Factors associated with the decline of psychological support in hospitalized patients with cancer

Many distressed cancer patients do not want or, finally, do not use psychological support. This study aimed at identifying factors associated with the decline of psychological support during hospital stay.


| BACKGROUND
A persistent issue in psycho-oncological care of patients with cancer is how to define indication for psychological support. This can be accomplished by expert ratings 1 or by patient self-report on distress or perceived need. 2,3 Typically, expert rating of distress, patient-reported level of psychosocial distress, and patient-reported perceived need correlate only weakly. 3,4 The German Clinical Practice Guideline for Psycho-Oncology, 5 thus, recommends to screen for psychosocial distress and to assess the subjective desire for psychological support.
Research shows that up to 52% of cancer patients indicate clinical levels of psychosocial distress. 6,7 Moreover, about one-third suffers from a mental disorder. 8,9 However, when psychological support is offered, more than half of the distressed cancer patients do not want or, finally, do not use psychological support. 10,11,13 These results suggest that the mere presence of elevated psychosocial distress does not drive patients to access psychological support. In addition, the subjective evaluation concerning the need and adequacy of psychological support has to be taken into account. 3 Studies on the subjective perspective of psychological support use have applied various concepts like subjective need, desire, or intent to use support. Female sex, younger age, and higher education were found to be associated with more needs, whereas living with a partner and being married were associated with fewer needs. 14 Merckaert et al 15 found an association between increased desire for support and younger age and female sex. Baker-Glenn et al 10 reported associations with distress, anxiety, and depression. Furthermore, they identified subjective reasons for not using psychological support: "receiving informal help elsewhere" and a "preference to manage on one's own" were the most common reasons for declining professional support among distressed patients. Studies examining the future intent to use psycho-oncological services suggest that subjective norms and outcome expectations and attitudes towards help seeking 16 and mental health literacy represent relevant factors. 17 Past research on the acceptance and decline of psychological support is characterized by the inclusion of different services, heterogeneity of samples, and variability of settings. Furthermore, large-scale studies that investigate a wide range of variables possibly associated with desire for psychological support are currently lacking.
Therefore, we investigated sociodemographic, clinical, and psychological factors and their association with the decline of psychological support in hospitalized patients with cancer.

| METHODS
A cross-sectional study was conducted in the two university hospitals of the Comprehensive Cancer Center Munich (CCC Munich), Germany. Local ethic committees approved the study (file numbers: 238/16S; 402-16). Participating departments at both hospitals were the departments for gynecology, urology, and radiation oncology.

| Participants and setting
Eligible participants were adult (greater than or equal to 18 years) German-speaking inpatients with a verified diagnosis of a malignant tumor. Exclusion criteria were physical, mental, or verbal impairments (clinical assessment by the treating physician) that interfered with the ability to give informed consent and to fill out the self-report questionnaire.
The study took place in the context of routine clinical care. In Germany, it is mandatory for a certified cancer center to provide a psycho-oncology service. In both university hospitals of the CCC Munich, patients are referred to the psycho-oncological service either by exceeding the cut-off for clinical distress on the distress screening measure or by ticking the box indicating a subjective need (regardless of the score on the distress screening). In addition, the treating physician refers the patient to the psycho-oncological service if she notices a need for clinical assessment or supportive care (regardless of the distress screening).

| Study procedure
An algorithm within the patient data base was implemented to identify all inpatients with a verified tumor diagnosis in the recruiting departments. A list was provided daily in order to document new admissions. Undergraduate students contacted the treating oncologists to exclude any contraindication for participation. Eligible patients were informed about the study. Those who agreed to participate signed an informed consent sheet. The questionnaire set was handed out to the patient for completion; sets were returned during hospital stay. The data collection took place between 01.08.2016 and 01.10.2017.

| Sociodemographic characteristics and clinical data
The following sociodemographic characteristics were assessed: age, sex, marital status, children, education, and work situation. Clinical data were assessed by patients' self-report and comprised tumor entity, date of initial diagnosis, disease status, metastases, and current treatment. Furthermore, patients rated their current physical condition using a visual analogue scale (1 = "excellent" to 10 = "very poor").

| Psychosocial distress
Psychosocial distress was assessed with two self-reporting questionnaires that are implemented as routine screening measures in the two hospitals of the CCCM: the Questionnaire on Stress in Cancer Patients-Revised (QSC-R10) 18 that is routinely used in one of the two university hospitals and the Distress Thermometer (DT) 19 that is used in the other one.
The QSC-R10 consists of 10 items. Patients answer whether or not each problem applies to them and-if it does-how distressed they feel (0 = "the problem does not apply to me", 5 = "the problem does apply and causes severe distress"). A cutoff ≥ 15 indicates clinical distress (Cronbach's alpha in the current sample: α = .87).
The DT assesses distress using a visual analogue scale from 0 ("no distress") to 10 ("extreme distress"). A cutoff ≥ 6 indicates clinically significant distress, based on previous research. 3,20

Depressive symptoms
The ultrashort version of the Patient Health Questionnaire (PHQ-2), 21,22 which consists of two items assessing depressive symptoms over the past 2 weeks was used. The items are scored from 0 ("not at all") to 3 ("nearly every day"). Cronbach's alpha in the current sample was α = .73. A sum score ≥ 3 indicates clinical depression. For use in the logistic regression, we classified patients as below or above the cutoff.

Self-efficacy
The short form of the German version of the General Self-Efficacy Scale (ASKU) 23 consists of three items, which are rated on a 5-point scale from 1 ("does not apply at all") to 5 ("applies completely"). A higher mean score indicates higher self-efficacy (Cronbach's alpha in the current sample: α = .89).

Personality
According to the Five-Factor Model of Personality, personality traits were measured using the short form of the Big Five Inventory (BFI-10). 24 The questionnaire consists of 10 items that measure extraversion, agreeableness, conscientiousness, neuroticism, and openness. 24 Higher scores suggest stronger manifestations of each of the personality traits.

Social support
Social support was assessed with one item asking for the presence of a confidant relationship (response options were "yes" and "no").

Information about psychological support and previous use of psychological treatments
Information about psychological support was assessed with the item "Do you feel well informed about the psychological support offered in this hospital?" Response options were "yes" and "no." We further asked patients about previous uptake of psychological treatments ("Have you ever been in psychological treatment?"). Here, response options were "yes, due to my cancer." "Yes, because of other problems" and "no." The answers were categorized in "yes" and "no."

| Desire for psychological support
To determine the desire for psychological support, participants were asked "Do you want psychological support during your stay in hospital?" Response options were "yes" and "no."

| Statistical analysis
Mean values, standard deviation, and frequencies were reported for descriptive purposes. Between-group comparisons were performed using chi-square test or independent t test. We used multivariable binary logistic regression analyses to identify variables associated with a decline of psychological support. Two models were performed, one for the total sample and a second one for the subsample of distressed patients only. For the total sample, the following variables were added: depression, attitudes, and information), based on previous research. 25 As the items assessing personal attitudes were newly developed, we refrained from computing a scale with unclear reliability and validity.
Instead, for each dimension, we used the one item with the strongest face validity because of consensus rating. All statistical tests were two-tailed. Results P < .05 were regarded as statistically significant.   (Table 1)

| Distress and decline of psychological support
Of 925 inpatients, 28

| Variables associated with decline: All patients (model 1)
Distress was the strongest predictor of decline of support, followed by perceived overload (Table 2, model 1). Men declined psychological support more than 2.5 times more often than women. Patients without depressive symptoms were nearly twice as likely to decline support as patients with depressive symptoms. Patients without previous uptake of psychological support declined more often than patients who had used psychological treatments. Patients who felt well informed about psychological support offers declined more frequently than patients who did not. The model for the total sample showed an explained variance of Nagelkerk's R 2 = .367.

| Variables associated with decline: Distressed patients (model 2)
In the group of highly distressed patients, perceived overload proved to be the strongest predictor of decline, followed by sex. Furthermore, feeling well informed and not feeling depressed also seemed to be associated with decline, although the overall test did not reach significance due to the category of missing responses (Table 2, model 2).
Additionally, patients with low agreeableness were more prone to decline support. Explained variance of this model was Nagelkerk's R 2 = .344.

| CONCLUSIONS
Many patients with cancer decline professional psychological support.
Screening for psychosocial distress and subsequent referral for psychological support has been regarded as a hallmark of timely patientcentered psychosocial care. Many efforts have been undertaken to implement distress screening and psychological support services in cancer centers. 26,27 However, research shows that there exists a complex interplay between distress, desire for, acceptance of, and uptake of psychological support. As the self-evaluation of the desire for support seems crucial, 3 we investigated variables associated with decline of psychological support offerings.
A total of 71.6% inpatients declined psychological support. In the group of highly distressed cancer patients, still 53.9% declined. It should be noted that we asked patients, "Do you want psychological support during your stay in hospital?" This narrow focus might have reduced the number of patients who agreed with the support offering, as some patients who declined might want psychological support after discharge from the hospital. Previous studies often did not specify a time frame. However, decline rates were similar across the different studies, with decline rates of 68% to 80% in unselected samples and 49% to 71% in patients with significant distress. [10][11][12][13][14][15] Thus, the decline rates of our study are quite compatible with the available evidence. Psychosocial distress showed the strongest association with the decline for psycho-oncological support, ie, low psychosocial distress was predictive of decline. This result is in line with the majority of studies 10,16,28,29 ; contradictory results, however, have also been reported. 30 In addition, not feeling depressed was also uniquely associated with decline. This is comparable with previous research investigating desire or uptake rates for psychological support. 10 These results are noteworthy and should be validated in future studies.

| Study limitations
The strengths of our study are the large sample size and the inclusion of various variables pertaining to different domains. However, some limitations have to be noted. The patients were recruited in two Note. Outcome variable: "Do you wish psychological support during your stay in hospital?" Response options were "yes" and "no." *"An additional appointment with a psychologist/psycho-oncologist would be too demanding, as I am very busy undergoing medical treatments." Abbreviations: CI, confidence intervals; OR, odds ratios; n.a., not applicable; P, P values.
hospitals of a comprehensive cancer center with well-established psycho-oncological care and thus may not be representative of other hospitals. Furthermore, as this study was conducted with patients undergoing inpatient treatment, the results cannot be generalized to the outpatient setting and to long-term cancer survivors. Moreover, there are some differences in routine screening and provision of psychological support in the two hospitals, which might have imposed some bias. Finally, this is a cross-sectional study that captures the desire for psychological support in acute care. We did not investigate current or future uptake of psychological support.

| Clinical implications
The decline of psychological support is primarily due to psychological factors, such as distress, feeling overload, and information status. As feeling well informed about support emerged as a relevant factor associated with decline, design of information material and education about available psychological services seem crucial. In further consequence, even patients who feel in overload should know that they can have a short contact and that psycho-oncologists are guided by patients' needs.