The number of cancer survivors living in the USA is expected to increase from 15.5 to 20 million during the next decade [2]. Progress in cancer treatment is improving the stage-specific survival for most types of cancer, yet many survivors develop a variety of physical and psychosocial long-term or late effects [3]. Despite the increasing number of cancer survivors, the transition out of active treatment, also referred to as “re-entry” [4], remains a relatively unstudied period. Although its boundaries are not sharply defined, re-entry can span from the point of medical treatment completion through the following year or more [5]. Research suggests that cancer survivors follow distinct adjustment trajectories after diagnosis [6,7,8]. For example, some survivors experience very low distress, other survivors experience distress only during active treatment, and others remain distressed during re-entry, and for some time beyond, a subset of whom experiences chronically elevated distress [6]. Thus, the transition from active to post-treatment care is individually variable and can pose considerable challenges in managing health and well-being [5, 8,9,10,11]. Improving physical and psychological cancer-related symptom management early on has been shown to improve overall distress levels, quality of life, and potentially increases survival [12].

Preparedness for the re-entry phase is defined as the extent to which the individual feels prepared for what to expect physically, emotionally, and behaviorally when primary cancer treatments are complete [13]. A review of research on the re-entry phase of survivorship suggests that while many survivors recover with limited need for support, others are unprepared for the transition and need psychological and physical aid, perhaps best delivered in a tailored approach [5]. Therefore, identifying who is—and who is not—prepared for re-entry, the primary aim of this study, has the potential to better identify which cancer survivors need tailored interventions to address unmet transition needs [8].

Being prepared for re-entry has many positive implications for cancer survivors. Existing studies suggest that being prepared may reduce uncertainty and thus positively influence physical and psychosocial outcomes [13, 14]. Furthermore, patients who feel less prepared are at increased risk for poor survivorship outcomes, such as mood disturbance [13]. Although the existing literature provides support for the value of preparing patients for re-entry, the number of studies examining this important construct remains small; further, studies that do examine preparedness are limited to a particular cancer type or treatment modality and thus have limited generalizability to other cancer populations [4, 5, 8]. Therefore, additional research on survivor preparedness is needed, particularly among survivors of cancers other than breast cancer.

Stanton and colleagues [1] proposed a conceptual framework for post-treatment adjustment, shown in Fig. 1, for identifying contributors to post-treatment adjustment based on stress and coping theories. The framework suggests that individuals completing treatment face challenges in four domains: emotional functioning, physical functioning, interpersonal relationships, and life perspectives/practical concerns. It further proposes that interpersonal/environmental, individual, and disease- or treatment-related characteristics influence these four domains through the individual’s appraisal of the situation and existing coping resources and skills. How a survivor appraises a situation is modified through enhancing his or her preparedness for the transition to post-treatment care; that is, augmenting coping resources and skills.

Fig. 1
figure 1

Conceptual framework for post-treatment adjustment. Source (with permission): Stanton AL, Ganz PA, Rowland JH, Meyerowitz BE, Krupnick JL, Sears SR. Promoting adjustment after treatment for cancer. Cancer. 2005; 104: 2608–13

Full size image

The current study sought to assess several of the relationships identified in Stanton and colleagues’ conceptual framework [1] and social-cognitive theory [15], and its aims were twofold. The primary goals were to examine levels of perceived preparedness among breast, colorectal, and prostate cancer survivors during the re-entry phase and to identify characteristics associated with relatively high and low preparedness. Second, extent of preparedness and other factors in the conceptual model were examined for their relations with cancer-related symptom management, as indicated by confidence in managing experienced physical symptoms and frequency of psychological symptoms. Indices for the interpersonal context included perceptions of social support, discussion of long-term side effects with provider, and quality of oncologist and PCP-provided care. The individual context was measured by socio-demographics, comorbidities, depressive symptomatology, and spirituality. The disease/treatment context was assessed by cancer type and cancer treatment. Appraisal was indicated by the level of perceived preparedness.

Method

Eligible participants for the American Cancer Society’s (ACS) Cancer Survivor Transition Study, approved by the Morehouse School of Medicine’s Institutional Review Board, included breast, colorectal, or prostate survivors who had completed curative treatment fewer than 13 months earlier, were 18 years of age or older, could communicate in English, and were part of the ACS constituent database (described below). Participants were recruited from the ACS constituent database that contains information on self-identified newly diagnosed cancer survivors. More than 100,000 new constituent records are added to the database each year, making it a cost-effective sampling frame. This database is used by the ACS for tracking and managing requests for information and services, volunteer activities, and participation in ACS events throughout the USA. A sample size of 1500 cancer cases, 500 of each cancer type, was needed to generate reasonable estimates of statistical power (80% power, 95% two-sided significance level) for testing the principal aims of the overall Cancer Survivor Transition Study which will involve three data collection points.

A short screener was sent to potentially eligible individuals to identify an eligible pool of participants from which to sample. Screener packets included a 1-page description and IRB-approved passive consent for the study, a 3-page screener survey to determine eligibility, a study FAQ, an ACS ribbon car magnet as an expression of gratitude, and a pre-paid return envelope. Although the constituent database includes self-reported date of cancer diagnosis, it does not contain information on treatment completion, making it difficult to estimate how many individuals would need to be contacted to achieve the 1500-participant target. This challenge was addressed by drawing the sample in two stages. In stage 1, responses from a small sample were used to estimate the proportion of screener respondents eligible for the study and the proportion of eligible screener respondents expected to return a completed survey. These parameters were then used to calculate how many additional individuals needed to be contacted in order to increase the likelihood of achieving the target analytic sample of 1500 completed surveys. Addresses and vital status were updated using National Change of Address and/or LexisNexis services. Once confirmed eligible through screener responses, survivors were sent a survey packet within 2 weeks. The packet contained a cover letter on ACS letterhead, the survey, and a pre-paid return envelope. Two reminders at roughly 3-week intervals were sent to increase recruitment. A dedicated toll-free number was included in recruitment materials for questions about the study.

A total of 4182 individuals returned the screener, 39.9% (n = 1670) of whom were assessed as eligible. Ineligible survivors were those in active treatment (n = 449) or those who had their last treatment more than 13 months ago (n = 1335). Of the 1670 surveys sent to eligible participants, 72.5% (n = 1211) completed surveys. For the current analysis, those reporting cancer recurrence (n = 12) or diagnosis of a new cancer (n = 11) were excluded, resulting in a final sample of 1188.

Measures

Questionnaire items were developed from several sources: the relevant literature on survivorship and re-entry; standard instruments used to test social-cognitive theory [14]; questions used previously in ACS’s Studies of Cancer Survivors I and II [16]; and findings from nine focus groups with breast, colorectal and prostate survivors which were conducted specifically to inform the development of this study’s survey. Cognitive interviewing [17] was used to pretest the screener and the survey. Scale scores were calculated only for individuals responding to at least 75% of each scale’s items.

Socio-demographic Factors

Participants reported their age in years; gender; education (some high school or less, high school graduate/GED, some college or more); race/ethnicity (non-Hispanic white, Hispanic or Latino, non-Hispanic black or African-American, other); and marital status (single/never married, married/living in a marriage-like relationship, separated/divorced/widowed).

Cancer-Related Factors

Participants specified their cancer type (breast, prostate, colorectal); month and year of cancer diagnosis; cancer stage (local, lymph node involvement/regional, metastasis/distant); and cancer treatments (surgery, radiation, chemotherapy, hormone treatment). Participants reported whether they were still receiving treatment or had completed each type of treatment (month and year of completion, except for hormonal treatment). For quantitative analyses, cancer type and gender were combined into a single factor with four levels: female breast, female colorectal, male colorectal, and male prostate.

Physical Comorbid Conditions

The Older American Resources and Services Comorbidity Scale [18] was used to compute the total number of reported chronic conditions. Comorbidities assessed included arthritis or rheumatism, glaucoma, emphysema or chronic bronchitis, high blood pressure, heart disease, circulation trouble in arms or legs, diabetes, stomach or intestinal problems, osteoporosis, chronic liver or kidney disease, and stroke. Cancer was excluded because, by selection, all respondents had cancer. Depressive symptomatology was measured separately since depression is not a physical comorbidity.

Body mass index

Self-reported height and current weight were used to compute current body mass index (BMI) (kg/m2).

Perceived Preparedness

A published [13] 4-item scale was used to measure preparedness for re-entry. The published 4-item scale was adapted from a previous 2-item measure of preparedness [8]. The scale uses the answers to the multi-part question: “When my cancer treatment ended, I felt prepared for” (1) what to expect physically, (2) what to expect emotionally, (3) making healthy eating choices, and (4) getting in touch with community services and programs. Response choices were “Strongly Disagree” (weight = 1) to “Strongly Agree” (weight = 5). The preparedness score is the average of response weights and ranged from 1 to 5, with higher scores indicating higher perceived preparedness. Calculated Cronbach’s α is 0.81, similar to the α of 0.83 reported in the study by Jones and colleagues [13] for breast cancer survivors.

Perceived Management of Cancer-Related Symptoms

A modified version of the Memorial Symptom Assessment Scale [19] was used to measure perceived cancer-related symptom management. Based on the symptoms most commonly reported by participants in our focus groups and reported in the literature on breast, prostate, and colorectal survivors, 12 of the original 32 symptoms were used. Participants were asked about their experience of eight physical symptoms (i.e., fatigue, loss of strength, problems with sexual interest or activity, shortness of breath, physical pain, incontinence or problems with urination, diarrhea/IBS/problems holding bowel, lymphedema, numbness or tingling) on two dimensions: (1) “In the past month, how often did you have this symptom?” (never, rarely, occasionally, frequently, almost constantly); and (2) “If you ever had this symptom, how confident are you that you can manage it?” (extremely, quite a bit, moderately, a little, not at all confident). They were also asked about frequency of four emotional/behavioral symptoms: difficulty concentrating, difficulty sleeping, feeling nervous or worrying, and feeling sad. The perceived cancer-related symptom management score was calculated in a similar way to the Memorial Symptom Assessment Scale total score [19]. Each physical symptom was scored by first dichotomizing whether or not the participant experienced the particular symptom (never vs. rarely/occasionally/frequently/almost constantly). Individuals reporting no experience with that symptom were given a score of zero, otherwise their confidence to manage that symptom was scored in increments of 0.8 (not at all = 4.0, a little = 3.2, moderately = 2.4, quite a bit = 1.6, extremely = 0.8). Each of the four psychological symptoms were scored on frequency using a scale of 0–4 (never = 0, rarely = 1.6, occasionally = 2.4, frequently = 3.2, almost constantly = 4). The 12 symptom scores were averaged to produce a total score ranging from 0 to 4; higher scores indicated poorer perceived cancer-related physical and psychological symptom management [18].

Social Support

The 6-item Interpersonal Support Evaluation List [20] assessed the perceived availability of functional social support (e.g., “When I feel lonely, there are several people I can talk to”). Items were rated on a 4-point scale (definitely false, probably false, probably true, definitely true) and summed to produce a total social support score. Scores ranged from 0 to 18, with higher scores indicating more perceived social support. This measure is well validated among cancer survivor populations [21] and has high internal consistency (Cronbach’s α of 0.89) [22].

Survivorship Care

The 7-item Primary Care Delivery of Survivorship Care Scale [23] assessed perceived provider behaviors and quality of transition support, completed separately for primary care providers and oncologists. The two scores each ranged from 0 to 100, with higher scores reflecting better perceived delivery of care. The Primary Care Delivery of Survivorship Care Scale was modified for use among breast cancer survivors in a prior study and was shown to have shown high internal consistency (Cronbach’s α = 0.89) with that population [23].

Long-Term Effect Discussion

Survivors rated the extent to which a doctor or other health care professional discussed potential long-term effects of cancer and its treatment, with response options: “Not at all,” “A Little,” “Somewhat,” and “A Lot.” This single item was developed specifically for the Cancer Survivor Transition Study based on the conceptual model.

Spiritual Well-Being

The Functional Assessment of Chronic Illness Therapy–Spiritual Well-Being Scale (FACIT-Sp-12) [24] is a 12-item measure of spiritual well-being in patients with chronic illness and includes three subscales: meaning, peace, and faith. Subscale scores ranged from 0 to 16, with higher scores indicating greater spiritual well-being. This measure demonstrated discriminant and convergent validity among breast cancer survivor populations as well as high reliability (Cronbach’s α = 0.81–0.88) [25].

Illness Perception

Acute versus chronic disease perception was measured using the three highest loading items from the 6-item disease timeline subscale of the revised Illness Perception Questionnaire [26]. Items were rated on a 5-point scale ranging from “strongly disagree” to “strongly agree.” The summed score ranged from 3 to 15, with higher scores representing more strongly held beliefs about cancer as a chronic (vs. acute) condition. This measure demonstrated construct, predictive and discriminant validity along with test-retest reliability [27]. The Cronbach’s α for each of the subscales ranged from 0.79 (timeline cyclical) to 0.89 (timeline acute/chronic) [26].

Depressive Symptomatology

The 10-item Center for Epidemiologic Studies–Depression Scale assesses depressive symptoms during the prior week [28]. Scores range from 0 to 30 with higher scores indicating greater depressive symptoms. This measure demonstrated concurrent validity by correlation with other mood state measures (Profile of Mood States-Short Form and Bradburn Positive and Negative Affect Scales) among cancer survivors [29]. Cronbach’s α was 0.86 [30].

Overall Physical Health

A single item assessing self-reported physical health (SF-1) [31] included the following response options: excellent, very good, good, fair, or poor.

Statistical Analyses

Participant characteristics were summarized with frequencies and valid percentages for categorical variables and means and standard deviations for continuous variables.

Perceived Preparedness

A nonlinear relationship between preparedness and the variables of interest was noted during preliminary analyses. For this reason, a nonparametric analysis approach, regression tree analysis, was used to examine preparedness response variability in terms of response groups and to identify the characteristics of these groups. Regression tree methodology is a common tool in health sciences and clinical research [32,33,34] and is increasingly used in public health research [35, 36]. Regression tree methodology uses recursive partitioning to stratify the respondents into relatively homogeneous response groups in situations where complex interactions among explanatory factors are likely [33]. The result of a regression tree analysis is a decision (or binary) tree that describes the steps used to create the homogeneous response groups. Starting with the whole sample (the root node), an attempt is made to split the sample into two groups in such a way that reduces within-group variability and minimizes error. The variable/value combination that produces the best split is identified and the sample is split into two groups. This process is repeated for each group (child node) identified in the previous set until further splits are deemed unproductive. Therefore, early nodes contain the strongest explanatory power in accounting for perceived preparedness, the dependent variable.

Driven by the study’s conceptual model, the regression tree methodology assessed the following variables in building the final decision tree: socio-demographic variables, cancer-related variables, number of comorbidities, depressive symptoms score, long-term side effect discussion, social support score, quality of oncology and primary care provider survivorship care scores, and overall physical health score. The regression tree analysis was conducted in R (version 3.0.2, http://cran.r-project.org/) using the rpart library. A complexity parameter of 0.01 was used as the stopping rule, meaning any further splitting of the nodes would result in a ≤0.01 improvement in relative error. The other independent variables were used to estimate values when the splitting factor was reported as missing.

Cancer-Related Symptom Management

The ability of preparedness and the relevant factors to explain variability in perceived cancer-related symptom management was examined using general linear models. Missing item responses were excluded from the model using listwise deletion. Preliminary one-factor-at-a-time modeling was used to reduce the number of variables included in the final multivariate model. Variables were: socio-demographic factors, cancer-related factors, social support, both survivorship care scores, illness perception as chronic, and the three spirituality subscales: meaning, peace, and faith. All factors for which the simple model reached a significance level of p < 0.15 (candidates) were included in the multivariate linear model. Interactions between candidate factors and perceived preparedness were also assessed. A linear model incorporating all candidate variables was further reduced using backward elimination of any variable whose marginal significance (last-in F test) exceeded the type I error rate of 0.05. Reported are results from a final linear model that includes all remaining candidate variables and the perceived preparedness score. Computations were performed in SAS v9.4 and all tests assumed two-sided alternatives.

Results

Characteristics of the 1188 cancer survivors (401 breast, 433 colorectal, 354 prostate) are summarized in Table 1. The average participant was 62.3 years old (SD = 10.93), female (57.1%), married (68.7%), and non-Hispanic white (84.2%), with some college education or higher (64.0%). Most survivors received surgery (81.0%) and reported being diagnosed with local (62.8%) or regional (30.0%) disease. On a scale of 1 to 5, on average, survivors felt moderately prepared (\( \overset{-}{x} \) = 3.6, SD = 0.84) for the transition to post-treatment care.

Table 1 Characteristics of cancer survivors (n = 1188)
Full size table

Survivors reported experiencing a number of symptoms (\( \overset{-}{x} \) = 7.1, range = 0–12) in the past month (Table 2), with an average cancer-related symptom management total score of 1.2 (SD = 0.81, range = 0–4). Of the 12 symptoms considered, 8 symptoms were reported by over 50% of respondents as ever experienced and 6 of these symptoms were reported by more than 20% of survivors as being experienced frequently. Fatigue and difficulty sleeping were experienced by more than 75% of respondents with approximately 30% experiencing those symptoms frequently. More than 30% reported frequent numbness and tingling and almost 40% reported frequent problems with sexual interest/activity.

Table 2 Cancer-related symptoms experienced in the past month
Full size table

Perceived Preparedness

One-factor-at-a-time linear model results (Table 3) indicated differences (p < 0.15) in perceived preparedness as a function of cancer type-gender, marital status, cancer stage, surgery, chemotherapy (yes/no), discussion of long-term effects, and overall physical health. Correlations with perceived preparedness were small-to-moderate and statistically significant with the continuous variables of age, social support, quality of oncologist-delivered survivorship care, quality of primary care provider-delivered survivorship care, chronic illness perception, the three spirituality subscales, and depressive symptoms. These variables were considered when estimating the final regression tree (Fig. 2). The regression tree identified eight distinct groups, terminal nodes 1 to 8, ordered from lowest average score to highest score and 7 decision rules.

Table 3 Level of preparedness and cancer-related symptom management by socio-demographic, medical, and other variables
Full size table
Figure 2
figure 2

Correlates of perceived preparedness, regression tree analysis

Full size image

Regression tree analysis identified “high-risk” subgroups of survivors who reported low preparedness. Following the regression tree, terminal node 1 had the lowest mean level of perceived preparedness (\( \overset{-}{x} \) = 2.75) and included 94 individuals who: (1) had a health professional discuss the possible long-term side effects with them not at all, a little, or somewhat; (2) had a depressive symptom score of 9.5 or higher; and (3) reported relatively low-quality survivorship care by their oncologist. In contrast, terminal node 8 had the highest mean level of perceived preparedness (\( \overset{-}{x} \) = 4.24) and included 109 individuals who: (1) had high health professional communication about long-term side effects and (2) had high perceived quality of primary care provider-provided survivorship care. Higher levels of social support were also associated with higher preparedness. The final model explained approximately 21% of the variability in preparedness.

Perceived Management of Cancer-Related Symptoms

Correlations between continuous variables and management of cancer-related symptoms are presented in Table 3. Individuals experiencing more problems managing cancer-related symptoms had lower perceived preparedness, social support, quality of oncologist-delivered survivorship care, and the three spirituality indices as well as more comorbidities and a higher BMI. Poorer symptom management also was associated significantly with viewing cancer as chronic rather than acute and younger age. Education was not significant in the univariate analysis and was not considered for the final model. The following constructs did not remain significant when considered for the multivariate model and were dropped: race/ethnicity, cancer type, gender, surgery, chemotherapy, months since last treatment, BMI, illness perception, meaning, discussion of long-term effects, and quality of oncologist and primary care provider-delivered survivorship care. No interaction between perceived preparedness and any other variable was statistically significant at p < 0.01.

In the assessment of model adequacy, residuals did not indicate any lack of fit, the normality assumption was satisfied, and no multi-collinearity was apparent. The final model, including age, marital status, cancer stage, number of comorbidities, social support, faith, peace, and perceived preparedness, was significant (F = 100.4, p < 0.0001) and accounted for 43% of the variance in symptom management (Table 4). A one-unit increase in perceived preparedness was associated with a 0.10 improvement in symptom management, after accounting for the other variables in the model.

Table 4 Multivariate modeling of cancer-related symptom management
Full size table

Discussion

The current study identified key factors associated with levels of preparedness for the transition to post-treatment care among breast, colorectal, and prostate survivors in the first year post-treatment, a significant yet relatively understudied phase of survivorship. Survivors felt moderately prepared for what to expect, which is consistent with prior findings [13]. Analyses identified eight subgroups of survivors with relatively high or low levels of preparedness. Having a health care professional thoroughly discuss long-term side effects with the patient was the most important variable in distinguishing high (vs. low) preparedness. This actionable result demonstrates the importance of health care professionals initiating such discussions with all patients, especially since there were no differences by cancer type. Providing a survivorship care plan, a document that contains specific, tailored information on an individual’s cancer history, summary of treatments received, potential long-term side effects, and recommendations for follow-up care and health behaviors, to a patient at the end of treatment is a step in the right direction. However, the survivorship care plan is only valuable to the extent that it is used as a communication tool. The most value, according to this study, appears to accrue when a health care professional devotes adequate time, in the eyes of survivors, to talk through the components of the survivorship care plan, including what to expect physically and emotionally, how to live a healthy life after cancer, and how to use relevant supportive resources.

Perceived high quality of survivorship care by the oncologist and by the primary care provider were key variables associated with greater preparedness. These findings support high-quality care provision within a “shared care” model, where coordinated care and clearly defined roles between primary care provider’s and oncologists can ensure that survivor information needs on the management of long-term and late effects are met and not “lost in transition” [37].

Survivors with moderate to high depressive symptoms who perceived an inadequate discussion of long-term effects with a health care provider and poor-quality survivorship care from the oncologist felt the least prepared. Depressive symptoms were a key and independent correlate of low preparedness and can be mitigated through biopsychosocial interventions [38]. Even if survivors experienced good oncology care and had satisfactory discussion of long-term effects with a health care provider, those with moderate to high levels of depressive symptoms felt much less prepared than did survivors with fewer depressive symptoms. This finding aligns with the proposed conceptual model, suggesting that individual-level factors can be important contributors to cancer-related appraisals and coping processes. Similar findings were reported by Jones and colleagues [13] when assessing the relationship of preparedness and mood (as measured by the Profile of Mood States-Short Form) [39, 40].

In comparison to other studies (e.g., 41), survivors in the ACS Cancer Survivor Transition Study reported higher incidence of several symptoms. Consistent with prior studies [41], the most problematic symptoms were fatigue, difficulty sleeping, feeling nervous/worrying, and physical pain. In addition, survivors in the present study reported problems with sexual interest or activity (68.1% ever, 39% almost constantly). The ACS Cancer Survivors Transition Study sample had higher reports of difficulty sleeping (75.3 vs. 49–53%), physical pain (67.8 vs. 37–44%), problems with sexual interest or activity (68.1 vs. 50–53%), feeling sad (64.1 vs. 35–39%), and numbness or tingling (63.7 vs. 36–39%) compared to a study of breast, colorectal, prostate, gynecologic, and lung cancer survivors [42]. Higher symptom reports may be the result of sampling survivors from the ACS constituent database, which may be more nationally representative than samples drawn from survivors affiliated with specific National Cancer Institute-designated Comprehensive Cancer Centers [42] or university cancer treatment centers [43]. The higher symptom prevalence may also be the result of survivors experiencing more symptoms being more likely to reach out to the ACS for help.

Despite the high prevalence of symptoms experienced in the past month, many survivors felt they could manage most of their symptoms, as demonstrated by a mean total cancer-related symptom management score of 1.2 (possible range 0 to 4, with higher scores indicating poorer symptom management). Prior studies using the Memorial Symptom Assessment Scale, with 32 assessed symptoms (excluding fatigue) reported a mean total symptom distress score of 0.74 and 0.76 [41] [43]. The difference is likely twofold. This study considered the 12 symptoms that might have been most bothersome to patients, as identified through our focus groups and the literature. Second, the current study combined how often a symptom was experienced with confidence in the ability to manage cancer-related symptoms rather than combining how often a symptom was experienced with how distressing survivors found each symptom. Given recent work in palliative care demonstrating the value of symptom management on quantity and quality of life [12], this study focused on an indicator of how well cancer survivors managed their symptoms.

Overall, the study demonstrated that the more a survivor feels prepared for the transition to post-treatment care, the better they are at managing their cancer-related symptoms. This finding supports the idea that preparedness includes both knowledge and an understanding of the skills necessary to manage post-treatment challenges. This finding supports the idea that preparedness includes both knowledge and an understanding of the skills necessary to manage post-treatment challenges. A similar construct to preparedness, uncertainty, was examined in a prior study; it concluded that uncertainty 1 month after treatment is associated with poorer quality of life, as measured by the Functional Assessment of Cancer Therapy-General, 3 months later [14]. If preparedness proves an equally effective predictor of outcomes in future studies, this 4-item measure may provide a rapid way, in a clinical setting, to identify survivors who have informational needs, and may be used to re-assess patients’ perceived preparedness after relevant discussions with their health care providers. The brief measure may also be used to screen for low levels of preparedness so appropriate referrals can be made to the appropriate healthcare professional in the interdisciplinary team.

Although they are not the focus of this paper, several additional factors, including younger age, advanced stage of disease, increased number of comorbid conditions, lower perceived social support, greater endorsement of faith, and lower peace were associated with poorer symptom management. Prior work on post-treatment psychological adjustment demonstrates an association between younger age [13] and depressive symptoms [44] with higher health-related distress as well as higher cancer-related symptom distress and higher numbers of comorbid conditions among cancer survivors [45] [46]. The findings also reinforce the critical role of social support in palliating survivors’ symptom experiences during the re-entry transition, when the health care team begins to play a lesser role and has less contact with survivors. Although having a sense of peace was negatively related to symptom management, as would be expected, the positive relationship of faith and symptom management obtained in the multivariate model is seemingly counter-intuitive. The negative zero-order correlation of the two variables (Table 3) suggests an interpretation: perhaps the variance unique to faith, controlling for level of peace, reflects an unfulfilled attempt to find peace through reliance on faith [47]. A prior study found that only those who indicated that they sought religion as a reaction to the cancer diagnosis had poorer outcomes [48]. The current study only assessed faith after cancer which precludes the ability to look at the potential interaction with faith before and after a cancer diagnosis.

Several limitations of the study require mention. The data were self-reported with no medical record abstraction. Survivors were mostly non-Hispanic white and highly educated. Sampling from the ACS constituent database likely produced a sample of information seekers (but note that approximately 20% of the sample were faxed referrals directly from physicians). The cross-sectional design does not allow causal inference; an additional wave of data will permit longitudinal analyses in the future. Longitudinal and experimental designs will be essential in verifying the utility of preparedness appraisals in promoting well-being and health during the re-entry phase of cancer survivorship. Education was used as a proxy for socioeconomic status in these analyses because other variables were not available from the baseline data. Future analyses will utilize a second time point of data that includes income and perceived financial strain.

This is the first report on the ACS’s Cancer Survivor Transition Study of approximately 1200 cancer survivors of prostate, colorectal, and breast cancer with representation from all 50 states. The study identified key factors associated with levels of preparedness for the transition to post-treatment care among survivors in the first year post-treatment, a significant yet relatively understudied phase of survivorship. Additionally, this study shows that preparedness is an important predictor of a future important outcome for cancer survivors, cancer-related symptom management. Hence, future studies are needed on the development of the best preparedness measures, their optimal integration in clinical practice as a risk stratification tool, and identification of the best communication channels and tailored interventions to prepare cancer survivors during the re-entry transition.