The results of this review are consonant with the Andersen Behavioural Model of health service utilisation. Younger, white and employed or student CSs were more predisposed to access and receive health care. Individuals over 65
years old represent an at-risk group and are encouraged by their health-care provider to receive vaccination annually [47
]. According to the results of this review, older CSs were more likely to receive influenza vaccination but not other types of care. This discrepancy between older and younger CSs in terms of receipt of preventive services may be explained partly by physicians making decisions about the utility of preventive care based on the life expectancy of CSs, [9
] – this may also explain the increased risk regarding inpatient hospitalisations among elderly CSs. The role of incentive payments to primary care physicians may play a part in the differential patterns of preventive services utilisation. Ethnicity was examined within US studies only and various explanations relating to economic differences, differences in health-seeking behaviours, preferences for treatment or perceptions of post-treatment cancer surveillance may account for reported ethnic disparities [19
Visits to primary care physicians appeared to enable the implementation of care recommendations for CSs. Other facilitating factors which were more pertinent in non-nationalized or universal health-care systems were additional health insurance, a higher education, income and living in an urban, affluent area. Many samples included CSs who had insurance (e.g. Medicare) and this factor may be related to a sense of self-efficacy in terms of seeking information and negotiating the health-care system [42
]. Cancer survivors had greater or at least similar frequency of contact with primary care physicians compared to the general population without cancer. Many cancer survivors may experience long-term health problems or further ill-health following treatment which require further care or specialist care. In many health-care systems a primary care physician may act as a gatekeeper for access to specialist services thus accounting for greater frequency of visits among the cancer survivor population.
Needs for care were also related to co-morbidity, later stage or undifferentiated tumour and menopausal or depressive symptoms. Younger, employed or student CSs expressed a need for mental health services. Overall, as CSs survived longer post-diagnosis they used less cancer-oriented care, with the exception of screening.
Although the review amalgamated literature regarding core health services, it excluded some health services such as complementary and alternative medicine due to resource restraints (and their perceived non-mainstream position). Services like complementary and alternative medicine are becoming increasingly important for CSs and require research attention. Limitations of the review include uncertainty about generalisability of findings. There is a need to give consideration to the merits of conducting comparative health care system research (including health service research in non-USA countries), particularly given the differing role of the oncologist between health-care systems and the role of insurance in obtaining access to care in USA studies. Fifteen studies comprised a secondary analysis of the SEER-Medicare database which included individuals over the age of 65
years and excluded individuals covered by other insurance plans or no insurance plan. Moreover, individuals were limited by their insurance plan regarding access to physicians and entitlement to receive particular procedures. For example, there are additional charges for receipt of procedures such as colonoscopy [33
]. A further issue for the review was the lack of generalisation of results to US cancer survivors without medical insurance and thus a primary care physician may not be the first point of contact for care for some individuals. The SEER-Medicare database also did not provide information regarding reasons for use of services (i.e. cancer-related follow-up or for another condition). None of the papers reported the reasons for health service contact and whilst CSs appeared to be using relevant health services, this did not equate to follow-up care. It may not have been appropriate to compare utilisation rates between CSs and ‘healthy’ individuals from the general population as they may also have chronic illnesses; this is an inherent limitation within the included primary studies. Unfortunately, none of the included studies examined the nature and extent of the coordination of different services provided for CSs and this needs to be empirically tested.
The Andersen Behavioural Model of health service utilisation was used as the structural framework to organise the review. Although it is limited in its scope regarding potentially important behavioural variables, it does take into account the health beliefs of an individual. Health beliefs were not extensively addressed by the studies in the review. Indeed, only one study looked at the perception of risk and only in respect of cancer recurrence. There is a need to give empirical attention to the role of research evidence-based behavioural and ‘cognitive-behavioural’ constructs in order to improve our understanding preventive procedure receipt and adoption of health promoting behaviours by cancer survivors [48
]. Future research efforts to understand health service use by cancer survivors should consider supplementing or expanding the Andersen Behavioural Model to include behavioural and cognitive components (e.g. subjective norms) from models such as the Theory of Planned Behaviour [49