Abstract

Background:

Cancer incidence, mortality and survival statistics for the UK are routinely available; however, data on prevalence, which is generally regarded as an important measure for health planning and resource allocation, are relatively scarce.

Materials and methods:

Eight cancer registries in the UK, covering more than half the population, provided data based on >1.5 million cases of cancer. Total prevalence was calculated using methods developed for the EUROPREVAL study, based on modelling incidence and survival trends. The prevalence of cancers of the stomach, colon, rectum, lung, breast (in females), cervix uteri, corpus uteri and prostate, melanoma of skin, Hodgkin’s disease, leukaemia and all malignant neoplasms combined, was estimated for the UK for the end of 1992.

Results:

Overall, ∼1.5% of males and 2.5% of females in the UK population at the end of 1992 were living with a diagnosis of cancer. These proportions increased steeply with age, with ∼7.5% (7.3% and 7.8%, in males and females, respectively) of people aged ≥65 years living with a diagnosis of cancer. Of the individual cancers, by far the highest prevalence (almost 1%) was seen for breast cancer in females; more than one in three of all living female cancer patients had been diagnosed with breast cancer. For males, around half of prevalent cases had been diagnosed >5 years previously and 30% >10 years previously; for females, these figures were both higher, at ∼60% and 40%, respectively.

Conclusions:

The estimates of prevalence presented here comprise: recently diagnosed patients in need of treatment and monitoring; long-term survivors, some of whom will nevertheless eventually die from the cancer, while others may be cured of the disease; and patients in the terminal phase who are dying from the cancer. Further work should attempt to identify the proportions of patients in the different phases of care in order to optimise the use of prevalence estimates in health care planning.

Received 26 August 2002; revised 25 November 2002; accepted 13 January 2003

Introduction

The prevalence of cancer in a population—the number or proportion of people alive at a specific date who have been diagnosed with cancer prior to that date—is generally regarded as an important measure for health planning and resource allocation purposes. While reliable cancer incidence, mortality and survival statistics for the UK are now routinely available from cancer registries, the Office for National Statistics (ONS) and the Scottish Cancer Intelligence Unit [13], prevalence data are relatively scarce. This is largely a result of methodological difficulties involved in the estimation of prevalence. To provide accurate figures, it is necessary to have either incidence data for very long time periods together with reliable follow-up for death, or incidence data combined with good estimates of survival for the specific cancer sites (or types).

Three- and ten-year cancer prevalence figures have recently been published for England and Wales for patients alive in January 1993 and diagnosed in 1990–1992 and 1983–1992, respectively [1], using incidence data and follow-up for death or emigration. This is the directly observed 3- or 10-year prevalence. Total prevalence is all those cases alive at a given date, i.e. those directly observed by a cancer registry plus those that were diagnosed before the registry started. The methodology developed by the collaborative EUROPREVAL study [4] provided an opportunity to estimate the total UK prevalence to the same follow-up date (the end of 1992) for 11 sites of cancer and for all malignant cancers combined. This enabled the estimation of the total burden of cancer within the UK by sex, age group and time since diagnosis that is presented in this paper. The EUROPREVAL estimates were validated by comparing them with the directly observed figures for England and Wales.

Materials and methods

Data

Eight cancer registries in the UK (seven regional registries from England together with the Scottish registry) provided the necessary incidence and survival data needed to estimate prevalence. The analysis of cancer prevalence in the UK was based on >1.5 million cases of cancer that occurred in the population of >30 million people in the areas covered by the eight registries (Table 1). All the registries provided data for which a high level of completeness could be assured and follow-up for vital status was complete until 31 December 1992 (the index date for the estimation of prevalence). The cases shown in Table 1 includes patients who died before the index date (and are, therefore, not included in the prevalence estimate).

Cases notified only through death certification and those diagnosed at autopsy were not included in the analyses. When more than one cancer was diagnosed in a patient, only the first diagnosis was included. These and further exclusions are discussed elsewhere [5]. For cases with missing data (on average, no more than 1% of the total number of cases) some imputation procedures were adopted [5].

The data from the seven geographically widespread registries in England related to 53% of the English population; this population had a very similar age distribution to that of England as a whole. The Scottish Cancer Registry covered the whole of the Scottish population. In order to obtain UK figures, we assumed first that the cancer prevalence in the English registries covered by EUROPREVAL was representative of England and Wales as a whole. For Northern Ireland, the prevalence estimates of cancer in Scotland were applied to the population estimates for Northern Ireland, which contains only ∼3% of the total UK population. The data were analysed in three age groups (0–44, 45–64 and ≥65 years) to enable adjustment of the overall prevalence estimates for any differences in age structure between populations.

Calculation of observed prevalence

Observed prevalence has been computed on the basis of the longest period of data available from each single registry, by means of the counting method implemented by PREVAL software [6]. Observed prevalence was then corrected to take account of the (small) contribution of cases lost to follow-up, by assigning to them a survival probability equal to that of patients successfully followed up of the same sex and with similar age and period of diagnosis [5, 6].

Calculation of total prevalence

Total prevalence figures for each cancer site were derived by applying an estimated correction factor, the completeness index (R) [7], defined as the proportion of the total prevalence represented by the observed prevalence. This varies according to the length of the registration period and the characteristics of the cancer being considered (mainly incidence and survival trends). Total prevalence derived by the completeness index method was recently validated using data from the Connecticut Cancer Registry for a group of different cancer sites [8].

The calculation of the completeness index is based on modelling incidence and survival trends with simple parametric functions [4, 9]. A single cancer-specific incidence trend was assumed for fitting data from the UK registries, while for the survival models, registry-specific relative risks were estimated. The periods of time covered by the EUROCARE data, used for estimation of survival trends, were not adequate to correctly estimate the required incidence and survival trends by age for cancer of the cervix uteri and for Hodgkin’s disease [4]; consequently, R values and hence the total prevalence could not be produced for these two malignancies, although the method allows an adjustment of the observed prevalence that furnished estimates of the 15-year prevalence.

The completeness indices that were used in the subsequent calculations of prevalence are shown in Table 2 for the different populations and types of cancer. They were generally higher for males than for females, e.g. for all cancers the indices were ∼0.8 and ∼0.7, respectively.

Validation

The cancer registration system in England and Wales incorporates two methods (in addition to information obtained locally by the regional cancer registries) by which the National Cancer Intelligence Centre (NCIC) at the ONS notifies deaths (or emigration) to the registries. Fuller details of the system have been published elsewhere [1]. Direct estimates of the prevalence of all malignant cancers (excluding non-melanoma skin cancer), and of 20 major cancers individually, were made at the NCIC by checking the vital status of all cases of cancer diagnosed in residents of England and Wales from 1971 (when the current system of follow-up began) to the end of 1992. For those patients who were not lost to follow up, the percentage who were still alive on 1st January 1993 was applied to the total number of cases to give the estimated number of prevalent cases.

Results

Overall, ∼1.5% of males and ∼2.5% of females in the UK population at the end of 1992 were living with a diagnosis of cancer. The estimated total prevalence rates per 100 000 of the population by cancer site, sex and age group are shown in Table 3. Prevalence was higher in Scotland than in England, in both sexes, for cancers of the colon and lung and for melanoma of the skin. Prevalence was higher in England for cancers of the rectum and prostate in males and cancers of the breast and uterus in females.

Cancer prevalence increased steeply with age, reaching values of 7.3% and 7.8% in males and females, respectively, in the ≥65 years age group. In this group, the most prevalent cancers were those of the prostate, lung, colon and rectum in males, and of the breast, colon and uterus in females. The pattern was similar among the middle aged population (45–64 years), although the absolute prevalence estimates were substantially lower, especially for prostate cancer in males. For female breast cancer in this age group, prevalence estimates were lower than for patients aged ≥65 years, but not as markedly as those for most other specific sites. The age related impact of breast cancer prevalence affected the corresponding all cancer prevalence estimates for ages 45–64 years, which were 1.7% in males and 3.8% in females. Prevalence for cervix cancer in females and for Hodgkin’s disease in males reached a maximum in this age group. In young persons (aged <45 years), the pattern of prevalence was somewhat different from that seen at older ages in males, being highest for haematological cancers and melanoma of skin. Although prevalence estimates for these cancers were also relatively high in young females, the highest estimates, as in middle aged females, were for breast and cervix cancers. The prevalence of Hodgkin’s disease in females reached a maximum in the youngest age group. Overall, the prevalence of breast cancer in females approached 1%, over four times the rate for prostate cancer, the most prevalent cancer in males.

The numbers of prevalent cases of cancers of stomach, colon, rectum, lung, uterus and prostate were all dominated by those in the oldest age group (≥65 years), >70% of cases being in this category (Table 4). This contrasts with the prevalence of melanoma, breast and cervix cancers that were more influenced by cases in middle age (45–64 years).

In males, about half the prevalent cases had been diagnosed >5 years before the index date (Table 5). In females, the corresponding figure was higher, ∼60%. For the most fatal cancers, i.e. lung and stomach cancer, high proportions of the prevalent cases (∼20–30%) were diagnosed within 1 year of the index date. A large proportion (∼70%) of prevalent cases of prostate cancer, which predominantly occurs in the elderly, was diagnosed within 5 years of the index date. Around 30% of male and 40% of female prevalent cancer cases had been diagnosed >10 years before the index date. For males, the proportion of these ‘long-term’ survivors was highest for lung and stomach cancer (38% and 32%, respectively), reflecting the previously high and decreasing incidence rates of these two cancers over time, whereas for females, the proportion was highest for uterus cancer and melanoma of skin (48% and 39%, respectively).

Table 6 shows a comparison of the total prevalence for all cancers and specific cancer sites at the end of 1992, of the EUROPREVAL estimates for England with the empirical ONS figures based on all cancer registrations in England and Wales made since 1971. With the exception of those for cervix cancer, where the EUROPREVAL estimates are for 15-year prevalence, the ONS rates were close to the estimates from EUROPREVAL; for leukaemia, in both sexes, the empirical figures were slightly higher. For all cancers, the direct results from ONS were reasonably close to the EUROPREVAL estimate in both males (1.45% compared with 1.50%) and females (2.3% compared with 2.5%).

Discussion

Approximately 1.2 million people, 1.5% of males and 2.5% of females, in the UK population at the end of 1992 were living with a diagnosis of cancer. As would be anticipated, these proportions were strongly age dependent, especially in males, and increased to 7–8% in those aged ≥65 years. It is probable that the prevalence has increased since 1992 for a combination of reasons. Incidence rates for some major cancers with poor survival have declined (male lung cancer and stomach cancer), while rates for those with a relatively good prognosis have increased (breast and prostate cancer) [1, 3] and, for many sites, survival rates have improved, particularly breast and colorectal cancers, which have high incidence rates [1012]. Even if age specific prevalence rates were not higher now, the gradual ageing of the population would have increased the absolute numbers of diagnosed cancer patients and hence also the numbers of prevalent cases.

Breast cancer dominates the pattern of cancer prevalence for females because it is the most commonly diagnosed tumour and, even before the introduction of breast screening in the UK in 1988, it had a five-year relative survival of 65% [10]. More than one in three of all living female cancer patients had been diagnosed with breast cancer. Although lung cancer has long been the most common cancer in UK males, the poor prognosis means that it makes only a small contribution (<10%) to overall cancer prevalence in males. No single site of cancer dominates the male prevalence pattern, although prostate cancer, which is the most common site overall, has by far the highest rate in the >65 years age group.

Approximately one half of all living male cancer patients had survived for >5 years and one-third had survived for >10 years. For females, these proportions were higher at ∼60% and ∼40%. The problems associated with cancer management are usually considered for the years immediately following diagnosis, whereas the majority of patients have been living with their disease for at least 5 years. Not all of these survivors are diagnosed with cancers that have good prognosis; thus, for both lung and stomach cancer, in both sexes, >30% of the prevalent cases had survived for >10 years. This apparent anomaly is, at least partly, explained by the relatively very high incidence rates for these cancers in the 1960s and 1970s such that a small fraction of long-term survivors can make a substantial contribution to current prevalence.

The annual number of new cancer cases diagnosed in the UK (excluding non-melanoma skin cancer) was ∼252 000 in 1992 and this increased to 263 000 cases by 1998. The corresponding numbers of cancer deaths were 164 000 and 154 000, respectively. The estimated total number of prevalent cases on 31 December 1992 was 1 178 000 (applying the all cancers, all ages prevalence proportions per 100 000, provided in Table 3, of 1506 for males and 2534 for females to the 1992 UK population of 28.362 million males and 29.645 million females). The ratio of prevalence to incidence was thus ∼5-fold, and the ratio of incidence to mortality ∼1.5-fold in 1992. With generally rising incidence, gradual improvement in cancer care and increasing survival of cancer patients, prevalence will gradually increase; mortality from cancer will continue to decrease if the improvements in survival more than counterbalance the increases in incidence.

The large and increasing size of the population of living cancer patients emphasises the magnitude of the impact of cancer on peoples’ lives, either as patients themselves or as relatives of cancer patients, but the relevance of prevalence figures to the planning of cancer care is less obvious. The number of new cases in a year is an indicator of the need for resources for diagnostic investigations and first-line treatment. Similarly, the annual number of cancer deaths is an indicator of the need for resources for palliation and terminal care. Prevalence is a composite measure of a highly heterogeneous population comprised of: recently diagnosed patients in need of treatment and monitoring; long-term survivors, some of whom will nevertheless eventually die from the cancer, while others may be cured of the disease; and patients in the terminal phase who are dying from the cancer. Direct estimates of the average costs in each phase of cancer care have been obtained by linking incidence registry data with administrative databases [1315]. Treatment costs are strongly related to age, cancer type and stage of disease at diagnosis [13]. The costs of surveillance and monitoring activities needed in the intermediate phase are much lower than costs for first and terminal treatment but, as they may continue for a long period of time, their impact on the total cost of cancer care is not negligible.

Further work on the development of the definition of prevalence should attempt to identify subgroups of patients in the different phases of care. The disaggregation of prevalence by time since diagnosis (as in this paper) is a first step in this direction. Prevalence by stage of disease at diagnosis would be an additional and very informative indicator of the need for care. Recent developments in cancer survival analysis [16] can, for some types of cancer, distinguish long-term survivors who are cured and those who will die from their cancer. This may lead to more detailed assessments of prevalence in the future.

Acknowledgements

This research was supported by the EUROPREVAL Biomed-2 Programme, Contract No. BH4 983899. Members of the UK EUROPREVAL Working Group. Scotland: R. Black, V. Harris, D. Stockton (Scottish Cancer Intelligence Unit). England: T.W. Davies, S. Godward (East Anglian Cancer Registry); M.P. Coleman, S. Harris (London School of Hygiene and Tropical Medicine); E.M.I. Williams (Merseyside and Cheshire Cancer Registry); D. Forman, R. Iddenden (Northern and Yorkshire Cancer Registry and Information Service and University of Leeds); M.J. Quinn, P.J. Babb (Office for National Statistics); M. Roche (Oxford Cancer Intelligence Unit); J. Smith (South and West Cancer Intelligence Unit); H. Møller (Thames Cancer Registry); P. Silcocks (Trent Cancer Registry); G. Lawrence, K. Hemmings (West Midlands Cancer Intelligence Unit). Italy: R. Capocaccia (Project Leader), I. Corazzaiari, R. De Angelis, S. Francisci, S. Hartley. F. Valente, A. Verdecchia, A. Zappone (National Institute of Health, Rome); F. Berrino, G. Gatta, A. Micheli, E. Mugno, M. Sant (National Institute for the Study and Cure of Tumours, Milan).

+

Correspondence to: Professor D. Forman, Arthington House, Cookridge Hospital, Leeds LS16 6QB, UK. Tel: +44-113-392-4309; Fax: +44-113-392-4164; E-mail: d.forman@leeds.ac.uk

Table 1.

Population covered, period of diagnosis and numbers of cases by cancer, and percentage of total cancers lost to follow-up by cancer registry

Registry area Population (1000s) Period All cancersa Lost to follow-up (%)  Cancer site and ICD9 code 
    Stomach 151 Colon 153 Rectum 154 Lung 162 Melanoma 172 Breast (female) 174 Cervix 180 Uterus 182 Prostate 185 Hodgkin’s 201 Leukaemia 204–208 
England                 
East Anglia 2089 1979–1992 95 452 0.16  4976 8807 5602 16 531 2052 14 262 1939 2332 7540 699 2377 
Merseyside and  Cheshire 2412 1985–1992 71 560 0.00  4286 6037 4024 15 208 1014 10 009 1961 1416 4164 413 1471 
Oxford 2582 1979–1992 119 304 0.00  6071 10 064 5795 20 231 2361 17 362 2307 2806 7488 944 2982 
South Thames 6756 1978–1992 313 481 0.47  15 344 25 409 15 663 57 732 5465 45 652 5918 7085 19 035 2364 7163 
Wessex 2993 1979–1992 163 325 0.00  7365 15 220 7873 24 389 3830 22 766 3286 3318 11 277 1040 4433 
West Midlands 5278 1978–1992 272 119 0.02  17 355 22 884 15 574 49 479 3957 37 528 6340 5728 15 110 1897 6431 
Yorkshire 3698 1978–1992 197 615 0.01  11 683 15 576 11 112 37 435 2775 25 257 5152 3487 11 241 1377 4667 
England sub-total 25 808  1 232 856   67 080 103 997 65 643 221 005 21 454 172 836 26 903 26 172 75 855 8734 29 524 
Scotland 5111 1978–1992 302 159 0.02  16 769 26 032 13 341 63 122 5575 37 120 6191 4410 15 432 2045 6660 
Total 30 919  1 535 015   83 849 130 029 78 984 284 127 27 029 209 956 33 094 30 582 91 287 10 779 36 184 
Registry area Population (1000s) Period All cancersa Lost to follow-up (%)  Cancer site and ICD9 code 
    Stomach 151 Colon 153 Rectum 154 Lung 162 Melanoma 172 Breast (female) 174 Cervix 180 Uterus 182 Prostate 185 Hodgkin’s 201 Leukaemia 204–208 
England                 
East Anglia 2089 1979–1992 95 452 0.16  4976 8807 5602 16 531 2052 14 262 1939 2332 7540 699 2377 
Merseyside and  Cheshire 2412 1985–1992 71 560 0.00  4286 6037 4024 15 208 1014 10 009 1961 1416 4164 413 1471 
Oxford 2582 1979–1992 119 304 0.00  6071 10 064 5795 20 231 2361 17 362 2307 2806 7488 944 2982 
South Thames 6756 1978–1992 313 481 0.47  15 344 25 409 15 663 57 732 5465 45 652 5918 7085 19 035 2364 7163 
Wessex 2993 1979–1992 163 325 0.00  7365 15 220 7873 24 389 3830 22 766 3286 3318 11 277 1040 4433 
West Midlands 5278 1978–1992 272 119 0.02  17 355 22 884 15 574 49 479 3957 37 528 6340 5728 15 110 1897 6431 
Yorkshire 3698 1978–1992 197 615 0.01  11 683 15 576 11 112 37 435 2775 25 257 5152 3487 11 241 1377 4667 
England sub-total 25 808  1 232 856   67 080 103 997 65 643 221 005 21 454 172 836 26 903 26 172 75 855 8734 29 524 
Scotland 5111 1978–1992 302 159 0.02  16 769 26 032 13 341 63 122 5575 37 120 6191 4410 15 432 2045 6660 
Total 30 919  1 535 015   83 849 130 029 78 984 284 127 27 029 209 956 33 094 30 582 91 287 10 779 36 184 

aAll malignancies, excluding non-melanoma skin cancer (ICD9 140–208 excluding 173).

Table 2.

Observation time and completeness index by cancer site (with ICD9 code), sex and cancer registry, all ages

Registry Observation time (years) All cancersa Stomach 151 Colon 153 Rectum 154 Lung 162 Melanoma 172 Prostate 185 Breast 174 Cervix 180b Uterus 182 Hodgkin’s 201b Leukaemia 204–208 
Males              
England              
East Anglia 14 0.80 0.77 0.85 0.85 0.70 0.89 0.95    0.94 0.92 
Merseyside and  Cheshire  8 0.63 0.63 0.68 0.69 0.58 0.71 0.85    0.58 0.75 
Oxford 14 0.79 0.77 0.85 0.86 0.71 0.89 0.96    0.94 0.88 
South Thames 15 0.81 0.79 0.87 0.87 0.73 0.91 0.96    1.00 0.92 
Wessex 14 0.80 0.77 0.85 0.85 0.70 0.89 0.95    0.95 0.90 
West Midlands 15 0.82 0.80 0.87 0.87 0.73 0.91 0.97    1.00 0.92 
Yorkshire 15 0.82 0.80 0.87 0.87 0.73 0.91 0.97    1.00 0.92 
Scotland 15 0.82 0.80 0.87 0.87 0.73 0.91 0.97    1.00 0.92 
Females              
England              
East Anglia 14 0.69 0.75 0.80 0.76 0.76 0.72  0.77 0.95 0.66 0.95 0.86 
Merseyside and  Cheshire  8 0.52 0.61 0.61 0.58 0.64 0.52  0.59 0.63 0.47 0.61 0.68 
Oxford 14 0.70 0.76 0.80 0.76 0.77 0.74  0.78 0.95 0.67 0.95 0.81 
South Thames 15 0.72 0.78 0.82 0.78 0.79 0.75  0.80 1.00 0.69 1.00 0.87 
Wessex 14 0.69 0.75 0.80 0.76 0.77 0.73  0.77 0.95 0.66 0.95 0.87 
West Midlands 15 0.72 0.78 0.82 0.78 0.79 0.76  0.80 1.00 0.71 1.00 0.88 
Yorkshire 15 0.72 0.78 0.82 0.78 0.79 0.76  0.80 1.00 0.70 1.00 0.88 
Scotland 15 0.72 0.79 0.82 0.78 0.79 0.76  0.80 1.00 0.72 1.00 0.87 
Registry Observation time (years) All cancersa Stomach 151 Colon 153 Rectum 154 Lung 162 Melanoma 172 Prostate 185 Breast 174 Cervix 180b Uterus 182 Hodgkin’s 201b Leukaemia 204–208 
Males              
England              
East Anglia 14 0.80 0.77 0.85 0.85 0.70 0.89 0.95    0.94 0.92 
Merseyside and  Cheshire  8 0.63 0.63 0.68 0.69 0.58 0.71 0.85    0.58 0.75 
Oxford 14 0.79 0.77 0.85 0.86 0.71 0.89 0.96    0.94 0.88 
South Thames 15 0.81 0.79 0.87 0.87 0.73 0.91 0.96    1.00 0.92 
Wessex 14 0.80 0.77 0.85 0.85 0.70 0.89 0.95    0.95 0.90 
West Midlands 15 0.82 0.80 0.87 0.87 0.73 0.91 0.97    1.00 0.92 
Yorkshire 15 0.82 0.80 0.87 0.87 0.73 0.91 0.97    1.00 0.92 
Scotland 15 0.82 0.80 0.87 0.87 0.73 0.91 0.97    1.00 0.92 
Females              
England              
East Anglia 14 0.69 0.75 0.80 0.76 0.76 0.72  0.77 0.95 0.66 0.95 0.86 
Merseyside and  Cheshire  8 0.52 0.61 0.61 0.58 0.64 0.52  0.59 0.63 0.47 0.61 0.68 
Oxford 14 0.70 0.76 0.80 0.76 0.77 0.74  0.78 0.95 0.67 0.95 0.81 
South Thames 15 0.72 0.78 0.82 0.78 0.79 0.75  0.80 1.00 0.69 1.00 0.87 
Wessex 14 0.69 0.75 0.80 0.76 0.77 0.73  0.77 0.95 0.66 0.95 0.87 
West Midlands 15 0.72 0.78 0.82 0.78 0.79 0.76  0.80 1.00 0.71 1.00 0.88 
Yorkshire 15 0.72 0.78 0.82 0.78 0.79 0.76  0.80 1.00 0.70 1.00 0.88 
Scotland 15 0.72 0.79 0.82 0.78 0.79 0.76  0.80 1.00 0.72 1.00 0.87 

aAll malignancies, excluding non-melanoma skin cancer (ICD9 140–208 excluding 173).

b15-year prevalence completeness.

Table 3.

Total prevalence by country, age, sex and cancer site (proportions per 100 000 population)

Cancer site England, all ages Scotland, all ages  UKa 
  0–44 years 45–64 years ≥65 years All ages 
Males        
Stomach 45 46  53 251 45 
Colon 142 159  146 832 144 
Rectum 108 92  119 601 106 
Lung 136 167  153 800 139 
Melanoma 42 56  15 81 124 44 
Prostate 210 198  <1 74 1484 209 
Hodgkin’sb 28 28  27 34 21 28 
Leukaemia 38 39  22 35 124 38 
All cancersc 1496 1582  296 1671 7269 1506 
Females        
Stomach 26 31  22 117 27 
Colon 176 204  146 776 179 
Rectum 97 91  86 410 96 
Lung 55 80  69 225 58 
Melanoma 106 129  37 187 246 108 
Breast 952 911  89 1688 2840 947 
Cervixb 124 122  70 225 177 124 
Uterus 180 132  216 674 175 
Hodgkin’sb 19 20  22 18 13 19 
Leukaemia 31 32  20 25 73 31 
All cancersc 2528 2581  444 3758 7817 2534 
Cancer site England, all ages Scotland, all ages  UKa 
  0–44 years 45–64 years ≥65 years All ages 
Males        
Stomach 45 46  53 251 45 
Colon 142 159  146 832 144 
Rectum 108 92  119 601 106 
Lung 136 167  153 800 139 
Melanoma 42 56  15 81 124 44 
Prostate 210 198  <1 74 1484 209 
Hodgkin’sb 28 28  27 34 21 28 
Leukaemia 38 39  22 35 124 38 
All cancersc 1496 1582  296 1671 7269 1506 
Females        
Stomach 26 31  22 117 27 
Colon 176 204  146 776 179 
Rectum 97 91  86 410 96 
Lung 55 80  69 225 58 
Melanoma 106 129  37 187 246 108 
Breast 952 911  89 1688 2840 947 
Cervixb 124 122  70 225 177 124 
Uterus 180 132  216 674 175 
Hodgkin’sb 19 20  22 18 13 19 
Leukaemia 31 32  20 25 73 31 
All cancersc 2528 2581  444 3758 7817 2534 

aAssuming that the UK registries covered by EUROPREVAL are representative of the UK as a whole; England study population prevalence used for Wales, Scotland prevalence used for Northern Ireland.

b15-year prevalence.

cAll malignancies, excluding non-melanoma skin cancer (ICD9 140–208 excluding 173).

Table 4.

Distribution of the proportions of prevalent cases of cancer in the UKa by age group

Cancer site  Age group (years) 
  0–44 45–64 ≥65 
Stomach  1.9 22.7 75.4 
Colon  2.0 19.9 78.2 
Rectum  1.6 22.3 76.2 
Lung  1.4 24.8 73.8 
Melanoma  20.8 38.5 40.6 
Breast (female)   5.6 38.7 55.7 
Cervixb  34.1 39.4 26.5 
Uterus  1.5 26.9 71.6 
Prostate  0.1  7.8 92.1 
Hodgkin’sb  65.3 23.8 10.9 
Leukaemia  38.1 19.1 42.8 
All cancersc  11.4 29.5 59.2 
Cancer site  Age group (years) 
  0–44 45–64 ≥65 
Stomach  1.9 22.7 75.4 
Colon  2.0 19.9 78.2 
Rectum  1.6 22.3 76.2 
Lung  1.4 24.8 73.8 
Melanoma  20.8 38.5 40.6 
Breast (female)   5.6 38.7 55.7 
Cervixb  34.1 39.4 26.5 
Uterus  1.5 26.9 71.6 
Prostate  0.1  7.8 92.1 
Hodgkin’sb  65.3 23.8 10.9 
Leukaemia  38.1 19.1 42.8 
All cancersc  11.4 29.5 59.2 

aAssuming that the UK registries covered by EUROPREVAL are representative of the UK as a whole; England study population prevalence used for Wales, Scotland prevalence used for Northern Ireland.

b15-year prevalence.

cAll malignancies, excluding non-melanoma skin cancer (ICD9 140–208 excluding 173).

Table 5.

UKa cancer prevalence per 100 000 population by time since diagnosis, cancer site and sex

Cancer site 1 year  1–5 years  5–10 years  >10 years 
 Rate  Rate  Rate  Rate 
Males            
Stomach 10.4 23.0  12.4 27.5  8.0 17.7  14.3 31.8 
Colon 22.6 15.7  51.8 36.1  31.6 22.0  37.5 26.1 
Rectum 18.4 17.4  37.0 34.9  23.9 22.5  26.8 25.3 
Lung 36.1 26.0  30.7 22.1  19.6 14.1  52.8 37.9 
Melanoma 6.3 14.3  17.4 39.7  11.0 25.1  9.1 20.9 
Prostate 48.1 23.0  94.2 45.1  42.6 20.4  24.0 11.5 
Hodgkin’sb 2.5 9.1  8.6 30.9  9.2 33.1  7.5 26.8 
Leukaemia 6.6 17.4  15.8 41.4  8.0 21.1  7.6 20.1 
All cancersc 247.2 16.4  481.8 32.0  313.4 20.8  463.2 30.8 
Females            
Stomach 5.6 20.7  7.3 27.0  5.1 19.0  9.0 33.3 
Colon 24.7 13.8  56.6 31.6  40.3 22.5  57.6 32.2 
Rectum 13.0 13.5  29.0 30.1  20.2 20.9  34.2 35.5 
Lung 17.3 29.8  14.4 24.8  8.7 14.9  17.8 30.6 
Melanoma 9.4 8.7  30.6 28.2  26.3 24.3  42.1 38.8 
Breast  105.4 11.1  301.7 31.8  214.0 22.6  326.2 34.4 
Cervixb 10.9 8.8  41.7 33.7  41.6 33.6  29.5 23.8 
Uterus 12.5 7.2  39.4 22.5  39.7 22.7  83.1 47.6 
Hodgkin’sb 1.9 9.5  6.1 31.6  6.5 33.6  4.9 25.3 
Leukaemia 5.0 16.2  11.3 36.2  6.8 22.0  8.0 25.6 
All cancersc 277.0 10.9  685.5 27.0  518.7 20.5  1053.1 41.6 
Cancer site 1 year  1–5 years  5–10 years  >10 years 
 Rate  Rate  Rate  Rate 
Males            
Stomach 10.4 23.0  12.4 27.5  8.0 17.7  14.3 31.8 
Colon 22.6 15.7  51.8 36.1  31.6 22.0  37.5 26.1 
Rectum 18.4 17.4  37.0 34.9  23.9 22.5  26.8 25.3 
Lung 36.1 26.0  30.7 22.1  19.6 14.1  52.8 37.9 
Melanoma 6.3 14.3  17.4 39.7  11.0 25.1  9.1 20.9 
Prostate 48.1 23.0  94.2 45.1  42.6 20.4  24.0 11.5 
Hodgkin’sb 2.5 9.1  8.6 30.9  9.2 33.1  7.5 26.8 
Leukaemia 6.6 17.4  15.8 41.4  8.0 21.1  7.6 20.1 
All cancersc 247.2 16.4  481.8 32.0  313.4 20.8  463.2 30.8 
Females            
Stomach 5.6 20.7  7.3 27.0  5.1 19.0  9.0 33.3 
Colon 24.7 13.8  56.6 31.6  40.3 22.5  57.6 32.2 
Rectum 13.0 13.5  29.0 30.1  20.2 20.9  34.2 35.5 
Lung 17.3 29.8  14.4 24.8  8.7 14.9  17.8 30.6 
Melanoma 9.4 8.7  30.6 28.2  26.3 24.3  42.1 38.8 
Breast  105.4 11.1  301.7 31.8  214.0 22.6  326.2 34.4 
Cervixb 10.9 8.8  41.7 33.7  41.6 33.6  29.5 23.8 
Uterus 12.5 7.2  39.4 22.5  39.7 22.7  83.1 47.6 
Hodgkin’sb 1.9 9.5  6.1 31.6  6.5 33.6  4.9 25.3 
Leukaemia 5.0 16.2  11.3 36.2  6.8 22.0  8.0 25.6 
All cancersc 277.0 10.9  685.5 27.0  518.7 20.5  1053.1 41.6 

aAssuming that the UK registries covered by EUROPREVAL are representative of the UK as a whole; England study population prevalence used for Wales, Scotland prevalence used for Northern Ireland.

b15-year prevalence.

cAll malignancies, excluding non-melanoma skin cancer (ICD9 140–208 excluding 173).

Table 6.

Comparison of Office for National Statistics (ONS) directly observed rates and EUROPREVAL estimates

Cancer sitea Prevalence per 100 000  Ratio 
 ONSb EUROPREVALc   
Males     
Stomach 41 45  0.92 
Colon 137 142  0.96 
Rectum 104 108  0.97 
Lung 116 136  0.85 
Melanoma 40 42  0.95 
Prostate 204 210  0.97 
Leukaemia 40 38  1.07 
All cancersd 1450 1496  0.97 
Females     
Stomach 26 26  0.99 
Colon 173 176  0.98 
Rectum 90 97  0.93 
Lung 52 55  0.93 
Melanoma  86 106  0.81 
Breast 872 952  0.92 
Cervixe 153 124  1.24 
Uterus 148 180  0.82 
Leukaemia 33 31  1.06 
All cancersd 2284 2528  0.90 
Cancer sitea Prevalence per 100 000  Ratio 
 ONSb EUROPREVALc   
Males     
Stomach 41 45  0.92 
Colon 137 142  0.96 
Rectum 104 108  0.97 
Lung 116 136  0.85 
Melanoma 40 42  0.95 
Prostate 204 210  0.97 
Leukaemia 40 38  1.07 
All cancersd 1450 1496  0.97 
Females     
Stomach 26 26  0.99 
Colon 173 176  0.98 
Rectum 90 97  0.93 
Lung 52 55  0.93 
Melanoma  86 106  0.81 
Breast 872 952  0.92 
Cervixe 153 124  1.24 
Uterus 148 180  0.82 
Leukaemia 33 31  1.06 
All cancersd 2284 2528  0.90 

aDirect estimates were not available for Hodgkin’s disease (ICD9 201).

bBased on the number of cases registered in England and Wales, 1971–1992 (see Materials and methods).

cRates for the England study population (Table 3).

dAll malignancies, excluding non-melanoma skin cancer (ICD9 140–208 excluding 173).

e15-year prevalence.

References

1.
Quinn MJ, Babb PJ, Brock A et al. Cancer Trends in England and Wales, 1950–1999. Studies in Medical and Population Subjects No. 66. London, UK: The Stationery Office
2001
.
2.
Booth H, Cooper N, Quinn MJ (eds): Cancer Statistics: Registrations, England, 1998. Series MB1 No. 29. London, UK: Office for National Statistics
2002
.
3.
Scottish Cancer Intelligence Unit. Scottish cancer statistics on the web. Information and Statistics Division [on-line] http://www.show.scot.nhs.uk/isd/cancer/cancer.htm (18 February 2003, date last accessed).
4.
Capocaccia R, Colonna M, Corazziari I et al. Measuring cancer prevalence in Europe: the EUROPREVAL project.
Ann Oncol
 
2002
;
13
:
831
–839.
5.
Micheli A, Mugno E, Krogh V et al. Cancer prevalence in European registry areas.
Ann Oncol
 
2002
;
13
:
840
–865.
6.
Krogh V, Micheli A. Measure of cancer prevalence with a computerized program: an example of larynx cancer.
Tumori
 
1996
;
82
:
1
–4.
7.
Capocaccia R, De Angelis R. Estimating the completeness of prevalence based on cancer registry data.
Stat Med
 
1997
;
16
:
425
–440.
8.
Merril RM, Feuer EJ, Capocaccia R, Mariotto A. Cancer prevalence estimates based on tumour registry data in the SEER program.
Int J Epidemiol
 
2000
;
29
:
197
–207.
9.
Capocaccia R, Gatta G, Chessa E, Valente F. The EUROCARE 2 Study. IARC Scientific Publication No. 151. Lyon, France: International Agency for Research on Cancer
1999
;
1
–40.
10.
Coleman MP, Babb P, Damiecki P et al. Cancer survival trends in England and Wales, 1971–1995: deprivation and NHS region. Studies in Medical and Population Subjects No. 61. London, UK: Office for National Statistics
1999
.
11.
Coleman MP, Babb P, Harris S et al. Cancer survival in England and Wales, 1991–1998.
Health Stat Q
 
2000
;
6
:
54
–63.
12.
Scottish Cancer Intelligence Unit. Trends in cancer survival in Scotland 1971–1995. Edinburgh, UK: Information & Statistics Division
2000
.
13.
Taplin SH, Barlow W, Urban N et al. Stage, age, comorbidity, and direct costs of colon, prostate, and breast cancer care.
J Natl Cancer Inst
 
1995
;
87
:
417
–426.
14.
Fireman BH, Quesenberry CP, Somkin CP et al. Cost of care for cancer in a health maintenance organization.
Health Care Financ Rev
 
1997
;
18
:
51
–76.
15.
Brown ML, Riley GF, Potosky AL, Etzioni RD. Obtaining long-term disease specific costs of care: application to Medicare enrollees diagnosed with colorectal cancer.
Med Care
 
1999
;
37
:
1249
–1259.
16.
Verdecchia A, De Angelis R, Capocaccia R et al. The cure for colon cancer: results from the EUROCARE study.
Int J Cancer
 
1998
;
77
:
322
–329.

Author notes

1Northern and Yorkshire Cancer Registry and Information Service, Leeds; 2University of Leeds, Leeds; 3Scottish Cancer Intelligence Unit, Information and Statistics Division, National Health Service Scotland, Edinburgh; 4Thames Cancer Registry, King’s College, London; 5London School of Hygiene and Tropical Medicine, London; 6National Cancer Intelligence Centre, Office for National Statistics, London, UK; 7Istituto Superiore di Sanità, Rome; 8Istituto Nazionale per lo Studio e la Cura dei Tumori, Milan, Italy