DISCUSSION

Section:

ChooseTop of pageAbstractINTRODUCTIONPATIENTS AND METHODSRESULTSDISCUSSION <<REFERENCES

It is estimated that there are over 10 million cancer survivors in the United States, and this segment of the population is expected to continue to grow.¹⁶ A landmark report published by the Institute of Medicine¹⁷ highlighted the diverse array of physical and psychosocial issues associated with cancer survivorship and underscored its importance as a research priority. However, there are concerns that many of these survivorship issues are frequently overlooked by the health care system.^18,19 We previously reported that expectations for survivorship care differed between patients and their physicians, and that a lack of clarity surrounding their respective roles may contribute to suboptimal levels of care.⁷ A few studies have described interventions to improve future survivorship care planning, but none have explored the role of patient-physician communication.^8–11 We analyzed surveys from cancer survivors and their providers and found that patient-physician discussions were associated with greater agreement in some survivorship care expectations and that concordance in these expectations was in turn correlated with a higher propensity for certain aspects of follow-up.

While we originally hypothesized that patients whose views on survivorship care are best aligned with those of their physicians are more likely to receive appropriate care, our stratified model confirmed this association to be true only for patients who reported having a conversation about cancer survivorship with their physician. In the absence of such a discussion, concordance in expectations failed to correlate with follow-up. In fact, there was a nonsignificant trend for possibly less care (eg, fewer influenza vaccinations). One interpretation of this finding is that similarities in patient-physician expectations by themselves are insufficient for ensuring survivorship care. Our multivariable and stratified analyses in Tables 5 and 6, respectively, suggest that a patient-physician discussion may contribute to survivorship care planning by acting indirectly through its primary influence on patient-physician expectations. Discussions about cancer follow-up appear to modify the effect of concordant expectations on survivorship care, which is consistent with prior reports demonstrating a positive impact of patient-physician communication on clinical decision making and coordination of care across various clinical settings in oncology.^12–15

Importantly, discussions mainly posed a positive effect on patient-oncologist expectations and not on patient-PCP or PCP-oncologist expectations. The mean CCS between patients and their oncologists was moderate when compared with those of the other matched pairs, proposing that relationships in which expectations were already mostly discordant or mostly concordant may be less susceptible to the effect of a patient-physician discussion. Another reason for this difference is that survivors may come to view their oncologists as their primary provider of healthcare and medical advice, considering the unique patient-physician rapport that often develops during and after cancer treatment. In support of this hypothesis, we observed that it was only discussions about cancer follow-up, which occurred more frequently between patients and oncologists, that resulted in more congruent expectations and increased receipt of certain follow-up care. Because this was a cross-sectional survey, we were unable to delineate the directionality of events. Therefore, we cannot exclude the possibility that individuals with lower CCS at the outset were the ones in whom personal motivations were also the strongest to seek formal discussions with their doctors. We are also uncertain of the reasons why some aspects of follow-up were unaltered by concordant expectations or the presence of patient-physician discussions. The performance of blood tests and imaging studies typically requires more coordination between physicians and other healthcare workers (eg, laboratory and radiology technicians), and thus may be less amenable to behavioral interventions, such as survivorship discussions. Alternately, these aspects of care may necessitate multidisciplinary interventions in order to alter their patterns of use.

There are several limitations to our study. First, we surveyed a geographically localized group of patients and physicians, and we analyzed only successfully returned questionnaires. Additionally, our sample consisted largely of white, well-educated, female breast cancer patients who may have had expectations for survivorship care that varied considerably from those in different demographic groups. For these reasons, our results may not be completely generalizable to other centers, nonrespondents, minorities, or survivors across different cancer types. Second, survivorship care patterns were based on patients' self-report only and were susceptible to recall bias. These data were neither corroborated by their providers nor verified by medical record review. Furthermore, we did not collect information regarding the context in which follow-up interventions were given, so we could not determine whether procedures, such as vaccinations, were administered to survivors for the appropriate clinical indications. Moreover, other important survivorship care areas, such as psychosocial and vocational issues, were not considered in the CCS. Third, each physician completed the questionnaire only once, and their questions were not directed to the care of a particular patient, but to any of their patients who might be a cancer survivor. Therefore, it is possible that some responses might have been different if they were answering about a specific patient. Fourth, we dichotomized survivorship care discussion into whether one was reported to have occurred or not. Survivorship care discussions between patients and physicians are likely quite complex and dynamic, and future work should focus on aspects of these discussions, such as content, format, and delivery, which may impact their overall effectiveness on survivorship care planning. Finally, we focused our analyses on evaluating statistically significant differences in CCS between matched survey pairs without considering clinically meaningful differences. Symptom and quality-of-life analyses in oncology have typically used a minimum difference of 5% between matched pairs as a threshold for clinical meaningfulness.^20–22 Because the CCS was created specifically for this analysis and it has not been used previously, whether this same definition can be applied to survivorship research warrants further study.

In summary, cancer survivorship care is likely driven by multiple factors. Our study findings provide preliminary support to suggest that survivorship care discussions may increase some aspects of follow-up care through a primary influence on patient-oncologist expectations. Which elements within these discussions are most important and whether they foster truly appropriate follow-up care will require more in-depth evaluations in prospective, longitudinal studies. Further research should work toward developing models for facilitating patient-physician discussions that would best promote and optimize care for future cancer survivors.