Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Comorbidity prevalence among cancer patients: a population-based cohort study of four cancers

Comorbidity prevalence among cancer patients: a population-based cohort study of four cancers Background: The presence of comorbidity affects the care of cancer patients, many of whom are living with multiple comorbidities. The prevalence of cancer comorbidity, beyond summary metrics, is not well known. This study aims to estimate the prevalence of comorbid conditions among cancer patients in England, and describe the association between cancer comorbidity and socio-economic position, using population-based electronic health records. Methods: We linked England cancer registry records of patients diagnosed with cancer of the colon, rectum, lung or Hodgkin lymphoma between 2009 and 2013, with hospital admissions records. A comorbidity was any one of fourteen specific conditions, diagnosed during hospital admission up to 6 years prior to cancer diagnosis. We calculated the crude and age-sex adjusted prevalence of each condition, the frequency of multiple comorbidity combinations, and used logistic regression and multinomial logistic regression to estimate the adjusted odds of having each condition and the probability of having each condition as a single or one of multiple comorbidities, respectively, by cancer type. Results: Comorbidity was most prevalent in patients with lung cancer and least prevalent in Hodgkin lymphoma patients. Up to two-thirds of patients within each of the four cancer patient cohorts we studied had at least one comorbidity, and around half of the comorbid patients had multiple comorbidities. Our study highlighted common comorbid conditions among the cancer patient cohorts. In all four cohorts, the odds of having a comorbidity and the probability of multiple comorbidity were consistently highest in the most deprived cancer patients. Conclusions: Cancer healthcare guidelines may need to consider prominent comorbid conditions, particularly to benefit the prognosis of the most deprived patients who carry the greater burden of comorbidity. Insight into patterns of cancer comorbidity may inform further research into the influence of specific comorbidities on socio-economic inequalities in receipt of cancer treatment and in short-term mortality. Keywords: Cancer, Comorbidity, Multimorbidity, Deprivation, Prevalence, England, Epidemiology * Correspondence: Helen.Fowler@lshtm.ac.uk Cancer Survival Group, Department of Non-Communicable Disease Epidemiology, London School of Hygiene & Tropical Medicine, Keppel Street, London WC1E 7HT, UK Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Fowler et al. BMC Cancer (2020) 20:2 Page 2 of 15 Background chronic obstructive pulmonary disease (COPD), rheum- Comorbidity refers to the existence of a long-term atological conditions, liver disease, diabetes, hemiplegia health condition in the presence of a primary disease of or paraplegia, renal disease, previous malignancy, obesity interest [1]. Having one or more comorbidities may in- or hypertension. The conditions, selected following a fluence the patient’s prognosis for a primary disease systematic search of the data, included conditions of the such as cancer. Comorbidity may influence the timing of Charlson Comorbidity Index [6] and any highly preva- cancer diagnosis, in either a positive or a negative way. lent conditions that may influence cancer management For example, the symptoms of comorbidity may drive a pa- alone or in combination with another condition. tient to seek medical care sooner, potentially leading to an earlier diagnosis. Alternatively, cancer symptoms may be Data mistakenly considered as symptoms of a pre-existing health This study used England National Cancer Registry data condition, and could delay diagnosis [2–4]. Following diag- of 331,655 patients aged 15–90 years at diagnosis with nosis, the presence of comorbidity may also influence tim- cancer of the colon, rectum, lung or Hodgkin’s lymph- ing, receipt, or outcome of treatment, with clear evidence oma, between 2009 and 2013. Registry data provided in- that those with comorbidity are less likely to receive curative formation on patient sex, age at diagnosis, site of cancer, treatment than those without, despite increasing evidence date of cancer diagnosis and area of residence at time of that many patients with comorbidity benefit from such diagnosis, which was used to derive socio-economic pos- treatment [3]. Although thepresenceof multipleco-existent ition, based on deprivation quintiles of the Income Do- health conditions is commonplace, the guidelines, funding main of the Indices of Multiple Deprivation [14]. The and structures of primary care may not support the care of five-level, ordinal variable indicates the level of more patients with multiple conditions [5], and care in sec- deprivation from 1 (least deprived) to 5 (most deprived). ondary and tertiary centres is typically highly siloed [3]. Areas of residence are defined at the Lower Super Out- Methods used in the scientific literature to describe, put Area level (mean population 1500). measure and quantify the status of comorbidity as an ex- Inpatient, outpatient and emergency hospital admis- planatory factor in adverse disease outcomes are varied. sions records (Hospital Episode Statistics, HES) [15] Many summarised metrics of comorbidity have been were successfully linked with over 99% of the cancer proposed, providing an overall picture of a patient’s co- registry records, using common unique variables present morbidity status, some specific to a primary disease in both data sources. The International Statistical Classi- while others are more general. For example, a widely fication of Diseases and Related Health Conditions tenth used metric of comorbidity in epidemiological studies is edition (ICD-10) [16] codes captured within the diagnos- the Charlson Comorbidity Index (CCI) [6], which tic fields of HES records provided information on health weights 19 long-term health conditions according to conditions recorded during hospital admissions. We their relative risk of one-year mortality, to produce an used the ICD-10 code groupings of health conditions overall index score. proposed by Quan and colleagues for defining comor- In this study, we firstly aimed to examine the preva- bidities using administrative data (see Additional file 1) lence of comorbid conditions in cancer patients using [17], and used an algorithm [18] to identify whether English population-based electronic health records of these conditions had been recorded in the six-year patients diagnosed with cancer of the colon, rectum, or period prior to cancer diagnosis. In contrast to the ap- lung or with Hodgkin lymphoma (HL). An association proach of Maringe and colleagues [18], we included between comorbidity (not specific to any primary disease diagnoses of conditions recorded up to 6 months prior of interest) and socio-economic position has been widely to cancer diagnosis. We anticipated that first-time diag- reported: the prevalence of certain specific comorbid noses of the conditions could occur in this period, and conditions [7–10] and general comorbidity prevalence wanted to obtain the most complete picture of patient being higher in deprived groups of patients [11–13]. Our comorbidity. We used cancer registry data to identify second aim was to describe patterns of comorbidities whether a patient had been diagnosed with an unrelated and multiple comorbidity in these cancer patient co- malignancy up to 6 years before their diagnosis with the horts, according to patient characteristics such as socio- cancer of interest. economic position (deprivation). Descriptive data analysis Methods We calculated the prevalence of a comorbid condition We defined a comorbid condition as one of the follow- within each of the four patient cohorts defined by cancer ing fourteen health conditions: myocardial infarction site, firstly as a crude measure, calculating the percent- (MI), congestive heart failure (CHF), peripheral vascular age of patients who had a recorded diagnosis of the co- disease (PVD), cerebrovascular disease (CVD), dementia, morbidity in HES records, and secondly adjusting for Fowler et al. BMC Cancer (2020) 20:2 Page 3 of 15 age and sex to account for the older age demographic of almost 30% of HL patients. Similar patterns in comor- cancer patient populations. Weights for this adjustment bidity prevalence were seen in males and females. The were obtained from 2011 UK census published popula- prevalence of either single or multiple comorbidity rose tion estimates of persons living in England [19]. with increasing age. Single comorbidity was more com- mon than multiple comorbidity in the younger age Statistical analysis groups, whereas in the older patients the opposite was Logistic regression models were used to estimate the observed. For example, approximately 29.2% of lung can- odds ratio (OR) of having each comorbidity by cancer cer patients aged 15–29 years had one comorbidity and site, adjusting for sex, age at cancer diagnosis and 3.4% had multiple comorbidities, while in lung cancer deprivation group. The binary outcome variable indi- patients aged 75–90 years the percentage of patients cated the presence of the comorbidity. To account for a with one comorbidity or with multiple comorbidities non-linear association between increasing age and the were 26.9 and 49.9%, respectively. presence of comorbidity, age was modelled as a continu- The prevalence of multiple comorbidity increased with ous variable using a restricted cubic spline with one knot deprivation level in colon, rectum and lung cancer pa- fixed at 70 years in analyses conducted for cancers of the tients, but there was no pattern with deprivation in HL colon, rectum and lung and at 45 years for HL (the knot patients or in the prevalence of one comorbidity. For ex- position was chosen as to be close to the mean age of ample, from 24.7 to 25.7% of rectal cancer patients had the patients in each of these cancer cohorts). To reduce one comorbidity, while 17.7 to 27.6% of patients had the risk of unstable models, we ensured there were at multiple comorbidities. least ten or more occurrences of a comorbidity within the specific cancer patient cohort for every parameter of Crude and adjusted prevalence of comorbidities at the the model (events per variable, EPV) [20]. time of cancer diagnosis Multinomial logistic regression was used to estimate Across all cancer patient cohorts, hypertension, COPD, the probability of having a given comorbidity, either in diabetes, CVD, CHF and PVD were among the most isolation, or as one of multiple comorbidities, according commonly recorded comorbid conditions. Adjusting for to cancer site. The three-category outcome variable indi- age and sex strongly impacted the prevalence of some cated whether the patient did not have the given comor- comorbid conditions in colon, rectum and lung cancer bidity, only had this comorbidity, or had this patients (Fig. 1). The three most prevalent comorbidities comorbidity with other comorbidities. Models were ad- in all four cancer patient cohorts were hypertension, justed for age, sex and deprivation, and were run for COPD and diabetes. The adjusted prevalence of hyper- each cancer site and comorbidity combination with at tension and of diabetes was similar among patients in least ten EPV. each of the four cohorts (approximately 15–20% of pa- All data analyses were conducted in STATA v.15.1 tients had hypertension while approximately 5% of pa- [21]. tients had diabetes). However, the adjusted prevalence of COPD was markedly higher in patients with lung cancer: Results approximately 25% of lung cancer patients had COPD Patient characteristics versus 10% of patients in the other patient cohorts. Simi- The characteristics of patients diagnosed with cancer of larly, in comparison between the four cohorts, the preva- the colon (N = 102,216), rectum (N = 56,342), lung (N = lence of several other conditions (CVD, CHF, PVD or 165,677) or with HL (N = 7420) between 2009 and 2013, previous malignancy) was highest among the lung cancer stratified by comorbidity status, are shown in Table 1. patients. The majority of patients in each cohort were male: ap- proximately 55% of colon, lung and HL patients and 63% of rectal cancer patients. At least 80% of colon, rec- Combinations of multiple comorbidity tum and lung cancer patients were in the two oldest age The relative frequency (%) in which five of the most group categories, while 50% of the HL patients were common conditions (COPD, diabetes, CVD, CHF and within the two youngest age groups. There was an even PVD) are present either as a single comorbidity or in distribution of patients among each of the deprivation combination with ten other common comorbid condi- groups, except among lung cancer patients, where the tions is shown in Fig. 2. For a given cancer (identified by percentage of patients in each group increased with colour), the denominator is the number of patients with deprivation level. the comorbid condition, as represented on the y-axis, Comorbidity was over twice as prevalent in lung can- and the numerator is the number of those patients who cer patients than in patients with HL: 67% of lung can- had the condition as a single comorbidity or who had cer patients had one or more comorbidities versus another condition, as depicted by the x-axis. Patients Fowler et al. BMC Cancer (2020) 20:2 Page 4 of 15 Table 1 Patient characteristics according to comorbidity status, by cancer Cancer Colon Rectum Lung Hodgkin lymphoma All patients Number of patient comorbidities All patients Number of patient comorbidities All patients Number of patient comorbidities All patients Number of patient comorbidities 0 1 2+ 0 1 2+ 0 1 2+ 0 1 2+ N % n% n % n% N % n% n % n % N % n% n % n% N % n % n % n % Sex Male 54,425 53.2 23,455 43.1 14,887 27.4 16,083 29.6 35,630 63.2 18,782 52.7 8850 24.8 7998 22.4 91,568 55.3 29,333 32.0 25,283 27.6 36,952 40.4 4163 56.1 2907 69.8 697 16.7 559 13.4 Female 47,791 46.8 21,339 44.7 13,878 29.0 12,574 26.3 20,712 36.8 11,044 53.3 5299 25.6 4369 21.1 74,109 44.7 24,661 33.3 21,536 29.1 27,912 37.7 3257 43.9 2307 70.8 573 17.6 377 11.6 Age at cancer diagnosis (years) 15–29 769 0.8 661 86.0 103 13.4 5 0.7 207 0.4 180 87.0 22 10.6 5 2.4 178 0.1 120 67.4 52 29.2 6 3.4 2111 28.5 1885 89.3 205 9.7 21 1.0 30–44 2666 2.6 2151 80.7 416 15.6 99 3.7 1552 2.8 1302 83.9 203 13.1 47 3.0 1757 1.1 1212 69.0 435 24.8 110 6.3 1660 22.4 1412 85.1 194 11.7 54 3.3 45–59 11,971 11.7 8035 67.1 2619 21.9 1317 11.0 9597 17.0 7143 74.4 1635 17.0 819 8.5 19,923 12.0 10,768 54.0 5574 28.0 3581 18.0 1461 19.7 1001 68.5 288 19.7 172 11.8 60–74 42,166 41.3 20,696 49.1 11,696 27.7 9774 23.2 25,230 44.8 14,073 55.8 6421 25.4 4736 18.8 75,085 45.3 25,973 34.6 22,241 29.6 26,871 35.8 1398 18.8 651 46.6 366 26.2 381 27.3 75–90 44,644 43.7 13,251 29.7 13,931 31.2 17,462 39.1 19,756 35.1 7128 36.1 5868 29.7 6760 34.2 68,734 41.5 15,921 23.2 18,517 26.9 34,296 49.9 790 10.6 265 33.5 217 27.5 308 39.0 Deprivation group (IMD income) Least 22,411 21.9 10,864 48.5 6331 28.2 5216 23.3 11,879 21.1 6839 57.6 2939 24.7 2101 17.7 23,066 13.9 8589 37.2 6528 28.3 7949 34.5 1339 18.0 980 73.2 217 16.2 142 10.6 deprived 2 22,623 22.1 10,484 46.3 6303 27.9 5836 25.8 12,222 21.7 6810 55.7 3031 24.8 2381 19.5 28,411 17.1 9913 34.9 8025 28.2 10,473 36.9 1428 19.2 1013 70.9 222 15.5 193 13.5 3 21,591 21.1 9460 43.8 6123 28.4 6008 27.8 11,750 20.9 6219 52.9 2970 25.3 2561 21.8 32,822 19.8 10,980 33.5 9365 28.5 12,477 38.0 1462 19.7 1018 69.6 263 18.0 181 12.4 4 19,940 19.5 8118 40.7 5614 28.2 6208 31.1 11,266 20.0 5646 50.1 2838 25.2 2782 24.7 39,220 23.7 12,356 31.5 10,885 27.8 15,979 40.7 1618 21.8 1132 70.0 277 17.1 209 12.9 Most 15,651 15.3 5868 37.5 4394 28.1 5389 34.4 9225 16.4 4312 46.7 2371 25.7 2542 27.6 42,158 25.4 12,156 28.8 12,016 28.5 17,986 42.7 1573 21.2 1071 68.1 291 18.5 211 13.4 deprived TOTAL 102,216 100.0 44,794 43.8 28,765 28.1 28,657 28.0 56,342 100.0 29,826 52.9 14,149 25.1 12,367 21.9 165,677 100.0 53,994 32.6 46,819 28.3 64,864 39.2 7420 100.0 5214 70.3 1270 17.1 936 12.6 Abbreviations - IMD Indices of Multiple Deprivation Fowler et al. BMC Cancer (2020) 20:2 Page 5 of 15 Colon Rectum % % 0 0 123456789 10 11 12 13 14 123456789 10 11 12 13 14 Adjusted Adjusted 95%CIs Lung Hodgkin lymphoma Crude % % 0 0 123456789 10 11 12 13 14 123456789 10 11 12 13 14 1: Liver disease; 2: Previous malignancy; 3: Diabetes; 4: Obesity; 5: Dementia; 6: Hemi / Paraplegia; 7: CVD; 8: Hypertension; 9: Renal disease; 10: MI; 11: COPD; 12: CHF; 13: PVD; 14: Rheumatological conditions Fig. 1 Crude and adjusted prevalence (%) of fourteen comorbidities among cancer patients in England, by cancer with two or more of the x-axis conditions are repre- Multivariate analysis sented in the numerator for each condition. The odds ratios derived from logistic regression of each Approximately one third of colorectal and lung cancer comorbid condition being present at the time of cancer patients with COPD, and over half of HL patients with diagnosis, by cancer site, for females relative to males, COPD, had this condition as a single comorbidity. By age (relative to age 70 in colon, rectal and lung cancer comparison, under one fifth of patients with diabetes, patients, and relative to age 45 in HL patients) and in- CVD, CHF and PVD had these conditions as a single co- creasing deprivation, adjusted for the other listed vari- morbidity. CHF was the condition least frequently ob- ables, are shown in Table 2. Analyses conducted for served as a single comorbidity across all four cancer patients with HL were restricted to the comorbidities of sites (89% or more of patients with CHF had additional diabetes, hypertension and COPD, as the prevalence comorbidities). counts of the other conditions did not adhere to the Hypertension was the condition most commonly minimum of ten EPV required for the analyses. present with each of comorbidities for which cross Female patients with colon, rectal or lung cancer had tabulations were investigated. In each of the cancer up to 29% increased adjusted odds of having dementia cohorts, approximately three-quarters of patients with (rectal cancer: OR 1.29; 95%CI 1.13, 1.48), up to 34% in- CHF, and a similar proportion with CVD, also had creased adjusted odds of having a previous malignancy hypertension. COPD was most commonly seen in (rectal cancer: OR 1.34; 1.23, 1.47) and approximately combination with diabetes, CVD, CHF or PVD in twice the adjusted odds of having rheumatological con- lung cancer patients: while over 50% of lung cancer ditions (colon cancer: OR 2.16; 1.98, 2.36) compared to patients with CHF also had COPD, around one third male patients. Conversely, compared with male patients of patients with HL, colon or rectal cancers with in their respective cohort, females had significantly re- CHFalsohad COPD. duced adjusted odds of having diabetes, hemiplegia or Fowler et al. BMC Cancer (2020) 20:2 Page 6 of 15 Fig. 2 Relative frequency (%) of five common conditions as a single comorbidity or with another comorbidity, by cancer paraplegia, CVD, renal disease, MI, CHF or PVD. Across cancer and colon cancer patients had approximately all four cancer cohorts, female patients had up to 38% twice the adjusted odds of having COPD compared with reduced odds of having diabetes (HL: OR 0.62; 95%CI the least deprived groups (OR 1.96; 1.89, 2.03 and OR 0.50, 0.77). 2.01; 1.89, 2.12 in the most deprived patients with lung The adjusted odds of dementia, CVD, hypertension, or colon cancer, respectively). No trend with deprivation renal disease, MI and CHF being present at diagnosis was seen with rheumatological conditions or with having consistently increased with age. For example, with 70- a previous malignancy. year old patients as the reference, colon cancer patients aged 45 had 87% reduced adjusted odds of CVD (OR 0.13; 0.13, 0.13) and 88% reduced adjusted odds of CHF Probability of having single or multiple comorbidity at (OR 0.12; 0.12, 0.12), while 90-year old patients had over the time of cancer diagnosis three times the adjusted odds of CVD (OR 3.27; 2.69, The graphs depicted in Fig. 3 show the adjusted prob- 3.99) and over four times the adjusted odds of CHF (OR ability of patients having one of the nine most common 4.72; 3.63, 6.13). There was no trend with age in colon, comorbid conditions recorded (hypertension, COPD, rectal or lung cancer patients for liver disease, having diabetes, CHF, CVD, PVD, MI, obesity or rheumato- had a previous malignancy, diabetes or obesity. In lung logical conditions) at the time of colon cancer diagnosis, cancer patients, no trend was observed with age for hav- either as a single comorbidity, or as one of multiple co- ing COPD. morbidities, according to age at cancer diagnosis and For at least eleven of the fourteen conditions, the ad- deprivation group (the least and most deprived groups), justed odds of having the comorbid condition increased as derived from multinomial logistic regression. with the level of deprivation in colon, rectal or lung can- With the exception of COPD, there was little differ- cer patients. Obesity, dementia, hemiplegia, CVD, hyper- ence between the most and least deprived groups in the tension, renal disease, MI, COPD, CHF and PVD were probability of having each of the conditions as a single associated with deprivation level in all three cancer co- comorbidity. Among those patients with COPD as a horts. For example, the most deprived groups of lung single comorbidity, the difference in probability between Fowler et al. BMC Cancer (2020) 20:2 Page 7 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Colon cancer Sex Male 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 [REF] Female 0.99 1.23 0.72 0.94 1.19 0.70 0.80 0.90 0.75 0.48 1.05 0.66 0.45 2.16 (0.91, 1.07) (1.16, 1.32) (0.69, 0.75) (0.87, 1.01) (1.09, 1.30) (0.62, 0.79) (0.75, 0.85) (0.87, 0.92) (0.71, 0.80) (0.45, 0.51) (1.01, 1.09) (0.63, 0.70) (0.42, 0.48) (1.98, 2.36) Age at cancer diagnosis (years) 45 1.18 0.40 0.27 0.79 0.02 0.39 0.13 0.14 0.15 0.12 0.55 0.12 0.12 0.22 (1.13, 1.23) (0.39, 0.41) (0.25, 0.29) (0.76, 0.82) (0.02, 0.02) (0.38, 0.39) (0.13, 0.13) (0.12, 0.16) (0.15, 0.15) (0.11, 0.12) (0.49, 0.62) (0.12, 0.12) (0.12, 0.13) (0.22, 0.22) 60 1.13 0.74 0.64 1.00 0.18 0.65 0.46 0.51 0.41 0.52 0.72 0.45 0.43 0.54 (1.08, 1.17) (0.70, 0.78) (0.55, 0.73) (0.95, 1.05) (0.17, 0.18) (0.65, 0.66) (0.45, 0.47) (0.33, 0.80) (0.40, 0.42) (0.50, 0.54) (0.62, 0.83) (0.43, 0.46) (0.41, 0.44) (0.53, 0.55) 70 [REF] 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 80 1.01 1.05 1.31 0.74 4.68 1.49 2.04 1.88 2.64 1.60 1.38 2.43 1.88 1.64 (0.98, 1.05) (0.97, 1.12) (1.00, 1.71) (0.72, 0.77) (4.49, 4.89) (1.46, 1.53) (1.80, 2.32) (0.68, 5.21) (2.30, 3.03) (1.42, 1.80) (1.05, 1.80) (2.10, 2.80) (1.66, 2.13) (1.58, 1.71) 90 0.91 0.71 0.82 0.17 13.95 1.42 3.27 2.13 4.35 2.26 1.04 4.72 1.70 1.31 (0.88, 0.94) (0.68, 0.75) (0.69, 0.98) (0.17, 0.17) (12.35, 15.75) (1.39, 1.45) (2.69, 3.99) (0.72, 6.26) (3.50, 5.40) (1.92, 2.65) (0.84, 1.29) (3.63, 6.13) (1.52, 1.91) (1.27, 1.36) Deprivation group Least 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 deprived [REF] 2 1.10 1.03 1.12 1.09 1.19 1.25 1.11 1.08 1.20 1.09 1.12 1.07 1.11 1.08 (0.96, 1.25) (0.93, 1.14) (1.05, 1.19) (0.95, 1.24) (1.03, 1.37) (1.02, 1.54) (1.01, 1.21) (1.03, 1.12) (1.09, 1.31) (0.99, 1.21) (1.06, 1.18) (0.98, 1.17) (1.00, 1.24) (0.96, 1.23) 3 1.13 0.99 1.25 1.56 1.29 1.47 1.20 1.18 1.34 1.21 1.24 1.16 1.26 0.95 (0.99, 1.29) (0.89, 1.09) (1.18, 1.33) (1.38, 1.76) (1.11, 1.48) (1.20, 1.79) (1.09, 1.31) (1.13, 1.23) (1.23, 1.47) (1.10, 1.34) (1.18, 1.32) (1.06, 1.27) (1.14, 1.40) (0.84, 1.08) 4 1.38 1.03 1.48 1.73 1.42 1.47 1.35 1.32 1.55 1.33 1.56 1.38 1.38 1.00 (1.22, 1.58) (0.93, 1.15) (1.39, 1.58) (1.53, 1.95) (1.23, 1.63) (1.20, 1.80) (1.23, 1.47) (1.27, 1.38) (1.42, 1.69) (1.21, 1.47) (1.47, 1.65) (1.26, 1.50) (1.25, 1.53) (0.88, 1.14) Most 1.50 1.13 1.75 1.91 1.70 2.29 1.60 1.54 1.83 1.55 2.01 1.67 1.59 0.99 deprived (1.31, 1.71) (1.02, 1.26) (1.64, 1.87) (1.68, 2.16) (1.47, 1.97) (1.88, 2.79) (1.46, 1.76) (1.47, 1.61) (1.67, 2.01) (1.40, 1.72) (1.89, 2.12) (1.52, 1.83) (1.42, 1.76) (0.86, 1.14) Fowler et al. BMC Cancer (2020) 20:2 Page 8 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) (Continued) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Rectal cancer Sex Male 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 [REF] Female 1.05 1.34 0.80 1.17 1.29 0.66 0.75 0.95 0.73 0.52 1.03 0.72 0.39 1.97 (0.91, 1.22) (1.23, 1.47) (0.75, 0.85) (1.04, 1.32) (1.13, 1.48) (0.54, 0.81) (0.69, 0.82) (0.91, 0.99) (0.66, 0.80) (0.46, 0.58) (0.97, 1.09) (0.65, 0.79) (0.34, 0.44) (1.73, 2.25) Age at cancer diagnosis (years) 45 0.93 0.42 0.27 0.69 0.06 0.39 0.12 0.13 0.12 0.11 0.47 0.12 0.07 0.30 (0.91, 0.95) (0.41, 0.43) (0.26, 0.29) (0.68, 0.71) (0.06, 0.06) (0.39, 0.40) (0.12, 0.12) (0.12, 0.15) (0.12, 0.13) (0.11, 0.11) (0.43, 0.51) (0.12, 0.12) (0.07, 0.07) (0.29, 0.30) 60 1.00 0.68 0.60 0.95 0.21 0.60 0.46 0.49 0.37 0.49 0.66 0.44 0.40 0.58 (0.98, 1.02) (0.65, 0.71) (0.54, 0.66) (0.92, 0.98) (0.21, 0.21) (0.60, 0.60) (0.45, 0.47) (0.34, 0.71) (0.36, 0.37) (0.48, 0.51) (0.59, 0.74) (0.43, 0.44) (0.39, 0.41) (0.57, 0.59) 70 [REF] 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 80 0.98 1.22 1.37 0.73 5.57 1.83 2.14 1.74 2.61 1.63 1.46 2.46 1.80 1.65 (0.96, 1.00) (1.13, 1.32) (1.11, 1.71) (0.71, 0.74) (5.32, 5.84) (1.81, 1.86) (1.92, 2.39) (0.71, 4.22) (2.35, 2.90) (1.48, 1.80) (1.17, 1.83) (2.21, 2.74) (1.62, 1.99) (1.60, 1.70) 90 1.11 0.96 0.83 0.17 13.80 1.82 4.07 2.51 5.14 2.06 1.29 4.96 2.24 1.91 (1.09, 1.13) (0.90, 1.02) (0.72, 0.95) (0.17, 0.17) (12.36, 15.40) (1.79, 1.85) (3.35, 4.94) (0.86, 7.29) (4.22, 6.26) (1.82, 2.32) (1.06, 1.58) (4.04, 6.10) (1.97, 2.54) (1.84, 1.99) Deprivation group Least 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 deprived [REF] 2 1.00 1.08 1.14 1.28 1.10 1.70 1.13 1.04 1.06 1.06 1.14 1.11 1.06 1.09 (0.77, 1.28) (0.94, 1.24) (1.04, 1.26) (1.06, 1.56) (0.89, 1.37) (1.20, 2.40) (0.98, 1.30) (0.99, 1.10) (0.91, 1.24) (0.91, 1.24) (1.04, 1.25) (0.95, 1.29) (0.90, 1.24) (0.89, 1.34) 3 1.50 0.91 1.29 1.27 1.21 2.13 1.39 1.14 1.26 1.17 1.36 1.23 1.11 1.13 (1.19, 1.89) (0.79, 1.05) (1.18, 1.42) (1.04, 1.54) (0.98, 1.50) (1.52, 2.98) (1.21, 1.60) (1.07, 1.20) (1.08, 1.46) (1.00, 1.36) (1.25, 1.49) (1.06, 1.42) (0.94, 1.30) (0.92, 1.39) 4 1.33 1.01 1.60 1.63 1.30 2.23 1.51 1.26 1.34 1.33 1.62 1.25 1.34 1.12 (1.05, 1.69) (0.88, 1.17) (1.46, 1.75) (1.35, 1.97) (1.05, 1.60) (1.59, 3.12) (1.32, 1.74) (1.19, 1.34) (1.15, 1.55) (1.14, 1.54) (1.49, 1.76) (1.07, 1.44) (1.14, 1.57) (0.91, 1.38) Most 1.77 1.14 1.80 1.80 1.63 2.66 1.78 1.48 1.57 1.52 2.25 1.65 1.59 1.24 deprived (1.39, 2.24) (0.99, 1.33) (1.63, 1.97) (1.48, 2.18) (1.31, 2.02) (1.89, 3.74) (1.54, 2.05) (1.39, 1.57) (1.34, 1.83) (1.30, 1.77) (2.07, 2.46) (1.42, 1.91) (1.36, 1.87) (1.00, 1.54) Fowler et al. BMC Cancer (2020) 20:2 Page 9 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) (Continued) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Lung cancer Sex Male 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 [REF] Female 0.87 1.11 0.75 1.12 1.22 0.81 0.83 0.98 0.72 0.60 1.09 0.76 0.50 1.98 (0.82, 0.93) (1.06, 1.16) (0.73, 0.78) (1.05, 1.20) (1.15, 1.30) (0.75, 0.88) (0.80, 0.87) (0.96, 1.00) (0.69, 0.75) (0.58, 0.63) (1.07, 1.11) (0.73, 0.80) (0.48, 0.52) (1.87, 2.09) Age at cancer diagnosis (years) 45 1.42 0.83 0.24 0.70 0.04 0.43 0.22 0.12 0.13 0.19 0.39 0.16 0.10 0.29 (1.35, 1.49) (0.76, 0.89) (0.23, 0.26) (0.68, 0.71) (0.04, 0.04) (0.43, 0.43) (0.21, 0.23) (0.11, 0.14) (0.13, 0.13) (0.19, 0.20) (0.31, 0.50) (0.16, 0.17) (0.09, 0.10) (0.28, 0.29) 60 1.35 0.93 0.60 0.97 0.28 0.75 0.55 0.49 0.38 0.62 0.69 0.47 0.47 0.70 (1.29, 1.42) (0.85, 1.01) (0.52, 0.69) (0.93, 1.00) (0.28, 0.28) (0.74, 0.76) (0.52, 0.59) (0.30, 0.80) (0.36, 0.39) (0.58, 0.66) (0.47, 1.03) (0.45, 0.50) (0.43, 0.51) (0.68, 0.73) 70 [REF] 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 80 0.80 1.02 1.24 0.67 4.37 1.10 1.56 1.59 2.48 1.38 1.10 1.93 1.47 1.07 (0.78, 0.82) (0.92, 1.12) (0.95, 1.63) (0.65, 0.69) (4.13, 4.63) (1.07, 1.12) (1.31, 1.85) (0.57, 4.41) (2.07, 2.97) (1.19, 1.61) (0.63, 1.94) (1.62, 2.29) (1.15, 1.87) (1.02, 1.13) 90 0.54 0.76 0.82 0.18 13.22 1.23 2.26 1.70 4.16 1.51 0.80 3.55 1.08 0.92 (0.53, 0.55) (0.71, 0.82) (0.68, 0.99) (0.18, 0.18) (11.20, 15.60) (1.20, 1.27) (1.77, 2.88) (0.59, 4.85) (3.13, 5.53) (1.28, 1.78) (0.51, 1.24) (2.63, 4.79) (0.90, 1.30) (0.88, 0.96) Deprivation group Least 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 deprived [REF] 2 1.28 0.95 1.10 1.19 1.13 1.38 1.09 1.02 1.07 1.08 1.17 1.11 1.12 1.09 (1.13, 1.46) (0.87, 1.03) (1.04, 1.16) (1.05, 1.36) (1.01, 1.27) (1.17, 1.62) (1.01, 1.17) (0.98, 1.05) (1.00, 1.16) (1.00, 1.17) (1.12, 1.22) (1.03, 1.19) (1.05, 1.20) (0.99, 1.20) 3 1.29 0.89 1.09 1.27 1.21 1.50 1.20 1.04 1.15 1.16 1.36 1.15 1.14 1.04 (1.14, 1.47) (0.82, 0.96) (1.03, 1.15) (1.12, 1.45) (1.09, 1.36) (1.28, 1.76) (1.12, 1.28) (1.01, 1.08) (1.07, 1.24) (1.08, 1.25) (1.31, 1.41) (1.07, 1.23) (1.07, 1.21) (0.95, 1.14) 4 1.44 0.84 1.25 1.51 1.38 1.76 1.36 1.13 1.21 1.28 1.59 1.30 1.17 1.08 (1.27, 1.62) (0.78, 0.91) (1.18, 1.32) (1.34, 1.70) (1.24, 1.53) (1.51, 2.04) (1.27, 1.45) (1.09, 1.17) (1.13, 1.30) (1.19, 1.38) (1.53, 1.65) (1.22, 1.39) (1.10, 1.25) (0.99, 1.18) Most 1.66 0.88 1.29 1.69 1.52 2.17 1.45 1.22 1.33 1.33 1.96 1.35 1.32 1.05 deprived (1.48, 1.87) (0.81, 0.95) (1.23, 1.36) (1.50, 1.90) (1.37, 1.69) (1.88, 2.52) (1.36, 1.54) (1.18, 1.27) (1.24, 1.43) (1.24, 1.43) (1.89, 2.03) (1.26, 1.44) (1.24, 1.41) (0.96, 1.14) Fowler et al. BMC Cancer (2020) 20:2 Page 10 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) (Continued) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Hodgkin lymphoma Sex Male –– 1.00 –– – – 1.00 –– 1.00 –– – [REF] Female –– 0.62 –– – – 0.88 –– 1.05 –– – (0.50, 0.77) (0.76, 1.02) (0.90, 1.23) Age at cancer diagnosis (years) 45 [REF] –– 1.00 –– – – 1.00 –– 1.00 –– – 60 –– 3.36 –– – – 4.46 –– 1.68 –– – (2.89, 3.91) (3.00, 6.63) (1.41, 2.00) 70 –– 5.03 –– – – 8.78 –– 2.48 –– – (4.04, 6.27) (4.57, 16.86) (1.94, 3.18) 80 –– 5.74 –– – – 13.86 –– 2.62 –– – (4.49, 7.33) (5.84, 32.91) (2.02, 3.39) 90 –– 5.17 –– – – 18.13 –– 1.43 –– – (4.13, 6.47) (6.70, 49.03) (1.23, 1.66) Deprivation group Least –– 1.00 –– – – 1.00 –– 1.00 –– – deprived [REF] 2 –– 1.19 –– – – 1.26 –– 1.17 –– – (0.83, 1.72) (1.00, 1.60) (0.90, 1.52) 3 –– 1.48 –– – – 1.43 –– 1.18 –– – (1.03, 2.11) (1.14, 1.81) (0.91, 1.54) 4 –– 1.89 –– – – 1.48 –– 1.37 –– – (1.34, 2.67) (1.18, 1.87) (1.07, 1.77) Most –– 2.39 –– – – 1.96 –– 1.86 –– – deprived (1.69, 3.37) (1.55, 2.48) (1.45, 2.38) Abbreviations - CI confidence intervals, CVD Cerebrovascular disease, MI Myocardial infarction, COPD Chronic obstructive pulmonary disease, CHF Congestive heart failure, PVD Pheripheral vascular disease, REF reference, Rheum. Rheumatological Fowler et al. BMC Cancer (2020) 20:2 Page 11 of 15 Hypertension COPD Diabetes 50 20 % % % 25 10 0 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 30 40 50 60 70 80 90 CHF CVD PVD 20 20 20 % % % 10 10 10 0 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 30 40 50 60 70 80 90 MI Obesity Rheumatological conditions 20 20 20 % % % 10 10 10 0 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 30 40 50 60 70 80 90 Age at cancer diagnosis (years) Note: Solid line represents most deprived patients, dashed line represents least deprived patients Single comorbidity Multiple comorbidity Fig. 3 Probability (%) of condition present as single or multiple comorbidity, by deprivation group (colon cancer) the most and least deprived groups decreased with age. patients had at least one long-term health condition at The most deprived patients had a higher probability of the time of their cancer diagnosis, and around half of having each of the conditions as one of multiple comor- these comorbid cancer patients had multiple long- bidities compared with the least deprived group, with term conditions. There was evidence that many of the one exception (rheumatological conditions). Generally, comorbid conditions we investigated were associated the difference in probability between the two deprivation with socio-economic deprivation, and the most de- groups was greatest in older age: it peaked at approxi- prived groups of patients had a higher probability of mately 80 years for hypertension, COPD, diabetes, PVD having multiple comorbidities compared with the less and obesity, while in patients with CHF, CVD, and MI deprived groups. the difference continued to increase with age. Having The choice of cancer sites we studied was based on rheumatological conditions was not associated with in- aetiology of the cancer: three of the cancer sites (colon, creasing age or deprivation level. rectum and lung) were associated with environmental Similar patterns in the probability of having a comor- risk factors including tobacco smoking [22, 23], alcohol bid condition according to deprivation group were ob- use and diet [24, 25]. Furthermore, tobacco smoking is served for patients with rectal or lung cancers associated with certain conditions, such as COPD [26–28] (Additional files 2 and 3). and Type 2 diabetes [29, 30], and is also associated with socioeconomic position [31]. HL is linked to infection ra- ther than environmental factors [22]. Discussion Hypertension, COPD and diabetes were the three most Our study is, to our knowledge, the first large-scale, prevalent comorbidities in all four cancer patient co- population-based study describing comorbidity preva- horts, with a higher prevalence in the most deprived pa- lence in cancer patient populations. Up to two-thirds of tients. The odds of having COPD from being in the Fowler et al. BMC Cancer (2020) 20:2 Page 12 of 15 most deprived group of lung cancer patients (compared comorbidity may have on cancer care, particularly where with being in the least deprived group – the ‘deprivation care is provided within the constraints of healthcare gap’) was 10% more than the deprivation gap in the ad- guidelines that are not designed for the simultaneous justed odds of having COPD in the Hodgkin lymphoma pa- management of two or more chronic conditions or mor- tients. This may be reflective of the role of smoking in the bidities (i.e. “multimorbidity”). Scientific studies indicate aetiology of both lung cancer and COPD, and the higher that multimorbidity is regularly observed in the popula- prevalence of smoking in the more deprived population. tion [37–39] and poses a challenge to health care sys- The association between smoking status and deprivation is tems, particularly those geared towards single disease not quantifiable in the cancer patient cohorts as we did not management [5, 40, 41]. Clinical guidelines in the United have information on smoking prevalence. Kingdom are not accommodating to the cumulative im- Similar work using administrative data to describe co- pact of treatment recommendations on those with mul- morbidity in cancer populations has been undertaken in tiple morbidities, and do not facilitate a comparison of New Zealand [32] and in Spain [33]. In the study of pa- potential benefits or risks [42]. Patients with multiple tients diagnosed with colon, rectal, breast, ovarian, uter- chronic conditions have higher rates of healthcare con- ine, stomach, liver, renal or bladder cancers in New sultations than those without [38, 43, 44]. Managing and Zealand (N = 14,096), commonly diagnosed comorbidi- treating comorbid conditions places an additional eco- ties among colon and rectal cancer patients were hyper- nomic burden on healthcare systems. In one study of the tension, cardiac conditions and diabetes. In the Spanish costs per capita of several comorbid conditions, renal cohort of colorectal cancer patients from the cancer disease was identified as one of the most costly condi- registries of Girona and Granada (N = 1061), diabetes, tions to manage among cancer patients (approximately COPD and CHF were the most common comorbidities. 174% of the costs of the cancer), while the cost of dia- Comparing our study with the study in New Zealand, betes or heart disease was substantially lower (approxi- there were similarities among colon cancer patients in mately 20% or 6% of cancer costs, respectively) [45]. The the age-sex adjusted prevalence of hypertension, while increase in costs also depends on the number and com- diabetes prevalence was higher in New Zealand. The ad- bination of comorbid conditions: among the cancer pa- justed prevalence of hypertension was 16.6%, uncompli- tients with diabetes in our study, between 10 and 15% of cated diabetes was 5.9% and diabetes with complications these patients also had renal disease. was 5.0% among patients in New Zealand, while in our In cancer patients, the presence of comorbidity can be study the adjusted prevalence of hypertension was 17.4% influential on cancer management and therapeutic op- and diabetes (with and without complications) was 5.7%. tions. Patients with comorbidity may be less likely than This supports our earlier assumption that less severe those without comorbidity to receive curative treatment diabetes may be underreported in hospital admissions [3]. Treatment decisions made by clinicians may be records. Given the ‘gatekeeper’ structure and functioning weighted by the type and severity of comorbidity, for ex- of the healthcare system in the UK [34] and the focus on ample, CHF has been reported to influence receipt of managing diabetes within primary care [35], cases of dia- surgery for non-small cell lung cancer [46], receipt of betes recorded in hospital admissions are possibly those adjuvant chemotherapy for colon cancer [47] and receipt that are not controlled within available primary care re- of any treatment for prostate cancer [48]. The presence sources [36] or present with complications. The Spanish of COPD influenced receipt of surgical treatment in study reported the crude prevalence of conditions non-small cell lung cancer patients [46] and adjuvant among colorectal cancer patients, which were generally therapy in colon cancer patients [47]. However, there is higher than the crude prevalence of conditions observed also evidence that comorbid patients who receive treat- in our study. Diabetes was prevalent in 23.6% of colorec- ment have better prognosis for survival than those who tal cancer patients in this study, while in our study the do not receive treatment, as shown with the receipt of crude prevalence of diabetes was 11.4% or 9.4% among adjuvant therapy for colon cancer [47, 49]. Moreover, colon or rectal cancer patients, respectively. Nonetheless, older cancer patients and patients with comorbidity have there was consistency between our study and both of historically been under-represented in cancer clinical tri- these other studies in terms of common comorbid con- als. This limits the applicability of cancer clinical trial ditions among the patient cohorts. results to a younger and healthier cohort of patients In our study, approximately 13% of the HL cohort, than clinicians are actually treating, meaning that while over 21% of the colorectal cancer cohorts and over 39% there is evidence suggesting that patients with comor- of the lung cancer cohort had multiple comorbidities, bidity as a group are not receiving optimal cancer treat- while from 17 to 28% of patients in each cohort had a ment, specific information required for clinical decision- single comorbidity at the time of their cancer diagnosis. making is often lacking [50]. We found a non-negligible These findings are important given the impact increase in the prevalence of comorbidities when we Fowler et al. BMC Cancer (2020) 20:2 Page 13 of 15 included diagnoses in the six-months prior to cancer obtain this information. The potential for measure- diagnoses. While some of these conditions may have ment error from the information recorded in the arisen in these months because of the cancer, their pres- diagnostic fields of hospital admissions records should ence will be as relevant when considering treatment, ir- also be acknowledged. However, we assume that the respective of the timing of their diagnosis. more severe conditions are likely to be captured Our study showed socio-economic position to be an within the diagnostic fields. Underreporting may important factor associated with having one or more co- occur in less severe conditions, such as obesity, that morbid conditions at the time of cancer diagnosis, with are unlikely to be the primary reason for the hospital comorbidity prevalence increasing with deprivation. It is admission, and may occur more frequently with eld- possible that mechanisms within clinical guidelines and erly patients or patients with more severe comorbidi- decision-making that lead to non-treatment of cancer ties, due to competing demands. Conditions such as patients with comorbidity disproportionately impact the less severe type II diabetes are possibly underreported. more deprived patients. An existence of socio-economic Further work comparing the prevalence of the condi- inequalities in receipt of treatment has been identified tions we studied in the cancer cohorts with the [51, 52]. Reviewing the treatment process of cancer pa- prevalence of these conditions in the general popula- tients with comorbidity may therefore have a beneficial tion in England, as reported in government publica- effect in reducing the socioeconomic inequalities in re- tions and scientific literature, would be useful step in ceipt of cancer treatment. Moreover, because cancer validating our results. data contains mainly cancer-related outcomes, how the Our study of over 300,000 patients is one of the largest cancer and related treatments impact patient comorbid- population-based studies of comorbidity prevalence ity and prognosis is not well known [3]. Having the re- among cancer patients, and one of the first such studies sources and guidelines within which to manage patient of patients in England. Using data from well-established comorbid conditions robustly during cancer treatment is sources, we were able to describe the prevalence of four- one strategy for mitigating the risk of adverse patient teen chronic health conditions among these cancer pa- outcomes occurring from comorbid disease. In England, tients, and highlight an association between socio- socio-economic inequalities in cancer survival have nar- economic position and prevalence of most of these rowed little, despite the implementation of government conditions. strategies that intended to reduce these inequalities [53]. Focusing on the management of comorbidity in cancer Conclusion patients could be one potential pathway to addressing This study underlines that many comorbid cancer pa- socio-economic inequalities in cancer outcomes. tients are living with multiple comorbidities, and that There are a variety of metrics of comorbidity in the the most deprived patients carry the greater burden of scientific literature that are used to study the relation- comorbidity. Healthcare guidelines may not always en- ship between comorbidity on cancer outcomes, although compass the simultaneous management of multiple no consensus has been reached on a gold standard chronic conditions, but guidelines for the management measure of comorbidity within the context of cancer of cancer may need to consider some prominent comor- [54]. Many of the approaches provide a summary meas- bid conditions. Insight into patterns of cancer comorbid- ure of the patient’s comorbid conditions and the severity ity informs further research into the influence of of these conditions. However, the prognostic impact of comorbidity - particularly the influence of specific co- comorbidity can depend on the type and stage of the morbid conditions - on outcomes following cancer diag- cancer [55]. In addition, the presence of comorbidity - nosis, including socio-economic inequalities in receipt of particularly certain comorbid conditions - adds com- treatment and short-term mortality. plexity to the provision of treatment for cancer. When investigating the relationship between comorbidity and cancer outcomes, a more granular approach investigat- Supplementary information ing specific comorbid conditions in turn, rather than Supplementary information accompanies this paper at https://doi.org/10. 1186/s12885-019-6472-9. using a summary measure of comorbidity, could be more appropriate and insightful. Additional file 1. Definition of the fourteen conditions, according to We acknowledge potential limitations in this study. ICD-10 code classification. Table of the fourteen conditions and the ICD- We capture comorbidity information based on diag- 10 code groupings used to define them. noses of health conditions recorded during hospital Additional file 2. Probability (%) of condition present as single or multiple admission(s) prior to cancer diagnosis, and are there- comorbidity, by deprivation group (lung cancer). Additional results in complement to those presented in Fig. 3: graphs representing the probability fore reliant on patients requiring hospital-based med- of having any of nine comorbid conditions in lung cancer patients. ical attention for their health condition(s) in order to Fowler et al. BMC Cancer (2020) 20:2 Page 14 of 15 Received: 8 May 2019 Accepted: 17 December 2019 Additional file 3. Probability (%) of condition present as single or multiple comorbidity, by deprivation group (rectal cancer). Additional results in complement to those presented in Fig. 3: graphs representing the probability of having any of nine comorbid conditions in rectal References cancer patients. 1. Porta MS, Greenland S, Last JM. A dictionary of epidemiology: Oxford University press; 2014. Available from: http://www.oxfordreference.com/ view/10.1093/acref/9780195314496.001.0001/acref-9780195314496-e-352 Abbreviations ?rskey=TmAWeL&result=351. Accessed 31 Oct 2018. CCI: Charlson Comorbidity Index; CHF: Congestive Heart Failure; 2. Gumey J, Sarfati D, Stanley J. The impact of patient comorbidity on cancer CI: Confidence Interval; COPD: Chronic Obstructive Pulmonary Disease; stage at diagnosis. BJC. 2015;113:1375–80. CVD: Cerebrovascular Disease; EPV: Events Per Variable; HES: Hospital Episode 3. Sarfati D, Koczwara B, Jackson C. The impact of comorbidity on cancer and Statistics; HL: Hodgkin lymphoma; ICD: International Classification of Diseases its treatment. CA Cancer J Clin. 2016;66(4):337–50. and Related Health Problems; MI: Myocardial infarction; OR: Odds Ratio; 4. Salika T, Lyratzopoulos G, Whitaker KL, Waller J, Renzi C. Do comorbidities PVD: Peripheral Vascular Disease influence help-seeking for cancer alarm symptoms? A population-based survey in England. J Public Health. 2018;40(2):340–9. 5. Tinetti ME, Fried TR, Boyd CM. Designing health care for the most common Acknowledgements chronic condition--multimorbidity. JAMA. 2012;307(23):2493–4. We would like to thank Adrian Turculet, Data Manager of the LSHTM Cancer 6. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying Survival Group, for his support and assistance with the data linkage. prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. Authors’ contributions 7. Connolly V, Unwin N, Sherriff P, Bilous R, Kelly W. Diabetes prevalence and CM, MALF and HF contributed to the conception of the study. BR, HF, AB socioeconomic status: a population based study showing increased and LE designed the study. AB and ENN provided advice on statistical prevalence of type 2 diabetes mellitus in deprived areas. J Epidemiol methods. HF conducted the analyses of the data and prepared the draft Community Health. 2000;54(3):173–7. of the manuscript, tables and figures. BR, AB, and LE supervised the study 8. Sundquist K, Malmstrom M, Johansson S. Neighbourhood deprivation and and provided comments on the manuscript draft. CM, MALF, NN incidence of coronary heart disease: a multilevel study of 2.6 million women and DS provided comments on the final draft of the manuscript. All and men in Sweden. J Epidemiol Community Health. 2004;58(1):71–7. authors read and approved the final manuscript. 9. Hossain MP, Palmer D, Goyder E, El Nahas AM. Social deprivation and prevalence of chronic kidney disease in the UK: workload implications for primary care. QJM. 2012;105(2):167–75. Funding 10. Fano V, Pezzotti P, Gnavi R, Bontempi K, Miceli M, Pagnozzi E, et al. The role This research was funded by Cancer Research UK grant number C7923/ of socio-economic factors on prevalence and health outcomes of persons A18525. The authors declare no support from any organisations for the with diabetes in Rome, Italy. Eur J Public Health. 2013;23(6):991–7. submitted work. The design of the study, the analyses and the writing of the 11. Macleod U, Mitchell E, Black M, Spence G. Comorbidity and socioeconomic manuscript were solely the responsibility of the authors. The findings and deprivation: an observational study of the prevalence of comorbidity in conclusions in this manuscript are those of the authors and do not general practice. Eur J Gen Pract. 2009;10(1):3. necessarily represent the views of Cancer Research UK. 12. McLean G, Gunn J, Wyke S, Guthrie B, Watt G, Blane D, et al. The influence of socioeconomic deprivation on multimorbidity at different ages: a cross- Availability of data and materials sectional study. Br J Gen Pract. 2014;64(624):8. The data that support the findings of this study are available via application 13. Moffat K, Mercer SW. Challenges of managing people with multimorbidity to the Public Health England Office for Data Release, but restrictions apply to in today's healthcare systems. BMC Fam Pract. 2015;16:129. the availability of these data. 14. Department for Communities and Local Government. The English Indices of Deprivation 2010. London: 2011. Available from: https://www.gov.uk/ government/statistics/english-indices-of-deprivation-2010 Ethics approval and consent to participate 15. Health and Social Care Information Centre. Hospital Episode Statistics 2013. We obtained the statutory approvals required for this research from the Available from: http://www.hscic.gov.uk/hes. Accessed 18 Nov 2017. Confidentiality Advisory Group (CAG) of the Health Research Authority (HRA): 16. World Health Organisation. International statistical classification of diseases and PIAG 1–05(c) 2007. related health problems. Tenth revision. Geneva: World Health Organization; 1994. Ethical approval was obtained from the Research Ethics Committee (REC) of 17. Quan HD, Li B, Couris CM, Fushimi K, Graham P, Hider P, et al. Updating and the Health Research Authority (HRA): 07/MRE01/52. validating the Charlson comorbidity index and score for risk adjustment in This work uses data provided by patients and collected by the National hospital discharge abstracts using data from 6 countries. Am J Epidemiol. Health Service as part of their care and support. We used anonymised 2011;173(6):676–82. National Cancer Registry and Hospital Episode Statistics data. No consent to 18. Maringe C, Fowler H, Rachet B, Luque-Fernández MA. Reproducibility, reliability participate was sought from patients. and validity of population-based administrative health data for the assessment of cancer non-related comorbidities. PLoS One. 2017;12(3):e0172814 Consent for publication 19. Office for National Statistics. 2011 Census: Population estimates by single Not applicable. year of age and sex for local authorities in the United Kingdom 2013. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/ Competing interests populationandmigration/populationestimates/datasets/2011 The authors declare that they have no competing interests. censuspopulationestimatesbysingleyearofageand sexforlocalauthoritiesintheunitedkingdom. Accessed 01 Aug 2018. Author details 20. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study Cancer Survival Group, Department of Non-Communicable Disease of the number of events per variable in logistic regression analysis. J Clin Epidemiology, London School of Hygiene & Tropical Medicine, Keppel Street, Epidemiol. 1996;49(12):1373–9. London WC1E 7HT, UK. Biomedical Research Institute of Granada, 21. StataCorp. STATA statistical software. 15th ed. College Station: Stata Non-Communicable and Cancer Epidemiology Group, University of Granada, Corporation; 2017. Granada, Spain. Department of Non-Communicable Disease Epidemiology, 22. Schottenfeld D, Fraumeni JF. Cancer epidemiology and prevention. 3rd ed. London School of Hygiene & Tropical Medicine, London, UK. UCL Oxford: Oxford University Press; 2006. Respiratory, University College London, London, UK. Department of Thoracic 23. Tindle HA, Stevenson Duncan M, Greevy RA, Vasan RS, Kundu S, Massion PP, Medicine, University College London Hospital, London, UK. Department of et al. Lifetime Smoking History and Risk of Lung Cancer: Results From the Public Health, University of Otago, Wellington, New Zealand. Framingham Heart Study. J Natl Cancer Inst. 2018;110(11):1201–7. Fowler et al. BMC Cancer (2020) 20:2 Page 15 of 15 24. Danaei G, Vander Hoorn S, Lopez AD, Murray CJ, Ezzati M. Causes of cancer 48. Bradley CJ, Dahman B, Anscher M. Prostate cancer treatment and survival: in the world: comparative risk assessment of nine behavioural and evidence for men with prevalent comorbid conditions. Med Care. 2014; environmental risk factors. Lancet. 2005;366(9499):1784–93. 52(6):482–9. 25. Haggar FA, Boushey RP. Colorectal cancer epidemiology: incidence, 49. Sarfati D, Hill S, Blakely T, Robson B, Purdie G, Dennett E, et al. The effect of mortality, survival, and risk factors. Clin Colon Rectal Surg. 2009;22(4):191–7. comorbidity on the use of adjuvant chemotherapy and survival from colon cancer: a retrospective cohort study. BMC Cancer. 2009;9:116. 26. Devereux G. ABC of chronic obstructive pulmonary disease - definition, 50. Aapro MS, Kohne CH, Cohen HJ, Extermann M. Never too old? Age should epidemiology, and risk factors. BMJ. 2006;332(7550):1142–4. not be a barrier to enrollment in cancer clinical trials. Oncologist. 2005;10(3): 27. Buist AS, McBurnie MA, Vollmer WM, Gillespie S, Burney P, Mannino DM, 198–204. et al. International variation in the prevalence of COPD (the BOLD study): a 51. Akinyemiju T, Meng Q, Vin-Raviv N. Race/ethnicity and socio-economic population-based prevalence study. Lancet. 2007;370(9589):741–50. differences in colorectal cancer surgery outcomes: analysis of the 28. Laniado-Laborín R. Smoking and chronic obstructive pulmonary disease nationwide inpatient sample. BMC Cancer. 2016;16:715. (COPD). Parallel epidemics of the 21(st) century. Int J Environ Res Public 52. Forrest LF, Sowden S, Rubin G, White M, Adams J. Socio-economic Health. 2009;6(1):209–24. inequalities in stage at diagnosis, and in time intervals on the lung cancer 29. Hu FB, Manson JE, Stampfer MJ, Colditz G, Liu S, Solomon CG, et al. Diet, pathway from first symptom to treatment: systematic review and meta- lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med. analysis. Thorax. 2017;72(5):430–6. 2001;345(11):790–7. 53. Exarchakou A, Rachet B, Belot A, Maringe C, Coleman MP. Impact of national 30. Wannamethee SG, Shaper AG, Perry IJ. Smoking as a modifiable risk factor cancer policies on cancer survival trends and socioeconomic inequalities in for type 2 diabetes in middle-aged men. Diabetes Care. 2001;24(9):1590–5. England, 1996-2013: population based study. BMJ. 2018;360:k764. 31. Hiscock R, Bauld L, Amos A, Platt S. Smoking and socioeconomic status in 54. Sarfati D. Review of methods used to measure comorbidity in cancer England: the rise of the never smoker and the disadvantaged smoker. J populations: no gold standard exists. J Clin Epidemiol. 2012;65(9):924–33. Public Health. 2012;34(3):390–6. 55. Read WL, Tierney RM, Page NC, Costas I, Govindan R, Spitznagel EL, et al. 32. Sarfati D, Gurney J, Lim BT, Bagheri N, Simpson A, Koea J, et al. Identifying Differential prognostic impact of comorbidity. J Clin Oncol. 2004;22(15): important comorbidity among cancer populations using administrative data: 3099–103. prevalence and impact on survival. Asia Pac J Clin Oncol. 2016;12(1):e47–56. 33. Luque-Fernandez MA, Redondo Sánchez D, Rodriguez-Barranco M, Carmona- Garcia MC, Marcos-Gragera R, Sanchez-Perez MJ. Multimorbidity by patient and Publisher’sNote tumor factors and time-to-surgery among colorectal cancer patients in Spain: a Springer Nature remains neutral with regard to jurisdictional claims in population-based study. Clinical Epidemiology. 2019 [in press]. published maps and institutional affiliations. 34. Loudon I. The principle of referral: the gatekeeping role of the GP. Br J Gen Pract. 2008;58(547):128–30 discussion 30. 35. Nicholson E, Cummings M, Cranston I, Meeking D, Kar P. The super six model of care: five years on. Diab Prim Care. 2016;18:6. 36. Saxena S, George J, Barber J, Fitzpatrick J, Majeed A. Association of population and practice factors with potentially avoidable admission rates for chronic diseases in London: cross sectional analysis. J R Soc Med. 2006; 99(2):81–9. 37. Fortin M, Bravo G, Hudon C, Vanasse A, Lapointe L. Prevalence of multimorbidity among adults seen in family practice. Ann Fam Med. 2005; 3(3):223–8. 38. Salisbury C, Johnson L, Purdy S, Valderas JM, Montgomery AA. Epidemiology and impact of multimorbidity in primary care: a retrospective cohort study. Br J Gen Pract. 2011;61(582):e12–21. 39. Schäfer I, Hansen H, Schön G, Höfels S, Altiner A, Dahlhaus A, et al. The influence of age, gender and socio-economic status on multimorbidity patterns in primary care. first results from the multicare cohort study. BMC Health Serv Res. 2012;12(1):89. 40. Boyd CM, Fortin M. Future of multimorbidity research: how should understanding of multimorbidity inform health system design? Public Health Rev. 2010;32(2):451–74. 41. Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study. Lancet. 2012;380(9836):37–43. 42. Hughes LD, McMurdo MET, Guthrie B. Guidelines for people not for diseases: the challenges of applying UK clinical guidelines to people with multimorbidity. Age Ageing. 2013;42(1):62–9. 43. Glynn LG, Valderas JM, Healy P, Burke E, Newell J, Gillespie P, et al. The prevalence of multimorbidity in primary care and its effect on health care utilization and cost. Fam Pract. 2011;28(5):516–23. 44. Bahler C, Huber CA, Brungger B, Reich O. Multimorbidity, health care utilization and costs in an elderly community-dwelling population: a claims data based observational study. BMC Health Serv Res. 2015;15:23. 45. Cortaredona S, Ventelou B. The extra cost of comorbidity: multiple illnesses and the economic burden of non-communicable diseases. BMC Med. 2017;15(1):216. 46. Belot A, Fowler H, Njagi EN, Luque-Fernandez MA, Maringe C, Magadi W, et al. Association between age, deprivation and specific comorbid conditions and the receipt of major surgery in patients with non-small cell lung cancer in England: a population-based study. Thorax. 2019;74(1):51–9. 47. Gross CP, McAvay GJ, Guo Z, Tinetti ME. The impact of chronic illnesses on the use and effectiveness of adjuvant chemotherapy for colon cancer. Cancer. 2007;109(12):2410–9. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Cancer Springer Journals

Comorbidity prevalence among cancer patients: a population-based cohort study of four cancers

Loading next page...
 
/lp/springer-journals/comorbidity-prevalence-among-cancer-patients-a-population-based-cohort-Z4bynlQhPD

References (62)

Publisher
Springer Journals
Copyright
Copyright © The Author(s). 2020
eISSN
1471-2407
DOI
10.1186/s12885-019-6472-9
Publisher site
See Article on Publisher Site

Abstract

Background: The presence of comorbidity affects the care of cancer patients, many of whom are living with multiple comorbidities. The prevalence of cancer comorbidity, beyond summary metrics, is not well known. This study aims to estimate the prevalence of comorbid conditions among cancer patients in England, and describe the association between cancer comorbidity and socio-economic position, using population-based electronic health records. Methods: We linked England cancer registry records of patients diagnosed with cancer of the colon, rectum, lung or Hodgkin lymphoma between 2009 and 2013, with hospital admissions records. A comorbidity was any one of fourteen specific conditions, diagnosed during hospital admission up to 6 years prior to cancer diagnosis. We calculated the crude and age-sex adjusted prevalence of each condition, the frequency of multiple comorbidity combinations, and used logistic regression and multinomial logistic regression to estimate the adjusted odds of having each condition and the probability of having each condition as a single or one of multiple comorbidities, respectively, by cancer type. Results: Comorbidity was most prevalent in patients with lung cancer and least prevalent in Hodgkin lymphoma patients. Up to two-thirds of patients within each of the four cancer patient cohorts we studied had at least one comorbidity, and around half of the comorbid patients had multiple comorbidities. Our study highlighted common comorbid conditions among the cancer patient cohorts. In all four cohorts, the odds of having a comorbidity and the probability of multiple comorbidity were consistently highest in the most deprived cancer patients. Conclusions: Cancer healthcare guidelines may need to consider prominent comorbid conditions, particularly to benefit the prognosis of the most deprived patients who carry the greater burden of comorbidity. Insight into patterns of cancer comorbidity may inform further research into the influence of specific comorbidities on socio-economic inequalities in receipt of cancer treatment and in short-term mortality. Keywords: Cancer, Comorbidity, Multimorbidity, Deprivation, Prevalence, England, Epidemiology * Correspondence: Helen.Fowler@lshtm.ac.uk Cancer Survival Group, Department of Non-Communicable Disease Epidemiology, London School of Hygiene & Tropical Medicine, Keppel Street, London WC1E 7HT, UK Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Fowler et al. BMC Cancer (2020) 20:2 Page 2 of 15 Background chronic obstructive pulmonary disease (COPD), rheum- Comorbidity refers to the existence of a long-term atological conditions, liver disease, diabetes, hemiplegia health condition in the presence of a primary disease of or paraplegia, renal disease, previous malignancy, obesity interest [1]. Having one or more comorbidities may in- or hypertension. The conditions, selected following a fluence the patient’s prognosis for a primary disease systematic search of the data, included conditions of the such as cancer. Comorbidity may influence the timing of Charlson Comorbidity Index [6] and any highly preva- cancer diagnosis, in either a positive or a negative way. lent conditions that may influence cancer management For example, the symptoms of comorbidity may drive a pa- alone or in combination with another condition. tient to seek medical care sooner, potentially leading to an earlier diagnosis. Alternatively, cancer symptoms may be Data mistakenly considered as symptoms of a pre-existing health This study used England National Cancer Registry data condition, and could delay diagnosis [2–4]. Following diag- of 331,655 patients aged 15–90 years at diagnosis with nosis, the presence of comorbidity may also influence tim- cancer of the colon, rectum, lung or Hodgkin’s lymph- ing, receipt, or outcome of treatment, with clear evidence oma, between 2009 and 2013. Registry data provided in- that those with comorbidity are less likely to receive curative formation on patient sex, age at diagnosis, site of cancer, treatment than those without, despite increasing evidence date of cancer diagnosis and area of residence at time of that many patients with comorbidity benefit from such diagnosis, which was used to derive socio-economic pos- treatment [3]. Although thepresenceof multipleco-existent ition, based on deprivation quintiles of the Income Do- health conditions is commonplace, the guidelines, funding main of the Indices of Multiple Deprivation [14]. The and structures of primary care may not support the care of five-level, ordinal variable indicates the level of more patients with multiple conditions [5], and care in sec- deprivation from 1 (least deprived) to 5 (most deprived). ondary and tertiary centres is typically highly siloed [3]. Areas of residence are defined at the Lower Super Out- Methods used in the scientific literature to describe, put Area level (mean population 1500). measure and quantify the status of comorbidity as an ex- Inpatient, outpatient and emergency hospital admis- planatory factor in adverse disease outcomes are varied. sions records (Hospital Episode Statistics, HES) [15] Many summarised metrics of comorbidity have been were successfully linked with over 99% of the cancer proposed, providing an overall picture of a patient’s co- registry records, using common unique variables present morbidity status, some specific to a primary disease in both data sources. The International Statistical Classi- while others are more general. For example, a widely fication of Diseases and Related Health Conditions tenth used metric of comorbidity in epidemiological studies is edition (ICD-10) [16] codes captured within the diagnos- the Charlson Comorbidity Index (CCI) [6], which tic fields of HES records provided information on health weights 19 long-term health conditions according to conditions recorded during hospital admissions. We their relative risk of one-year mortality, to produce an used the ICD-10 code groupings of health conditions overall index score. proposed by Quan and colleagues for defining comor- In this study, we firstly aimed to examine the preva- bidities using administrative data (see Additional file 1) lence of comorbid conditions in cancer patients using [17], and used an algorithm [18] to identify whether English population-based electronic health records of these conditions had been recorded in the six-year patients diagnosed with cancer of the colon, rectum, or period prior to cancer diagnosis. In contrast to the ap- lung or with Hodgkin lymphoma (HL). An association proach of Maringe and colleagues [18], we included between comorbidity (not specific to any primary disease diagnoses of conditions recorded up to 6 months prior of interest) and socio-economic position has been widely to cancer diagnosis. We anticipated that first-time diag- reported: the prevalence of certain specific comorbid noses of the conditions could occur in this period, and conditions [7–10] and general comorbidity prevalence wanted to obtain the most complete picture of patient being higher in deprived groups of patients [11–13]. Our comorbidity. We used cancer registry data to identify second aim was to describe patterns of comorbidities whether a patient had been diagnosed with an unrelated and multiple comorbidity in these cancer patient co- malignancy up to 6 years before their diagnosis with the horts, according to patient characteristics such as socio- cancer of interest. economic position (deprivation). Descriptive data analysis Methods We calculated the prevalence of a comorbid condition We defined a comorbid condition as one of the follow- within each of the four patient cohorts defined by cancer ing fourteen health conditions: myocardial infarction site, firstly as a crude measure, calculating the percent- (MI), congestive heart failure (CHF), peripheral vascular age of patients who had a recorded diagnosis of the co- disease (PVD), cerebrovascular disease (CVD), dementia, morbidity in HES records, and secondly adjusting for Fowler et al. BMC Cancer (2020) 20:2 Page 3 of 15 age and sex to account for the older age demographic of almost 30% of HL patients. Similar patterns in comor- cancer patient populations. Weights for this adjustment bidity prevalence were seen in males and females. The were obtained from 2011 UK census published popula- prevalence of either single or multiple comorbidity rose tion estimates of persons living in England [19]. with increasing age. Single comorbidity was more com- mon than multiple comorbidity in the younger age Statistical analysis groups, whereas in the older patients the opposite was Logistic regression models were used to estimate the observed. For example, approximately 29.2% of lung can- odds ratio (OR) of having each comorbidity by cancer cer patients aged 15–29 years had one comorbidity and site, adjusting for sex, age at cancer diagnosis and 3.4% had multiple comorbidities, while in lung cancer deprivation group. The binary outcome variable indi- patients aged 75–90 years the percentage of patients cated the presence of the comorbidity. To account for a with one comorbidity or with multiple comorbidities non-linear association between increasing age and the were 26.9 and 49.9%, respectively. presence of comorbidity, age was modelled as a continu- The prevalence of multiple comorbidity increased with ous variable using a restricted cubic spline with one knot deprivation level in colon, rectum and lung cancer pa- fixed at 70 years in analyses conducted for cancers of the tients, but there was no pattern with deprivation in HL colon, rectum and lung and at 45 years for HL (the knot patients or in the prevalence of one comorbidity. For ex- position was chosen as to be close to the mean age of ample, from 24.7 to 25.7% of rectal cancer patients had the patients in each of these cancer cohorts). To reduce one comorbidity, while 17.7 to 27.6% of patients had the risk of unstable models, we ensured there were at multiple comorbidities. least ten or more occurrences of a comorbidity within the specific cancer patient cohort for every parameter of Crude and adjusted prevalence of comorbidities at the the model (events per variable, EPV) [20]. time of cancer diagnosis Multinomial logistic regression was used to estimate Across all cancer patient cohorts, hypertension, COPD, the probability of having a given comorbidity, either in diabetes, CVD, CHF and PVD were among the most isolation, or as one of multiple comorbidities, according commonly recorded comorbid conditions. Adjusting for to cancer site. The three-category outcome variable indi- age and sex strongly impacted the prevalence of some cated whether the patient did not have the given comor- comorbid conditions in colon, rectum and lung cancer bidity, only had this comorbidity, or had this patients (Fig. 1). The three most prevalent comorbidities comorbidity with other comorbidities. Models were ad- in all four cancer patient cohorts were hypertension, justed for age, sex and deprivation, and were run for COPD and diabetes. The adjusted prevalence of hyper- each cancer site and comorbidity combination with at tension and of diabetes was similar among patients in least ten EPV. each of the four cohorts (approximately 15–20% of pa- All data analyses were conducted in STATA v.15.1 tients had hypertension while approximately 5% of pa- [21]. tients had diabetes). However, the adjusted prevalence of COPD was markedly higher in patients with lung cancer: Results approximately 25% of lung cancer patients had COPD Patient characteristics versus 10% of patients in the other patient cohorts. Simi- The characteristics of patients diagnosed with cancer of larly, in comparison between the four cohorts, the preva- the colon (N = 102,216), rectum (N = 56,342), lung (N = lence of several other conditions (CVD, CHF, PVD or 165,677) or with HL (N = 7420) between 2009 and 2013, previous malignancy) was highest among the lung cancer stratified by comorbidity status, are shown in Table 1. patients. The majority of patients in each cohort were male: ap- proximately 55% of colon, lung and HL patients and 63% of rectal cancer patients. At least 80% of colon, rec- Combinations of multiple comorbidity tum and lung cancer patients were in the two oldest age The relative frequency (%) in which five of the most group categories, while 50% of the HL patients were common conditions (COPD, diabetes, CVD, CHF and within the two youngest age groups. There was an even PVD) are present either as a single comorbidity or in distribution of patients among each of the deprivation combination with ten other common comorbid condi- groups, except among lung cancer patients, where the tions is shown in Fig. 2. For a given cancer (identified by percentage of patients in each group increased with colour), the denominator is the number of patients with deprivation level. the comorbid condition, as represented on the y-axis, Comorbidity was over twice as prevalent in lung can- and the numerator is the number of those patients who cer patients than in patients with HL: 67% of lung can- had the condition as a single comorbidity or who had cer patients had one or more comorbidities versus another condition, as depicted by the x-axis. Patients Fowler et al. BMC Cancer (2020) 20:2 Page 4 of 15 Table 1 Patient characteristics according to comorbidity status, by cancer Cancer Colon Rectum Lung Hodgkin lymphoma All patients Number of patient comorbidities All patients Number of patient comorbidities All patients Number of patient comorbidities All patients Number of patient comorbidities 0 1 2+ 0 1 2+ 0 1 2+ 0 1 2+ N % n% n % n% N % n% n % n % N % n% n % n% N % n % n % n % Sex Male 54,425 53.2 23,455 43.1 14,887 27.4 16,083 29.6 35,630 63.2 18,782 52.7 8850 24.8 7998 22.4 91,568 55.3 29,333 32.0 25,283 27.6 36,952 40.4 4163 56.1 2907 69.8 697 16.7 559 13.4 Female 47,791 46.8 21,339 44.7 13,878 29.0 12,574 26.3 20,712 36.8 11,044 53.3 5299 25.6 4369 21.1 74,109 44.7 24,661 33.3 21,536 29.1 27,912 37.7 3257 43.9 2307 70.8 573 17.6 377 11.6 Age at cancer diagnosis (years) 15–29 769 0.8 661 86.0 103 13.4 5 0.7 207 0.4 180 87.0 22 10.6 5 2.4 178 0.1 120 67.4 52 29.2 6 3.4 2111 28.5 1885 89.3 205 9.7 21 1.0 30–44 2666 2.6 2151 80.7 416 15.6 99 3.7 1552 2.8 1302 83.9 203 13.1 47 3.0 1757 1.1 1212 69.0 435 24.8 110 6.3 1660 22.4 1412 85.1 194 11.7 54 3.3 45–59 11,971 11.7 8035 67.1 2619 21.9 1317 11.0 9597 17.0 7143 74.4 1635 17.0 819 8.5 19,923 12.0 10,768 54.0 5574 28.0 3581 18.0 1461 19.7 1001 68.5 288 19.7 172 11.8 60–74 42,166 41.3 20,696 49.1 11,696 27.7 9774 23.2 25,230 44.8 14,073 55.8 6421 25.4 4736 18.8 75,085 45.3 25,973 34.6 22,241 29.6 26,871 35.8 1398 18.8 651 46.6 366 26.2 381 27.3 75–90 44,644 43.7 13,251 29.7 13,931 31.2 17,462 39.1 19,756 35.1 7128 36.1 5868 29.7 6760 34.2 68,734 41.5 15,921 23.2 18,517 26.9 34,296 49.9 790 10.6 265 33.5 217 27.5 308 39.0 Deprivation group (IMD income) Least 22,411 21.9 10,864 48.5 6331 28.2 5216 23.3 11,879 21.1 6839 57.6 2939 24.7 2101 17.7 23,066 13.9 8589 37.2 6528 28.3 7949 34.5 1339 18.0 980 73.2 217 16.2 142 10.6 deprived 2 22,623 22.1 10,484 46.3 6303 27.9 5836 25.8 12,222 21.7 6810 55.7 3031 24.8 2381 19.5 28,411 17.1 9913 34.9 8025 28.2 10,473 36.9 1428 19.2 1013 70.9 222 15.5 193 13.5 3 21,591 21.1 9460 43.8 6123 28.4 6008 27.8 11,750 20.9 6219 52.9 2970 25.3 2561 21.8 32,822 19.8 10,980 33.5 9365 28.5 12,477 38.0 1462 19.7 1018 69.6 263 18.0 181 12.4 4 19,940 19.5 8118 40.7 5614 28.2 6208 31.1 11,266 20.0 5646 50.1 2838 25.2 2782 24.7 39,220 23.7 12,356 31.5 10,885 27.8 15,979 40.7 1618 21.8 1132 70.0 277 17.1 209 12.9 Most 15,651 15.3 5868 37.5 4394 28.1 5389 34.4 9225 16.4 4312 46.7 2371 25.7 2542 27.6 42,158 25.4 12,156 28.8 12,016 28.5 17,986 42.7 1573 21.2 1071 68.1 291 18.5 211 13.4 deprived TOTAL 102,216 100.0 44,794 43.8 28,765 28.1 28,657 28.0 56,342 100.0 29,826 52.9 14,149 25.1 12,367 21.9 165,677 100.0 53,994 32.6 46,819 28.3 64,864 39.2 7420 100.0 5214 70.3 1270 17.1 936 12.6 Abbreviations - IMD Indices of Multiple Deprivation Fowler et al. BMC Cancer (2020) 20:2 Page 5 of 15 Colon Rectum % % 0 0 123456789 10 11 12 13 14 123456789 10 11 12 13 14 Adjusted Adjusted 95%CIs Lung Hodgkin lymphoma Crude % % 0 0 123456789 10 11 12 13 14 123456789 10 11 12 13 14 1: Liver disease; 2: Previous malignancy; 3: Diabetes; 4: Obesity; 5: Dementia; 6: Hemi / Paraplegia; 7: CVD; 8: Hypertension; 9: Renal disease; 10: MI; 11: COPD; 12: CHF; 13: PVD; 14: Rheumatological conditions Fig. 1 Crude and adjusted prevalence (%) of fourteen comorbidities among cancer patients in England, by cancer with two or more of the x-axis conditions are repre- Multivariate analysis sented in the numerator for each condition. The odds ratios derived from logistic regression of each Approximately one third of colorectal and lung cancer comorbid condition being present at the time of cancer patients with COPD, and over half of HL patients with diagnosis, by cancer site, for females relative to males, COPD, had this condition as a single comorbidity. By age (relative to age 70 in colon, rectal and lung cancer comparison, under one fifth of patients with diabetes, patients, and relative to age 45 in HL patients) and in- CVD, CHF and PVD had these conditions as a single co- creasing deprivation, adjusted for the other listed vari- morbidity. CHF was the condition least frequently ob- ables, are shown in Table 2. Analyses conducted for served as a single comorbidity across all four cancer patients with HL were restricted to the comorbidities of sites (89% or more of patients with CHF had additional diabetes, hypertension and COPD, as the prevalence comorbidities). counts of the other conditions did not adhere to the Hypertension was the condition most commonly minimum of ten EPV required for the analyses. present with each of comorbidities for which cross Female patients with colon, rectal or lung cancer had tabulations were investigated. In each of the cancer up to 29% increased adjusted odds of having dementia cohorts, approximately three-quarters of patients with (rectal cancer: OR 1.29; 95%CI 1.13, 1.48), up to 34% in- CHF, and a similar proportion with CVD, also had creased adjusted odds of having a previous malignancy hypertension. COPD was most commonly seen in (rectal cancer: OR 1.34; 1.23, 1.47) and approximately combination with diabetes, CVD, CHF or PVD in twice the adjusted odds of having rheumatological con- lung cancer patients: while over 50% of lung cancer ditions (colon cancer: OR 2.16; 1.98, 2.36) compared to patients with CHF also had COPD, around one third male patients. Conversely, compared with male patients of patients with HL, colon or rectal cancers with in their respective cohort, females had significantly re- CHFalsohad COPD. duced adjusted odds of having diabetes, hemiplegia or Fowler et al. BMC Cancer (2020) 20:2 Page 6 of 15 Fig. 2 Relative frequency (%) of five common conditions as a single comorbidity or with another comorbidity, by cancer paraplegia, CVD, renal disease, MI, CHF or PVD. Across cancer and colon cancer patients had approximately all four cancer cohorts, female patients had up to 38% twice the adjusted odds of having COPD compared with reduced odds of having diabetes (HL: OR 0.62; 95%CI the least deprived groups (OR 1.96; 1.89, 2.03 and OR 0.50, 0.77). 2.01; 1.89, 2.12 in the most deprived patients with lung The adjusted odds of dementia, CVD, hypertension, or colon cancer, respectively). No trend with deprivation renal disease, MI and CHF being present at diagnosis was seen with rheumatological conditions or with having consistently increased with age. For example, with 70- a previous malignancy. year old patients as the reference, colon cancer patients aged 45 had 87% reduced adjusted odds of CVD (OR 0.13; 0.13, 0.13) and 88% reduced adjusted odds of CHF Probability of having single or multiple comorbidity at (OR 0.12; 0.12, 0.12), while 90-year old patients had over the time of cancer diagnosis three times the adjusted odds of CVD (OR 3.27; 2.69, The graphs depicted in Fig. 3 show the adjusted prob- 3.99) and over four times the adjusted odds of CHF (OR ability of patients having one of the nine most common 4.72; 3.63, 6.13). There was no trend with age in colon, comorbid conditions recorded (hypertension, COPD, rectal or lung cancer patients for liver disease, having diabetes, CHF, CVD, PVD, MI, obesity or rheumato- had a previous malignancy, diabetes or obesity. In lung logical conditions) at the time of colon cancer diagnosis, cancer patients, no trend was observed with age for hav- either as a single comorbidity, or as one of multiple co- ing COPD. morbidities, according to age at cancer diagnosis and For at least eleven of the fourteen conditions, the ad- deprivation group (the least and most deprived groups), justed odds of having the comorbid condition increased as derived from multinomial logistic regression. with the level of deprivation in colon, rectal or lung can- With the exception of COPD, there was little differ- cer patients. Obesity, dementia, hemiplegia, CVD, hyper- ence between the most and least deprived groups in the tension, renal disease, MI, COPD, CHF and PVD were probability of having each of the conditions as a single associated with deprivation level in all three cancer co- comorbidity. Among those patients with COPD as a horts. For example, the most deprived groups of lung single comorbidity, the difference in probability between Fowler et al. BMC Cancer (2020) 20:2 Page 7 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Colon cancer Sex Male 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 [REF] Female 0.99 1.23 0.72 0.94 1.19 0.70 0.80 0.90 0.75 0.48 1.05 0.66 0.45 2.16 (0.91, 1.07) (1.16, 1.32) (0.69, 0.75) (0.87, 1.01) (1.09, 1.30) (0.62, 0.79) (0.75, 0.85) (0.87, 0.92) (0.71, 0.80) (0.45, 0.51) (1.01, 1.09) (0.63, 0.70) (0.42, 0.48) (1.98, 2.36) Age at cancer diagnosis (years) 45 1.18 0.40 0.27 0.79 0.02 0.39 0.13 0.14 0.15 0.12 0.55 0.12 0.12 0.22 (1.13, 1.23) (0.39, 0.41) (0.25, 0.29) (0.76, 0.82) (0.02, 0.02) (0.38, 0.39) (0.13, 0.13) (0.12, 0.16) (0.15, 0.15) (0.11, 0.12) (0.49, 0.62) (0.12, 0.12) (0.12, 0.13) (0.22, 0.22) 60 1.13 0.74 0.64 1.00 0.18 0.65 0.46 0.51 0.41 0.52 0.72 0.45 0.43 0.54 (1.08, 1.17) (0.70, 0.78) (0.55, 0.73) (0.95, 1.05) (0.17, 0.18) (0.65, 0.66) (0.45, 0.47) (0.33, 0.80) (0.40, 0.42) (0.50, 0.54) (0.62, 0.83) (0.43, 0.46) (0.41, 0.44) (0.53, 0.55) 70 [REF] 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 80 1.01 1.05 1.31 0.74 4.68 1.49 2.04 1.88 2.64 1.60 1.38 2.43 1.88 1.64 (0.98, 1.05) (0.97, 1.12) (1.00, 1.71) (0.72, 0.77) (4.49, 4.89) (1.46, 1.53) (1.80, 2.32) (0.68, 5.21) (2.30, 3.03) (1.42, 1.80) (1.05, 1.80) (2.10, 2.80) (1.66, 2.13) (1.58, 1.71) 90 0.91 0.71 0.82 0.17 13.95 1.42 3.27 2.13 4.35 2.26 1.04 4.72 1.70 1.31 (0.88, 0.94) (0.68, 0.75) (0.69, 0.98) (0.17, 0.17) (12.35, 15.75) (1.39, 1.45) (2.69, 3.99) (0.72, 6.26) (3.50, 5.40) (1.92, 2.65) (0.84, 1.29) (3.63, 6.13) (1.52, 1.91) (1.27, 1.36) Deprivation group Least 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 deprived [REF] 2 1.10 1.03 1.12 1.09 1.19 1.25 1.11 1.08 1.20 1.09 1.12 1.07 1.11 1.08 (0.96, 1.25) (0.93, 1.14) (1.05, 1.19) (0.95, 1.24) (1.03, 1.37) (1.02, 1.54) (1.01, 1.21) (1.03, 1.12) (1.09, 1.31) (0.99, 1.21) (1.06, 1.18) (0.98, 1.17) (1.00, 1.24) (0.96, 1.23) 3 1.13 0.99 1.25 1.56 1.29 1.47 1.20 1.18 1.34 1.21 1.24 1.16 1.26 0.95 (0.99, 1.29) (0.89, 1.09) (1.18, 1.33) (1.38, 1.76) (1.11, 1.48) (1.20, 1.79) (1.09, 1.31) (1.13, 1.23) (1.23, 1.47) (1.10, 1.34) (1.18, 1.32) (1.06, 1.27) (1.14, 1.40) (0.84, 1.08) 4 1.38 1.03 1.48 1.73 1.42 1.47 1.35 1.32 1.55 1.33 1.56 1.38 1.38 1.00 (1.22, 1.58) (0.93, 1.15) (1.39, 1.58) (1.53, 1.95) (1.23, 1.63) (1.20, 1.80) (1.23, 1.47) (1.27, 1.38) (1.42, 1.69) (1.21, 1.47) (1.47, 1.65) (1.26, 1.50) (1.25, 1.53) (0.88, 1.14) Most 1.50 1.13 1.75 1.91 1.70 2.29 1.60 1.54 1.83 1.55 2.01 1.67 1.59 0.99 deprived (1.31, 1.71) (1.02, 1.26) (1.64, 1.87) (1.68, 2.16) (1.47, 1.97) (1.88, 2.79) (1.46, 1.76) (1.47, 1.61) (1.67, 2.01) (1.40, 1.72) (1.89, 2.12) (1.52, 1.83) (1.42, 1.76) (0.86, 1.14) Fowler et al. BMC Cancer (2020) 20:2 Page 8 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) (Continued) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Rectal cancer Sex Male 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 [REF] Female 1.05 1.34 0.80 1.17 1.29 0.66 0.75 0.95 0.73 0.52 1.03 0.72 0.39 1.97 (0.91, 1.22) (1.23, 1.47) (0.75, 0.85) (1.04, 1.32) (1.13, 1.48) (0.54, 0.81) (0.69, 0.82) (0.91, 0.99) (0.66, 0.80) (0.46, 0.58) (0.97, 1.09) (0.65, 0.79) (0.34, 0.44) (1.73, 2.25) Age at cancer diagnosis (years) 45 0.93 0.42 0.27 0.69 0.06 0.39 0.12 0.13 0.12 0.11 0.47 0.12 0.07 0.30 (0.91, 0.95) (0.41, 0.43) (0.26, 0.29) (0.68, 0.71) (0.06, 0.06) (0.39, 0.40) (0.12, 0.12) (0.12, 0.15) (0.12, 0.13) (0.11, 0.11) (0.43, 0.51) (0.12, 0.12) (0.07, 0.07) (0.29, 0.30) 60 1.00 0.68 0.60 0.95 0.21 0.60 0.46 0.49 0.37 0.49 0.66 0.44 0.40 0.58 (0.98, 1.02) (0.65, 0.71) (0.54, 0.66) (0.92, 0.98) (0.21, 0.21) (0.60, 0.60) (0.45, 0.47) (0.34, 0.71) (0.36, 0.37) (0.48, 0.51) (0.59, 0.74) (0.43, 0.44) (0.39, 0.41) (0.57, 0.59) 70 [REF] 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 80 0.98 1.22 1.37 0.73 5.57 1.83 2.14 1.74 2.61 1.63 1.46 2.46 1.80 1.65 (0.96, 1.00) (1.13, 1.32) (1.11, 1.71) (0.71, 0.74) (5.32, 5.84) (1.81, 1.86) (1.92, 2.39) (0.71, 4.22) (2.35, 2.90) (1.48, 1.80) (1.17, 1.83) (2.21, 2.74) (1.62, 1.99) (1.60, 1.70) 90 1.11 0.96 0.83 0.17 13.80 1.82 4.07 2.51 5.14 2.06 1.29 4.96 2.24 1.91 (1.09, 1.13) (0.90, 1.02) (0.72, 0.95) (0.17, 0.17) (12.36, 15.40) (1.79, 1.85) (3.35, 4.94) (0.86, 7.29) (4.22, 6.26) (1.82, 2.32) (1.06, 1.58) (4.04, 6.10) (1.97, 2.54) (1.84, 1.99) Deprivation group Least 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 deprived [REF] 2 1.00 1.08 1.14 1.28 1.10 1.70 1.13 1.04 1.06 1.06 1.14 1.11 1.06 1.09 (0.77, 1.28) (0.94, 1.24) (1.04, 1.26) (1.06, 1.56) (0.89, 1.37) (1.20, 2.40) (0.98, 1.30) (0.99, 1.10) (0.91, 1.24) (0.91, 1.24) (1.04, 1.25) (0.95, 1.29) (0.90, 1.24) (0.89, 1.34) 3 1.50 0.91 1.29 1.27 1.21 2.13 1.39 1.14 1.26 1.17 1.36 1.23 1.11 1.13 (1.19, 1.89) (0.79, 1.05) (1.18, 1.42) (1.04, 1.54) (0.98, 1.50) (1.52, 2.98) (1.21, 1.60) (1.07, 1.20) (1.08, 1.46) (1.00, 1.36) (1.25, 1.49) (1.06, 1.42) (0.94, 1.30) (0.92, 1.39) 4 1.33 1.01 1.60 1.63 1.30 2.23 1.51 1.26 1.34 1.33 1.62 1.25 1.34 1.12 (1.05, 1.69) (0.88, 1.17) (1.46, 1.75) (1.35, 1.97) (1.05, 1.60) (1.59, 3.12) (1.32, 1.74) (1.19, 1.34) (1.15, 1.55) (1.14, 1.54) (1.49, 1.76) (1.07, 1.44) (1.14, 1.57) (0.91, 1.38) Most 1.77 1.14 1.80 1.80 1.63 2.66 1.78 1.48 1.57 1.52 2.25 1.65 1.59 1.24 deprived (1.39, 2.24) (0.99, 1.33) (1.63, 1.97) (1.48, 2.18) (1.31, 2.02) (1.89, 3.74) (1.54, 2.05) (1.39, 1.57) (1.34, 1.83) (1.30, 1.77) (2.07, 2.46) (1.42, 1.91) (1.36, 1.87) (1.00, 1.54) Fowler et al. BMC Cancer (2020) 20:2 Page 9 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) (Continued) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Lung cancer Sex Male 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 [REF] Female 0.87 1.11 0.75 1.12 1.22 0.81 0.83 0.98 0.72 0.60 1.09 0.76 0.50 1.98 (0.82, 0.93) (1.06, 1.16) (0.73, 0.78) (1.05, 1.20) (1.15, 1.30) (0.75, 0.88) (0.80, 0.87) (0.96, 1.00) (0.69, 0.75) (0.58, 0.63) (1.07, 1.11) (0.73, 0.80) (0.48, 0.52) (1.87, 2.09) Age at cancer diagnosis (years) 45 1.42 0.83 0.24 0.70 0.04 0.43 0.22 0.12 0.13 0.19 0.39 0.16 0.10 0.29 (1.35, 1.49) (0.76, 0.89) (0.23, 0.26) (0.68, 0.71) (0.04, 0.04) (0.43, 0.43) (0.21, 0.23) (0.11, 0.14) (0.13, 0.13) (0.19, 0.20) (0.31, 0.50) (0.16, 0.17) (0.09, 0.10) (0.28, 0.29) 60 1.35 0.93 0.60 0.97 0.28 0.75 0.55 0.49 0.38 0.62 0.69 0.47 0.47 0.70 (1.29, 1.42) (0.85, 1.01) (0.52, 0.69) (0.93, 1.00) (0.28, 0.28) (0.74, 0.76) (0.52, 0.59) (0.30, 0.80) (0.36, 0.39) (0.58, 0.66) (0.47, 1.03) (0.45, 0.50) (0.43, 0.51) (0.68, 0.73) 70 [REF] 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 80 0.80 1.02 1.24 0.67 4.37 1.10 1.56 1.59 2.48 1.38 1.10 1.93 1.47 1.07 (0.78, 0.82) (0.92, 1.12) (0.95, 1.63) (0.65, 0.69) (4.13, 4.63) (1.07, 1.12) (1.31, 1.85) (0.57, 4.41) (2.07, 2.97) (1.19, 1.61) (0.63, 1.94) (1.62, 2.29) (1.15, 1.87) (1.02, 1.13) 90 0.54 0.76 0.82 0.18 13.22 1.23 2.26 1.70 4.16 1.51 0.80 3.55 1.08 0.92 (0.53, 0.55) (0.71, 0.82) (0.68, 0.99) (0.18, 0.18) (11.20, 15.60) (1.20, 1.27) (1.77, 2.88) (0.59, 4.85) (3.13, 5.53) (1.28, 1.78) (0.51, 1.24) (2.63, 4.79) (0.90, 1.30) (0.88, 0.96) Deprivation group Least 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 deprived [REF] 2 1.28 0.95 1.10 1.19 1.13 1.38 1.09 1.02 1.07 1.08 1.17 1.11 1.12 1.09 (1.13, 1.46) (0.87, 1.03) (1.04, 1.16) (1.05, 1.36) (1.01, 1.27) (1.17, 1.62) (1.01, 1.17) (0.98, 1.05) (1.00, 1.16) (1.00, 1.17) (1.12, 1.22) (1.03, 1.19) (1.05, 1.20) (0.99, 1.20) 3 1.29 0.89 1.09 1.27 1.21 1.50 1.20 1.04 1.15 1.16 1.36 1.15 1.14 1.04 (1.14, 1.47) (0.82, 0.96) (1.03, 1.15) (1.12, 1.45) (1.09, 1.36) (1.28, 1.76) (1.12, 1.28) (1.01, 1.08) (1.07, 1.24) (1.08, 1.25) (1.31, 1.41) (1.07, 1.23) (1.07, 1.21) (0.95, 1.14) 4 1.44 0.84 1.25 1.51 1.38 1.76 1.36 1.13 1.21 1.28 1.59 1.30 1.17 1.08 (1.27, 1.62) (0.78, 0.91) (1.18, 1.32) (1.34, 1.70) (1.24, 1.53) (1.51, 2.04) (1.27, 1.45) (1.09, 1.17) (1.13, 1.30) (1.19, 1.38) (1.53, 1.65) (1.22, 1.39) (1.10, 1.25) (0.99, 1.18) Most 1.66 0.88 1.29 1.69 1.52 2.17 1.45 1.22 1.33 1.33 1.96 1.35 1.32 1.05 deprived (1.48, 1.87) (0.81, 0.95) (1.23, 1.36) (1.50, 1.90) (1.37, 1.69) (1.88, 2.52) (1.36, 1.54) (1.18, 1.27) (1.24, 1.43) (1.24, 1.43) (1.89, 2.03) (1.26, 1.44) (1.24, 1.41) (0.96, 1.14) Fowler et al. BMC Cancer (2020) 20:2 Page 10 of 15 Table 2 Odds ratios of condition being present, by cancer (adjusted for other listed variables) (Continued) Liver Previous Diabetes Obesity Dementia Hemi- or CVD Hyper- Renal MI COPD CHF PVD Rheum. disease malignancy paraplegia tension disease conditions OR (95% CIs) Hodgkin lymphoma Sex Male –– 1.00 –– – – 1.00 –– 1.00 –– – [REF] Female –– 0.62 –– – – 0.88 –– 1.05 –– – (0.50, 0.77) (0.76, 1.02) (0.90, 1.23) Age at cancer diagnosis (years) 45 [REF] –– 1.00 –– – – 1.00 –– 1.00 –– – 60 –– 3.36 –– – – 4.46 –– 1.68 –– – (2.89, 3.91) (3.00, 6.63) (1.41, 2.00) 70 –– 5.03 –– – – 8.78 –– 2.48 –– – (4.04, 6.27) (4.57, 16.86) (1.94, 3.18) 80 –– 5.74 –– – – 13.86 –– 2.62 –– – (4.49, 7.33) (5.84, 32.91) (2.02, 3.39) 90 –– 5.17 –– – – 18.13 –– 1.43 –– – (4.13, 6.47) (6.70, 49.03) (1.23, 1.66) Deprivation group Least –– 1.00 –– – – 1.00 –– 1.00 –– – deprived [REF] 2 –– 1.19 –– – – 1.26 –– 1.17 –– – (0.83, 1.72) (1.00, 1.60) (0.90, 1.52) 3 –– 1.48 –– – – 1.43 –– 1.18 –– – (1.03, 2.11) (1.14, 1.81) (0.91, 1.54) 4 –– 1.89 –– – – 1.48 –– 1.37 –– – (1.34, 2.67) (1.18, 1.87) (1.07, 1.77) Most –– 2.39 –– – – 1.96 –– 1.86 –– – deprived (1.69, 3.37) (1.55, 2.48) (1.45, 2.38) Abbreviations - CI confidence intervals, CVD Cerebrovascular disease, MI Myocardial infarction, COPD Chronic obstructive pulmonary disease, CHF Congestive heart failure, PVD Pheripheral vascular disease, REF reference, Rheum. Rheumatological Fowler et al. BMC Cancer (2020) 20:2 Page 11 of 15 Hypertension COPD Diabetes 50 20 % % % 25 10 0 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 30 40 50 60 70 80 90 CHF CVD PVD 20 20 20 % % % 10 10 10 0 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 30 40 50 60 70 80 90 MI Obesity Rheumatological conditions 20 20 20 % % % 10 10 10 0 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 30 40 50 60 70 80 90 Age at cancer diagnosis (years) Note: Solid line represents most deprived patients, dashed line represents least deprived patients Single comorbidity Multiple comorbidity Fig. 3 Probability (%) of condition present as single or multiple comorbidity, by deprivation group (colon cancer) the most and least deprived groups decreased with age. patients had at least one long-term health condition at The most deprived patients had a higher probability of the time of their cancer diagnosis, and around half of having each of the conditions as one of multiple comor- these comorbid cancer patients had multiple long- bidities compared with the least deprived group, with term conditions. There was evidence that many of the one exception (rheumatological conditions). Generally, comorbid conditions we investigated were associated the difference in probability between the two deprivation with socio-economic deprivation, and the most de- groups was greatest in older age: it peaked at approxi- prived groups of patients had a higher probability of mately 80 years for hypertension, COPD, diabetes, PVD having multiple comorbidities compared with the less and obesity, while in patients with CHF, CVD, and MI deprived groups. the difference continued to increase with age. Having The choice of cancer sites we studied was based on rheumatological conditions was not associated with in- aetiology of the cancer: three of the cancer sites (colon, creasing age or deprivation level. rectum and lung) were associated with environmental Similar patterns in the probability of having a comor- risk factors including tobacco smoking [22, 23], alcohol bid condition according to deprivation group were ob- use and diet [24, 25]. Furthermore, tobacco smoking is served for patients with rectal or lung cancers associated with certain conditions, such as COPD [26–28] (Additional files 2 and 3). and Type 2 diabetes [29, 30], and is also associated with socioeconomic position [31]. HL is linked to infection ra- ther than environmental factors [22]. Discussion Hypertension, COPD and diabetes were the three most Our study is, to our knowledge, the first large-scale, prevalent comorbidities in all four cancer patient co- population-based study describing comorbidity preva- horts, with a higher prevalence in the most deprived pa- lence in cancer patient populations. Up to two-thirds of tients. The odds of having COPD from being in the Fowler et al. BMC Cancer (2020) 20:2 Page 12 of 15 most deprived group of lung cancer patients (compared comorbidity may have on cancer care, particularly where with being in the least deprived group – the ‘deprivation care is provided within the constraints of healthcare gap’) was 10% more than the deprivation gap in the ad- guidelines that are not designed for the simultaneous justed odds of having COPD in the Hodgkin lymphoma pa- management of two or more chronic conditions or mor- tients. This may be reflective of the role of smoking in the bidities (i.e. “multimorbidity”). Scientific studies indicate aetiology of both lung cancer and COPD, and the higher that multimorbidity is regularly observed in the popula- prevalence of smoking in the more deprived population. tion [37–39] and poses a challenge to health care sys- The association between smoking status and deprivation is tems, particularly those geared towards single disease not quantifiable in the cancer patient cohorts as we did not management [5, 40, 41]. Clinical guidelines in the United have information on smoking prevalence. Kingdom are not accommodating to the cumulative im- Similar work using administrative data to describe co- pact of treatment recommendations on those with mul- morbidity in cancer populations has been undertaken in tiple morbidities, and do not facilitate a comparison of New Zealand [32] and in Spain [33]. In the study of pa- potential benefits or risks [42]. Patients with multiple tients diagnosed with colon, rectal, breast, ovarian, uter- chronic conditions have higher rates of healthcare con- ine, stomach, liver, renal or bladder cancers in New sultations than those without [38, 43, 44]. Managing and Zealand (N = 14,096), commonly diagnosed comorbidi- treating comorbid conditions places an additional eco- ties among colon and rectal cancer patients were hyper- nomic burden on healthcare systems. In one study of the tension, cardiac conditions and diabetes. In the Spanish costs per capita of several comorbid conditions, renal cohort of colorectal cancer patients from the cancer disease was identified as one of the most costly condi- registries of Girona and Granada (N = 1061), diabetes, tions to manage among cancer patients (approximately COPD and CHF were the most common comorbidities. 174% of the costs of the cancer), while the cost of dia- Comparing our study with the study in New Zealand, betes or heart disease was substantially lower (approxi- there were similarities among colon cancer patients in mately 20% or 6% of cancer costs, respectively) [45]. The the age-sex adjusted prevalence of hypertension, while increase in costs also depends on the number and com- diabetes prevalence was higher in New Zealand. The ad- bination of comorbid conditions: among the cancer pa- justed prevalence of hypertension was 16.6%, uncompli- tients with diabetes in our study, between 10 and 15% of cated diabetes was 5.9% and diabetes with complications these patients also had renal disease. was 5.0% among patients in New Zealand, while in our In cancer patients, the presence of comorbidity can be study the adjusted prevalence of hypertension was 17.4% influential on cancer management and therapeutic op- and diabetes (with and without complications) was 5.7%. tions. Patients with comorbidity may be less likely than This supports our earlier assumption that less severe those without comorbidity to receive curative treatment diabetes may be underreported in hospital admissions [3]. Treatment decisions made by clinicians may be records. Given the ‘gatekeeper’ structure and functioning weighted by the type and severity of comorbidity, for ex- of the healthcare system in the UK [34] and the focus on ample, CHF has been reported to influence receipt of managing diabetes within primary care [35], cases of dia- surgery for non-small cell lung cancer [46], receipt of betes recorded in hospital admissions are possibly those adjuvant chemotherapy for colon cancer [47] and receipt that are not controlled within available primary care re- of any treatment for prostate cancer [48]. The presence sources [36] or present with complications. The Spanish of COPD influenced receipt of surgical treatment in study reported the crude prevalence of conditions non-small cell lung cancer patients [46] and adjuvant among colorectal cancer patients, which were generally therapy in colon cancer patients [47]. However, there is higher than the crude prevalence of conditions observed also evidence that comorbid patients who receive treat- in our study. Diabetes was prevalent in 23.6% of colorec- ment have better prognosis for survival than those who tal cancer patients in this study, while in our study the do not receive treatment, as shown with the receipt of crude prevalence of diabetes was 11.4% or 9.4% among adjuvant therapy for colon cancer [47, 49]. Moreover, colon or rectal cancer patients, respectively. Nonetheless, older cancer patients and patients with comorbidity have there was consistency between our study and both of historically been under-represented in cancer clinical tri- these other studies in terms of common comorbid con- als. This limits the applicability of cancer clinical trial ditions among the patient cohorts. results to a younger and healthier cohort of patients In our study, approximately 13% of the HL cohort, than clinicians are actually treating, meaning that while over 21% of the colorectal cancer cohorts and over 39% there is evidence suggesting that patients with comor- of the lung cancer cohort had multiple comorbidities, bidity as a group are not receiving optimal cancer treat- while from 17 to 28% of patients in each cohort had a ment, specific information required for clinical decision- single comorbidity at the time of their cancer diagnosis. making is often lacking [50]. We found a non-negligible These findings are important given the impact increase in the prevalence of comorbidities when we Fowler et al. BMC Cancer (2020) 20:2 Page 13 of 15 included diagnoses in the six-months prior to cancer obtain this information. The potential for measure- diagnoses. While some of these conditions may have ment error from the information recorded in the arisen in these months because of the cancer, their pres- diagnostic fields of hospital admissions records should ence will be as relevant when considering treatment, ir- also be acknowledged. However, we assume that the respective of the timing of their diagnosis. more severe conditions are likely to be captured Our study showed socio-economic position to be an within the diagnostic fields. Underreporting may important factor associated with having one or more co- occur in less severe conditions, such as obesity, that morbid conditions at the time of cancer diagnosis, with are unlikely to be the primary reason for the hospital comorbidity prevalence increasing with deprivation. It is admission, and may occur more frequently with eld- possible that mechanisms within clinical guidelines and erly patients or patients with more severe comorbidi- decision-making that lead to non-treatment of cancer ties, due to competing demands. Conditions such as patients with comorbidity disproportionately impact the less severe type II diabetes are possibly underreported. more deprived patients. An existence of socio-economic Further work comparing the prevalence of the condi- inequalities in receipt of treatment has been identified tions we studied in the cancer cohorts with the [51, 52]. Reviewing the treatment process of cancer pa- prevalence of these conditions in the general popula- tients with comorbidity may therefore have a beneficial tion in England, as reported in government publica- effect in reducing the socioeconomic inequalities in re- tions and scientific literature, would be useful step in ceipt of cancer treatment. Moreover, because cancer validating our results. data contains mainly cancer-related outcomes, how the Our study of over 300,000 patients is one of the largest cancer and related treatments impact patient comorbid- population-based studies of comorbidity prevalence ity and prognosis is not well known [3]. Having the re- among cancer patients, and one of the first such studies sources and guidelines within which to manage patient of patients in England. Using data from well-established comorbid conditions robustly during cancer treatment is sources, we were able to describe the prevalence of four- one strategy for mitigating the risk of adverse patient teen chronic health conditions among these cancer pa- outcomes occurring from comorbid disease. In England, tients, and highlight an association between socio- socio-economic inequalities in cancer survival have nar- economic position and prevalence of most of these rowed little, despite the implementation of government conditions. strategies that intended to reduce these inequalities [53]. Focusing on the management of comorbidity in cancer Conclusion patients could be one potential pathway to addressing This study underlines that many comorbid cancer pa- socio-economic inequalities in cancer outcomes. tients are living with multiple comorbidities, and that There are a variety of metrics of comorbidity in the the most deprived patients carry the greater burden of scientific literature that are used to study the relation- comorbidity. Healthcare guidelines may not always en- ship between comorbidity on cancer outcomes, although compass the simultaneous management of multiple no consensus has been reached on a gold standard chronic conditions, but guidelines for the management measure of comorbidity within the context of cancer of cancer may need to consider some prominent comor- [54]. Many of the approaches provide a summary meas- bid conditions. Insight into patterns of cancer comorbid- ure of the patient’s comorbid conditions and the severity ity informs further research into the influence of of these conditions. However, the prognostic impact of comorbidity - particularly the influence of specific co- comorbidity can depend on the type and stage of the morbid conditions - on outcomes following cancer diag- cancer [55]. In addition, the presence of comorbidity - nosis, including socio-economic inequalities in receipt of particularly certain comorbid conditions - adds com- treatment and short-term mortality. plexity to the provision of treatment for cancer. When investigating the relationship between comorbidity and cancer outcomes, a more granular approach investigat- Supplementary information ing specific comorbid conditions in turn, rather than Supplementary information accompanies this paper at https://doi.org/10. 1186/s12885-019-6472-9. using a summary measure of comorbidity, could be more appropriate and insightful. Additional file 1. Definition of the fourteen conditions, according to We acknowledge potential limitations in this study. ICD-10 code classification. Table of the fourteen conditions and the ICD- We capture comorbidity information based on diag- 10 code groupings used to define them. noses of health conditions recorded during hospital Additional file 2. Probability (%) of condition present as single or multiple admission(s) prior to cancer diagnosis, and are there- comorbidity, by deprivation group (lung cancer). Additional results in complement to those presented in Fig. 3: graphs representing the probability fore reliant on patients requiring hospital-based med- of having any of nine comorbid conditions in lung cancer patients. ical attention for their health condition(s) in order to Fowler et al. BMC Cancer (2020) 20:2 Page 14 of 15 Received: 8 May 2019 Accepted: 17 December 2019 Additional file 3. Probability (%) of condition present as single or multiple comorbidity, by deprivation group (rectal cancer). Additional results in complement to those presented in Fig. 3: graphs representing the probability of having any of nine comorbid conditions in rectal References cancer patients. 1. Porta MS, Greenland S, Last JM. A dictionary of epidemiology: Oxford University press; 2014. Available from: http://www.oxfordreference.com/ view/10.1093/acref/9780195314496.001.0001/acref-9780195314496-e-352 Abbreviations ?rskey=TmAWeL&result=351. Accessed 31 Oct 2018. CCI: Charlson Comorbidity Index; CHF: Congestive Heart Failure; 2. Gumey J, Sarfati D, Stanley J. The impact of patient comorbidity on cancer CI: Confidence Interval; COPD: Chronic Obstructive Pulmonary Disease; stage at diagnosis. BJC. 2015;113:1375–80. CVD: Cerebrovascular Disease; EPV: Events Per Variable; HES: Hospital Episode 3. Sarfati D, Koczwara B, Jackson C. The impact of comorbidity on cancer and Statistics; HL: Hodgkin lymphoma; ICD: International Classification of Diseases its treatment. CA Cancer J Clin. 2016;66(4):337–50. and Related Health Problems; MI: Myocardial infarction; OR: Odds Ratio; 4. Salika T, Lyratzopoulos G, Whitaker KL, Waller J, Renzi C. Do comorbidities PVD: Peripheral Vascular Disease influence help-seeking for cancer alarm symptoms? A population-based survey in England. J Public Health. 2018;40(2):340–9. 5. Tinetti ME, Fried TR, Boyd CM. Designing health care for the most common Acknowledgements chronic condition--multimorbidity. JAMA. 2012;307(23):2493–4. We would like to thank Adrian Turculet, Data Manager of the LSHTM Cancer 6. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying Survival Group, for his support and assistance with the data linkage. prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. Authors’ contributions 7. Connolly V, Unwin N, Sherriff P, Bilous R, Kelly W. Diabetes prevalence and CM, MALF and HF contributed to the conception of the study. BR, HF, AB socioeconomic status: a population based study showing increased and LE designed the study. AB and ENN provided advice on statistical prevalence of type 2 diabetes mellitus in deprived areas. J Epidemiol methods. HF conducted the analyses of the data and prepared the draft Community Health. 2000;54(3):173–7. of the manuscript, tables and figures. BR, AB, and LE supervised the study 8. Sundquist K, Malmstrom M, Johansson S. Neighbourhood deprivation and and provided comments on the manuscript draft. CM, MALF, NN incidence of coronary heart disease: a multilevel study of 2.6 million women and DS provided comments on the final draft of the manuscript. All and men in Sweden. J Epidemiol Community Health. 2004;58(1):71–7. authors read and approved the final manuscript. 9. Hossain MP, Palmer D, Goyder E, El Nahas AM. Social deprivation and prevalence of chronic kidney disease in the UK: workload implications for primary care. QJM. 2012;105(2):167–75. Funding 10. Fano V, Pezzotti P, Gnavi R, Bontempi K, Miceli M, Pagnozzi E, et al. The role This research was funded by Cancer Research UK grant number C7923/ of socio-economic factors on prevalence and health outcomes of persons A18525. The authors declare no support from any organisations for the with diabetes in Rome, Italy. Eur J Public Health. 2013;23(6):991–7. submitted work. The design of the study, the analyses and the writing of the 11. Macleod U, Mitchell E, Black M, Spence G. Comorbidity and socioeconomic manuscript were solely the responsibility of the authors. The findings and deprivation: an observational study of the prevalence of comorbidity in conclusions in this manuscript are those of the authors and do not general practice. Eur J Gen Pract. 2009;10(1):3. necessarily represent the views of Cancer Research UK. 12. McLean G, Gunn J, Wyke S, Guthrie B, Watt G, Blane D, et al. The influence of socioeconomic deprivation on multimorbidity at different ages: a cross- Availability of data and materials sectional study. Br J Gen Pract. 2014;64(624):8. The data that support the findings of this study are available via application 13. Moffat K, Mercer SW. Challenges of managing people with multimorbidity to the Public Health England Office for Data Release, but restrictions apply to in today's healthcare systems. BMC Fam Pract. 2015;16:129. the availability of these data. 14. Department for Communities and Local Government. The English Indices of Deprivation 2010. London: 2011. Available from: https://www.gov.uk/ government/statistics/english-indices-of-deprivation-2010 Ethics approval and consent to participate 15. Health and Social Care Information Centre. Hospital Episode Statistics 2013. We obtained the statutory approvals required for this research from the Available from: http://www.hscic.gov.uk/hes. Accessed 18 Nov 2017. Confidentiality Advisory Group (CAG) of the Health Research Authority (HRA): 16. World Health Organisation. International statistical classification of diseases and PIAG 1–05(c) 2007. related health problems. Tenth revision. Geneva: World Health Organization; 1994. Ethical approval was obtained from the Research Ethics Committee (REC) of 17. Quan HD, Li B, Couris CM, Fushimi K, Graham P, Hider P, et al. Updating and the Health Research Authority (HRA): 07/MRE01/52. validating the Charlson comorbidity index and score for risk adjustment in This work uses data provided by patients and collected by the National hospital discharge abstracts using data from 6 countries. Am J Epidemiol. Health Service as part of their care and support. We used anonymised 2011;173(6):676–82. National Cancer Registry and Hospital Episode Statistics data. No consent to 18. Maringe C, Fowler H, Rachet B, Luque-Fernández MA. Reproducibility, reliability participate was sought from patients. and validity of population-based administrative health data for the assessment of cancer non-related comorbidities. PLoS One. 2017;12(3):e0172814 Consent for publication 19. Office for National Statistics. 2011 Census: Population estimates by single Not applicable. year of age and sex for local authorities in the United Kingdom 2013. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/ Competing interests populationandmigration/populationestimates/datasets/2011 The authors declare that they have no competing interests. censuspopulationestimatesbysingleyearofageand sexforlocalauthoritiesintheunitedkingdom. Accessed 01 Aug 2018. Author details 20. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study Cancer Survival Group, Department of Non-Communicable Disease of the number of events per variable in logistic regression analysis. J Clin Epidemiology, London School of Hygiene & Tropical Medicine, Keppel Street, Epidemiol. 1996;49(12):1373–9. London WC1E 7HT, UK. Biomedical Research Institute of Granada, 21. StataCorp. STATA statistical software. 15th ed. College Station: Stata Non-Communicable and Cancer Epidemiology Group, University of Granada, Corporation; 2017. Granada, Spain. Department of Non-Communicable Disease Epidemiology, 22. Schottenfeld D, Fraumeni JF. Cancer epidemiology and prevention. 3rd ed. London School of Hygiene & Tropical Medicine, London, UK. UCL Oxford: Oxford University Press; 2006. Respiratory, University College London, London, UK. Department of Thoracic 23. Tindle HA, Stevenson Duncan M, Greevy RA, Vasan RS, Kundu S, Massion PP, Medicine, University College London Hospital, London, UK. Department of et al. Lifetime Smoking History and Risk of Lung Cancer: Results From the Public Health, University of Otago, Wellington, New Zealand. Framingham Heart Study. J Natl Cancer Inst. 2018;110(11):1201–7. Fowler et al. BMC Cancer (2020) 20:2 Page 15 of 15 24. Danaei G, Vander Hoorn S, Lopez AD, Murray CJ, Ezzati M. Causes of cancer 48. Bradley CJ, Dahman B, Anscher M. Prostate cancer treatment and survival: in the world: comparative risk assessment of nine behavioural and evidence for men with prevalent comorbid conditions. Med Care. 2014; environmental risk factors. Lancet. 2005;366(9499):1784–93. 52(6):482–9. 25. Haggar FA, Boushey RP. Colorectal cancer epidemiology: incidence, 49. Sarfati D, Hill S, Blakely T, Robson B, Purdie G, Dennett E, et al. The effect of mortality, survival, and risk factors. Clin Colon Rectal Surg. 2009;22(4):191–7. comorbidity on the use of adjuvant chemotherapy and survival from colon cancer: a retrospective cohort study. BMC Cancer. 2009;9:116. 26. Devereux G. ABC of chronic obstructive pulmonary disease - definition, 50. Aapro MS, Kohne CH, Cohen HJ, Extermann M. Never too old? Age should epidemiology, and risk factors. BMJ. 2006;332(7550):1142–4. not be a barrier to enrollment in cancer clinical trials. Oncologist. 2005;10(3): 27. Buist AS, McBurnie MA, Vollmer WM, Gillespie S, Burney P, Mannino DM, 198–204. et al. International variation in the prevalence of COPD (the BOLD study): a 51. Akinyemiju T, Meng Q, Vin-Raviv N. Race/ethnicity and socio-economic population-based prevalence study. Lancet. 2007;370(9589):741–50. differences in colorectal cancer surgery outcomes: analysis of the 28. Laniado-Laborín R. Smoking and chronic obstructive pulmonary disease nationwide inpatient sample. BMC Cancer. 2016;16:715. (COPD). Parallel epidemics of the 21(st) century. Int J Environ Res Public 52. Forrest LF, Sowden S, Rubin G, White M, Adams J. Socio-economic Health. 2009;6(1):209–24. inequalities in stage at diagnosis, and in time intervals on the lung cancer 29. Hu FB, Manson JE, Stampfer MJ, Colditz G, Liu S, Solomon CG, et al. Diet, pathway from first symptom to treatment: systematic review and meta- lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med. analysis. Thorax. 2017;72(5):430–6. 2001;345(11):790–7. 53. Exarchakou A, Rachet B, Belot A, Maringe C, Coleman MP. Impact of national 30. Wannamethee SG, Shaper AG, Perry IJ. Smoking as a modifiable risk factor cancer policies on cancer survival trends and socioeconomic inequalities in for type 2 diabetes in middle-aged men. Diabetes Care. 2001;24(9):1590–5. England, 1996-2013: population based study. BMJ. 2018;360:k764. 31. Hiscock R, Bauld L, Amos A, Platt S. Smoking and socioeconomic status in 54. Sarfati D. Review of methods used to measure comorbidity in cancer England: the rise of the never smoker and the disadvantaged smoker. J populations: no gold standard exists. J Clin Epidemiol. 2012;65(9):924–33. Public Health. 2012;34(3):390–6. 55. Read WL, Tierney RM, Page NC, Costas I, Govindan R, Spitznagel EL, et al. 32. Sarfati D, Gurney J, Lim BT, Bagheri N, Simpson A, Koea J, et al. Identifying Differential prognostic impact of comorbidity. J Clin Oncol. 2004;22(15): important comorbidity among cancer populations using administrative data: 3099–103. prevalence and impact on survival. Asia Pac J Clin Oncol. 2016;12(1):e47–56. 33. Luque-Fernandez MA, Redondo Sánchez D, Rodriguez-Barranco M, Carmona- Garcia MC, Marcos-Gragera R, Sanchez-Perez MJ. Multimorbidity by patient and Publisher’sNote tumor factors and time-to-surgery among colorectal cancer patients in Spain: a Springer Nature remains neutral with regard to jurisdictional claims in population-based study. Clinical Epidemiology. 2019 [in press]. published maps and institutional affiliations. 34. Loudon I. The principle of referral: the gatekeeping role of the GP. Br J Gen Pract. 2008;58(547):128–30 discussion 30. 35. Nicholson E, Cummings M, Cranston I, Meeking D, Kar P. The super six model of care: five years on. Diab Prim Care. 2016;18:6. 36. Saxena S, George J, Barber J, Fitzpatrick J, Majeed A. Association of population and practice factors with potentially avoidable admission rates for chronic diseases in London: cross sectional analysis. J R Soc Med. 2006; 99(2):81–9. 37. Fortin M, Bravo G, Hudon C, Vanasse A, Lapointe L. Prevalence of multimorbidity among adults seen in family practice. Ann Fam Med. 2005; 3(3):223–8. 38. Salisbury C, Johnson L, Purdy S, Valderas JM, Montgomery AA. Epidemiology and impact of multimorbidity in primary care: a retrospective cohort study. Br J Gen Pract. 2011;61(582):e12–21. 39. Schäfer I, Hansen H, Schön G, Höfels S, Altiner A, Dahlhaus A, et al. The influence of age, gender and socio-economic status on multimorbidity patterns in primary care. first results from the multicare cohort study. BMC Health Serv Res. 2012;12(1):89. 40. Boyd CM, Fortin M. Future of multimorbidity research: how should understanding of multimorbidity inform health system design? Public Health Rev. 2010;32(2):451–74. 41. Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study. Lancet. 2012;380(9836):37–43. 42. Hughes LD, McMurdo MET, Guthrie B. Guidelines for people not for diseases: the challenges of applying UK clinical guidelines to people with multimorbidity. Age Ageing. 2013;42(1):62–9. 43. Glynn LG, Valderas JM, Healy P, Burke E, Newell J, Gillespie P, et al. The prevalence of multimorbidity in primary care and its effect on health care utilization and cost. Fam Pract. 2011;28(5):516–23. 44. Bahler C, Huber CA, Brungger B, Reich O. Multimorbidity, health care utilization and costs in an elderly community-dwelling population: a claims data based observational study. BMC Health Serv Res. 2015;15:23. 45. Cortaredona S, Ventelou B. The extra cost of comorbidity: multiple illnesses and the economic burden of non-communicable diseases. BMC Med. 2017;15(1):216. 46. Belot A, Fowler H, Njagi EN, Luque-Fernandez MA, Maringe C, Magadi W, et al. Association between age, deprivation and specific comorbid conditions and the receipt of major surgery in patients with non-small cell lung cancer in England: a population-based study. Thorax. 2019;74(1):51–9. 47. Gross CP, McAvay GJ, Guo Z, Tinetti ME. The impact of chronic illnesses on the use and effectiveness of adjuvant chemotherapy for colon cancer. Cancer. 2007;109(12):2410–9.

Journal

BMC CancerSpringer Journals

Published: Jan 28, 2020

There are no references for this article.