The magnitude and characteristics of the population of cancer survivors: using population-based estimates of cancer prevalence to inform service planning for survivorship care

In most countries the number of people diagnosed with cancer is rising steadily. In Europe, an estimated 3.2 million cancers were diagnosed in 2006; by 2012, this had risen to 3.5 million [1, 2]. This is a result of both demographic changes (i.e. population ageing) and changes in underlying risk. At the same time, survival for many cancers is improving – by 1-2% per annum - and 5-year relative survival for all cancers combined now exceeds 50% [3, 4]. These trends mean that there are increasing numbers of “cancer survivors” (i.e. people living with and beyond a diagnosis of cancer [5]) in the population.

In the past, the life trajectory for most people diagnosed with cancer was one of inexorable decline. Nowadays, therapeutic advances, and a greater focus on addressing the side-effects and toxicities of treatment, mean that many people successfully complete primary treatment, recover substantial functional capacity, and can - potentially - resume everyday activities [6]. However, this cannot be taken to suggest that survivors simply return to “normal life”; instead they must find a “new normal” that recognizes and accommodates cancer and its consequences [7‐9]. For example, many survivors have complex health conditions which co-exist with their cancer [10], or arise as a consequence of treatment (see, for example, [11‐16]). They also have an elevated risk of developing a second cancer [16]. In addition, evidence is accumulating that many survivors experience: significant functional and psychological problems and limitations; social, sexual and relationship difficulties; and financial/economic problems due to cancer and its treatment [17]. Taken together these issues mean that survivors may have considerable ongoing needs for medical and non-medical support and care long after their initial diagnosis and treatment. The prevalence, nature and extent of these supportive care needs vary by survivors’ socio-demographic and clinical characteristics (see, for example, [18‐23]). In addition, research from several healthcare systems suggests these needs often go unmet [21].

The resource requirements for supporting cancer survivors are likely to be quite different to those necessary for treating newly diagnosed cancers [24]. Therefore, estimates of the number of people living with cancer in the population - and a description of their characteristics - could aid decision-makers, planners and service providers (both statutory and voluntary, in health and social care) in developing services, supports and other initiatives to meet survivors’ needs.

Population-based cancer registries, which record and follow-up every cancer diagnosed within a defined population, operate in many countries worldwide; in the Europe Union, 26 countries have either regional or national registries [25]. Such registries are considered an essential component of any comprehensive cancer control programme. They provide robust data on cancer incidence, survival, and mortality in the population, and the only truly valid data for: monitoring and projecting the population-level burden of cancer (to inform service planning); assessing variations in incidence (to reveal possible differences in exposure to risk factors and provide information on effectiveness of prevention strategies); and examining patterns and trends in clinical outcomes (to evaluate quality of, and equity of access to, services) [26, 27]. To date, however, registry data has not been extensively used to inform resource requirements for follow-up and after-care services. In order to illustrate how cancer registry data may be useful in survivorship care service planning, we aimed to: (i) generate population-based estimates of cancer prevalence in Ireland; and (ii) describe the socio-demographic and clinical characteristics of the population of people living with cancer. Our secondary aim was to identify subgroups of survivors who might have specific needs in terms of follow-up or after-care services.

In total, 178,813 people in Ireland were diagnosed with an invasive cancer during 1994–2011 and were still alive on 31/12/2011. 90,835 (51%) of these survivors were female and 87,978 (49%) were male. When NMSC was excluded, there were 112,610 cancer survivors at 31/12/2011, of whom 58,054 (52%) were female and 54,556 (48%) were male.

Prevalence is increasingly considered an important measure of the population-level cancer burden. Although some recent studies have estimated prevalence (see [29] and references therein), few (if any) set out explicitly to examine socio-demographic and clinical heterogeneity within the survivor population in order to inform the development of strategies for service provision around cancer follow-up and supportive care.

We estimated that, in total, 178,813 people had been diagnosed with an invasive primary cancer in Ireland during 1994–2011 and were alive at the end of 2011. Expressed crudely in relation to the 2011 population, this means that at least 3.9% of the population of Ireland are cancer survivors. In addition, it implies that each general practitioner has, on average, 60 patients who are cancer survivors. The NCR is unusual in that it aims to record all NMSC cases. Since most other registries do not, in order to permit international comparisons, we repeated our analysis excluding NMSC. When this was done, there were 112,610 survivors, representing 2.4% of the male, and 2.5% of the female, population. These figures are very close to the estimate of 18-year prevalence at the end of 2010 in Northern Ireland (2.5%) [30]. In contrast they are somewhat lower than estimates for Switzerland, Italy and the United Kingdom (UK) [31‐33], but these studies estimated complete prevalence (i.e. all survivors) whereas we considered limited duration (i.e. 17-year) prevalence because national cancer registration was not established until 1994. Therefore, the figures in this paper provide a lower bound for the total survivor population in Ireland, and thus the number of people who may require some level of access to cancer follow-up and/or after-care services.

Because prevalence is a function of incidence and survival, there are some notable differences between the cancers which rank highest when incidence or mortality are considered and those which are most prevalent. For example, testicular cancer, which is relatively uncommon but has good survival (175 new cases and 5 deaths per annum in Ireland [4]), is the 14^th most commonly-diagnosed cancer among men but ranked 7^th in terms of prevalence. In contrast, lung cancer - the 3^rd most commonly diagnosed cancer and most common cause of cancer death in both sexes [4] - ranked 8^th (females) and 9^th (males) when prevalence was considered. Similar patterns are evident in Northern Ireland [30] and the United States [34]. The implication of these observations for service providers is that the cancer-specific composition (or configuration) of services for diagnosis and treatment should be different from that for rehabilitation, follow-up and after-care services (because the composition of the types of cancers requiring diagnosis/treatment and follow-up/after-care differs).

The rationale for examining prevalence by time since diagnosis is that survivors at different “phases” of follow-up/survivorship may have different needs. For example, many of those who have survived the first year post-diagnosis may be in need of (or benefit from) rehabilitation services [35]; the higher prevalence of anxiety in those who have survived two or more years post-diagnosis than among controls [36], suggests psychological support services could be of benefit to notable proportions of intermediate-term survivors; while long-term survivors (i.e. survived ≥10 years) may be in need of services focused on the detection and management of late-effects of treatment and/or second primaries. The distribution of time since diagnosis for all invasive cancers in this analysis was almost identical to that seen in Northern Ireland [30]. The observed higher percentage of longer-term survivors among females is driven in large-part by breast cancer: almost 6,500 women had survived for ≥10 years after a breast cancer diagnosis. Moreover, two-thirds of breast cancer survivors had received radiotherapy and half had received chemotherapy. Radiotherapy to the breast and trastuzamab are associated with risk of late cardiac complications [11, 12] and those responsible for follow-up care for long-term breast cancer survivors need to be alert to these risks.

We found that 15% of survivors were younger than 50 and a further 17% were aged 50–59. The concerns, and hence burden of supportive care needs, of younger and older survivors may be quite different [23]. For example, younger survivors may be more concerned about employment and related financial matters, relationships, and fertility and sexuality issues [37‐39]. In addition, they may adapt less well to having cancer than older survivors [40]. Our findings with regard to the age distribution of different cancers suggest that follow-up and after-care services and supports for survivors of leukaemia, lymphoma and melanoma and testicular and cervical cancer should encompass these types of issues.

At the other end of the age spectrum, four in every 10 survivors were aged 70 or older. This means that at least 12.6% of this age-group in Ireland (16.1% of males; 9.9% of females) has a history of invasive cancer other than NMSC. Older survivors tend to have more comorbidities than younger survivors [40]. They also have high levels of psychological distress related to the continuing effects of cancer and its treatment, significant limitations in physical functioning, and higher rates of frailty than the general population [41‐43]. These issues are inter-related and may be exacerbated by lifestyle factors [42, 44]. Our results therefore suggest that follow-up and after-care services for survivors of bladder, colorectal and lung cancer, in particular, should be linked closely to geriatric and other non-cancer specialties, and should be broad ranging. It is noteworthy, however, that while there have been trials of alternative models of follow-up for some of these cancers [45, 46], models specifically focused on older survivors needs, or which explicitly seek to involve geriatric and other-specialties, do not appear to have been evaluated.

In some studies unmarried survivors have higher supportive care needs [23]. Thirty-two percent of our survivor population was not married at the time of diagnosis, and this figure was higher among older survivors (≥70, 35%; ≥80, 47%), providing further evidence to suggest that the elderly survivor population may have pronounced needs. In Ireland, among the elderly, living alone is a marker of poverty and experiencing multiple types of enforced deprivation [47]. Since socio-economic status influences access to cancer care [48‐51], in developing supportive care services it will be important to consider strategies and approaches for minimizing inequalities in access.

The higher proportion of female to male survivors in this study echoes findings elsewhere [30, 34] and is a function of the dominance of breast cancer among women. As in the United States and UK [33, 34], more than 40% of female survivors in Ireland had a history of breast cancer. Among breast cancer survivors unmet needs are common, are present across the survivorship continuum, and often relate to emotional or existential concerns (see, for example, [52, 53]), suggesting that after-care services for both shorter-term and longer-term breast cancer survivors should encompass these types of issues.

Just under half of survivors in Ireland were male (48%), higher than figures from Northern Ireland (43%) and the UK as a whole (41%) [30, 33]. This is due to prostate cancer incidence in Ireland, which was estimated to be the highest in Europe in 2008 [54]. This high incidence is a consequence of widespread prostate specific antigen testing in primary care [55]. Indeed, 42% of the male survivors in Ireland had been diagnosed with prostate cancer, almost identical to the US (43%) [34]. Prostate cancer treatments are associated with significant side-effects which can impact adversely on health-related quality-of-life [56]. Prevalence of related side-effects is high, even years after treatment [57] and, although there are interventions available for the management of these, men often do not receive information about what is available (M Hennessy, personal communication). This suggests that there may be significant needs for physical, psychological and psychosexual support among prostate cancer survivors.

This study was based on high-quality cancer registration data. Although registration completeness is high [28], a small proportion of cases is missed by the Registry, meaning that these figures slightly under-estimate true 17-year prevalence; the extent of this under-estimation is likely to vary by site. Accurate prevalence estimates also require comprehensive death registration and the ability to perform accurate linkage between death certificates and cancer registrations. While death ascertainment is likely to be high in Ireland, it is possible that some people diagnosed with cancer left Ireland and subsequently died before the end of 2011; these deaths would not be known to the Registry. As noted earlier, these figures do not claim to be estimates of the total number of survivors – in particular, they underestimate the true number of long-term survivors (i.e. they do not include people diagnosed with cancer before 1994 and who were still alive at the end of 2011). While methods are available for estimating total prevalence [58, 59], these require assumptions which may not be valid. Moreover, it might be argued that many of those who have survived cancer for at least 18 years are likely to be at low risk of recurrence and have little need for active follow-up. Therefore, it has been suggested that limited duration prevalence is likely to be more pertinent for estimating the needs for cancer services according to specific phases of cancer care [29]. In terms of other limitations, these analyses do not identify those survivors whose cancer was cured, those in active therapy or those dying from cancer; this information is not available through the Registry. Nor do they reveal anything directly about the health status, or unmet supportive care needs, of survivors. Evidence is accruing that aspects of health-related quality-of–life may vary by survivors’ socio-economic, urban/rural, or immigrant status [60‐64] suggesting that supportive care needs may also vary and, in turn, that estimates of the prevalence of survivors in different socio-economic groups, for example, could be valuable for service planning. We did not consider these characteristics and this is a limitation of the study. Finally, these figures are a snap-shot of prevalence at one point in time. As in other developed countries, cancer incidence in Ireland will continue to rise in coming years [65]; even if survival does not improve, the numbers of cancer survivors will grow, particularly in the older age groups. Estimates of this future burden are also needed to help providers make provision for this booming population.

In conclusion, this study shows that data from cancer registries can provide a population-based estimate of the number of cancer survivors, information likely to be of considerable value to service planners and providers in the statutory and voluntary sectors. They also reveal important heterogeneity within the survivor population - which is likely to determine (at least in part) their ongoing medical and supportive care needs and hence influence service requirements - and provide an indication of the likely magnitude of groups of survivors who may have specific service and support needs (e.g. testicular cancer survivors, survivors of leukaemia and lymphoma, long-term breast cancer survivors, unmarried survivors, elderly survivors). Figures such as these provide an important first step in informing development of follow-up and after-care services, supports and other initiatives which will effectively meet survivors’ needs.