The sociodemographic and clinical characteristics of the sample (N = 140) are included in Table 1. The majority of participants were female (65.7%), Black (79.3%), and not working (88.6%). The median monthly income was US$700 with almost 50% of participants with a monthly income of less than US$700. The majority of patients were breast cancer (38%), hematologic (13.1%) and colorectal (12.4%) cancer survivors. Table 2 shows information need scores at baseline, immediately postintervention, and 6 months postintervention. Across time, need for information decreases across both groups although it appears that the decrease is somewhat larger for the ICC group (Table 2). Across time, decreases in information needs were observed for Understanding cancer treatments, p = .04; Keeping well in mind and body, p = .04; and Overall needs score, p = .04. Significant decreases were not seen for Learning about cancer, p = .12 or Relieving cancer pain, p = .26.

Our primary hypothesis, that individuals randomized to ICC would have less postintervention information needs was not supported (p > .10). Several covariates however, were significant. For all five outcomes (learning about cancer, understanding cancer treatment, keeping well in mind and body; relieving cancer pain, and the overall needs score), there was an inverse relationship between self-efficacy and the information needs scores. Age, education, and income were also significant covariates for selected outcomes (Table 3). Older individuals had greater treatment information needs. Individuals with more education or with more financial resources had lower information needs. When we included the total number of sessions attended as a covariate, the pattern of findings was similar.

The pattern of results was the same when we conducted our subgroup analysis that included the three groups (ICC, wellness, and control). Group was not associated with the study outcomes. As with the previous analyses, several covariates were significant and self-efficacy was significantly associated with all five outcomes (Table 4). Again, self-efficacy was inversely related to information need scores. Adding the number of sessions attended as a covariate in this analysis did not alter the results. No association was found in an additional analysis that examined attendance to specific ICC sessions and change in the information need score associated with that given session.

Sessions were favorably received as reflected both by the quantitative assessment as well as the narrative responses provided by participants (Table 5). Participants in the ICC arms enjoyed the sessions, the information learned, and the opportunity to hear from others. Some also thought the sessions would be beneficial early during treatment or soon after diagnosis. Participants also enjoyed the sessions of the wellness arm and felt they helped them release stress and temporarily forget their problems.

Our study is the first to evaluate the American Cancer Society’s ICC program among low-income, primarily minority cancer survivors. Consistent with ICC as offered in community settings, community experts delivered the ICC sessions. The hypothesis that ICC would result in less information needs as it related to cancer when compared to the wellness condition was not supported. However, participant ratings of both the ICC and wellness sessions obtained immediately after the sessions suggested that the intervention was meaningful for participants and of value.

Independent of the intervention, information needs decreased over time for most of our information need outcomes. These findings suggest that information needs may be met to some degree without formal intervention as patients navigate the cancer journey. While the decrease in information needs may be encouraging, it should be noted that need scores at 6 months indicated that not all their information needs were met. Thus, cancer patients continue to seek information after the time of their initial diagnosis and completion of treatment. This finding is consistent with the work by Heese and colleagues that demonstrated that information-seeking behavior among cancer survivors does not decrease over time (Hesse, Arora, Burke Beckjord, & Finney Rutten, 2008).

It is important to note that pain management was one area where we did not observe a decrease in information needs. This is in line with the continued difficulty of pain control in individuals with a history of cancer, especially minority populations. For example, a recent study found that compared to whites, Black patients with cancer had more pain interference and disability and Black race predicted pain severity (Green, Hart-Johnson, & Loeffler, 2011). Efficacious interventions to provide cancer survivors with the needed information to minimize or eliminate cancer-related pain are needed, and especially for minority, medically underserved populations.

We found that the cancer patient’s confidence that they could obtain advice about cancer was consistently associated with lower information needs. This finding is important from both a theoretical and practical perspective. In theory, self-efficacy is a central construct of several health behavior theories (e.g., social cognitive theory and the health belief model). In our study, it is possible that those with more confidence in obtaining advice about cancer also tended to ask more questions of their physician and thus had lower information needs. We did not assess question-asking behavior; future studies would have to test this hypothesis. Interestingly, it is not the case that those with higher self-efficacy were simply looking for more information beyond their doctor, nurse, or healthcare provider: when we controlled for information-seeking behavior, self-efficacy continued to have a significant relation with met informational needs. On a practical level, given the apparent role of confidence, researchers might use qualitative methods to better understand the sources of self-efficacy in this context and population and its relationship to meeting their informational needs. Information learned from this in-depth exploration can then inform interventions that aim to increase information-seeking confidence in cancer survivors.

Our finding that older cancer patients had more information needs related to treatment was also of great interest. In part, this finding might be explained by previous work that suggested that younger cancer survivors (<60 years at diagnosis) sought more cancer information from all sources when compared to older cancer survivors (Bennett, Cameron, Whitehead, & Porter, 2009). The greater need for treatment related information may reflect older individuals not seeking treatment-related information on their own. Alternatively, the fact that the age effect was specific to treatment is intriguing: studies suggest that older cancer patients are less likely to receive recommended treatment (Pisu et al., 2010). One hypothesis perhaps is that in addition to being under treated, older cancer patients may be provided less information by their physician as it relates to cancer treatment compared to younger cancer patients. Future research should seek to better understand the treatment-related information needs of older cancer patients.

A limitation of the study is that nearly one third of participants did not attend both the ICC and wellness sessions. Although this percentage of nonattendance may be considered high, reports from our advisory group suggest it is not uncommon for ICC sessions delivered in community settings. The low attendance rate may also be a reflection of our population. In an evaluation of ICC, those with less education were more likely to not complete the ICC program (Diekmann, 1988). In follow-up qualitative interviews with individuals who did not attend any sessions, we learned that reasons for nonattendance included illness burden (cancer or other), treatment side effects, or caregiving responsibilities. Notably there were some non-attendees that we could not engage even in the qualitative study, suggesting that there may be a subset of individuals who choose to, or are unable to, participate in some programs. Alternatively, these patients may be individuals who prefer to not seek information. For example, Miller suggests that under threatening conditions some individuals engage in monitoring (i.e., attending to information) while others engage in blunting (i.e., avoiding information; Miller, 1995). A recent study highlighted the heterogeneity of information-seeking among cancer patients (Protiere, Moumjid, Bouhnik, Le Corroller Soriano, & Moatti, 2012). It may be erroneous to assume that everyone desires cancer information (Leydon et al., 2000) and even the best educational programs may not be able to engage these patients.

A second limitation may be our measure for information needs. Because the content of our outcome measure had to match the content included in the sessions, an appropriate assessment tool did not exist in the literature. Thus, we created one for the specific purposes of the project. Furthermore, the wording of the questions may have been somewhat difficult (e.g., rating one’s level of need for information). Nonetheless, the fact that our measure was significantly correlated with education, income, and self-efficacy in the expected direction inspires some confidence in our assessment tool.

Our study also has several strengths. First, we evaluated a national cancer program among low-income, primarily African American cancer survivors. Traditionally, ICC attendees do not reflect this demographic, and our engagement of this population is of great significance. In addition, previous evaluations of ICC were primarily internal to the organization (e.g., ICC facilitators provided feedback about the program). The evaluations that were completed by attendees were more limited in scope and a published evaluation included responses from only a small number of racial=ethnic minorities (Diekmann, 1988). Our study may become even more relevant with recent shifts in the United States that include a decline in median household income and an increase in the poverty rate. (DeNavas-Walt, Proctor, & Smith, 2011) Second, we delivered ICC sessions that resonated with the population. The narrative comments from study participants and their rating of the sessions suggested that the intervention was meaningful to participants. In fact, participants expressed an interest for a longer intervention. We believe having cancer survivors from the county hospital choose the topics of most interest for both intervention arms was thus an important formative step (Martin et al., 2011). In addition, it highlighted the potential need for such programs in county hospitals. Third, we facilitated participation by providing transportation and off-setting in part the costs incurred by transportation for those who provided their transportation to the study. This was an important barrier to overcome. Approximately 50% of study participants used our transportation service. A fourth strength of our study was its design. This is the first evaluation of ICC using a randomized controlled design and including a 6-month follow-up. A fifth strength of our study is that we were able to engage African American cancer patients in an intervention in which they were willing and in fact did discuss their own personal experiences with cancer openly in a group setting. This support was important given the lack of support that is often reported by African American cancer survivors (Hamilton, Moore, Powe, Agarwal, & Martin, 2010; Mosavel & Sanders, 2011). Our study also provides important information for effect size. In planning the study, there was no information upon which to estimate effect. In a post hoc power analysis, in light of the changes in information needs observed postintervention, investigators would need over 400 cancer survivors to find a significant difference between groups. As this sample size may be difficult to attain with fluctuating patient flow at a single county hospital, researchers should consider multi-site studies that engage several public hospitals. An important and final strength of our study was our ability to design and deliver a comparison intervention that was well-received by study participants. The quantitative ratings and the narrative responses speak to the importance of addressing the multidimensional aspect of cancer including the usefulness of teaching techniques that can help with stress management and coping. Moreover, the positive assessment of our comparison intervention highlights the potential of less traditional approaches to interventions for cancer survivors. In our study, the comparison condition was novel both in content (e.g., humor and music were stand alone sessions) and in approach (we partnered with community organizations and a county hospital, and their staff served as the interventionists). Building on the assets and skill set already in place in the community to collectively offer a program for cancer survivors is important for leveraging resources and producing interventions that may be sustainable beyond initial grant funding. Future research should continue to explore how interventions such as our wellness intervention, further impacts patient-centered outcomes over the course of coping with cancer.

In this study, we found that although many had been coping with cancer for some time, cancer survivors presented with a moderate level of information needs at baseline. The consistent finding that those with less education and income had greater information needs suggests that our educational interventions should continue to prioritize cancer survivors drawn from this population. Our findings also suggest that delivering quality cancer care involves ensuring that cancer survivors have the skills and confidence needed to obtain desired information. Because information needs vary greatly across patients and the need for information is dynamic, the approach used in this study may better meet the needs of cancer survivors compared to programs delivered with a standard, predetermined educational content that cannot be tailored to individual patient needs, and=or may not delivered at a time when cancer survivors have need for or the capacity to process the information (e.g., during times of distress, during active treatment). There are many opportunities to further understand how best to support cancer patients and their information needs. Other types of benefits from the group environment and the structure of ICC have not been rigorously measured (e.g., question-and-answer sessions with an expert and information sharing with others in the group). From the comments provided to us from participants and our own observations during intervention delivery, these appeared to be components that were important to patients. We challenge future researchers to consider varied ways to deliver information to survivors. Typically, research interventions evaluate one delivery method (e.g., group sessions, Internet, print). While these studies have advantages, a one-size-fits-all approach with regard to delivery method may not be ideal. Future studies should explore individual patient preferences with regard to how they would like to receive the information and, potentially, evaluate whether the preferred educational approach provides advantages and improves outcomes to a greater degree than delivery channels not tailored to patient preferences. Similarly, the preferred delivery channel may also depend on the type of information that is being imparted. For example, learning about cancer treatment may be a topic best served by an in-person conversation while other cancer related topics, could be easily and effectively delivered through print interventions.

In conclusion, minority low income cancer survivors appreciated the ICC sessions offered in one arm of the trial. While their information needs did not change compared to participants in the control group, the ICC program is important in this population, and represents an important tool among those available to improve the survivorship experience of this population. Several lessons learned from our trial can be applied to improve the program and assure its maximum benefits for disadvantaged cancer survivors. Despite the complexity of tailoring the information to the patient at the appropriate time and the difficult task of delivery, information is necessary for the survivor to manage physical aspects of the disease, participate in medical decisions to be made, and cope with the mental and emotional toll of the disease and uncertainty of outcomes (Beckjord et al., 2008; Sheehan, Sherman, Lam, & Boyages, 2007). For lower income and minority survivors who continue to have worse cancer outcomes than the higher income majority population (Adler & Rehkopf, 2008; Institute of Medicine, 1999; Kagawa-Singer, Dadia, Yu, & Surbone, 2010), the need for information remains critically important.