Cancer Epidemiology 39 (2015) 126–132

Comparison of expected health impacts for major cancers: Integrationof incidence rate and loss of quality-adjusted life expectancy

Mei-Chuan Hung a, Wu-Wei Lai b, Helen H.W. Chen c, Wu-Chou Su d, Jung-Der Wang a,d,e,*a Department of Public Health, National Cheng Kung University College of Medicine, Tainan 704, Taiwanb Department of Surgery, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan 704, Taiwanc Department of Radiation Oncology, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan 704, Taiwand Department of Internal Medicine, National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Tainan 704, Taiwane Department of Occupational and Environmental Medicine, National Cheng Kung University Hospital, Tainan 704, Taiwan

A R T I C L E I N F O

Article history:

Received 26 July 2014

Received in revised form 14 November 2014

Accepted 8 December 2014

Available online 30 December 2014

Keywords:

Health impacts

Incidence rate

Loss-of-QALE (quality-adjusted life

expectancy)

Cancer prevention

A B S T R A C T

Purpose: The study aims to quantify the expected impacts of different cancers through multiplying the

incidence rate by loss-of-QALE (quality-adjusted life expectancy), with QALY (quality-adjusted life year)

as the common unit, to aid prevention policy decisions.

Methods: 464,722 patients with pathologically verified cancer registered in the Taiwan Cancer Registry

during 1998–2009 were used to estimate lifetime survival through Kaplan–Meier estimation combined

with a semi-parametric method. A convenience sample for measuring the utility value with EQ-5D was

conducted with 11,453 cancer patients, with the results then multiplied by the survival functions to

estimate QALE. The loss-of-QALE was calculated by subtracting the QALE of each cancer cohort from the

life expectancy of the corresponding age- and gender-matched reference population. The cumulative

incidence rates from age 20 to 79 (CIR20–79) were calculated to estimate the lifetime risk of cancer for

each organ-system.

Results: Liver and lung cancer were found the highest expected lifetime health impacts in males and

females, or expected lifetime losses of 0.97 and 0.41 QALYs that could be averted, respectively. While the

priority changes for prevention based on expected health impacts were slightly different for females

based on standardized mortality rates, those of males involve a broader spectrum, including oral,

colorectal, esophageal and stomach cancer.

Conclusion: The integration of incidence rate with loss-of-QALE could be used to represent the expected

losses that could be averted by prevention, which may be useful in prioritizing strategies for cancer

control.

� 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND

license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Contents lists available at ScienceDirect

Cancer EpidemiologyThe International Journal of Cancer Epidemiology, Detection, and Prevention

jou r nal h o mep age: w ww.c an cer ep idem io log y.n et

1. Introduction

While applications of epidemiologic methods to health policydecisions were once popular before the 19th century, they havebecome less widely used since the beginning of the second half ofthe 20th century. Although estimates of standardized mortalityrates can still provide useful information on the impacts of diseasesor technologies, which can then aid health policy decision, such

* Corresponding author at: Department of Public Health, National Cheng Kung

University College of Medicine, No. 1, University Road, Tainan 701, Taiwan.

Tel.: +886 6 2353535x5600; fax: +886 6 2359033.

E-mail address: jdwang121@gmail.com (J.-D. Wang).

http://dx.doi.org/10.1016/j.canep.2014.12.004

1877-7821/� 2014 The Authors. Published by Elsevier Ltd. This is an open access articl

3.0/).

comparisons for different illnesses have become less efficient,especially because for most cancers patients usually survive formore than 5–10 years, and thus one must wait for a long period oftime for mortality to occur [1–6]. The current coding of oneunderlying cause of death might also underestimate the impact ofother causes, including cancer. At the same time, the sustainabilityof current healthcare systems has become questionable under thefinancial pressures of an aging population and development of newtechnologies [7]. From 1971 to 2012, the number of cancersurvivors in the United States increased from 3.0 million to13.4 million, representing 4.6% of the population [8,9]. As thepopulation of cancer survivors increases, the economic impact ofcancer for patients, caregivers, employers, the health-care systemand society overall is expected to grow [10,11]. There is thus a need

e under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/

0 100 200 300 400 500 6000.0

0.2

0.4

0.6

0.8

1.0

Age, sex-matched referents ____ Colorec tal cancer ----- -

Loss-of-QALE=5.19 QA LY

Time in month after diagnosis

Util

ity

Fig. 1. Estimation of loss-of-QALE (quality-adjusted life expectancy): the dashed

line represents the QALE of male patients with colorectal cancer, while the solid line

represents the QALE of corresponding age and sex-matched referents. The area

between solid and dashed lines would be the loss-of-QALE, or 5.19 QALY (quality-

adjusted life years).

M.-C. Hung et al. / Cancer Epidemiology 39 (2015) 126–132 127

and opportunity to consider alternative outcome measures whenmaking policy decisions.

In clinical medicine, QALY (quality-adjusted life years) is ameasure that incorporates both quality of life and length of survivalinto one common unit, and this has been widely applied in cost-utility and/or cost-effectiveness analysis over the last few decades[12–14]. Recently, the authors of this study have developedestimates of the expected years of life lost and loss-of-QALE thatquantify the consequences of illness with life-year and QALY,respectively, which make possible direct comparisons of preventionwith clinical care [1–4,15] if incidence rates can also be incorporatedtogether [16]. It may thus be possible to quantify both incidencerates and how much utility is lost from developing a specific illness,rather than simply making policy decisions based on mortality rates.

Unfortunately, randomized controlled trials are generallyexpensive, and the financial costs for those focused on preventionare even higher because of the low occurrence rates of mostdiseases [17,18]. Therefore, we attempted to use the big data and/or databases of electronic medical records available in Taiwan tointegrate the likelihood of an event, or incidence rate andconsequence of an event, or loss-of-QALE, to estimate the lifetimehealth impacts of major cancers to aid prevention policydecisions, and further compare the results of this with thestandardized mortality rates in order to decide priorities forcancer prevention efforts. In other words, we quantified thelifetime risk (or, cumulative incidence rate for age 20–79) andthen multiplied this by the loss-of-QALE to estimate the healthimpacts for different cancers, and compared the results with therelated standardized mortality rates.

2. Method

2.1. Study population and datasets

The study commenced after gaining approval of the Institu-tional Review Board of the National Cheng Kung UniversityHospital (NCKUH, IRB number: ER-102-034). Written informedconsent was obtained from every interviewed cancer patient. Atotal of 464,722 patients with pathologically diagnosed canceraged 20–79 years at diagnosis were abstracted and verified fromthe Taiwan Cancer Registry for 1998–2009, which contains data onthe patients’ demographics, date of diagnosis, cancer site, histologyand vital status, and followed up until 2011 [19]. The identificationnumbers of all individuals were encrypted to protect their privacy.The nine major cancers included lung (ICD-9-CM code: 162),esophagus (ICD-9-CM code: 150), liver (ICD-9-CM code: 155),stomach (ICD-9-CM code: 151), colorectum (ICD-9-CM code: 153–154), oral (ICD-9-CM code: 140–141), nasopharynx (ICD-9-CMcode: 147), cervix (ICD-9-CM code: 180), and breast (ICD-9-CMcode: 174).

2.2. Survival analysis and extrapolation to estimate life expectancy for

different cancers

All of the above patients were linked to the Taiwan MortalityRegistry during 1998–2011 to obtain their survival functions viathe Kaplan–Meier estimation method up to the end of the follow-up period. These were further extrapolated to lifetime based on asemi-parametric method using the age- and sex-matched referentssimulated from the life tables of the Taiwan Vital Statistics, whichonly requires an assumption of constant excess hazards after theend of follow-up [20,21]. The computation of these estimates werecarried out using iSQoL software [22]. Detailed methodsand mathematical proofs are described in our previous studies[1–4,20,21].

2.3. Measurements of quality of life (QOL) via EQ-5D for estimation of

quality-adjusted life expectancies (QALE) and loss-of-QALE

To estimate the utility value of QOL for cancer patients atdifferent duration-to-dates (namely, from diagnosis of disease upto the date of interview), we recruited a convenient, cross-sectionalsample with 11,453 measurements aged 20–79 years at diagnosisbetween 2011 and 2014 at the Oncology Center of NCKUH. Patientswere not included if they had disturbances in consciousness,moderate to severe dementia or aphasia, or an inability tocommunicate with research assistants. We assessed their QOLby administering the Taiwanese version of EQ-5D [23] throughdirect face-to-face interviews. EQ-5D is a preference-based,generic instrument [24,25] that provides a utility value basedon survey carried out in Taiwan [26]. A kernel-type smoothingmethod (using a moving average of the nearby 10%) was performedto estimate the mean QOL across time [27,28]. The lifetime survivalprobabilities over the course of a cancer were multiplied (oradjusted) with the QOL utility values to obtain a quality-adjustedsurvival curve. The loss-of-QALE for these patients was calculatedby subtracting the QALE of cancer patients from the QALE of theage- and gender- matched referents simulated from the hazardfunctions of vital statistics with the EQ-5D utility values borrowedfrom those of the 2009 National Health Interview Survey inTaiwan, as illustrated in Fig. 1 with colorectal cancer as anexample.

2.4. Estimation of cumulative incidence rates (CIR) of major cancers

We used data from the Catastrophic Illnesses Registry databaseof Taiwan for the period 2008–2010 [29] to calculate the incidencerates of major cancers. The CIR was calculated from age 20 to 79(CIR20–79) to estimate the lifetime risk of a specific cancer, asfollows

CIR ¼ 1 � exp �X

iðIRiÞðDtiÞ

h i

where IRi represents the age-specific incidence rate and Dti

indicates the range of each age stratum.

2.5. Expected health impacts for cancer

The health impact of cancer at each organ-system wasestimated by multiplying the CIR with loss-of-QALE of nine majorcancers. We first estimated the lifetime risks by CIR20–79, or agefrom 20 to 79, followed by CIR20–44, and CIR45–64, and CIR65–79, and

Table 1Comparison of frequency distributions of cancer patients registered in the Cancer

Registry of Taiwan and a cross-sectional sample at oncology clinics.

Cancer

patients (1998–2009)

Cross-sectional

sample (2011–2014)

Liver, no. case 91,747 1612

Sex (% male) 67,461 (73.5) 1182 (73.3)

Mean age (years, SD) 60.8 (11.8) 58.7 (10.4)

Lung, no. case 74,720 3073

Sex (% male)* 49,598 (66.4) 1567 (51)

Mean age (years, SD)* 64.4 (11.0) 60.9 (10.3)

Oral, no. case 38,662 1559

Sex (% male) 35,222 (91.1) 1445 (92.7)

Mean age (years, SD) 52 (11.4) 52 (9.2)

Colorectum, No. case 89,276 2763

Sex (% male) 51,190 (57.3) 1552 (56.2)

Mean age (years, SD) 62 (12.2) 60 (11.3)

Esophagus, no. case 15,924 284

Sex (% male) 14,931 (93.7) 266 (93.7)

Mean age (years, SD) 58 (11.2) 53.4 (9.1)

Stomach, no. case 35,720 336

Sex (% male) 22,978 (64.3) 198 (58.9)

Mean age (years, SD) 62.8 (12.5) 58.3 (12.3)

Nasopharynx, no. case 17,118 284

Sex (% male) 12,750 (74.5) 214 (75.3)

Mean age (years, SD) 49.6 (12.4) 48 (10.9)

Cervix, no. case 24,525 689

Mean age (years, SD) 53.9 (12.8) 50.1 (10)

Female breast, no. case 77,030 991

Mean age (years, SD) 50.9 (11.2) 48.7 (9.5)

* p < 0.05.

M.-C. Hung et al. / Cancer Epidemiology 39 (2015) 126–132128

the results were then multiplied by the corresponding expectedloss-of-QALE within different age ranges.

2.6. Validation of the extrapolation method and sensitivity analysis

for loss-of-QALE

To show that our extrapolation method is valid, we selectedsub-cohorts of cancer patients between 1998 and 2004,extrapolated these results to the end of 2011, and comparedthe extrapolated results with the Kaplan–Meier (K–M) estimatesof the actual 14 years of follow-up. Assuming that the K–M

Table 2Comparison of expected lifetime impacts and standardized mortality rates adjusted by

expected lifetime risk (CIR20–79)a, loss-of-QALE (quality-adjusted life expectancy) unde

Sites of

cancer

Gender 2008–2010

(CIR20–79)a (%)

Loss-of-QALE Loss-of-QALE

under lifetime risk

Liver Male 6.80 14.22 (0.03) 0.97

Female 3.12 11.44 (0.07) 0.36

Lung Male 6.05 11.66 (0.03) 0.71

Female 3.15 13.11 (0.05) 0.41

Oral Male 2.99 13.54 (0.12) 0.40

Female 0.38 5.07 (0.51) 0.02

Colorectum Male 6.30 5.19 (0.08) 0.33

Female 4.36 5 (0.19) 0.22

Esophagus Male 1.55 16.69 (0.06) 0.26

Female 0.12 11.41 (0.78) 0.01

Stomach Male 2.08 8.26 (0.16) 0.17

Female 1.13 8.14 (0.19) 0.09

Nasopharynx Male 0.88 13.75 (0.27) 0.12

Female 0.29 11.54 (1.67) 0.03

Female breast Female 6.52 5.22 (0.16) 0.34

Cervix Female 1.37 3.17 (0.26) 0.04

a CIR20–79, cumulative incidence rates from age 20 to 79.b Cited from: https://cris.hpa.gov.tw/pagepub/Home.aspx?itemNo=cr.q.50.

estimates are the gold standard, this study calculated the relativebiases for sub-cohorts with different cancers. The relative bias(RB) is defined as follows: RB = (estimate from extrapolation � K–M estimate)/K–M estimate. In addition, we applied a sensitivityanalysis by assuming a uniform utility value of 1 for referents totest if the final conclusions of decision making are affected.

2.7. Other statistical analysis

The differences in the frequency distributions of cancer patientsbetween the cohort and cross-sectional samples at clinics weretested with the chi-squared and two-sample t test, withp < 0.05 regarded as significant. The analysis was carried outusing SAS software (ver. 9.4). The standardized mortality rateswere calculated by adjusting the 2000 world standard populationfor nine major cancers for different age periods with the agespecific cancer rates of Taiwan for the period 2008–2010, stratifiedby gender [30].

3. Results

The basic demographic details of the cancer patients arecategorized by cancer sites in Table 1, which shows that the sexdistributions and mean ages at diagnosis of the cross-sectionalsample interviewed at the Oncology Center were similar to those inthe national data, except for lung cancer.

The average estimated loss-of-QALE were the highest foresophageal cancer in males (16.69 QALYs) and lung cancer infemales (13.11 QALYs) followed by other cancers, while those ofmale colorectal cancer and female cervical cancer had the lowestestimates, at 5.19 and 3.17 QALYs, respectively (Table 2). Aftermultiplication with the incidence rates in Taiwan for the specificcancer sites during 2008–2010, we found the highest expectedhealth impacts in males and females were liver and lung cancer,with expected losses of 0.97 and 0.41 QALYs, respectively. Assummarized in Fig. 2, the priorities for prevention based onexpected health impacts are different from those based onstandardized mortality rates. While the priority changes in femalesmay be less affected except breast and colorectal cancers, those ofmales involve a broader spectrum, including oral, colorectal,

2000 world standard population for major cancers in Taiwan stratified by gender:

r unit of QALY (quality-adjusted life year).

Loss-of-QALE

(with referents’

utility as 1)

Loss-of-QALE under

lifetime risk

(with referents’ utility as 1)

2008–2010 standardized

mortality rates20–79b

17.2 (0.04) 1.17 37.53 � 10�5

16.01 (0.07) 0.50 14.33 � 10�5

14.29 (0.03) 0.86 35.22 � 10�5

18.14 (0.04) 0.57 16.49 � 10�5

16.68 (0.1) 0.50 14.56 � 10�5

10.68 (0.53) 0.04 0.96 � 10�5

8.05 (0.08) 0.51 17.04 � 10�5

10.1 (0.16) 0.44 11.84 � 10�5

19.64 (0.07) 0.30 9.35 � 10�5

16.39 (0.66) 0.02 –

11.05 (0.15) 0.23 9.49 � 10�5

13.33 (0.19) 0.15 4.98 � 10�5

17.1 (0.31) 0.15 3.73 � 10�5

17.28 (1.49) 0.05 –

10.8 (0.16) 0.70 10.61 � 10�5

8.76 (0.24) 0.12 4.39 � 10�5

Fig. 2. Comparison of priority ratings for cancer prevention policy decisions between lifetime expected loss-of-QALE (quality-adjusted life expectancy) using the QALY

(quality-adjusted life years) as the common unit and standardized mortality rate adjusted for the 2000 world standard population stratified by gender.

M.-C. Hung et al. / Cancer Epidemiology 39 (2015) 126–132 129

esophageal and stomach cancers (see also Table 2). The abovetrends remained unchanged for the age groups of 45–64 afterstratification. Lung cancer became the leading expected loss of0.272 QALYs in the age group of 65–79 (Table 3).

The results of the validation of the extrapolation method forestimates of survival for each cancer subcohort showed that all therelative biases were less than 6.82%, despite the high censoringrates (more than 80% for breast cancer), as summarized in theAppendix 1. We expect that the possible bias would be even lessafter 14 years of follow-up, because the life expectancy for eachcancer was usually below 15 years, except breast and cervicalcancer. If we assume a uniform utility value of one for referents inthe sensitivity analysis of loss-of-QALE, the priority setting forprevention based on expected health impacts would be quitedifferent, with breast cancer becoming the top priority for cancercontrol in females (Table 2).

4. Discussion

This study is the first that integrates incidence rates with loss-of-QALE to estimate the health impacts and cancer risks to aidcancer prevention policy decisions using QALY (quality-adjustedlife years) as the common unit. The results show that liver and lungcancer had the highest expected health impacts in males andfemales, or expected losses of 0.97 and 0.41 QALY, respectively. Thepriorities for prevention based on expected health impacts aredifferent from those based on standardized mortality rates, withoral, colorectal, esophageal and stomach cancer becoming the thirdto sixth priority in cancer control for males (Table 2 and Fig. 2).However, we must carefully examine the accuracy of our estimatesbefore making further inferences. First, since we only includedpatients with the nine cancers verified with histopathologicalevidence (except liver cancer) [31], the diagnoses were all valid.

Table 3Estimation of expected risk, loss-of-QALE (quality-adjusted life expectancy) for an average case and expected loss-of-QALE for different major cancers stratified by age

periods.

Sites of cancer 2008–2010

(CIR20–44)

Loss-of-QALE

(20–44 years)

Expected loss-of-

QALE for

CIR20–44

2008–2010

(CIR45–64)

Loss-

of-QALE

(45–64 years)

Expected loss-

of-QALE for

CIR45–64

2008–2010

(CIR65–79)

Loss-of-

QALE

(65–79 years)

Expected loss-

of-QALE for

CIR65–79

Liver 0.0020 28.31 (0.14) 0.057 0.0142 17.21 (0.05) 0.244 0.0322 8.15 (0.04) 0.262

Lung 0.0011 29.53 (0.17) 0.032 0.0104 17.85 (0.05) 0.186 0.0326 8.34 (0.03) 0.272

Oral 0.0027 21.42 (0.29) 0.058 0.0086 12.91 (0.12) 0.111 0.0057 5.04 (0.14) 0.029

Colorectum 0.0021 16.03 (0.33) 0.034 0.0143 7.33 (0.12) 0.0143 0.0355 3.27 (0.07) 0.116

Esophagus 0.0006 29.3 (0.21) 0.018 0.0039 19.33 (0.09) 0.075 0.0037 8.34 (0.07) 0.031

Stomach 0.0006 21.59 (0.54) 0.013 0.0046 11.42 (0.12) 0.053 0.0147 5.69 (0.2) 0.084

Nasopharynx 0.0014 18.08 (0.87) 0.025 0.0027 12.62 (0.44) 0.034 0.0017 6.42 (0.38) 0.011

Female breast 0.0118 13.49 (0.44) 0.159 0.0352 4.98 (0.19) 0.175 0.0211 0.92 (0.44) 0.019

Cervix 0.0019 5.22 (0.23) 0.010 0.0059 5.82 (0.32) 0.034 0.0060 2.42 (0.57) 0.015

CIR, cumulative incidence rates.

M.-C. Hung et al. / Cancer Epidemiology 39 (2015) 126–132130

Second, all the extrapolations of survival functions are based on thevalidated assumption of ‘‘constant excess hazard’’ after the end offollow-up, which can be obtained by showing a straight line aftertaking the logit transform of the survival ratio between the indexand age- and gender-matched referents [20,21]. In addition, sinceour validation study of this method showed that the relative biaseswere all below 6.8% after seven years of follow-up, it is reasonableto assume that those of the estimation of life expectancies based on14 years of follow-up would be even smaller. Third, we collectedthe QOL measure using the EQ-5D from all patients attending theoncology clinic of NCKUH, and obtained their utility values atdifferent time points after diagnosis to integrate with the survivalfunction at time t. In total we excluded approximately 1% of theinvited patients who were unable to communicate with researchassistants, plus 9% who refused to be interviewed, and the genderand age distributions at diagnosis seemed similar to the nationaldata (Table 1). Therefore, our convenience samples might not bebiased too much, or at least are more accurate than simply makinginferences based on the mean values. Finally, CIR is an indicatorthat is inherently standardized for the specified age period, andthere is no need to invoke any outside weights for comparison. Inaddition, everyone’s QALY would be the same across differentillnesses and technologies, and could also be directly comparedwith clinical care services [32]. The combination of these twoindicators would thus be more efficient in quantifying the loss ofhealth after the occurrence of a specific illness. We thus tentativelyconclude that our results are relatively accurate and useful forassisting policy decisions related to cancer prevention.

Based on the above analysis, oral and esophageal cancer becamethe third and fifth priority in cancer control for males followed bycolorectal and stomach cancer, especially for the age group of 45–64, owing to its high loss-of-QALE; breast cancer became the thirdpriority for females followed by colorectal cancer, because of itshigh CIR and loss-of-QALE. Had we assumed a uniform utility valueof 1 for the referents simulated from the general population, thepriority for prevention based on expected health impacts wouldhave been quite different, with breast cancer becoming the toppriority for cancer control in females because of their longsurvivorship (Table 2). Because cancer patients may die of othercauses, the health effects attributed to cancer would be under-estimated based on standardized mortality rate. In the U.S., anestimated 64% of cancer survivors live for more than five yearsafter their diagnosis [33], and this usually requires continualmonitoring with regard to recurrence and long-term care for thelate effects of cancer [11,33], issues that must be addressed whendeveloping of cancer control policies. Therefore, counting thepotential savings with regard to loss-of-QALE, costs of lifetimehealthcare and long term care, and productivity loss for society[34] when estimating the total health impact would be morecomprehensive and efficient than considering the standardized

mortality rate or CIR alone, and it would be better if futuredecisions cancer prevention policies would take all these potentialimpacts into consideration.

5. Limitations

Our study has following limitations: first, the lifetimeextrapolation is based on current and prior experiences, especiallythe national life tables. However, such an ex post approach couldeasily underestimate the actual survival of future cancer popula-tions, because of the development and adoption of newertechnologies for cancer care. Therefore, our estimation of thelifetime survival of cancer patients may be a conservative one,while the loss-of-QALE and lifetime health impacts might beoverestimated. Second, during the lifetime extrapolation of theQOL function, it was assumed that the patients continued to stay atthe same level of QOL as near the end of follow-up. Thisassumption could have resulted in an overestimation, becausethe real QOL might gradually decline as the patient ages [35]. Third,as most of interviewed subjects were recruited from clinics, theirgeneral conditions were probably better than those confined athome or institutions, which might have led to over-estimating theQOL and under-estimating the prevention effects.

6. Conclusion

The integration of incidence rates with loss-of-QALE could beused to represent the expected losses that could be averted byprevention. This estimate can become a summary index of costeffectiveness for cancer prevention, and may be useful inprioritizing strategies for cancer control after considering otherrelated parameters.

Competing interests

The authors declare that they have no competing interests.

Authorship contribution

Conceived and designed the experiments: J.-D.W., M.-C.H.,W.W.L., H.H.W.C. and W.C.S. Performed the data analysis: M.-C.H.Wrote the paper: M.-C.H., J.-D.W. All authors read and approvedthe final manuscript.

Acknowledgements

This work was supported by grants from the National ScienceCouncil of Taiwan (Grant Nos. NSC 102-2314-B-006-029-MY2, NSC102-2811-B-006-018 and NSC 101-314-Y-006-001), Ministry of

M.-C. Hung et al. / Cancer Epidemiology 39 (2015) 126–132 131

Health and Welfare of Taiwan (DOH 102-TD-C-111-003) and theTop University Project to the National Cheng Kung University fromthe Ministry of Education of Taiwan. The funding bodies had norole in the study design, data collection and analysis, decision to

Appendix 1

Cancer Gender Cohort size Age at diagnosis (SD) Censored rate

Lung Male 30,231 67.92 (11.39) 23.03

Female 13,756 64.59 (13.1) 29.68

Esophagus Male 7515 60.29 (12.6) 26.88

Female 624 68.36 (13.24) 33.33

Liver Male 38,770 60.27 (13.39) 30.15

Female 14,275 65.59 (12.63) 33.22

Stomach Male 16,160 67.23 (13.17) 42.57

Female 8418 63.58 (15.2) 47.96

Colorectum Male 29,017 64.96 (13.29) 63.14

Female 21,856 64.03 (14.18) 65.67

Oral Male 17,310 51.57 (11.99) 63.26

Female 1839 58.74 (15.63) 68.24

Nasopharynx Male 7329 49.63 (13.21) 69.04

Female 2553 48.77 (13.72) 75.52

Breast Female 3,6874 50.63 (12.16) 87.61

Cervix Female 17,269 55.26 (14.28) 78.82

SD, standard deviation; SE, standard error of mean.a Relative bias = (estimate from extrapolation � K–M estimate)/K–M estimate.

publish, or preparation of the manuscript. We are grateful to all thepatients who participated in this study, and to Drs. Jenq-Chang Lee,Yih-Jyh Lin, Jenn-Ren Hsiao, Ya-Min Cheng and Yan-Shen Shan forhelping us in the collection of QOL data from their clinics.

Estimates of mean survival after 14 years of follow-up using the semi-parametric method of extrapolation based on the first seven yearsof follow-up data with a high censored rate and compared with Kaplan–Meier (K–M) estimates of 14 years of follow-up for patients withmajor cancers

(%) 14-year survival

based on K–M estimate

(SE) (month)

Extrapolation from

first 7 years of follow

up (SE) (month)

Relative bias (%)a

1.98 (0.02) 2.08 (0.01) 4.83

2.57 (0.04) 2.52 (0.02) �2.07

2.25 (0.04) 2.4 (0.02) 6.82

3.53 (0.21) 3.61 (0.08) 2.19

2.71 (0.02) 2.65 (0.02) �2.12

2.92 (0.03) 2.83 (0.02) �3.23

4.53 (0.05) 4.56 (0.03) 0.74

5.43 (0.06) 5.41 (0.03) �0.43

6.83 (0.03) 6.82 (0.02) �0.18

7.44 (0.04) 7.32 (0.02) �1.56

6.95 (0.05) 7.01 (0.02) 0.86

8.02 (0.14) 8.17 (0.07) 1.90

7.85 (0.07) 7.71 (0.04) �1.71

9.27 (0.11) 8.88 (0.05) �4.14

11.03 (0.03) 10.54 (0.01) �4.39

9.96 (0.04) 10.01 (0.02) 0.51

References

[1] Liu PH, Wang JD, Keating NL. Expected years of life lost for six potentiallypreventable cancers in the United States. Prev Med 2013. http://dx.doi.org/10.1016/j.ypmed.

[2] Chu PC, Wang JD, Hwang JS, Chang YY. Estimation of life expectancy and theexpected years of life lost in patients with major cancers: extrapolation ofsurvival curves under high-censored rates. Value Health 2008;11:1102–9.

[3] Hung MC, Liu MT, Cheng YM, Wang JD. Estimation of savings of life-years andcost from early detection of cervical cancer: a follow-up study using nationwidedatabases for the period 2002–2009. BMC Cancer 2014;14:505.

[4] Yang SC, Lai WW, Su WC, Wu SY, Chen Helen HW, Wu YL, et al. Estimating thelifelong health impact and financial burdens of different types of lung cancer.BMC Cancer 2013;13:579.

[5] Arrington AK, Goldstein L, Kruper L, Vito C, Yim J, Chen SL. Life expectancy aftercurative-intent treatment of breast cancer: impact on long-term follow-upcare. Am Surg 2014;80(6):604–9.

[6] Gross CP, McAvay GJ, Krumholz HM, Paltiel AD, Bhasin D, Tinetti ME. The effectof age and chronic illness on life expectancy after a diagnosis of colorectalcancer: implications for screening. Ann Intern Med 2006;145(9):646–53.

[7] Black N. Can England’s NHS survive? N Engl J Med 2013;369:1.[8] Howlander N, Noone AM, Krapcho M, Garshell J, Miller D, Altekruse SF, et al.,

eds. SEER cancer statistics review, 1975–2011. Bethesda, MD: National CancerInstitute, 2014. Available at http://seer.cancer.gov/statfacts/html/all.html.

[9] De Moor JS, Mariotto AB, Parry C, Alfano CM, Padgett L, Kent EE, et al. Cancersurvivors in the United States: prevalence across the survivorship trajectoryand implications for care. Cancer Epidemiol Biomark Prev 2013;22:561–70.

[10] Institute of Medicine Delivering Affordable Cancer Care in the 21st Century:Workshop Summary; 2014, Available at: http://www.nap.edu/catalog.php?record_id=18273 [cited 16.06.14.].

[11] Ekwueme DU, Yabroff KR, Guy Jr GP, Banegas MP, de Moor JS, Li C, et al.Division of Cancer Prevention and Control, National Center for Chronic DiseasePrevention and Health Promotion, CDC. MMWR Morb Mortal Wkly Rep2014;63(23):505–10.

[12] Weinstein MC, Torrance G, McGuire A. QALYs: the basics. Value Health2009;12(Suppl. 1):S5–9.

[13] Lipscomb J, Drummond M, Fryback D, Gold M, Revicki D. Retaining andenhancing the QALY. Value Health 2009;12(Suppl. 1):S18–26.

[14] Kind P, Lafata JE, Matuszewski K, Raisch D. The use of QALYs in clinical andpatient decision-making: issues and prospects. Value Health 2009;12(Suppl.1):S27–30.

[15] Burnet NG, Jefferies SJ, Benson RJ, Hunt DP, Treasure FP. Years of life lost (YLL)from cancer is an important measure of population burden and should beconsidered when allocating research funds. Br J Cancer 2005;92:241–5.

[16] Raiffa H. Decision analysis introductory lectures on choices under uncertainty.America: Addison-Wesley Publishing Company, 1968.

[17] Bridges CB, Thompson WW, Meltzer MI, Reeve GR, Talamonti WJ, Cox NJ, et al.Effectiveness and cost-benefit of influenza vaccination of healthy workingadults: a randomized controlled trial. JAMA 2000;284(13):1655–63.

[18] Herrero R, Quint W, Hildesheim A, Gonzalez P, Struijk L, Katki HA, et al.Reduced prevalence of oral human papillomavirus (HPV) 4 years after bivalentHPV vaccination in a randomized clinical trial in Costa Rica. PLoS ONE2013;8(7):e68329.

[19] Taiwan Cancer Registry Cancer Registry Dataset; 2014, Available at: http://tcr.cph.ntu.edu.tw/main.php?Page=N1 [accessed 11.07.14.].

[20] Hwang JS, Wang JD. Monte Carlo estimation of extrapolation of quality-adjusted survival for follow-up studies. Stat Med 1999;18:1627–40.

[21] Fang CT, Chang YY, Hsu HM, Twu SJ, Chen KT, Lin CC, et al. Life expectancy ofpatients with newly diagnosed HIV infection in the era of highly activeantiretroviral therapy. QJM 2007;100:97–105.

[22] iSQoL Package; 2014, Available at: http://www.stat.sinica.edu.tw/isqol/[accessed 11.07.14.].

[23] Chang TJ, Tarn YH, Hsieh CL, Liou WS, Shaw JW, Chiou XG. Taiwanese version ofthe EQ-5D: validation in a representative sample of the Taiwanese population.J Formos Med Assoc 2007;106(12):1023–31.

[24] EuroQol Group. EuroQol-a new facility for the measurement of health-relatedquality of life. Health Policy 1990;16(3):199–208.

[25] Szende A, Oppe M, Devlin N. EQ-5D value sets: inventory, comparative reviewand user guide. Dordrecht, the Netherlands: Springer, 2007.

[26] Lee HY, Hung MC, Hu FC, Chang YY, Hsieh CL, Wang J-D, et al. Estimatingquality weights for EQ-5D health states with the time trade-off method inTaiwan. J Formos Med Assoc 2013;112(11):699–702.

[27] Hwang JS, Tsauo JY, Wang JD. Estimation of expected quality adjusted survivalby cross sectional survey. Stat Med 1996;15(1):93–102.

[28] Hwang JS, Wang JD. Integrating health profile with survival for quality of lifeassessment. Qual Life Res 2004;13(1):1–10.

M.-C. Hung et al. / Cancer Epidemiology 39 (2015) 126–132132

[29] National Health Research Institutes in Taiwan: National Health InsuranceResearch database; 2003, Available at: http://www.nhri.org.tw/nhird/date_01.html#_edn1 [accessed 11.07.14.].

[30] Health Promotion Administration, Ministry of Health and Welfare of Taiwan.Standardized mortality rates of cancer; 2010, Available at: https://cris.hpa.gov.tw/pagepub/Home.aspx?itemNo=cr.q.50 [in Chinese].

[31] Bruix J, Sherman M. Management of hepatocellular carcinoma: an update.Hepatology 2011;53:1020–2.

[32] Weinstein MC. A QALY is a QALY is a QALY – or is it. J Health Econ 1988;7289–90.

[33] Centers for Disease Control and Prevention. Cancer Survivors — United States,2007. MMWR Morb Mortal Wkly Rep 2011;60:269–72.

[34] Neiger BL, Thackeray R, Fagen MC. Basic priority rating model 2.0: currentapplications for priority setting in health promotion practice. Health PromotPract 2011;12(4):630.

[35] Fryback DG, Dunham NC, Palta M. US norms for six generic health-relatedquality-of-life indexes from the National Health Measurement Study. MedCare 2007;45:1162–70.