Addressing fear of cancer recurrence among women with cancer: a feasibility and preliminary outcome study

At the time we started the study, there were only 4 longer (i.e. 10+ items) FCR instruments that were available: the concerns about recurrence scale (CARS) [3], the fear of recurrence questionnaire (FRQ) [31], the fear of cancer recurrence inventory (FCRI) [56], and the fear of disease progression questionnaire (FoP-Q) [57]. However, the CARS has been validated with women with breast cancer only. The FCRI and the FoP-Q have been validated with French and German cancer patients, respectively, and while English translations are available for both instruments, empirical validations of these translated versions have yet to be published. This left the FRQ as the most suitable longer measure of FCR because it is applicable to both breast and ovarian cancer patients and because it is validated in English. However, the FRQ does not have a clinical cut-off score. We attempted to select participants with moderate to high FCF by using a score of 4 or more on 50 % of the 22 items of the FRQ as an inclusion criterion. Recognizing that this method had limitations, we also used the validated cut-off score of the Impact of Events Scale (IES) to reliably identify participants who could be considered to have elevated cancer-specific distress.

We compared the effects of the intervention by cancer diagnosis (breast vs. ovarian) using repeated measures ANOVAs with cancer diagnosis as a between factor. Results revealed few differences among the two groups of participants. There was a TimeXDiagnosis difference in the use of active coping (F _{(2, 34)} = 3.31, p < 0.05). Univariate analyses revealed that the use of active coping increased over time only among women with ovarian cancer. There were also main effects of diagnosis on the use of instrumental support (F _{(1, 36)} = 5.35, p < 0.05) and planning coping (F _{(1, 36)} = 5.66, p < 0.05). For both coping strategies, women with breast cancer reported more use of these strategies than women with ovarian cancer but neither group showed changes over time on either strategy.

At the suggestion of one of the reviewers, we performed additional analyses to see if time since diagnosis had any impact on our results. First, we ran bivariate correlations between time since diagnosis and the variables that were significantly impacted by the intervention (FCR, cancer-specific distress, uncertainty, use of emotional and instrumental support coping, positive reframing coping, acceptance coping, and negative quality of life) for each of the three time points. Of possible 24 comparisons, only two were significant (greater time since diagnosis was associated with greater use of acceptance coping at T1 and higher levels of cancer-specific distress at T3). We also recalculated the repeated measures ANOVAs to examine the time effects for FCR, cancer-specific distress, uncertainty, use of emotional and instrumental support coping, positive reframing coping, acceptance coping, and negative quality of life controlling for time since diagnosis as a covariate. Time since diagnosis was not a significant covariate in any of these analyses and the significance of the results remained unchanged. It thus appears that the intervention is equally effective for cancer survivors who have finished treatment, regardless of time since diagnosis.