PhD, Division of Cancer Prevention and Control, Centers for Disease Control and Prevention, 4770
*This is a US government work and, as such, is in the public domain in the United States of America.
Published 2008 by the American Cancer Society* DOI 10.1002/cncr.23761 Published online 3 November 2008 in Wiley InterScience (www.interscience.wiley.com).
YPLL, Productivity Costs, and Mortality/Ekwueme et al
C
ancer adversely affects the nation’s health and welfare through premature death, disability, medical costs, and lost productivity. Studies have reported that cancer is associated with the highest prevalence of impairment and amount of work loss of any chronic disease.1,2 Therefore, the indirect costs of cancer (eg, lost productivity because of morbidity and mortality) probably exceed the direct medical costs of cancer by a wide margin. Reducing the indirect costs of human papillomavirus (HPV)-associated cancers is an important potential benefit of public health interventions like cervical cancer screening and HPV vaccination. HPV, a sexually transmitted infection, plays a causal role in nearly 100% of cervical cancers, and it is believed that HPV also is associated causally with approximately 90% of anal cancers; 40% of vaginal, vulvar, and penile cancers; 25% of oral cavity cancers; and 35% of oropharyngeal cancers.3-6 In this report, we estimate the burden of specific cancer sites where HPV may be a risk factor for carcinogenesis (called HPV-associated cancers in this report) and all malignant cancers in terms of the years of potential life lost (YPLL) and the lifetime future productivity losses associated with cancer mortality in the United States in 2003. Although several studies have estimated various components of the health and economic burden of cancer,7-24 to our knowledge, this is the first study to examine the mortality costs of all HPV-associated cancers by race/ethnicity and sex. We believe that it is also the first to compare the HPV-associated cancer mortality burden to that of all other malignant cancers. Quantifying the burden of cancer mortality in the population may provide useful information for understanding the full potential benefit of prevention efforts.
MATERIALS AND METHODS We estimated the burden of selected HPV-associated cancers (cervical, anal, vaginal, vulvar, and penile cancers and a subset of oral cavity and pharyngeal cancers) without regard to the percentage of each of these cancers that could be attributed to HPV. We used 3 measures to quantify the societal burden of mortality because of HPV-associated cancers and all malignant cancers. These burden measures included mortality, YPLL, and the value of productivity loss from premature mortality. Mortality associated with all malignant cancers and HPV-associated cancers in 2003 was used in conjunction with such other data as life expectancy to estimate YPLL. We used spreadsheet software to estimate lost productivity in terms of incidence costs,25,26 focusing on the foregone lifetime productivity of those who died of cancer (HPV-
2937
associated cancers and all malignant cancers) in 2003.25,26 We estimated ‘incidence costs’ (the lifetime lost productivity of those who died of cancer in 2003), as opposed to ‘prevalence costs’ (the lost productivity in 2003 of those who died of cancer in 2003 or in previous years and who otherwise would have been alive in 2003). Productivity losses were measured by using the human capital approach,16-18 in which productivity losses resulting from premature mortality are valued by using market wage rates, and household productivity losses (child care, food preparation and clean-up, housework, home and car maintenance, etc) because of premature mortality are imputed by using replacement workers’ wage rates. All costs presented in this article were standardized to 2003 US dollars.
Data Sources and Extraction Data for this study were obtained from several sources. We used Surveillance, Epidemiology, and End Results software (SEER*Stat, version 6.3.6) to extract cancer mortality and associated population data, which covered 100% of US deaths,27,28 to estimate YPLL, average YPLL, and YPLL rates for all malignant cancers and for all HPV-associated cancers. Mortality data were obtained from the Centers for Disease Control and Prevention’s National Center for Health Statistics (NCHS) through the National Vital Statistics System.29 Unlike cancer incidence data, cancer mortality data do not provide any information on histology or detailed information other than standard anatomic site categories. For cancers of the cervix, vagina, vulva, penis, and anus, we examined all deaths from these cancers regardless of histology. For cancers of the oral cavity and pharynx, we chose 3 anatomic sites (tongue, tonsil, and other oral cavity and pharynx) as a proxy for HPV-associated cancers. These are identified as ‘oral cavity and pharyngeal’ cancers in this report. The methodology in this study was different from that described in the article on methods in this Supplement of Cancer.30 Cancers were coded as the primary cause of death according to 10 International Classification of Diseases for Oncology categories (cervix, C53; vagina, C52; vulva, C51; penis, C60; anus, C21; and a subset of oral cavity and pharyx: tongue, C01-C02; tonsil, C09; other oral cavity and pharynx, C14) that were in use at the time of death (available at: http://seer.cancer.gov/codrecode/ 19691_d09172004/index.html accessed on December 10, 2007).31 Population denominator data used for calculating rates and age-adjusted rates were obtained from the 2000 US Census.28 All rates are expressed per
2938
CANCER Supplement
November 15, 2008 / Volume 113 / Number 10
100,000 population. In each of these databases, we extracted data by 5-year age groups and sex, and we created 4 mutually exclusive race/ethnicity categories (ie, white non-Hispanic, black non-Hispanic, nonHispanic other, and Hispanic). We included individuals aged 15 years, because there were no deaths from HPV-associated cancers reported from ages birth to 14 years in 2003. We used 5-year age groups (ie, 15-19 years, 20–24 years, etc) through age 84 years, with a final age category for individuals aged 85 years. Estimates of the present value of future lifetime earnings (PVFLE) were based on the work of Grosse in 2003 and were updated from 2000 dollars to 2003 dollars by using currently available estimates of annual mean earnings from the US Census, estimates of the mean value of household services, and life table data.32–34
Estimation of YPLL, Average YPLL, YPLL Rates, and Age-adjusted YPLL Rates We used the life-expectancy method to calculate YPLL.12,35,36 We assumed that all cancer-related deaths were premature and that they contributed to YPLL, regardless of age. We calculated YPLL by multiplying the number of deaths in each age group by the corresponding remaining life expectancy obtained from the US life tables.34 The estimated number of YPLL in each age group was summed to obtain the total YPLL for all malignant cancers and all HPV-associated cancers in individuals aged 15 years. The estimation was stratified by race/ethnicity and sex. Because current life tables do not report life expectancy for American Indians/Alaska Natives, Asian/Pacific Islanders, or Hispanics, we used the life table data on all races as a proxy for non-Hispanic others and Hispanics.34 To compare YPLL across racial/ethnic groups, we estimated crude and age-adjusted rates (YPLL per 100,000 population). Age-adjusted rates were calculated by the direct method, using the 2000 US standard population by 5-year groups.37 In addition to the estimation of YPLL, average YPLL, and YPLL rates, we also estimated the ratio of the relative contribution of each HPV-associated cancer to YPLL and mortality. The numerator of the ratio was the YPLL for the specific HPV-associated cancer divided by the YPLL for all HPV-associated cancers combined and is expressed as a percentage; the denominator was the number of deaths from the specific HPV-associated cancer divided by the number of deaths from all HPV-associated cancers combined and is expressed as a percentage (ie, %YPLL/%mortality). This ratio was normalized to sum to 100%.
Estimation of Mortality-related Productivity Costs We followed previously published methods to estimate lifetime mortality-related productivity costs for all malignant cancers and for HPV-associated cancers.18,23,24,26,38 Productivity costs because of premature mortality were estimated by multiplying the number of deaths in 2003 (stratified by age, sex, and race/ethnicity) by the PVFLE stratified by age and sex. The PVFLE estimates that we applied33 took into account factors like life expectancy, the labor force participation rate, the future growth rate in productivity, and the imputed value of housekeeping services (eg, cooking, cleaning, childcare). In the base case, we applied PVFLE estimates based on a 3% discount rate, although we also examined how our results changed when we applied PVFLE estimates based on 0% and 5% discount rates. The discount rate was used to convert future earnings to present value.39
RESULTS Estimates of the number of YPLL and mortalityrelated productivity costs are described below for HPV-associated cancers and for all malignant cancers. It is important to clarify that the estimates of the burden of HPV-associated cancers reflects the burden of all cervical, anal, vaginal, vulvar, and penile cancers and a subset of oral cavity and pharyngeal cancers, without regard to the percentages of each of these cancers that could be attributed to HPV.
Years of Potential Life Lost In total, 8302 individuals died from all HPV-associated cancers in 2003. These deaths accounted for a loss of an estimated 181,026 YPLL at an average of 21.8 YPLL per death (Table 1). Men accounted for 29% of YPLL from HPV-associated cancers, with an average of 19.3 YPLL per death, and women accounted for 71%, with an average of 23.1 YPLL per death. In total, 462,618 individuals died from all malignant cancers in 2003, and they accounted for the loss of an estimated 7.5 million YPLL at an average of 16.3 YPLL per death. Thus, HPV-associated cancers accounted for 2.4% of the estimated YPLL among all malignant cancers. Cancer-specific estimates of YPLL because of HPV-associated cancer mortality ranged from 3654 YPLL for penile cancer to 45,815 YPLL for oral cavity/pharyngeal cancer in men and from 5199 YPLL for vaginal cancer to 89,936 YPLL for cervical cancer in women. The estimated age-adjusted YPLL rate for all HPV-associated
YPLL, Productivity Costs, and Mortality/Ekwueme et al
2939
TABLE 1 The Number of Deaths and Years of Potential Life Lost Associated With All Malignant Cancers and Human Papillomavirus-associated Cancers by Sex—United States, 2003* YPLL HPV-Associated and All Malignant Cancers by Sex Males HPV-associated cancers§ Penis Anus Oral cavity/pharynxk All HPV-associated cancers All malignant cancers Females HPV-associated cancers§ Cervix Vagina Vulva Anus Oral cavity/pharynxk All HPV-associated cancers All malignant cancers Males and females HPV-associated cancers§ Anus Oral cavity/pharynxk All HPV-associated cancers All malignant cancers
No. of Deaths
Population
Rate per 100,000
Total
Per Deathsy
Rate per 100,000
Age-Adjusted Rate{
212 156 2377 2745 238,377
92,821,488 92,821,488 92,821,488 92,821,488 92,821,488
0.23 0.17 2.56 2.96 256.81
3654 3401 45,815 52,870 3,651,468
17.24 21.80 19.27 19.26 15.32
3.94 3.66 49.36 56.96 3933.86
4.13 3.71 50.13 57.97 4191.69
3259 325 674 297 1002 5557 224,241
97,994,875 97,994,875 97,994,875 97,994,875 97,994,875 97,994,875 97,994,875
3.33 0.33 0.69 0.30 1.02 5.67 228.83
89,936 5199 9347 5901 17,773 128,155 3,877,885
27.60 16.00 13.87 19.87 17.74 23.06 17.29
91.78 5.31 9.54 6.02 18.14 130.78 3957.23
91.24 4.94 8.76 5.71 17.08 127.73 3724.44
453 3379 8302 462,618
190,816,363 190,816,363 190,816,363 190,816,363
0.24 1.77 4.35 242.44
9302 63,587 181,026 7,529,354
20.54 18.82 21.81 16.28
4.88 33.32 94.87 3945.86
4.80 32.71 93.98 3911.43
YPLL indicates years of potential life lost; HPV, human papillomavirus. *Data presented are for all races/ethnicities aged 15 years to 85 years. y YPLL per death was calculated by dividing the estimated number of YPLL in each cancer site by the number of deaths that occurred prematurely. { Age-adjusted rates per 100,000 population were calculated by the direct method using the 2000 US standard population in 5-year groups (ie, ages 15-19 years, 20-24 years, . . ., 85 years; see Surveillance Epidemiology, and End Results Program 200737). § HPV-associated cancers were defined as those cancer sites where HPV may be a risk factor for carcinogenesis without regard to the fraction of those cancers that can be attributed to HPV. k The subset of oral cavity and pharyngeal cancers included as HPV-associated cancers: cancers of the tonsil, tongue, and other oral cavity/pharyngeal cancers.
cancers was 94 per 100,000 men and women, 58 per 100,000 men, and 127.7 per 100,000 women. The proportion of YPLL because of HPV-associated cancers relative to YPLL because of all malignant cancers varied by age and sex (Fig. 1). For HPVassociated cancers, greater proportionate contributions to YPLL were observed in the groups between ages 25 to 29 years and ages 55 to 59 years overall, in the groups between ages 40 to 44 years and ages 55 to 59 years for men, and in the groups between ages 25 to 29 years and ages 55 to 59 years for women. HPV-associated cancers had the largest relative contribution to YPLL among women ages 30 to 34 years. In all HPV-associated cancers, as expected, YPLL per death decreased as age at death increased (data not shown). Figure 2 demonstrates the YPLL and mortality ratios for specific HPV-associated cancers. The line y 5 x represents the line of equality and indicates a
cancer site’s equal relative contribution to YPLL and mortality for all HPV-associated cancers. HPV-associated cancers that fall below the line of equality contributed more to mortality than to YPLL, whereas cancers that fell above the line of equality contributed more to YPLL than to mortality. For example, cervical cancer contributed more to YPLL relative to mortality (ratio, 1.27), whereas vulvar cancer contributed more to mortality than to YPLL (ratio, 0.64).
Mortality-related Productivity Costs The estimated total lifetime mortality cost from all malignant cancer mortality in 2003 was $133.5 billion, or $288,644 per death (Table 2). When discounted at an annual rate from 0% to 5% (rather than 3%), the estimated costs ranged from $111.7 billion to $185.1 billion. Deaths among men accounted for $74.2 billion of lost productivity (or 55.6% of the total), with an average cost per death of $311,255.
2940
CANCER Supplement
November 15, 2008 / Volume 113 / Number 10
FIGURE 1. This chart illustrates the proportion of years of potential life lost (YPLL) because of human papillomavirus (HPV)-associated cancers relative to YPLL because of all malignant cancers, by age at death and sex, in the United States in 2003. HPV-associated cancers were defined as those cancer sites where HPV may be a risk factor for carcinogenesis without regard to the fraction of those cancers that can be attributed to HPV. The subset of oral cavity and pharyngeal cancers included as HPV-associated cancers: cancers of the tonsil, tongue, and other oral cavity/pharyngeal cancers. (Data are from population-based national mortality data collected by the Centers for Disease Control and Prevention’s National Center for Health 29
Statistics, 2007. )
FIGURE 2. This chart illustrates the years of potential life lost (YPLL) and mortality ratios for specific human papillomavirus (HPV)-associated cancers in the United States in 2003. The ratios indicated by asterisks were calculated as follows: The numerator of the ratio is the YPLL for the specific HPVassociated cancer divided by the YPLL for all HPV-associated cancers combined and is expressed as a percentage; the denominator is the number of deaths caused by the specific HPV-associated cancer divided by the number of deaths from all HPV-associated cancers combined and also is expressed as a percentage (ie, % YPLL/% mortality), which indicates the relative contribution of each cancer to these 2 measures of burden of HPV-associated cancer. For example, cervical cancer contributes relatively more to YPLL than to mortality, whereas vulvar cancer contributes relatively more to mor-
Deaths among women accounted for $59.3 billion of lost productivity (or 44.4% of the total), for an average cost per death of $264,608. The estimated total lifetime mortality costs from HPV-associated cancer mortality in 2003 were $3.6 billion, or $437,223 per death (Table 2). The estimated mortality cost accounted for 2.7% of the mortality cost of all malignant cancers (Fig. 3). When discounted at an annual rate from 0% to 5%, the estimated mortality cost for HPV-associated cancer ranged from $2.9 billion to $5.3 billion (Table 2). The estimated mortality cost accounted for $1.3 billion or $465,957 cost per death among men and for $2.4 billion or $423,030 cost per death among women. For all HPV-associated cancers, white non-Hispanics had the highest estimated mortality cost, $2.5 billion (37.4% among men and 62.6% among women). NonHispanics of other races had the lowest estimated cost at $117.0 million (21.7% among men and 78.3% among women). For HPV-associated cancers among men, penile cancer had the lowest estimated cost at $86.5 million ($407,986 per death), whereas oral cavity/pharyngeal cancer had the highest estimated cost at $1.1 billion ($463,623 per death). Among women, vaginal cancer had the lowest estimated cost at $83.3 million ($256,241 per death), and cervical cancer had the highest estimated cost at $1.8 billion ($541,576 per death).
tality than to YPLL. All ratios were normalized to sum to 100%. HPV-associated cancers were defined as those cancer sites where HPV may be a risk factor for carcinogenesis without regard to the fraction of those cancers that can be attributed to HPV. The subset of oral cavity and pharyngeal cancers that were included as HPV-associated cancers are indicated by a dagger and included cancers of the tonsil, tongue, and other oral cavity/ pharyngeal cancers.
The percentage contributions of each HPV-associated cancer by race/ethnicity to the total lifetime mortality costs for all malignant cancers are summarized in Figure 3. The proportional burden of HPV-associated cancer was greatest for Hispanics (4.1%), followed by black non-Hispanics (3.6%), nonHispanics of other races (2.9%), and white non-Hispanics (2.4%). The percentage distributions of the total lifetime mortality costs of all HPV-associated cancers by race/ethnicity varied widely and are presented in Figure 4. For example, among white nonHispanics, cervical cancer accounted for 44.2% of the total mortality costs of all HPV-associated cancers, oral cavity/pharyngeal accounted for 40.9%, and anal cancer accounted for 5.8%. The observed higher proportional burden of HPV-associated cancer for Hispanics and non-Hispanics of other races may be attributable to the small sample size. For example, non-Hispanics of other races and Hispanics
YPLL, Productivity Costs, and Mortality/Ekwueme et al
2941
TABLE 2 Present Future Value of Lifetime Lost Productivity Because of Mortality From All Malignant Cancers and Human Papillomavirus-associated Cancers by Sex and Race/Ethnicity—United States, 2003 Males
Females
Discounted Total Cost ($ 3103) Cancer Site By Race/Ethnicity Cervix All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic Vagina All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic Vulva All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic Penis All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic Anus All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic Oral cavity/pharynx* All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic All HPV-associated cancersy All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic All malignant cancers All races/ethnicities White non-Hispanic Black non-Hispanic Other non-Hispanic Hispanic
Males and Females
Discounted Total Cost ($ 3103)
Discounted Total Cost ($ 3103)
3%
5-0%
Per Death
3%
5-0%
Per Death
5-0%
Per Death
— — — — —
— — — — —
— — — — —
1,764,998 1,115,935 331,119, 74,768 238,143
1,403,287-2,714,162 890,432-1,702,429 261,948-514,818 59,452-115,069 187,419-374,297
541,576 — 520,007 — 560,270 — 522,852 — 641,895 —
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
83,278 61,773 13,863 4252 3310
68,996-117,423 51,374-86,358 11,390-19,870 3317-6791 2842-4309
256,241 — 244,161 — 308,069 — 708,694 — 165,482 —
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
133,476 115,909 10,394 1086 5905
112,415-181,683 97,874-156,894 8517-15,017 917-1465 4949-8079
198,035 — 188,470 — 358,418 — 217,110 — 246,030 —
— — — — —
— — — — —
86,493 51,294 6261 4537 21,281
71,810-121,067 42,942-70,562 5334-8303 6306-3773 17,202-31,437
407,986 344,254 260,885 756,209 709,366
— — — — —
— — — — —
— — — — —
— — — — —
— — — — —
90,526 64,486 16,791 1843 7406
74,601-128,570 53,493-90,362 13,490-25,097 1537-2550 6082-10,562
580,292 520,047 883,726 614,279 740,614
98,974 83,040 9556 3145 3233
81,155-142,480 68,121-119,386 7705-14,346 2607-4388 2722-4361
333,246 189,500 329,524 147,526 398,180 26,347 349,394 4987 269,427 10,639
155,756-271,051 121,614-209,748 21,195-39,443 4144-6937 8803-14,922
418,321 392,356 612,724 415,615 483,603
1,102,032 828,083 186,510 19,036 66,559
924,061-1,511,053 695,190-1,132,342 155,817-257,486 15,969-26,119 55,511-92,671
463,623 446,888 550,178 432,634 489,401
270,049 205,535 43,807 8320 12,205
224,338-377,663 171,418-285,068 35,720-63,661 6974-11,383 10,067-17,322
269,510 1,372,081 250,958 1,033,618 401,902 230,318 268,386 27,356 290,589 78,763
1,148,399-1,888,7162 866,609-1,417,410 191,537-321,147 22,944-37,502 65,578-109,993
406,061 386,833 514,102 364,745 442,491
1,279,051 943,863 209,563 25,416 95,246
1,070,472-1,760,690 791,625-1,293,266 174,641-290,885 21,279-34,975 78,794-134,669
465,957 443,962 548,593 479,547 541,169
2,350,776 1,582,191 408,740 91,570 262,795
1,890,191-3,533,411 1,279,220-2,350,136 325,280-627,713 73,268-139,095 207,999-408,367
423,030 387,317 512,206 472,011 560,331
2,960,663-5,294,102 2,070,844-3,643,402 499,921-918,598 94,547-174,070 286,793-543,036
437,223 406,707 523,985 473,628 555,103
74,196,015 57,359,510 9,447,749 2,225,370 4,973,058
62,383,812-101,534,904 48,404,105-77,801,408 7,893,563-13,092,294 1,842,100-3,148,645 4,076,314-7,217,416
311,255 291,360 400,057 389,187 426,031
59,335,958 46,015,634 7,527,722 1,845,846 3,820,851
49,184,073-83,533,223 38,322,872-64,096,229 6,157,135-10,898,526 1,506,124-2,689,720 3,098,791-5,671,123
264,608 133,531,973 247,497 103,366,144 341,393 16,975,471 344,825 4,071,216 364,168 8,793,909
111,567,885-185,068,134 86,726,976-141,897,637 14,050,698-23,990,820 3,349,124-5,838,367 7,170,104-12,888,539
288,644 270,054 371,731 367,737 396,748
3%
— — — — —
3,629,826 2,526,054 618,303 116,986 358,041
HPV indicates human papillomavirus. *The subset of oral cavity and pharyngeal cancers included as HPV-associated cancers: cancers of the tonsil, tongue, and other oral cavity/pharyngeal cancers. y HPV-associated cancers were defined as those cancer sites where HPV may be a risk factor for carcinogenesis without regard to the fraction of those cancers that can be attributed to HPV.
2942
CANCER Supplement
November 15, 2008 / Volume 113 / Number 10
FIGURE 3. This chart illustrates the percentage contribution of each
FIGURE 4. This chart illustrates the percentage contribution of each
human papillomavirus (HPV)-associated cancer to the total lifetime mortality
human papillomavirus (HPV)-associated cancer to the total lifetime mortality
costs of all malignant cancers by race/ethnicity in the United States in 2003. HPV-associated cancers (asterisks) were defined as cancer sites at which
costs of all HPV-associated cancers by race/ethnicity in the United States in 2003. HPV-associated cancers were defined as those cancer sites where
HPV may be a risk factor for carcinogenesis without regard to the fraction of
HPV may be a risk factor for carcinogenesis without regard to the fraction
that could be attributed to HPV. The subset of oral cavity and pharyngeal
of those cancers that can be attributed to HPV. A dagger indicates the sub-
cancers that were included as HPV-associated cancers are indicated by a dagger and included cancers of the tonsil, tongue, and other oral cavity/pha-
set of oral cavity and pharyngeal cancers that were included as HPV-asso-
ryngeal cancers.
pharyngeal cancers.
accounted for 3% and 8% of all HPV-associated mortality, respectively. The estimated PVFLE is sensitive to the discount rate chosen. For example, each percentage point increase in the discount rate causes the total mortality-related productivity loss from all malignant cancers to decrease by $16.9 billion to $20.6 billion (or 7.3%-7.7%). For HPV-associated cancers, the total mortality-related productivity costs decreased by $543.4 million to $679.1 million (or 7.9%-8.2%).
DISCUSSION HPV-associated cancer mortality in 2003 accounted for 181,026 YPLL or 2.4% of the estimated 7.5 million YPLL attributable to mortality from all malignant cancers in 2003. The average number of YPLL was 21.8 per HPV-associated cancer death and 16.3 per malignant cancer death, indicating the extent to which HPV-associated cancers and cancers in general reduce life expectancy. HPV-associated cancers accounted for 2.7% of $133.5 billion in lost productivity from all malignant cancer mortality in 2003. Our estimates suggest that each HPV-associated cancer death was associated with $437,223 in forgone lifetime earnings expressed in 2003 US dollars. For all malignant cancers, the productivity loss per death was $288,644. In addition, our estimates suggest that HPV-associated cancers among younger individuals, and among women in particular (eg, cervical cancer),
ciated cancers: cancers of the tonsil, tongue, and other oral cavity/
contribute more to YPLL than estimates of HPV-associated cancers among older individuals. It is important to note that our estimates reflect the burden of selected cancers that often are associated with HPV. Therefore, we are not taking into account the attributable fraction for each selected cancer. If we were to consider the attributable fraction, then the burden estimates would be lower than those presented here. The results presented in this report help to quantify the burden of each HPV-associated cancer compared with all malignant cancers in terms of YPLL and mortality-related productivity costs in the United States. Another special feature of this report is that we quantify the burden of these cancers by race/ethnicity. These estimates can be valuable for assessing the relative burden of HPVassociated cancers compared with the overall burden of cancer; the estimates can inform national decisions about how to prioritize cancer prevention programs for populations with the greatest need for prevention services, and they can provide a benchmark for efforts to quantify the costs and benefits of prevention strategies for HPV-associated cancers. Our estimates of YPLL and mortality costs for all malignant cancers and for some HPV-associated cancers in the United States are not the first. Previous studies have reported that cancer deaths were responsible for nearly 8.6 million YPLL in 2004, with an average of 15.5 YPLL per death.40 This previous
YPLL, Productivity Costs, and Mortality/Ekwueme et al
estimate is roughly comparable to our estimates for all malignant cancers (7.5 million YPLL with an average of YPLL at 16.3 years). One possible explanation for the slight difference in our estimated YPLL, compared with that from a previous study, may be that the previous study used 1-year intervals, whereas we used 5-year age-group intervals. To our knowledge, there is no previous study on YPLL for all HPV-associated cancers with which to compare our YPLL estimate of 181,026. However, studies have reported average YPLL for various HPV-associated cancers, such as cervical cancer (25.9-26.1 YPLL per death) and oral cavity/pharyngeal cancer (16.4-17.4 YPLL per death).7,14 Our estimated averages of 27.6 YPLL for cervical cancer and 18.8 YPLL for oral cavity/pharyngeal cancers were slightly higher than results from other studies. Similarly, although there is no study reporting the mortality costs of all HPV-associated cancers, a few studies have estimated mortality costs of cervical and oral cavity/pharyngeal cancers.15,23,24 For cervical cancer, a recent study reported that annual mortality costs in the United States were $1.9 billion in 2000 US dollars.24 Although there are methodological differences between that study and ours, our $1.8 billion estimate is consistent with the previous study. Other studies that have estimated mortality costs for cervical cancer were conducted more than 20 years ago, probably precluding a comparison of those results and the results from the current study.18,19 For oral cavity/pharyngeal cancer, 1 study reported that mortality costs were $943 million in 1997 US dollars, discounted at 4%.15 If adjusted for inflation to 2003 dollars, then the mortality costs would be equivalent to $1.1 billion, an amount that is consistent with our estimated mortality costs of $1.4 billion when discounted at an annual rate of 3%. Several studies have reported aggregate costs of all cancers in the United States. However, those studies were conducted more than 10 years ago.15,18-21 Consequently, the estimates reported in these studies may not be comparable to our estimated mortality costs of cancer. Therefore, we compared our results with those in the most recent report from the National Institutes of Health (NIH), which estimated that mortality costs in 2003 for all cancers were $109.0 billion.22 Our estimated mortality cost was $133.5 billion, which is 18.4% more than that reported by the NIH.22 The difference in our mortality cost estimate and that of the NIH probably is attributable to differences in methodology and data sources. For example, we excluded deaths of individuals aged <15 years. Furthermore, we used 2003 data on the US mortality, life tables, annual earnings,
2943
household services, and labor force participation rate to estimate mortality cost. In contrast, the estimate reported by the NIH used the 1999 mortality data, estimated PVFLE in 1999, and updated those data to 2003 dollars. The results reported in this article are based on the most current and reliable population-based national death certificate registry and other national databases. However, there are some limitations to this study. First, productivity losses are based on average wages, and the reported productivity losses may be overestimated if HPV-associated cancers or cancers in general are more likely to occur in individuals with lower socioeconomic status (SES). Benard et al in this Supplement report evidence that individuals of lower SES have a higher incidence of HPVassociated cancers, with the exception of anal and oral cavity/oropharyngeal cancers.41 Similarly, our analysis did not incorporate information on racespecific earnings, so that actual productivity losses may be overestimated for racial groups with belowaverage earnings and underestimated for racial groups with above-average earnings. Second, the human capital approach has several limitations.26,42,43 For example, this approach does not recognize the costs of pain and suffering or the psychosocial consequences of cancer disease. In addition, the value of an individualâ&#x20AC;&#x2122;s lifetime earnings does not reflect the full value that society places on the individual. Third, our use of life table data on all races as a proxy for life expectancy of Hispanics and â&#x20AC;&#x2DC;non-Hispanic othersâ&#x20AC;&#x2122; may have overestimated (or underestimated) the burden of cancer for a given group if that group had a lower (or higher) than average life expectancy. Fourth, because mortality data did not have histology or subsite specified, we may have overestimated the mortality burden, especially for oral cavity and pharyngeal cancers. Fifth, as noted above, our estimates apply to HPV-associated cancers without regard to the percentage of these cancers that may be attributable to HPV. Finally, our analysis excluded the productivity losses associated with cancer morbidity, direct medical costs of treatment, and nonmedical costs (such as those associated with time spent in seeking cancer treatment and care). The estimated direct treatment costs of all cancers in the United States in 2004 was $72.2 billion.44,45 The nonmedical costs associated with cancer have been estimated at $2.3 billion per year.46 Furthermore, our analyses excluded productivity losses for caregivers. These losses include the time that family, friends, and volunteers spend caring for an individual with cancer, and the associated cost can be substantial.47-50
2944
CANCER Supplement
November 15, 2008 / Volume 113 / Number 10
Despite these limitations, the health and economic burden estimates presented in this report illustrate the significant numbers of premature deaths and significant mortality costs associated with all malignant cancers as well as HPV-associated cancers in the United States. In the past, estimates of economic burden of diseases, including cancer, have proven useful as guidance for healthcare policy debates in the general population.18,20 The current findings underscore the need for HPV vaccination, cancer screening, and other effective early detection programs for HPV-associated cancers. It is through these interventions that there can be a reduction in premature deaths and the economic burden associated with these cancers.
REFERENCES 1.
Kessler RC, Greenberg PE, Mickelson KD, Meneades LM, Wang PS. The effects of chronic medical conditions on work loss and work cutback. J Occup Environ Med. 2001;43:218-225. 2. Chirikos TN, Russell-Jacobs A, Cantor AB. Indirect economic effects of long-term breast cancer survival. Cancer Pract. 2002;10:248-255. 3. Walboomers JMM, Jacobs MV, Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical worldwide. J Pathol. 1999;189:12-19. 4. Franco EL, Rohan TE, Villa LL. Epidemiologic evidence and human papillomavirus infection as a necessary cause of cervical cancer. J Natl Cancer Inst. 1999;91:506-511. 5. Kreimer AR, Clifford GM, Boyle P, Franceschi S. Human papillomavirus types in head and neck squamous cell carcinomas worldwide: a systematic review. Cancer Epidemiol Biomarkers Prev. 2005;14:467-475. 6. Parkin DM, Bray F. Chapter 2: the burden of HPV-related cancers. Vaccine. 2006;24(suppl 3):S11-S25. 7. Ries LAG, Harkins D, Krapcho M, et al. SEER Cancer Statistics Review, 1975-2003. Bethesda, Md: National Cancer Institute, SEER Program; 2006, Available at: http://seer.cancer.gov/csr/1975_2003/. Based on November 2005 SEER data submission, posted to the SEER website, 2006. Accessed December 10, 2007. 8. Horm JW, Sondik EJ. Person-years of life lost due to cancer in the United States, 1970 and 1984. Am J Public Health. 1989;79:1490-1493. 9. Rainbolt T. Years of potential life lost: an evaluation of premature cancer deaths in Texas from 1981 through 1988. Tex Med. 1992;88:74-76. 10. Firman PC, Finney JW, Leibowitz JM. Years of potential life lost: evaluating premature cancer death in men. J Commun Health. 1989;14:101-106. 11. Albert VA, Koh HK, Geller AC, Miller DR, Prout MN, Lew RA. Years of potential life lost: another indicator of the impact of cutaneous malignant melanoma on society. J Am Acad Dermatol. 1990;23(2 pt 1):308-310. 12. Burnet NG, Jefferies SJ, Benson RJ, Hunt DP, Treasure FP. Years of life lost (YLL) from cancer is an important measure of population burdenâ&#x20AC;&#x201D;and should be considered when allocating research funds. Br J Cancer. 2005;92:241245.
13. Wong MD, Shapiro MF, Boscardin WJ, Ettner SL. Contribution of major diseases to disparities in mortality. N Engl J Med. 2002;347:1585-1592. 14. Brown ML, Lipscomb J, Snyder C. The burden of illness of cancer: economic cost and quality of life. Annu Rev Public Health. 2001;22:91-13. 15. Brown ML, Hodgson TA, Rice DP. Chapter 13. Economic impact of cancer in the United States. In: Schottenfeld D, Fraumini JF, eds. Cancer Epidemiology and Prevention, 2nd ed. New York, NY: Oxford University Press;1996:255266. 16. Rice DP. Estimating the Cost of Illness. Health Economics Series, No. 6. Public Health Service Publication No. 947-6. Washington, DC: US Department of Health, Education, and Welfare; 1966. 17. Cooper BS, Rice DP. The economic cost of illness revisited. Soc Secur Bull. 1976;39:21-36. 18. Rice DP, Hodgson TA. Social and economic implications of cancer in the United States. Vital and Health Statistics, Series 3, No. 20. DHHS Publication No. PHS-81-1404. Public Health Service. Washington, DC: US Government Printing Office; 1981. 19. Hodgson TA, Rice DP. Chapter 12. Economic impact of cancer in the United States. In: Schottenfeld D, Fraumeni JF, eds. Cancer Epidemiology and Prevention. Philadelphia, Pa: Saunders; 1982:208-228. 20. Rice DO. Hodgson TA, Kopstein AN. The economic costs of illness: a replication and update. Health Care Financ Rev. 1985;7:61-80. 21. Rice DP, Hodgson TA, Capell F. The economic burden of cancer, 1985: United States and California. In: Scheffler R, Andrews NC, eds. Cancer Care and Cost; DRGs and Beyond, Chapter 3. Ann Arbor, MI: Health Administration Press Perspectives; 1986:39-59. 22. National Heart, Lung, and Blood Institute. Fact Book: Fiscal Year 2002. Bethesda, Md: National Institutes of Health; 2003. Available at: http://www.nhlbi.nih.gov/about/02factbk. pdf. Accessed December 10, 2007. 23. Max W, Rice DP, Sung HY, Michel M, Breuer W, Zhang X. The economic burden of gynecologic cancers in California, 1998. Gynecol Oncol. 2003;88:96-03. 24. Insinga RP. Annual productivity costs due to cervical cancer mortality in the United States. Womens Health Issues. 2006;16:236-242. 25. Rice DP, Kelman S, Miller LS, Dunmeyer S. The Economic Costs of Alcohol and Drug Abuse and Mental Illness. Rockville, Md: National Institute on Drug Abuse; 1990. 26. Hodgson TA. The state of the art of cost-of-illness estimates. Adv Health Econ Health Serv Res. 1983;4:129-164. 27. Surveillance, Epidemiology, and End Results (SEER) Program. SEER*Stat Software [computer program]. Version 6.3.6. Bethesda, Md: National Cancer Institute; 2007. Available at: http://seer.cancer.gov/seerstat/. Accessed December 10, 2007. 28. US Cancer Statistics Working Group. United States Cancer Statistics: 2003 Incidence and Mortality. Atlanta, Ga: Department of Health and Human Services, Centers for Disease Control and Prevention, and National Cancer Institute; 2006. Available at: http://elib2.cdc.gov:2087/cancer/ npcr/uscs. Accessed December 10, 2007. 29. National Center for Health Statistics. National Vital Statistics System. Atlanta, Ga: Centers for Disease Control and Prevention, National Center for Health Statistics; 2007. Available at: http://www.cdc.gov/nchs/nvss.htm. Accessed December 10, 2007.
YPLL, Productivity Costs, and Mortality/Ekwueme et al 30. Watson M, Saraiya M, Ahmed F, et al. Using populationbased cancer registry data to assess the burden of human papillomavirus -associated cancers in the United States: overview of methods. Cancer. 2008;113(10 suppl):2841-2854. 31. Fritz A, Percy C, Jack A, et al, eds. International Classification of Diseases for Oncology, 3rd ed. Geneva, Switzerland: World Health Organization; 2000. 32. US Census Bureau. Current Population Survey, 2001-2004. Annual Social and Economic Supplements. Special Income Tabulation for the Year 2003. Washington, DC: Government Printing Office; 2004. 33. Grosse SD. Appendix I. Productivity loss tables. In: Haddix AC, Teutsch SM, Corso PS, eds. Prevention Effectiveness: A Guide to Decision Analysis and Economic Evaluation, 2nd ed. New York, NY: Oxford University Press; 2003:245-257. 34. Arias E. United States Life Tables, 2003. National Statistics Reports, 54. Hyattsville Md: National Center for Health Statistics; 2006. 35. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA. 2003;289:187193. 36. Rosenberg DC, Buescher PA. Years of Potential Life Lost by Sex, Race, and Ethnicity: North Carolina, 2000. A Special Report Series by the State Center for Health Statistics. No. 130, February 2002. Available at: http://www.schs.state. nc.us/SCHS/pdf/SCHS130.pdf. Accessed December 10, 2007. 37. Surveillance Epidemiology and End Results. 2000 US Standard Population Versus Standard Million. Bethesda, Md: National Cancer Institute, SEER Program; 2007; Available at: http://seer.cancer.gov/stdpopulations/single_age. html. Accessed December 10, 2007. 38. Max W, Rice DP, Sung H-Y, Michel M, Breuer W, Zhang X. The economic burden of prostate cancer, California, 1998. Cancer. 2002;94:2906-2913. 39. Corso PA, Haddix AC. Time effects. In: Haddix AC, Teutsch SM, Corso PS, eds. Prevention Effectiveness: A Guide to Decision Analysis and Economic Evaluation, 2nd ed. New York, NY: Oxford University Press; 2003:92-102.
2945
40. Ries LAG, Eisner MP, Kosary CL, et al. SEER Cancer Statistics Review, 1975-2004. Bethesda, Md: National Cancer Institute; 2007. Available at:http://seer.cancer.gov/csr/1975_ 2004/. Accessed December 10, 2007. 41. Benard V, Thompson T, Johnson C, et al. Examining the association between socio-economic factors and potentially human papillomavirus-associated cancers. Cancer. 2008;113(10 suppl):2910-2918. 42. Drummond M. Cost-of-illness studies: a major headache? Pharmacoeconomics. 1992;2:1-4. 43. Hodgson TA, Meiners MR. Cost-of-illness methodology: a guide to current practices and procedures. Milbank Mem Fund Q Health Soc. 1982;60:429-462. 44. National Cancer Institute (NIH). Cancer Trends Progress Reportâ&#x20AC;&#x201D;2005 Update. Available at: http://progressreport. cancer.gov/doc_detail.asp?pid51&did52005&chid525&coid5 226&mid5. Accessed December 10, 2007. 45. Brown ML, Yabroff KR. Economic impact of cancer in the United States. In: Schottenfeld D, Faumeni JF, eds. Cancer Epidemiology and Prevention. New York, NY: Oxford University Press; 2006:202-214. 46. Yabroff KR, Davis WW, Lamont EB, et al. Patient time costs associated with cancer care. J Natl Cancer Inst. 2007;3:99: 14-23. 47. Hayman JA, Langa KM, Kabeto MU, et al. Estimating the cost of informal caregiving for elderly patients with cancer. J Clin Oncol. 2001;19:3219-3225. 48. Emanuel EJ, Fairclough DL, Slutsman J, Alpert H, Baldwin D, Emanuel LL. Assistance from family members, friends, paid care givers, and volunteers in the care of terminally ill patients. N Engl J Med. 1999;341:956-963. 49. Siegel K, Raveis VH, Houts P, Mor V. Caregiver burden and unmet patient needs. Cancer. 1991;68:1131-1140. 50. Covinsky KE, Goldman L, Cook EF, et al. The impact of serious illness on patientsâ&#x20AC;&#x2122; families. SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment. JAMA. 1994;272:18391844.