http://orcid.org/0000-0002-9684-0789
Figures
Abstract
Evidence is limited regarding the effect of diagnosis-to-treatment interval (DTI) on the survival of colorectal cancer (CRC) patients. In addition, previous studies on treatment delay and CRC survival have largely grouped patients from all stages (I-IV) into one cohort. Our study provides analysis on each stage individually. We conducted a retrospective cohort study with 39,000 newly diagnosed CRC patients obtained from the Taiwan Cancer Registry Database from 2004–2010 to examine the effect of DTIs on overall survival. DTIs were divided into 3 groups: ≤ 30 days (36,115 patients, 90.5% of study patients), 31–150 days (2,533, 6.4%), and ≥ 151 days (1,252, 3.15%). Risk of death was increased for DTI 31–150 days (hazard ratio 1.51; 95% confidence interval 1.43–1.59) and DTI ≥ 151 days (1.64; 1.54–1.76) compared to DTI ≤ 30. This risk was consistent across all cancer stages. Additional factors that increased risk of death include male gender, age >75, Charlson Comorbidity Index ≥7, other catastrophic illnesses, lack of multidisciplinary team involvement, and treatment in a low volume center. From these results, we advise that the DTI for all CRC patients, regardless of cancer staging, should be 30 days or less.
Citation: Lee Y-H, Kung P-T, Wang Y-H, Kuo W-Y, Kao S-L, Tsai W-C (2019) Effect of length of time from diagnosis to treatment on colorectal cancer survival: A population-based study. PLoS ONE 14(1): e0210465. https://doi.org/10.1371/journal.pone.0210465
Editor: Surinder K. Batra, University of Nebraska Medical Center, UNITED STATES
Received: March 27, 2018; Accepted: December 25, 2018; Published: January 14, 2019
Copyright: © 2019 Lee et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Regarding the data availability, data were obtained from the National Health Insurance Research Database published by the Ministry of Health and Welfare, Taiwan. Due to legal restrictions imposed by the Taiwan government related to the Personal Information Protection Act, the database cannot be made publicly available. All researchers can apply for using the databases to conduct their studies. Requests for data can be sent as a formal proposal to the Health and Welfare Data Science Center of the Ministry of Health and Welfare (http://www.mohw.gov.tw/EN/Ministry/Index.aspx). Any raw data are not allowed to be brought out from the Health and Welfare Data Science Center. The restrictions prohibited the authors from making the minimal data set publicly available.
Funding: This study was supported by grants (MOST 104-2410-H-039 -002 to WCT; ASIA104-CMUH-17 to PTK) from the Ministry of Science and Technology (https://www.most.gov.tw/) and China Medical University (http://www.cmu.edu.tw/) and Asia University (http://www.asia.edu.tw/EN/), Taiwan. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: AJCC, American Joint Committee on Cancer; CCI, Charlson Comorbidity Index; CI, confidence interval; CRC, colorectal cancer; DTI, diagnosis-to-treatment interval; HR, Hazard ratio; MDT, multidisciplinary team; SES, socioeconomic status
Introduction
There is an estimated 50,630 deaths from colorectal cancer (CRC) in 2018 in the United States, second only to lung cancer [1]. The Taiwan Cancer Registry report showed that in 2014, CRC was the most frequent cancer in men and second most frequent in women [2]. According to the latest updated data from Taiwan Office of Statistics under the Ministry of Health and Welfare, CRC ranked third among men and women in terms of mortality rate in 2017. In 2007, 19.5 out of every 100,000 people died of CRC, which increased to 24.7 out of 100,000 in 2017. In addition, the mortality rate was higher in men (28.1 per 100,000) than women (21.2 per 100,000) [3].
Delays in diagnosis and treatment can happen any time from the onset of cancer symptoms to start of treatment. Many studies have explored whether this delay (diagnosis-to-treatment interval, DTI) would significantly affect the stage of CRC at diagnosis, prognosis, and overall survival. However, DTIs were often inconsistently defined and measured across studies, leading to inconclusive results. For example, Facione took the time from onset of cancer symptoms to start of treatment and divided it into two intervals: patient delay and provider delay. Patient delay was defined as the period from onset of symptoms to the first medical visit, while provider delay was defined as the period from first visit to start of treatment [4]. Dwivedi et al. divided the same period into three intervals: primary delay, the time from onset of symptoms to the first medical visit; secondary delay, the time from first visit to a confirmed diagnosis; and tertiary delay, the time from a confirmed diagnosis to start of treatment [5].
Treatment delays are postulated to be more likely in countries such as the United States and the United Kingdom due to the wait times inherent to their referral systems [6–8]. Taiwan does not have a mandatory referral system, and once patients apply for a Catastrophic Illness Card, any cancer-related treatment costs are waived. This removes financial barriers and improves medical accessibility. Thus, treatment delays secondary to long wait times or financial difficulties are relatively uncommon. Despite this, there are still patients in Taiwan who refuse to accept treatment, or fail to seek medical attention in a timely manner [9, 10]. No conclusions have been reached regarding the effect of DTI on survival of CRC patients in existing literature [11–17], and no studies examine the survival impact stratified by cancer staging. Thus, the purpose of this study is to determine the effect of DTI on the overall risk of death in patients with CRC and the risk of death for each stage of CRC.
Materials and methods
Study design
We designed this study as a retrospective cohort study, utilizing patients newly diagnosed with CRC between 2004 and 2010 as our parent population; these patients were followed to the end of 2012. We obtained the study subjects from the Taiwan Cancer Registry Database, and the contributing variables from the Taiwan’s National Health Insurance Research Database and the administrative data set from the Taiwan Death Registry between 2002 and 2012. The Statistics Center of Ministry of Health and Welfare, Taiwan, provided us with the data sets. Before we used all data, all personal identification information had been deleted by the Ministry of Health and Welfare. All data had been fully anonymized before we accessed them and personal privacy was under protection from using these data. The study was approved by the Institutional Review Board of Cheng Ching Hospital. The IRB number is HP150003.
Patient selection
We selected patients with Stage I to IV CRC whose ICD-O-3 was C180—C218 as our study subjects. Exclusion criteria were as follows: unknown stage, carcinoma in situ, multiple cancers, having received only palliative treatment for the past year, and missing personal information (such as monthly salary or area of residence) and unknown healthcare provider. The Taiwan Death Registry was used to determine whether a study subject was dead or alive.
Study variables
The main independent variable in this study is the diagnosis-to-treatment interval (DTI). Control variables include patient profile (gender and age), economic status (monthly income), environmental factor (degree of urbanization), patient’s health condition (catastrophic illnesses except for cancer, Charlson Comorbidity Index [CCI], cancer staging), hospital attributes (hospital level and ownership, service volume), and involvement of a multidisciplinary team (MDT). The dependent variable is overall survival.
We defined DTI as the time interval from a confirmed, pathologic diagnosis of CRC from biopsy to start of treatment (surgery, chemotherapy, or radiotherapy). In this study, we divided DTIs into three categories: ≤ 30 days, 31–150 days, and ≥ 151 days.
Monthly salary was used as a marker of a patient’s economic status. We divided income into 7 levels: low-income households, <NT$17,800, NT$17,281–22,800, NT$22,801–28,000, NT$28,801–36,300, NT$36,301–45,800, and >NT$45,801. For environmental factors, we assigned each patient’s place of residence a score indicating degree of urbanization, with Level 1 as the highest degree of urbanization and Level 7 the lowest [18]. Our assessment of each patient’s health status included the presence of catastrophic illnesses except for cancer, CCI, and cancer staging. We defined ‘catastrophic illness’ according to the National Health Insurance Administration definition, which includes 30 diseases such as malignant neoplasm, hemophilia, hemolytic anemias, chronic renal failure, stroke, insulin dependent diabetes mellitus, systemic sclerosis, and systemic lupus erythematosus. For patients with a qualifying health condition, expenses related to medical treatment are waived. In addition, we measured a patient’s degree of burden from comorbidities with the Deyo-modified Charlson Comorbidity Index. The CCI was shown to be significantly correlated with factors such as postoperative complications, death rate, admission days and cost of hospitalization and is divided into three levels: ≤ 3, 4–6, and ≥ 7 [19]. For cancer staging, we used the American Joint Committee on Cancer (AJCC) sixth edition staging guidelines for CRC [20].
We defined the primary healthcare facility as a patient’s main treatment center, and divided them into four tiers: major medical center, regional hospital, local hospital, and other healthcare facilities. We also determined whether the hospital ownership was public or non-public, and evaluated their annual service volume according to quartiles: low, <25%; mid, 25–75%; and high, >75%. With regards to the MDT, we defined this as a multidisciplinary group including surgeons, medical oncologists, radiation oncologists, and members of allied health such as psychologists, physical therapists, dieticians, pharmacists, cancer nurses, cancer case managers, hospices, and discharge planning staff members who all work together to provide comprehensive cancer care [21].
Statistical analysis
In this study, we utilized descriptive statistics to analyze different DTIs and the distribution of control variables (including patient profile, economic status, environmental factor, health condition of the patient, hospital attributes, and MDT involvement). We used the log-rank test for inferential statistical analysis to test for statistically significant differences (P<0.05) between different DTIs and the respective control variables on the survival of CRC patients.
The Cox proportional hazards model was used to analyze the effect of different DTIs on the risk of death of patients with CRC as a whole, and that of patients with CRC in respective stages after controlling for related variables (including patient profile, economic status, environmental factor, health conditions, hospital attributes, and MDT involvement). We used months as the unit of measurement for survival. When an observed patient died before the end of 2012, we defined this as an event; if they were alive by the end of 2012 or had withdrawn from National Health Insurance coverage without subsequent data available for observation, we defined this as censored.
We constructed the survival curve of different DTIs versus patients with CRC as a whole and patients with CRC in different stages after controlling for confounding variables as listed above.
Finally, we designated our study population for each CRC stage as the derivation cohort, and randomly selected 10% from each stage as our validation cohort. Bivariate correlation analyses were performed between the derivation cohort and the validation cohort for each cancer stage. We draw the receiver operating characteristic (ROC) curves and calculated the area under curves (AUCs) for the validation cohort by different cancer stage.
In this study, all statistical analyses were performed using the SAS software, version 9.3 (SAS Institute Inc., Cary, NC). All tests were two-tailed, with α = 0.05 as the confidence interval.
Results
Enrollment and patient characteristics
In the Taiwan Cancer Registry Database, there were a total of 64,241 patients with newly confirmed diagnosis of CRC between 2004 and 2010. After excluding those with unknown disease stage (18,612), carcinoma in situ (3,070), multiple cancers (973), palliative treatment in the past year (360), and missing personal information or unknown healthcare provider (1,326), a total of 39,900 patients were obtained as the study population, with an average age of 65.1±13.4 years. These patients were followed until 2012; a total of 22,614 patients (56.7%) survived and 17,286 (43.3%) died (Table 1).
The DTIs for CRC in this study were divided into three groups: ≤ 30 days, 31–150 days, and ≥ 151 days. 36,115 study subjects (90.5%) received treatment within 30 days of a confirmed diagnosis; 2533 (6.4%) did so between 31–150 days; and 1,252 (3.1%) did not receive treatment until more than 151 days after diagnosis. In addition, patient survival was significantly impacted (p < 0.05) by variables such as gender, age, monthly income, degree of urbanization in area of residence, Charlson Comorbidity Index (CCI), presence of catastrophic illnesses except for cancer, cancer staging, involvement of a multidisciplinary team (MDT), and hospital factors (hospital level and ownership, responsibilities, and service volume).
Factors associated with overall survival in colorectal cancer patients.
Table 2 shows the risk of death in terms of hazard ratios (HR) with different DTIs as determined by the Cox proportional hazards model, controlling for all other variables. Our results show that patients with longer DTIs had greater risk of death. With patients receiving treatment within 30 days of a confirmed diagnosis as the reference group, the risk of death associated with starting treatment between 31–150 days and after 151 days of a confirmed CRC diagnosis were 1.51 times (95% confidence interval: 1.43–1.59) and 1.64 times (95% CI: 1.54–1.76) greater, respectively. Examining other variables, the risk of death was 1.1 times (95% CI: 1.08–1.15) greater in men than in women. The risk of death among patients aged 45–64 was lower than those younger than 44 (HR = 0.90). For patients older than 75, however, the risk of death was 1.77 times (95% CI: 1.66–1.89) greater than those aged 44 or below. In addition, a higher salary was associated with lower risk of death; for example, patients with monthly income greater than NTD 45801 had a risk of death only 0.64 times that of patients with lower monthly income. As well, the higher the CCI score, the higher the risk of death; a CCI score of ≥7 was associated with a risk of death 1.68 times (95% CI: 1.62–1.74) higher than a CCI score of ≤3. Similarly, presence of catastrophic illnesses except for cancer was related to a greater risk of death (HR 1.7, 95% CI: 1.59–1.83). As expected, a higher cancer stage was associated with greater risk of death. The risk of death from Stage IV disease was 8.86 times (95% CI: 8.32–9.45) that of Stage I disease. Lastly, lower risk of death was seen with involvement of an MDT (HR: 0.94; 95% CI: 0.89–0.99) and high hospital service volume (HR: 071; 95% CI: 0.67–0.74).
The Fig 1 shows the risk of death with different DTIs as determined by the Cox proportional hazards model, controlling for all other variables. Our results show that patients with longer DTIs had greater risk of death. Patients receiving treatment within 30 days of a confirmed diagnosis had the highest survival rate.
Overall survival of different DTIs based on cancer stage.
After controlling for each variable, we found that longer DTIs increased the risk of death in Stages I, II, and III through regression analysis (Fig 2).
In patients with Stage I disease, with DTI ≤ 30 days as the reference group, the risk of death was 1.41 times greater at DTI 31–150 days (95% CI: 1.25–1.47), and 2.66 times greater at DTI ≥ 151 days (95% CI: 2.09–3.40). In patients with Stage IV disease, the risk of death at DTI 31–150 days was the same as DTI ≥ 151 days (1.37, 95% CI: 1.28–1.47, and 1.36, 95% CI: 1.25–1.47, respectively) (Table 3).
Analysis of the validation cohort based on cancer stage.
We performed bivariate correlation analyses between the derivation cohort and the validation cohort for each cancer stage. Both cohorts were found to be similar in distribution across all cancer stages (p value > 0.05) (S1 and S2 Tables). We constructed receiver operating characteristic (ROC) curves and calculated the area under curves (AUCs) for the validation cohort for each tumor stages. The AUC for Stage I to IV were 0.80, 0.74, 0.71, and 0.72, respectively (Fig 3), which were over 0.70.
We also used the validation cohort to plot the survival curves of different diagnosis-to-treatment intervals by cancer stages (Fig 4). We found that longer DTIs increased the risk of death in the validation cohort (Fig 4), which was similar to those in the derivation cohort (Fig 2).
Discussion
There have been multiple studies exploring the impact of delay in treatment on risk of death in CRC; however, no conclusions have been definitively reached. In many publications, treatment delay has not been found to decrease survival of CRC patients [11, 14–17]. However, those studies did not include a sufficiently large sample size, and currently there are no studies that investigate the effect of DTIs on CRC survival for each stage of disease. In this study, we accessed the nationwide database to obtain a large enough sample size to conduct robust statistical analysis. A total of 39,900 patients with CRC were analyzed and results showed that longer DTI was associated with higher risk of death. The risk of death with treatment between 31–150 days and after 151 days were 1.51 times and 1.64 times higher, respectively, than those who received treatment within 30 days following confirmed diagnosis of cancer. Decreased survival with longer DTIs was again seen when stratified according to stage of disease, confirming that the risk of death was lowest for any patient if they receive treatment within 30 days of confirmed diagnosis. This finding is in keeping with the NHS Cancer Plan executed by the United Kingdom in 2000, which proposed the goal of reducing wait times from diagnosis to treatment for all cancers to one month or less by 2005 [22]. In addition, timely treatment provided to patients after a confirmed diagnosis helps to reduce their psychological stress and anxiety, and improve quality of life [23]. Therefore, we recommend that the DTI for all CRC patients, regardless of disease staging, should be 30 days.
Existing literature defines ‘treatment delay’ as DTI longer than 30 days [24, 25]. Irene et al. found that treatment delay occurred in 65.5% of CRC patients in Spain [24]. Abu-Helalah et al. found that treatment delay occurred in 32.6% of CRC patients in Jordan [25]. Studies in the USA have also shown that the average waiting time has grown longer and longer in recent years [26]. As shown in our study, 90.5% of Taiwanese CRC patients started treatment within 30 days of a confirmed diagnosis, which meant treatment delay only occurred in 9.5% of patients. Explanations for this include the existence of National Health Insurance in Taiwan, and the ability of cancer patients to apply for a Catastrophic Illness Card, which waives any medical expenses involved in the cancer treatment. In addition, no mandatory referral system exists in Taiwan, thereby minimizing delays; and tertiary care centers in Taiwan handle a high density of cancer cases, which translate to greatly improve medical accessibility for patients. As a result, patients in Taiwan are less likely to experience delays in their cancer treatment.
A large European study (EUROCARE) found that the risk of death was higher in men than women with CRC (HR = 1.05–1.08) [27]. In the US, population data from 2009–2013 showed that the rate of death from CRC was 1.4 times higher in men than women (95% CI: 1.42–1.44) [1]. The results obtained in our study are similar, with men having a greater risk of death (HR = 1.11), and match the findings in our previous study [21]. In terms of the effect of age on the risk of death, however, the results of this study differed from the EUROCARE study [27]. EUROCARE determined that as age increased, the risk of death increased as well (HR = 1.20–2.63). In our study, the lowest risk of death was not found within the youngest age group (age<45) but within those aged 45–64 (HR = 0.90). This finding is similar to that of the SEER (Surveillance, Epidemiology, and End Results) registry study between 2001 and 2010 in the US [28]. The SEER registry study showed that CRC ranked first in prevalence and death rate among all cancers in Americans younger than 50, and the rates are rising. High death rates in young patients with CRC are thought to be due to longer treatment delays and higher staging at time of diagnosis. Similar studies in Taiwan have also shown a higher stage and worse prognosis among young patients with CRC [21, 29, 30].
Studies have indicated that patients with higher socioeconomic status (SES) have better health awareness and are more involved in cancer screening. As a result, their stage of cancer at time of diagnosis are often less advanced [31]. Compared to those with high SES, patients with low SES are less likely to receive cancer screening and adjuvant treatments such as radiotherapy or chemotherapy, and have greater risk of death (HR = 1.3–1.8) [32]. In addition, they are at higher risk of death from any cause compared to those with high SES [33]. Our current study also showed similar findings; with low income earners as the reference group, the higher the monthly salary, the lower the risk of death. CRC patients with a monthly salary greater than NTD 45801 have a risk of death 0.64 times that of those with lower income. In addition, studies have found correlation between monthly salary and degree of co-morbidities. The risk that cancer patients with low SES have at least one co-morbidity is 50% higher than patients with high SES [34]. Co-morbidities affect the stage of cancer at diagnosis, the treatment options available, and the overall prognosis, leading to an increased risk of death [35, 36]. Results of this study showed that the risk of death among patients with higher CCI scores was also greater. A CCI score of ≥ 7 conferred a risk of death 1.68 times higher than a CCI score of ≤ 3.
Studies have shown that patients with CRC who receive treatment or surgery at a large volume center may have better overall survival [21, 37, 38]. Large volume centers are thought to be better trained and equipped to provide higher quality care, including better surgical resections, more comprehensive multidisciplinary team involvement, more frequent follow up and surveillance, and fewer treatment complications [37, 38]. A Taiwanese study compared outcomes between low and high service volume hospitals and found lower rates of death in high volume centers for patients with CRC (HR = 0.55) [21]. Similarly, our study showed the risk of death among patients treated in high service volume facilities was indeed significantly lower (HR = 0.67).
In 2003, the Taiwan Health Promotion Administration, Ministry of Health and Welfare introduced a plan called “Cancer Centers for a Great Improvement in the Quality of Cancer Care” which promoted the MDT framework for care of cancer patients. After patients with CRC join the MDT cancer care network, the medical team will consist of not only the attending physician but also radiologists, psychologists, physical therapists, dieticians, pharmacists, cancer nurses, cancer case managers, hospice care staff and discharge planning staff. The MDT provides integrated patient-centered care resulting in more comprehensive treatment plans and better quality of care. The benefit of MDT policies on the prognosis of patients with CRC was studied in literature and results conclusively show significantly higher survival in CRC patients treated with a MDT [39–41]. A Taiwanese study with 25,766 CRC patients found that the risk of death among CRC patients with MDT involvement was lower than patients without MDT care (HR = 0.91) [21]. Our study revealed similar findings, with MDT involvement decreasing the risk of death (HR = 0.94).
Finally, we found that the AUC for Stages I to IV were 0.80, 0.74, 0.71, and 0.72, respectively. An AUC above 0.8 indicates excellent predictive power, whereas AUCs between 0.7–0.8 are considered moderate and 0.6–0.7 are considered poor. Through the above validation cohort analyses, we conclude that our study results show acceptable discrimination, which further strengthens the findings and conclusion of our study.
Limitations of this study stem from the fact that data were obtained from a secondary database. Therefore, potentially confounding factors such as presenting symptoms and signs, family history, and lifestyle factors including diet and physical activity were unavailable, and could not be included in the study analysis. We selected 64,241 patients with Stage I to IV CRC as our study subjects, but 24,341 patients with unknown stage, carcinoma in situ, multiple cancers, having received only palliative treatment for the past year, and missing personal information and unknown healthcare provider were excluded. Although the dropout rate reaches 37.9%, the sufficiently large sample size and robust analysis lends confidence that the final results are not significantly affected.
In conclusion, this study found that a longer interval from a confirmed diagnosis of CRC to start of treatment was associated with a significantly higher risk of death. This result was consistent across all cancer stages. As our data show, the sooner patients with CRC receive treatment, the better the survival rate. Contributing factors include the patient’s age and gender, their economic status and overall health condition, as well as attributes of the treating hospital and involvement of a multidisciplinary team. It is advised through this study that the DTI for CRC patients should be within 30 days. In addition, as the factors underlying treatment delay were not included in this study, we believe future studies examining this area may be beneficial.
Supporting information
S1 Table. Bivariate correlation analysis between the derivation cohort and the validation cohort for cancer stage 1 & 2.
https://doi.org/10.1371/journal.pone.0210465.s001
(DOCX)
S2 Table. Bivariate correlation analysis between the derivation cohort and the validation cohort for cancer stage 3 & 4.
https://doi.org/10.1371/journal.pone.0210465.s002
(DOCX)
Acknowledgments
This study was partially based on data from the NHIRD provided by the Bureau of National Health Insurance, Department of Health, and managed by the National Health Research Institutes. Interpretation and conclusions reported herein do not represent the views of the Bureau of National Health Insurance.
References
- 1. Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2018. CA Cancer J Clin. 2018;68(1):7–30. pmid:29313949.
- 2.
Taiwan Health Promotion Administration. Taiwan Cancer Registry, Cancer Statistics: Incidence and mortality rates for the top 10 cancer in Taiwan, 2008–2014 http://tcr.cph.ntu.edu.tw/uploadimages/Top%2010%20cancer%20in%20Taiwan%202008-2014.pdf (accessed Nov 24, 2018).
- 3.
Ministry of Health and Welfare and Statistics Department: Taiwan’s Leading Causes of Death in 2017 https://www.mohw.gov.tw/cp-3961-42866-2.html (accessed Nov 24, 2018)
- 4. Facione NC. Delay versus help seeking for breast cancer symptoms: a critical review of the literature on patient and provider delay. Soc Sci Med. 1993;36(12):1521–34. pmid:8327915.
- 5. Alok Kumar Dwivedi SN D, Suryanarayana Deo, Rakesh Shukla, Arvind Pandey, Durgesh Kumar Dwivedi. An epidemiological study on delay in treatment initiation of cancer patients. Health 2012;4(2):66–79.
- 6. Afzelius P, Zedeler K, Sommer H, Mouridsen HT, Blichert-Toft M. Patient’s and doctor’s delay in primary breast cancer. Prognostic implications. Acta Oncol. 1994;33(4):345–51. pmid:8018364.
- 7. Fisher ER, Redmond C, Fisher B. A perspective concerning the relation of duration of symptoms to treatment failure in patients with breast cancer. Cancer. 1977;40(6):3160–7. pmid:589573.
- 8. Montella M, Crispo A, D’Aiuto G. Influence of delay to diagnosis on prognostic indicators of screen-detected breast carcinoma. Cancer. 2002;95(10):2254–5; author reply 5. pmid:12412183.
- 9. Huang HL, Kung PT, Chiu CF, Wang YH, Tsai WC. Factors associated with lung cancer patients refusing treatment and their survival: a national cohort study under a universal health insurance in Taiwan. PLoS One. 2014;9(7):e101731. pmid:24999633
- 10. Liu CY, Chen WT, Kung PT, Chiu CF, Wang YH, Shieh SH, et al. Characteristics, survival, and related factors of newly diagnosed colorectal cancer patients refusing cancer treatments under a universal health insurance program. BMC Cancer. 2014;14:446. pmid:24938667
- 11. Ramos M, Esteva M, Cabeza E, Campillo C, Llobera J, Aguilo A. Relationship of diagnostic and therapeutic delay with survival in colorectal cancer: a review. Eur J Cancer. 2007;43(17):2467–78. pmid:17931854.
- 12. Wattacheril J, Kramer JR, Richardson P, Havemann BD, Green LK, Le A, et al. Lagtimes in diagnosis and treatment of colorectal cancer: determinants and association with cancer stage and survival. Aliment Pharmacol Ther. 2008;28(9):1166–74. pmid:18691351
- 13. Iversen LH, Antonsen S, Laurberg S, Lautrup MD. Therapeutic delay reduces survival of rectal cancer but not of colonic cancer. Br J Surg. 2009;96(10):1183–9. pmid:19787765.
- 14. Helewa RM, Turner D, Park J, Wirtzfeld D, Czaykowski P, Hochman D, et al. Longer waiting times for patients undergoing colorectal cancer surgery are not associated with decreased survival. J Surg Oncol. 2013;108(6):378–84. pmid:24037666.
- 15. Pruitt SL, Harzke AJ, Davidson NO, Schootman M. Do diagnostic and treatment delays for colorectal cancer increase risk of death? Cancer Causes Control. 2013;24(5):961–77. pmid:23446843
- 16. Murchie P, Raja EA, Brewster DH, Campbell NC, Ritchie LD, Robertson R, et al. Time from first presentation in primary care to treatment of symptomatic colorectal cancer: effect on disease stage and survival. Br J Cancer. 2014;111(3):461–9. pmid:24992583
- 17. Neal RD, Tharmanathan P, France B, Din NU, Cotton S, Fallon-Ferguson J, et al. Is increased time to diagnosis and treatment in symptomatic cancer associated with poorer outcomes? Systematic review. Br J Cancer. 2015;112 Suppl 1:S92–107. pmid:25734382
- 18. Liu CY H Y, Chuang YL, Chen YJ, Weng WS, Liu JS, Liang KY. Incorporating Development Stratification of Taiwan Townships into Sampling Design of Large Scale Health Interview Survey. Journal of Health Management. 2006;4(1):1–22.
- 19. Chu YT W S. A review of claims-based comorbidity measures. Taiwan J Public Health. 2010;29(1):8–21.
- 20.
AJCC 6th Ed Cancer Staging Manual American Joint Committee on Cancer2002 [cited 2018 February 22]. https://cancerstaging.org/references-tools/deskreferences/Pages/default.aspx.
- 21. Hsu YH, Kung PT, Wang ST, Fang CY, Tsai WC. Improved patient survivals with colorectal cancer under multidisciplinary team care: A nationwide cohort study of 25,766 patients in Taiwan. Health Policy. 2016;120(6):674–81. pmid:27131976.
- 22.
The NHS Cancer Plan: a plan for investment, a plan for reform. Sep. 2000. https://wwwthhnhsuk/documents/_departments/cancer/nhscancerplanpdf (accessed May 10, 2017).
- 23. Visser MR, van Lanschot JJ, van der Velden J, Kloek JJ, Gouma DJ, Sprangers MA. Quality of life in newly diagnosed cancer patients waiting for surgery is seriously impaired. J Surg Oncol. 2006;93(7):571–7. pmid:16705725.
- 24. Zarcos-Pedrinaci I, Fernandez-Lopez A, Tellez T, Rivas-Ruiz F, Rueda A, Morales Suarez-Varela MM, et al. Factors that influence treatment delay in patients with colorectal cancer. Oncotarget. 2017;8(22):36728–42. pmid:27888636
- 25. Abu-Helalah MA, Alshraideh HA, Da’na M, Al-Hanaqtah M, Abuseif A, Arqoob K, et al. Delay in Presentation, Diagnosis and Treatment for Colorectal Cancer Patients in Jordan. J Gastrointest Cancer. 2016;47(1):36–46. pmid:26615546.
- 26. Bilimoria KY, Ko CY, Tomlinson JS, Stewart AK, Talamonti MS, Hynes DL, et al. Wait times for cancer surgery in the United States: trends and predictors of delays. Ann Surg. 2011;253(4):779–85. pmid:21475020.
- 27. Brenner H, Bouvier AM, Foschi R, Hackl M, Larsen IK, Lemmens V, et al. Progress in colorectal cancer survival in Europe from the late 1980s to the early 21st century: the EUROCARE study. Int J Cancer. 2012;131(7):1649–58. pmid:21607946.
- 28. Bhandari A, Woodhouse M, Gupta S. Colorectal cancer is a leading cause of cancer incidence and mortality among adults younger than 50 years in the USA: a SEER-based analysis with comparison to other young-onset cancers. J Investig Med. 2017;65(2):311–5. pmid:27864324.
- 29. Chiang JM, Chen MC, Changchien CR, Chen JS, Tang R, Wang JY, et al. Favorable influence of age on tumor characteristics of sporadic colorectal adenocarcinoma: patients 30 years of age or younger may be a distinct patient group. Dis Colon Rectum. 2003;46(7):904–10. pmid:12847364.
- 30. Chou CL, Chang SC, Lin TC, Chen WS, Jiang JK, Wang HS, et al. Differences in clinicopathological characteristics of colorectal cancer between younger and elderly patients: an analysis of 322 patients from a single institution. Am J Surg. 2011;202(5):574–82. pmid:21872205.
- 31. Parikh-Patel A BJ, Campleman S. Colorectal cancer stage at diagnosis by socioeconomic and urban/rural status in California, 1988–2000. Cancer. 2006;107:1189–95. pmid:16835910
- 32. Aarts MJ, Lemmens VE, Louwman MW, Kunst AE, Coebergh JW. Socioeconomic status and changing inequalities in colorectal cancer? A review of the associations with risk, treatment and outcome. Eur J Cancer. 2010;46(15):2681–95. pmid:20570136.
- 33. Byers TE, Wolf HJ, Bauer KR, Bolick-Aldrich S, Chen VW, Finch JL, et al. The impact of socioeconomic status on survival after cancer in the United States: findings from the National Program of Cancer Registries Patterns of Care Study. Cancer. 2008;113(3):582–91. pmid:18613122.
- 34. Louwman WJ, Aarts MJ, Houterman S, van Lenthe FJ, Coebergh JW, Janssen-Heijnen ML. A 50% higher prevalence of life-shortening chronic conditions among cancer patients with low socioeconomic status. Br J Cancer. 2010;103(11):1742–8. pmid:20978508
- 35. Sogaard M, Thomsen RW, Bossen KS, Sorensen HT, Norgaard M. The impact of comorbidity on cancer survival: a review. Clin Epidemiol. 2013;5(Suppl 1):3–29. pmid:24227920
- 36. Sarfati D, Koczwara B, Jackson C. The impact of comorbidity on cancer and its treatment. CA Cancer J Clin. 2016;66(4):337–50. pmid:26891458.
- 37. Schrag D, Cramer LD, Bach PB, Cohen AM, Warren JL, Begg CB. Influence of hospital procedure volume on outcomes following surgery for colon cancer. JAMA. 2000;284(23):3028–35. pmid:11122590.
- 38. Yun YH, Kim YA, Min YH, Park S, Won YJ, Kim DY, et al. The influence of hospital volume and surgical treatment delay on long-term survival after cancer surgery. Ann Oncol. 2012;23(10):2731–7. pmid:22553194.
- 39. Lordan JT, Karanjia ND, Quiney N, Fawcett WJ, Worthington TR. A 10-year study of outcome following hepatic resection for colorectal liver metastases—The effect of evaluation in a multidisciplinary team setting. Eur J Surg Oncol. 2009;35(3):302–6. pmid:18328668.
- 40. van de Velde CJ, Boelens PG, Tanis PJ, Espin E, Mroczkowski P, Naredi P, et al. Experts reviews of the multidisciplinary consensus conference colon and rectal cancer 2012: science, opinions and experiences from the experts of surgery. Eur J Surg Oncol. 2014;40(4):454–68. pmid:24268926.
- 41. Keating J, Yong D, Cutler G, Johnston J. Multidisciplinary treatment of colorectal cancer in New Zealand: survival rates from 1997–2002. N Z Med J. 2006;119(1242):U2238. pmid:16998579.