University of Groningen Supportive care needs and psychological complaints among Mexican breast cancer patients Perez Fortis, Adriana

(1)

Supportive care needs and psychological complaints among Mexican breast cancer patients

Perez Fortis, Adriana

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Perez Fortis, A. (2018). Supportive care needs and psychological complaints among Mexican breast cancer patients. University of Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

Affective forecasting accuracy and psychological care

needs among Mexican breast cancer patients

Pérez-Fortis A, Schroevers MJ, Sánchez Sosa JJ, Fleer J, Ranchor AV. Submitted

CH

APT

ER

(3)

ABSTRACT

Objective: Affective forecasting is a cognitive process little explored in the medical scenario, though its relevance in terms of psychological adaptation to a chronic disease. We explored (a) whether Mexican breast cancer patients accurately predict their positive and negative affect on their first medical treatment after diagnosis, and (b) whether the psychological care needs post-surgery differed by the affective forecasting accuracy. Methods: After receiving breast cancer diagnosis and before starting treatment (T1), 120 Mexican breast cancer patients were asked to forecast how they would feel after surgery or after finishing with the neoadjuvant treatment, respectively. After the treatment (T2), patients were asked to report their affect and psychological care needs. Affect was measured with the Positive and Negative Affect Schedule (PANAS). Affective forecasting accuracy was indexed from an absolute and relative sense.

Results: From an absolute sense, patients within the neoadjuvant treatment group were accurate predicting their positive and negative affect. Patients within the surgery group overestimated the negative affect (r = -.25), and accurately predicted the positive affect. From a relative sense, only 37 % and 25% of the patients within the surgery group accurately predicted the negative and positive affect, respectively. Patients who underestimated the negative affect to the surgery showed the highest psychological care needs at post-surgery.

Conclusions: Mexican breast cancer patients are both accurate and inaccurate when predicting positive and negative affect to their first medical treatment after diagnosis. Affective forecasting accuracy might identify differences in the psychological care needs post-surgery among Mexican breast cancer patients.

Key words: Affective forecasting, breast cancer, psychological care needs, Latinas.

INTRODUCTION

Affective forecasting refers to the cognitive anticipation of a person’s emotional reactions to specific future events (Wilson & Gilbert, 2003). Affective forecasting has been widely studied in the context of diverse social events (Buehler & McFarland, 2001; Hoerger, Chapman, & Duberstein, 2016), but it has been little explored in the medical context (Halpern & Arnold, 2008).

Affective forecasting literature suggests that people, in general, are inaccurate when predicting their affect regarding a specific event in the future; i.e., people tend to overestimate the affective impact of future events, especially those considered negative (Ayton, Pott, & Elwakili, 2007; Wilson & Gilbert, 2005). More recent evidence, however, suggest that people are both accurate and inaccurate when predicting their emotions to focal events in the future, but depends on how accuracy is indexed, in a relative or absolute sense. The absolute sense refers to the mathematical difference between the predicted and the actual affect (individuals will have affects as intense as they predicted), whereas the relative sense refers to the accuracy taking into account the direction of the prediction, (persons who predict more intense affect in fact will experience more intense affect and vice-versa) (Coteţ & David, 2016; Mathieu & Gosling, 2012). Nonetheless, research in the health context considering both ways of indexing accuracy have not been explored, so far. Despite the potential relevance of affective forecasting accuracy within the medical setting (Halpern & Arnold, 2008; Rhodes & Strain, 2008), so far only one study has examined the relation between affective forecasting accuracy and psychological adjustment, showing that an underestimation of distress among people undergoing genetic testing for cancer susceptibility was related to an increase of general psychological distress in the long-term (Dorval et al., 2000). Previous research has shown that breast cancer patients with a poor psychological adjustment to cancer usually show high levels of negative psychological outcomes over the disease treatment, which are positively associated to higher psychological care needs (Brandão, Schulz, & Matos, 2017; Pérez-Fortis, Fleer, et al., 2017). Therefore, the question arises whether affective forecasting accuracy is also related to psychological care needs. This would give insight into how affective forecasting accuracy influences the psychological care needs of the patients. Given the high emotional implications of a breast cancer diagnosis among Mexican women (Pérez-Fortis, Schroevers, et al., 2017), it is relevant to identify the psychological care needs of these patients and whether the accuracy of their anticipated affect is related to these needs.

In this study, we first examined whether Mexican breast cancer patients accurately predict their negative and positive affect about their first medical treatment, either neoadjuvant treatment or surgery. Secondly, we examined whether the psychological care needs at post-surgery were different according to their affective forecasting accuracy.For

(4)

6

ABSTRACT

Objective: Affective forecasting is a cognitive process little explored in the medical scenario, though its relevance in terms of psychological adaptation to a chronic disease. We explored (a) whether Mexican breast cancer patients accurately predict their positive and negative affect on their first medical treatment after diagnosis, and (b) whether the psychological care needs post-surgery differed by the affective forecasting accuracy. Methods: After receiving breast cancer diagnosis and before starting treatment (T1), 120 Mexican breast cancer patients were asked to forecast how they would feel after surgery or after finishing with the neoadjuvant treatment, respectively. After the treatment (T2), patients were asked to report their affect and psychological care needs. Affect was measured with the Positive and Negative Affect Schedule (PANAS). Affective forecasting accuracy was indexed from an absolute and relative sense.

Results: From an absolute sense, patients within the neoadjuvant treatment group were accurate predicting their positive and negative affect. Patients within the surgery group overestimated the negative affect (r = -.25), and accurately predicted the positive affect. From a relative sense, only 37 % and 25% of the patients within the surgery group accurately predicted the negative and positive affect, respectively. Patients who underestimated the negative affect to the surgery showed the highest psychological care needs at post-surgery.

Conclusions: Mexican breast cancer patients are both accurate and inaccurate when predicting positive and negative affect to their first medical treatment after diagnosis. Affective forecasting accuracy might identify differences in the psychological care needs post-surgery among Mexican breast cancer patients.

Key words: Affective forecasting, breast cancer, psychological care needs, Latinas.

INTRODUCTION

Affective forecasting refers to the cognitive anticipation of a person’s emotional reactions to specific future events (Wilson & Gilbert, 2003). Affective forecasting has been widely studied in the context of diverse social events (Buehler & McFarland, 2001; Hoerger, Chapman, & Duberstein, 2016), but it has been little explored in the medical context (Halpern & Arnold, 2008).

Affective forecasting literature suggests that people, in general, are inaccurate when predicting their affect regarding a specific event in the future; i.e., people tend to overestimate the affective impact of future events, especially those considered negative (Ayton, Pott, & Elwakili, 2007; Wilson & Gilbert, 2005). More recent evidence, however, suggest that people are both accurate and inaccurate when predicting their emotions to focal events in the future, but depends on how accuracy is indexed, in a relative or absolute sense. The absolute sense refers to the mathematical difference between the predicted and the actual affect (individuals will have affects as intense as they predicted), whereas the relative sense refers to the accuracy taking into account the direction of the prediction, (persons who predict more intense affect in fact will experience more intense affect and vice-versa) (Coteţ & David, 2016; Mathieu & Gosling, 2012). Nonetheless, research in the health context considering both ways of indexing accuracy have not been explored, so far. Despite the potential relevance of affective forecasting accuracy within the medical setting (Halpern & Arnold, 2008; Rhodes & Strain, 2008), so far only one study has examined the relation between affective forecasting accuracy and psychological adjustment, showing that an underestimation of distress among people undergoing genetic testing for cancer susceptibility was related to an increase of general psychological distress in the long-term (Dorval et al., 2000). Previous research has shown that breast cancer patients with a poor psychological adjustment to cancer usually show high levels of negative psychological outcomes over the disease treatment, which are positively associated to higher psychological care needs (Brandão, Schulz, & Matos, 2017; Pérez-Fortis, Fleer, et al., 2017). Therefore, the question arises whether affective forecasting accuracy is also related to psychological care needs. This would give insight into how affective forecasting accuracy influences the psychological care needs of the patients. Given the high emotional implications of a breast cancer diagnosis among Mexican women (Pérez-Fortis, Schroevers, et al., 2017), it is relevant to identify the psychological care needs of these patients and whether the accuracy of their anticipated affect is related to these needs.

In this study, we first examined whether Mexican breast cancer patients accurately predict their negative and positive affect about their first medical treatment, either neoadjuvant treatment or surgery. Secondly, we examined whether the psychological care needs at post-surgery were different according to their affective forecasting accuracy. For

(5)

the first aim, we hypothesized that on average patients would overestimate the negative affect that they would experience after the first medical treatment. As for the second aim, we hypothesized that there would be significant differences in the levels of psychological care needs after the surgery, according to the level of affective forecasting accuracy that patients made of the negative affect. Regarding the positive affect, we kept an exploratory approach.

METHOD

Participants and procedure

Data were obtained from a larger observational longitudinal study with newly diagnosed Mexican breast cancer patients. Eligible patients were women aged 18 to 75, who received a cancer diagnosis for the first time and had not started treatment at time of enrollment in the study. Further details about selection and recruitment of patients can be consulted elsewhere (Pérez-Fortis, Fleer, et al., 2017). A total of 173 patients participated in the baseline measure (67% response rate), and complete data for the present study were obtained from 120 patients. Individuals excluded because of incomplete data (N = 53) were older (mean age 55.2 years vs. 51.8 years, p = .05), and a lower proportion had middle or higher education (52.8% vs. 76%, p < 0.01), than patients with complete data. There were no statistically significant differences on the other sociodemographic characteristics between both groups. Sociodemographic characteristics of the participants were similar in the surgery and neoadjuvant treatment group, with the exception of breast cancer stage. Most of the patients in the neoadjuvant treatment group were in stage III, whereas most of the patients in the surgery group were in stage I or II (p<.001). More details of the descriptive characteristics of the sample are presented in supplemental Table 1.

Patients were classified into two groups: (1) patients who underwent surgery and (2) patients who followed neoadjuvant treatment after the breast cancer diagnosis. A description of the participants’ flow in the study is shown in Figure 1. The study was approved by the local ethics committee of the hospital (R-2014-3504-40), and the patients who participated in the study provided informed consent. Data were collected through a semi-structured interview by three trained psychologists. Patients were evaluated two times, the first measure (T1) was made after receiving the breast cancer diagnosis and before the start of treatment (either neoadjuvant treatment or surgery). The second measure (T2) was conducted after the end of the treatment.

Figure 1. Participants in the study

Measures

Affective forecasting was measured with the Spanish version of the Positive and Negative Affect Schedule (PANAS) (Robles & Páez, 2003; Watson, Clark, & Tellegen, 1988). This scale consists of 20 items to evaluate both positive and negative affect at the present moment (Cronbach alpha = .82 to .86). It employs a 5-point Likert scale ranging from 1 = very slightly or not at all to 5 = extremely. At T1 predicted affect was assessed by asking patients: imagine how will you feel after the surgery/ after finishing with the neoadjuvant treatment? (depending on the treatment that they would follow). At T2 we assessed affect at the present moment. Regarding the accuracy of affective forecasting, previous research on affective forecasting has stressed that the accuracy or inaccuracy of the forecast depends on how accuracy is approached (Mathieu & Gosling, 2012). Thus, in this study accuracy of affective forecasting was computed considering both an absolute and relative sense. To compute the affective forecasting accuracy from an absolute sense, we subtracted actual scores at T2 from predicted scores at T1 (T1-T2) separately for negative and positive affect in each group. To compute the affective forecasting accuracy from a relative sense, we focus only on patients who following surgery after the diagnosis, since the patients following neoadjuvant treatment were a small group. First, we identified patients who predicted a low affect and a high affect, based on the median of the predicted negative affect (median = 19.00) and the predicted positive affect (median = 30.00),

Surgery (n=99) Neoadjuvant treatment (n=55) - 5 dropouts - 7 refused to predict - 3 initial treatment was changed - 1 assessment not meeting validity checks - 9 dropouts - 4 refused to predict - 3 palliative patients - 2 assessments not meeting validity checks Cases analyzed (n=83) Cases analyzed (n=37) Baseline measure (T1) n=173

Knowledge next treatment

(6)

6

the first aim, we hypothesized that on average patients would overestimate the negative affect that they would experience after the first medical treatment. As for the second aim, we hypothesized that there would be significant differences in the levels of psychological care needs after the surgery, according to the level of affective forecasting accuracy that patients made of the negative affect. Regarding the positive affect, we kept an exploratory approach.

METHOD

Participants and procedure

Data were obtained from a larger observational longitudinal study with newly diagnosed Mexican breast cancer patients. Eligible patients were women aged 18 to 75, who received a cancer diagnosis for the first time and had not started treatment at time of enrollment in the study. Further details about selection and recruitment of patients can be consulted elsewhere (Pérez-Fortis, Fleer, et al., 2017). A total of 173 patients participated in the baseline measure (67% response rate), and complete data for the present study were obtained from 120 patients. Individuals excluded because of incomplete data (N = 53) were older (mean age 55.2 years vs. 51.8 years, p = .05), and a lower proportion had middle or higher education (52.8% vs. 76%, p < 0.01), than patients with complete data. There were no statistically significant differences on the other sociodemographic characteristics between both groups. Sociodemographic characteristics of the participants were similar in the surgery and neoadjuvant treatment group, with the exception of breast cancer stage. Most of the patients in the neoadjuvant treatment group were in stage III, whereas most of the patients in the surgery group were in stage I or II (p<.001). More details of the descriptive characteristics of the sample are presented in supplemental Table 1.

Patients were classified into two groups: (1) patients who underwent surgery and (2) patients who followed neoadjuvant treatment after the breast cancer diagnosis. A description of the participants’ flow in the study is shown in Figure 1. The study was approved by the local ethics committee of the hospital (R-2014-3504-40), and the patients who participated in the study provided informed consent. Data were collected through a semi-structured interview by three trained psychologists. Patients were evaluated two times, the first measure (T1) was made after receiving the breast cancer diagnosis and before the start of treatment (either neoadjuvant treatment or surgery). The second measure (T2) was conducted after the end of the treatment.

Figure 1. Participants in the study Measures

Affective forecasting was measured with the Spanish version of the Positive and Negative Affect Schedule (PANAS) (Robles & Páez, 2003; Watson, Clark, & Tellegen, 1988). This scale consists of 20 items to evaluate both positive and negative affect at the present moment (Cronbach alpha = .82 to .86). It employs a 5-point Likert scale ranging from 1 = very slightly or not at all to 5 = extremely. At T1 predicted affect was assessed by asking patients: imagine how will you feel after the surgery/ after finishing with the neoadjuvant treatment? (depending on the treatment that they would follow). At T2 we assessed affect at the present moment. Regarding the accuracy of affective forecasting, previous research on affective forecasting has stressed that the accuracy or inaccuracy of the forecast depends on how accuracy is approached (Mathieu & Gosling, 2012). Thus, in this study accuracy of affective forecasting was computed considering both an absolute and relative sense. To compute the affective forecasting accuracy from an absolute sense, we subtracted actual scores at T2 from predicted scores at T1 (T1-T2) separately for negative and positive affect in each group. To compute the affective forecasting accuracy from a relative sense, we focus only on patients who following surgery after the diagnosis, since the patients following neoadjuvant treatment were a small group. First, we identified patients who predicted a low affect and a high affect, based on the median of the predicted negative affect (median = 19.00) and the predicted positive affect (median = 30.00),

Surgery (n=99) Neoadjuvant treatment (n=55)

- 5 dropouts - 7 refused to predict - 3 initial treatment was changed - 1 assessment not meeting validity checks - 9 dropouts - 4 refused to predict - 3 palliative patients - 2 assessments not meeting validity checks Cases analyzed (n=83) Cases analyzed (n=37) Baseline measure (T1) n=173

Knowledge next treatment

(7)

separately. This method was employed given that there are no defined cutoff scores for the PANAS. Then, we got a discrepancy score based on the subtraction of predicted scores at T1 minus actual scores at T2. Based on the results of the absolute sense approach, we settled a threshold of 3 points in the discrepancy score (statistically significant difference), to classify participants into 6 groups: underestimators of a low score, underestimators of a high score, overestimators of a low score, overestimators of a high score, predicted an accurate low score, and predicted an accurate high score.

Psychological care needs were measured at T2 with the psychological dimension of the Supportive Care Needs Survey short-form (SCNS-SF34) (Boyes, Girgis, & Lecathelinais, 2009). This dimension employs 10 items to assess the supportive psychological care needs of cancer patients during the last two weeks. The response format is a 5-point Likert scale type ranging from 1 = not applicable to 5 = high need. Details about the adaptation of the instrument into the Spanish language have been reported elsewhere (Pérez-Fortis, Fleer, et al., 2017). Cronbach alpha of this dimension was .91

Demographic and medical data included age, education, marital and working status, time since diagnosis, breast cancer stage, comorbidities, lapse between the application of the first interview and the surgery, and the lapse between the surgery and the application of the second interview.

Statistical Analyses

We identified one case with missing data in four items of the PANAS (two items from the negative and two items from the positive subscale), and one case with a missing item in the psychological care needs questionnaire. Both cases were imputed with the mean of each case. Descriptive statistics were computed on the characteristics of each sample group. Given that the psychological care needs of the patients did not follow a normal distribution and considering the small size of the groups, we used nonparametric tests to test the accuracy of affecting forecasting. First, Wilcoxon signed ranks tests were computed to explore whether participants accurately predicted their positive and negative affect from an absolute sense in both neoadjuvant treatment and surgery groups. Effect sizes were computed by converting a z-score into r estimate with the equation proposed by Rosenthal (1991). Second, Kruskal-Wallis tests were run to test whether the psychological care needs at post surgery differ according the affecting forecasting accuracy, from a relative sense, i.e., among the 6 groups. Effect sizes were computed by the epsilon-squared estimate (Tomczak & Tomczak, 2014). Pairwise multiple comparisons (Dunn, 1964) were computed whenever the Kruskal-Wallis test was significant. We used the SPSS software to analyze the data and we considered a p-value of 0.05 (two-sided) to test our hypothesis.

RESULTS

Affective forecasting accuracy

The results of the Wilcoxon signed ranks tests showed that, from an absolute sense approach, participants within the surgery group showed significantly lower levels of negative affect after the surgery (Mdn = 16.00) than what they predicted before the surgery (Mdn = 19.00), z = -3.17, p < .01, r = -.25 (Table 1). However, for positive affect there were no significant changes. Patients who followed neoadjuvant treatment did not show significant changes, for either positive or negative affect, between the predicted levels before starting the treatment and the subsequent actual levels after finishing the treatment. From a relative sense approach, within the surgery group 45% of the patients overestimated, 18% underestimated, and 37 % accurately predicted negative affect (Table 2). Regarding the positive affect 35% of the patients overestimated, 40% underestimated, and 25% accurately predicted it.

Table 1. Pairwise comparison of predicted and actual affect of participants, using Wilcoxon signed ranks tests.

Predicted at T1 Median Actual at T2 Median z p Effect size (r) Surgery (n = 83) Positive affect 30.00 32.00 -1.20 .231 -0.09 Negative affect 19.00 16.00 -3.17 .002 -0.25 Neoadjuvant treatment (n = 37) Positive affect 35.00 35.00 -1.82 .068 -0.21 Negative affect 15.00 16.00 -0.23 .816 -0.03

Differences in psychological care needs by affective forecasting accuracy

Table 2 shows the results of Kruskal-Wallis tests. We observed that the psychological care needs of the patients at post-surgery were significantly different according to their affective forecasting accuracy for negative affect H (5) = 18.912, p < .01, e2_{= .23. Pairwise} comparisons showed that patients who significantly underestimated a low (Mdn = 50.00) ora high score (Mdn = 68.75) of negative affect, that is, those patients who after the surgery had significantly higher scores than they expected, showed higher psychological care needs after the surgery than patients who overestimated a low negative affect (Mdn = 2.50), that is, those patients who after the surgery had significantly lower scores than they anticipated. Psychological care needs of the patients at post-surgery did not show significant differences by the affective forecasting accuracy for positive affect.

Table 2. Levels of psychological care needs after surgery according to the affective forecasting accuracy.

(8)

6

separately. This method was employed given that there are no defined cutoff scores for the PANAS. Then, we got a discrepancy score based on the subtraction of predicted scores at T1 minus actual scores at T2. Based on the results of the absolute sense approach, we settled a threshold of 3 points in the discrepancy score (statistically significant difference), to classify participants into 6 groups: underestimators of a low score, underestimators of a high score, overestimators of a low score, overestimators of a high score, predicted an accurate low score, and predicted an accurate high score.

Psychological care needs were measured at T2 with the psychological dimension of the Supportive Care Needs Survey short-form (SCNS-SF34) (Boyes, Girgis, & Lecathelinais, 2009). This dimension employs 10 items to assess the supportive psychological care needs of cancer patients during the last two weeks. The response format is a 5-point Likert scale type ranging from 1 = not applicable to 5 = high need. Details about the adaptation of the instrument into the Spanish language have been reported elsewhere (Pérez-Fortis, Fleer, et al., 2017). Cronbach alpha of this dimension was .91

Demographic and medical data included age, education, marital and working status, time since diagnosis, breast cancer stage, comorbidities, lapse between the application of the first interview and the surgery, and the lapse between the surgery and the application of the second interview.

Statistical Analyses

We identified one case with missing data in four items of the PANAS (two items from the negative and two items from the positive subscale), and one case with a missing item in the psychological care needs questionnaire. Both cases were imputed with the mean of each case. Descriptive statistics were computed on the characteristics of each sample group. Given that the psychological care needs of the patients did not follow a normal distribution and considering the small size of the groups, we used nonparametric tests to test the accuracy of affecting forecasting. First, Wilcoxon signed ranks tests were computed to explore whether participants accurately predicted their positive and negative affect from an absolute sense in both neoadjuvant treatment and surgery groups. Effect sizes were computed by converting a z-score into r estimate with the equation proposed by Rosenthal (1991). Second, Kruskal-Wallis tests were run to test whether the psychological care needs at post surgery differ according the affecting forecasting accuracy, from a relative sense, i.e., among the 6 groups. Effect sizes were computed by the epsilon-squared estimate (Tomczak & Tomczak, 2014). Pairwise multiple comparisons (Dunn, 1964) were computed whenever the Kruskal-Wallis test was significant. We used the SPSS software to analyze the data and we considered a p-value of 0.05 (two-sided) to test our hypothesis.

RESULTS

Affective forecasting accuracy

The results of the Wilcoxon signed ranks tests showed that, from an absolute sense approach, participants within the surgery group showed significantly lower levels of negative affect after the surgery (Mdn = 16.00) than what they predicted before the surgery (Mdn = 19.00), z = -3.17, p < .01, r = -.25 (Table 1). However, for positive affect there were no significant changes. Patients who followed neoadjuvant treatment did not show significant changes, for either positive or negative affect, between the predicted levels before starting the treatment and the subsequent actual levels after finishing the treatment. From a relative sense approach, within the surgery group 45% of the patients overestimated, 18% underestimated, and 37 % accurately predicted negative affect (Table 2). Regarding the positive affect 35% of the patients overestimated, 40% underestimated, and 25% accurately predicted it.

Table 1. Pairwise comparison of predicted and actual affect of participants, using Wilcoxon signed ranks tests.

Predicted at T1 Median Actual at T2 Median z p Effect size (r) Surgery (n = 83) Positive affect 30.00 32.00 -1.20 .231 -0.09 Negative affect 19.00 16.00 -3.17 .002 -0.25 Neoadjuvant treatment (n = 37) Positive affect 35.00 35.00 -1.82 .068 -0.21 Negative affect 15.00 16.00 -0.23 .816 -0.03

Differences in psychological care needs by affective forecasting accuracy

Table 2 shows the results of Kruskal-Wallis tests. We observed that the psychological care needs of the patients at post-surgery were significantly different according to their affective forecasting accuracy for negative affect H (5) = 18.912, p < .01, e2_{= .23. Pairwise} comparisons showed that patients who significantly underestimated a low (Mdn = 50.00) or a high score (Mdn = 68.75) of negative affect, that is, those patients who after the surgery had significantly higher scores than they expected, showed higher psychological care needs after the surgery than patients who overestimated a low negative affect (Mdn = 2.50), that is, those patients who after the surgery had significantly lower scores than they anticipated. Psychological care needs of the patients at post-surgery did not show significant differences by the affective forecasting accuracy for positive affect.

Table 2. Levels of psychological care needs after surgery according to the affective forecasting accuracy.

(9)

Psychological care needs after surgery

Pairwise comparisons n (%) Median _SquareChi- a

Effec t size

(e2) Negative affect accuracy 18.912** .23 Accurate Low score (AL) 22 (27) 20.00

Underestimate Low score (UL) 9 (11) 50.00 UL-OL*

Overestimate Low score (OL) 7 (8) 2.50 Accurate High score (AH) 8 (10) 26.25

Underestimate High score (UH) 6 (7) 68.75 UH-OL**

Overestimate High score (OH) 31 (37) 25.00

Positive affect accuracy 10.659 .13 Accurate Low score (AL) 9 (11) 52.50

Underestimate Low score (UL) 24 (29) 27.50 Overestimate Low score (OL) 5 (6) 70.00 Accurate High score (AH) 12 (14) 12.50 Underestimate High score (UH) 9 (11) 20.00 Overestimate High score (OH) 24 (29) 18.75

a_{From Kruskal-Wallis H-test statistic. *p< .05 **p< .01}

DISCUSSION

In general, Mexican breast cancer patients accurately (absolute sense) predicted positive and negative affect that they would experience after finishing neoadjuvant treatment. Patients who followed surgery after breast cancer diagnosis, on average, overestimated (absolute sense) the negative affect, yet, the effect was small. From a relative sense, most of the patients who followed surgery were inaccurate in predicting their negative and positive affect. Particularly for negative affect almost half of the sample overestimated their predictions. Patients who underestimated the negative affect to the surgery showed the highest psychological care needs after the surgery.

Our results partially confirmed our first hypothesis, only patients who followed surgery as first treatment after diagnosis significantly overestimated the negative affect, which is in line with previous studies suggesting that, in general, people tend to overestimate the negative affect, especially when they predict emotions regarding an event they consider negative (Ayton et al., 2007; Wilson & Gilbert, 2005). Overestimation of negative affect might be related with two cognitive failures that patients make when predicting their affective states (Stiggelbout & de Vogel-Voogt, 2008; Wilson & Gilbert, 2005), which could play a role in the reactions of cancer patients to their diagnosis and treatment (Rhodes & Strain, 2008). The first one is focalism, in which individuals disregard the influence of other events on their cognitions and emotions (Wilson, Wheatley, Meyers, Gilbert, & Axsom, 2000). Thus, cancer patients might mainly focus on how the disease will change their life and they might disregard what will remain equal. The second one is unanticipated adaptation, in which people do not anticipate an eventual adaptation to their

new status (Damschroder, Zikmund-Fisher, & Ubel, 2008; Stiggelbout & de Vogel-Voogt, 2008). Thus, cancer patients might not be aware of their underlying cognitive resources to cope with changed circumstances or to emotionally adapt to their condition over time. Surprisingly, patients who followed neoadjuvant treatment predicted a rather accurate negative affect. This finding is interesting because overall affective forecasting literature has suggested that individuals tend to be inaccurate in the absolute sense (Mathieu & Gosling, 2012), but in the present study we did not find that pattern. We cannot certainty state what factors influence some patients to be more accurate than others, but we anticipate that it might be related with the type of focal event, the previous information that patients have over the focal event, or specifically in this study, the stage of the disease, or the representation that patients have over the illness and treatment. Thus, our results are in favor of recent studies suggesting that people are both accurate and inaccurate when predicting affect for a focal event in the future (Coteţ & David, 2016; Levine, Lench, Kaplan, & Safer, 2012), and opposite to the claim that people are, in general, poor predicting their affect (Ayton et al., 2007; Wilson & Gilbert, 2003). Overall, these findings suggest that there are some Mexican breast cancer patients who may be overwhelmed by the diagnosis and the previous distress associated with the surgery; thus, they make an impact bias when predicting their affective states (Wilson & Gilbert, 2003). However, there were also patients who anticipated their adaptation in a positive or negative way to their condition. Previous studies have also showed that patients are able to anticipated adaptation, especially at the onset of the disease (Peeters, Ranchor, & Stiggelbout, 2010; Stiggelbout & de Vogel-Voogt, 2008).

Regarding our second aim, as hypothesized, there were differences in psychological care needs at post-surgery by the affective forecasting accuracy in negative affect (relative sense), although the effect was small. Regardless a low or high score in negative affect, patients who underestimated the negative affect that they would experience after the surgery showed higher psychological care needs compared particularly to the patients who overestimated a low score of negative affect. It seems that the patients who overestimated the negative affect that they would have post-surgery were maybe pessimistic and anticipated a poor adaptation, but had more psychological resources to cope with the recovery, therefore they showed the lowest psychological care needs. This finding is interesting because the previous literature on affective forecasting has only highlighted the drawbacks of overestimating negative affective states (Wilson & Gilbert, 2005). However, in the context of the psychological adaptation to a chronic disease, it was interesting to see that the overestimators of negative affect had lower psychological care needs after the surgery.

Our findings might be somewhat limited by (a) the sample size which was rather small when we classified the patients in terms of accuracy level, then we could have lost power to detect significant effects, (b) univariate analyses were conducted to test our

(10)

6

Psychological care needs after surgery

Pairwise comparisons n (%) Median _SquareChi- a

Effec t size

(e2)

Negative affect accuracy 18.912** .23

Accurate Low score (AL) 22 (27) 20.00

Underestimate Low score (UL) 9 (11) 50.00 UL-OL*

Overestimate Low score (OL) 7 (8) 2.50

Accurate High score (AH) 8 (10) 26.25

Underestimate High score (UH) 6 (7) 68.75 UH-OL**

Positive affect accuracy 10.659 .13

Accurate Low score (AL) 9 (11) 52.50

Underestimate Low score (UL) 24 (29) 27.50

Overestimate Low score (OL) 5 (6) 70.00

Accurate High score (AH) 12 (14) 12.50

Underestimate High score (UH) 9 (11) 20.00

a_{From Kruskal-Wallis H-test statistic. *p< .05 **p< .01}

DISCUSSION

In general, Mexican breast cancer patients accurately (absolute sense) predicted positive and negative affect that they would experience after finishing neoadjuvant treatment. Patients who followed surgery after breast cancer diagnosis, on average, overestimated (absolute sense) the negative affect, yet, the effect was small. From a relative sense, most of the patients who followed surgery were inaccurate in predicting their negative and positive affect. Particularly for negative affect almost half of the sample overestimated their predictions. Patients who underestimated the negative affect to the surgery showed the highest psychological care needs after the surgery.

Our results partially confirmed our first hypothesis, only patients who followed surgery as first treatment after diagnosis significantly overestimated the negative affect, which is in line with previous studies suggesting that, in general, people tend to overestimate the negative affect, especially when they predict emotions regarding an event they consider negative (Ayton et al., 2007; Wilson & Gilbert, 2005). Overestimation of negative affect might be related with two cognitive failures that patients make when predicting their affective states (Stiggelbout & de Vogel-Voogt, 2008; Wilson & Gilbert, 2005), which could play a role in the reactions of cancer patients to their diagnosis and treatment (Rhodes & Strain, 2008). The first one is focalism, in which individuals disregard the influence of other events on their cognitions and emotions (Wilson, Wheatley, Meyers, Gilbert, & Axsom, 2000). Thus, cancer patients might mainly focus on how the disease will change their life and they might disregard what will remain equal. The second one is unanticipated adaptation, in which people do not anticipate an eventual adaptation to their

new status (Damschroder, Zikmund-Fisher, & Ubel, 2008; Stiggelbout & de Vogel-Voogt, 2008). Thus, cancer patients might not be aware of their underlying cognitive resources to cope with changed circumstances or to emotionally adapt to their condition over time. Surprisingly, patients who followed neoadjuvant treatment predicted a rather accurate negative affect. This finding is interesting because overall affective forecasting literature has suggested that individuals tend to be inaccurate in the absolute sense (Mathieu & Gosling, 2012), but in the present study we did not find that pattern. We cannot certainty state what factors influence some patients to be more accurate than others, but we anticipate that it might be related with the type of focal event, the previous information that patients have over the focal event, or specifically in this study, the stage of the disease, or the representation that patients have over the illness and treatment. Thus, our results are in favor of recent studies suggesting that people are both accurate and inaccurate when predicting affect for a focal event in the future (Coteţ & David, 2016; Levine, Lench, Kaplan, & Safer, 2012), and opposite to the claim that people are, in general, poor predicting their affect (Ayton et al., 2007; Wilson & Gilbert, 2003). Overall, these findings suggest that there are some Mexican breast cancer patients who may be overwhelmed by the diagnosis and the previous distress associated with the surgery; thus, they make an impact bias when predicting their affective states (Wilson & Gilbert, 2003). However, there were also patients who anticipated their adaptation in a positive or negative way to their condition. Previous studies have also showed that patients are able to anticipated adaptation, especially at the onset of the disease (Peeters, Ranchor, & Stiggelbout, 2010; Stiggelbout & de Vogel-Voogt, 2008).

Regarding our second aim, as hypothesized, there were differences in psychological care needs at post-surgery by the affective forecasting accuracy in negative affect (relative sense), although the effect was small. Regardless a low or high score in negative affect, patients who underestimated the negative affect that they would experience after the surgery showed higher psychological care needs compared particularly to the patients who overestimated a low score of negative affect. It seems that the patients who overestimated the negative affect that they would have post-surgery were maybe pessimistic and anticipated a poor adaptation, but had more psychological resources to cope with the recovery, therefore they showed the lowest psychological care needs. This finding is interesting because the previous literature on affective forecasting has only highlighted the drawbacks of overestimating negative affective states (Wilson & Gilbert, 2005). However, in the context of the psychological adaptation to a chronic disease, it was interesting to see that the overestimators of negative affect had lower psychological care needs after the surgery.

Our findings might be somewhat limited by (a) the sample size which was rather small when we classified the patients in terms of accuracy level, then we could have lost power to detect significant effects, (b) univariate analyses were conducted to test our

(11)

hypothesis, and other third variables such as the physical symptoms of the disease and psychological complaints might influence the results. Therefore, the findings of the present study must be interpreted cautiously.

To the best of our knowledge, this study is the first examining the role of a cognitive process, such as affective forecasting accuracy in relation to psychological care needs among breast cancer patients. Future work on the field of emotional anticipation should explore the psychological and physical characteristics of the patients that are associated with the accuracy of their predictions, using multivariate statistical techniques and incorporating measures of the cognitive mechanisms that might influence the accuracy of the prediction. We also suggest replicating this study with larger sample sizes and perhaps using samples with different clinical impairments, since this topic has been little explored within the field of health psychology. The findings of the present study contribute to the knowledge on affective forecasting accuracy within the field of health psychology. Overall, the preliminary results of this study suggest that (1) not all the patients unanticipated the adaptation to their new condition when they predict their affective states, and (2) the psychological care needs of Mexican breast cancer patients might differ by the cognitive process of accurately forecasting affects regarding a specific event during the cancer treatment trajectory.

REFERENCES

Ayton, P., Pott, A., & Elwakili, N. (2007). Affective forecasting: Why can’t people predict their emotions? Thinking & Reasoning, 13(1), 62-80. https://doi.org/10.1080/13546780600872726

Boyes, A., Girgis, A., & Lecathelinais, C. (2009). Brief assessment of adult cancer patients’ perceived needs: development and validation of the 34-item Supportive Care Needs Survey (SCNS-SF34). Journal of Evaluation in Clinical Practice, 15(4), 602-606. https://doi.org/10.1111/j.1365-2753.2008.01057.x Brandão, T., Schulz, M. S., & Matos, P. M. (2017). Psychological adjustment after breast cancer: a systematic

review of longitudinal studies. Psycho-Oncology, 26(7), 917-926. https://doi.org/10.1002/pon.4230 Buehler, R., & McFarland, C. (2001). Intensity Bias in Affective Forecasting: The Role of Temporal Focus.

Personality and Social Psychology Bulletin, 27(11), 1480-1493. https://doi.org/10.1177/01461672012711009

Coteţ, C. D., & David, D. (2016). The truth about predictions and emotions: Two meta-analyses of their relationship. Personality and Individual Differences, 94, 82-91. https://doi.org/10.1016/j.paid.2015.12.046

Damschroder, L. J., Zikmund-Fisher, B. J., & Ubel, P. A. . (2008). Considering adaptation in preference elicitations. Health Psychology, 27(3), 394-399. https://doi.org/10.1037/0278-6133.27.3.394

Dorval, M., Patenaude, A. F., Schneider, K. A., Kieffer, S. A., DiGianni, L., Kalkbrenner, K. J., … others. (2000). Anticipated versus actual emotional reactions to disclosure of results of genetic tests for cancer susceptibility: findings from p53 and BRCA1 testing programs. Journal of Clinical Oncology, 18(10), 2135– 2142.

Dunn, O. J. (1964). Multiple Comparisons Using Rank Sums. Technometrics, 6(3), 241-252. https://doi.org/10.2307/1266041

Halpern, J., & Arnold, R. M. (2008). Affective Forecasting: An Unrecognized Challenge in Making Serious Health Decisions. Journal of General Internal Medicine, 23(10), 1708-1712. https://doi.org/10.1007/s11606-008-0719-5

Hoerger, M., Chapman, B., & Duberstein, P. (2016). Realistic affective forecasting: The role of personality. Cognition and Emotion, 30(7), 1304-1316. https://doi.org/10.1080/02699931.2015.1061481

Levine, L. J., Lench, H. C., Kaplan, R. L., & Safer, M. A. (2012). Accuracy and artifact: Reexamining the intensity bias in affective forecasting. Journal of Personality and Social Psychology, 103(4), 584-605. https://doi.org/10.1037/a0029544

Mathieu, M. T., & Gosling, S. D. (2012). The Accuracy or Inaccuracy of Affective Forecasts Depends on How Accuracy Is Indexed: A Meta-Analysis of Past Studies. Psychological Science, 23(2), 161-162. https://doi.org/10.1177/0956797611427044

Peeters, Y., Ranchor, A., & Stiggelbout, A. M. (2010). Direct and indirect effects of adaptation on health state utilities. Psychology & Health, 25, 70-70.

Pérez-Fortis, A., Fleer, J., Sánchez-Sosa, J. J., Veloz-Martínez, M. G., Alanís-López, P., Schroevers, M. J., & Ranchor, A. V. (2017). Prevalence and factors associated with supportive care needs among newly diagnosed Mexican breast cancer patients. Supportive Care in Cancer, 25(10), 3273-3280. https://doi.org/10.1007/s00520-017-3741-5

Pérez-Fortis, A., Schroevers, M. J., Fleer, J., Alanís-López, P., Veloz-Martínez, M. G., Ornelas-Mejorada, R. E., … Sánchez Sosa, J. J. (2017). Psychological burden at the time of diagnosis among Mexican breast

(12)