Screening and treatment of psychological distress in patients with metastatic colorectal cancer: study protocol of the TES trial

The study is designed as a multicenter cluster randomized controlled clinical trial with two treatment arms. Sixteen hospitals, each as a separate unit, are randomly assigned to either the TES program for screening and treatment of psychological distress or usual care. Psychological distress, secondary outcomes and costs are evaluated at baseline (T0), after 3 weeks (T3), 10 weeks (T10), 24 weeks (T24) and 48 weeks (T48). The design is illustrated in Figure 1.

The trial has been approved by the Medical Ethics Committee of the VU university medical center and the institutional ethics committees of the other participating sites. The study is conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines.

This multicenter study is carried out in one large Dutch university medical center and 15 teaching hospitals. Patients are recruited from the outpatient clinics and inpatient wards at the Departments of Medical Oncology. Patients are eligible to participate in the study if they meet the following inclusion criteria: diagnosis of metastatic colorectal disease, start of treatment with 1st line chemotherapy, and life expectancy of more than 3 months. Exclusion criteria for participation are: age < 18 or > 85 years, insufficient command of the Dutch language, recent psychotherapy (in the past 3 months, at least one session every 2 weeks), severe psychopathology, and no informed consent.

To allocate patients to either the TES program or to care as usual, cluster randomization is used, with hospitals as the unit of randomization to avoid contamination between groups. The randomization procedure is performed by a statistician who is blinded for hospital characteristics to ensure concealment of treatment allocation. The random allocation of hospitals is performed prior to patient recruitment. Accordingly, all participants within the same hospital are randomized into the same treatment condition (i.e. TES or usual care). Due to the nature of the intervention, neither oncologists, nurses, psychologists or their patients can be blinded. However, scoring of outcome and statistical analysis is performed blindly.

Targeted selection is performed by a trained nurse/trained clinical nurse specialist. Screening for psychological distress is implemented for patients starting with chemotherapy, and 10 and 18 weeks later (S0, S10, and S18 – see Figure 2). The HADS (Hospital Anxiety and Depression Scale) and “Lastmeter” (Distress Thermometer and corresponding Problem List) are used as screening tools. The HADS consists of 14 questions. There are two subscales assessing anxiety and depression, which can be combined into one scale assessing psychological distress, with scores ranging from 0 to 42. The cut off score for the combined anxiety and depression scale in cancer patients is ≥ 13 [29]. The “Lastmeter” (a combination of the Distress Thermometer ranging from 0 to 10 with the Problem List and a single question on the need to talk to a professional)[30] evaluates a wide range of problems and is recommended in the Dutch guideline “Detecting the need for psychological care” [31]. In line with this Dutch guideline, and supported by various studies [30,32] we apply a cut off score of 5. The trained nurse evaluates the distress scores and offers treatment to patients who score above the cut off points on either one or both of the tools, and to patients expressing the need to talk to a professional. Furthermore, the trained nurse informs the oncologist when a patient has psychological distress or starts with stepped care.

The entire treatment process is managed by the trained nurse, who coordinates screening of all patients at the appropriate moments in time (see above); evaluates the distress scores, discusses with the patient whether he/she wants to be involved in the treatment program (regardless of the distress score) and informs the oncologist about the outcome; manages and monitors the various steps in the treatment process, as well as the transition of the patient from one step to the next step; and provides guidance for the patient during the entire treatment process.

Treatment for psychological distress is delivered according to the stepped care principle [33]. The steps include: (I) watchful waiting, (II) a guided self-help program, (III) face to face problem solving treatment (PST), and (IV) psychotherapy, medication or a referral to other services (e.g. social work). Patients are screened after each step, and if psychological distress persists or if patients express the need for further guidance, a next treatment step is offered (Figure 2).

Preferably, treatment starts at step I, and then progresses to step II, III or IV, as needed. However, if the evaluation of the screening results indicates so, step I or any other step can be skipped and treatment may start at step II, III or IV. Criteria for skipping steps include major depressive disorder, family problems, severe risk of mood disturbance (risk factors include other major life events, history of psychiatric problems, young age, and little social support), the wish of the patient to skip a step, or the opinion of the oncologist that a step needs to be skipped.

Patients screening positive for distress at S10 or S18 are still allowed the full stepped care treatment.

The first step consists of 3 weeks “watchful waiting”, as patients with psychological distress often recover spontaneously over time. This step can be chosen twice.

If distress still persists, a guided self-help program is offered to the patient. A brief intervention for problem solving in adults experiencing common psychological symptoms such as anxiety and depression [34] has been adjusted for patients with mCRC. Both international and national randomized controlled studies have shown that problem solving treatment is effective in treating psychological distress [34-36]. Moreover, there is evidence from a growing number of trials that psychological treatments can be effectively delivered over the internet. A meta-analysis revealed that the effects of internet-based treatments for depression and anxiety disorders are comparable to those of face to face treatments [37,38]. The intervention is available as a web based version and as a booklet. The intervention takes approximately 5–7 weeks in total. In this period, patients describe what they think is important in their lives, make a list of their problems and concerns, and divide these into three categories: important and solvable problems, unimportant problems (problems that are not related to what is important in their lives); and important but unsolvable problems (such as losing someone through death). For the important and solvable problems a six-step procedure of problem solving is provided. The patient analyzes the problem and generates alternative solutions; selects and implements the chosen solution; and evaluates the results and prepares for the future. For the important but unsolvable problems, the patient receives advice and suggestions to better cope with these. Coaching is performed by the researcher or research assistant and consists of brief, weekly contacts by e-mail or telephone, which take about 10 to 15 minutes per week. The coaching is meant to give support in working through the self-help method and is not aimed at developing a patient-therapist relationship.

After completing step II, all patients are screened for psychological distress. In patients still meeting the criterion for psychological distress or with the desire of psychosocial support, a diagnostic evaluation and needs assessment is made by the psychologist. The results are used by the psychologist and patient to analyze problems and to evaluate further need for treatment. Together, the patient and psychologist decide on the next treatment step: face-to-face problem solving treatment (step III) or psychotherapy, medication or a referral (step IV), depending on needs and preferences of the patient. This problem analysis is performed in a maximum of 4 weeks.

In addition the Composite International Diagnostic Interview (CIDI) is used for the classification of symptoms. This is a comprehensive, fully-structured interview for the assessment of mental disorders according to the definitions and criteria of ICD-10 and DSM-IV [39] and is conducted by the research assistant by telephone.

Problem solving treatment has shown to be effective in treating psychological distress in a wide range of somatic conditions, including cancer [40,41]. Problem solving treatment (face-to-face) is provided by the trained nurse, with the individual treatment plan serving as a guide. The treatment has a maximum of six sessions which take place in approximately 10 weeks. The treatment plan consists of identified problems, need for care, personalized goals, treatment and time for evaluation specified by the psychologist. The nurse is thoroughly trained in problem solving treatment, and follows a manual in delivering treatment [42]. If indicated, the partner is involved in one of the sessions.

If psychological distress persists after step (III) or if indicated after diagnostic evaluation by the psychologist, the patient is offered psychotherapy, medication prescribed by the consultant psychiatrist, or a referral to other services (e.g. social work). Psychotherapy is delivered according to Dutch guidelines [43,44]. Suggested medication includes, among others, SSRI’s and benzodiazepines. Occasionally, antipsychotics or mood stabilizers may be needed.

To create a collaborative process the medical oncologist is informed and updated about inclusion of patients in the study and treatment provided. Besides, all psychologists and nurses are thoroughly trained in the TES protocol, including the procedures for targeted selection, enhanced care and stepped care. This is done in a 1-day TES workshop. Furthermore, the approach to targeted selection, enhanced care and the various steps in the stepped care program are supervised by the psychologist from VU University medical center.

Detection and management of psychological distress in hospitals assigned to the control group is not restricted in any way. If the patient mentions psychological distress, the oncologist interviews the patient (on an ad hoc basis, no formal screening moments for distress). During regular visits to the department of oncology, oncologists and nurses routinely ask for psychological distress and provide emotional support and advise patients on how to cope with psychological distress, on an ad hoc basis. If urgent problems emerge, the patient is referred to a psychological or psychiatric service.

A marked contrast between the TES program and usual care exists. Key elements of the contrast include: formal screening at regular intervals for psychological distress versus ad hoc interview if a patients brings up any problems or if an oncologist or nurse recognize problems; enhanced care in a collaborative team coordinated by a trained nurse versus regular care delivered by oncologist and nurse; diagnostic evaluation with a standardized interview assessing psychological distress and problem analysis by a psychologist versus non-standardized interview by oncologist and nurse; guided self-help program, individual face-to-face counseling, medication, or planned referral to other services versus ad hoc support, advice and referral to other services if needed.

Assessments are done in both treatment arms at T0, T3, T10, T24 and T48.

The following sociodemographic data are collected: age, gender, social status, employment, Dutch vs non-Dutch origin, and lifestyle characteristics.

Collected medical data are: diagnosis, interval since diagnosis, initial TNM stage (classification of tumors using the size and extension of the primary tumor (T), its lymphatic involvement (N), and the presence of metastases (M)), cancer treatment, comorbidity, survival, relapsed disease with concomitant treatment and death.

Primary outcome is the difference in treatment effect over time in reduction of psychological distress between the TES program and usual care arm. Psychological distress is measured by the Hospital Anxiety and Depression Scale (HADS). The HADS has shown to be reliable, valid, and responsive, and has been widely used in research on cancer patients (e.g. [45]). A review showed sensitivity and specificity values for the HADS of 0.8 and higher [46].

Quality of life (QoL) is measured by the EORTC-QLQ-C30 version 3.0. This is an integrated system for assessing the health related quality of life of cancer patients[47]. The questionnaire incorporates five functional scales (physical, role, cognitive, emotional, and social), three symptom scales (fatigue, pain, and nausea and vomiting), a global health status / QoL scale, and a number of single items assessing additional symptoms commonly reported by cancer patients (dyspnoea, loss of appetite, insomnia, constipation and diarrhoea) and perceived financial impact of the disease. The instrument has been shown to be valid, reliable, and responsive to change [33,46].

In addition to the EORTC-QLQ-C30, the RAND-36 is used as a generic measure of health-related quality of life [48]. The RAND-36 is a generic, short-form health survey with 36 questions and consists of 8 sub-scales: physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, and mental health. Patients’ evaluation of care is assessed with the Client Satisfaction Questionnaire – 8 [49]. The CSQ-8 is a validated instrument to assess patient’s satisfaction with mental health care.

To evaluate recognition of and referral related to psychological distress (e.g. to psychological or psychiatric services) by clinicians in both treatment arms, patients’ medical records are reviewed by a blinded assistant. A structured form is used to extract data from the records.

The economic evaluation is conducted from both the societal and the hospital perspective. An estimation of all relevant costs will be made in both TES and usual care: intervention costs, healthcare costs not related to the oncological treatment, informal care costs, and costs of production losses due to absenteeism from paid work. Costs of the TES intervention will be estimated using a bottom-up approach. Costs that will be included in this calculation are: 1) costs of the development of the intervention; 2) costs of training the nurses implementing the intervention; 3) costs of implementing the intervention. A detailed description of resources involved in implementing the TES intervention will be based on the logbook, prospectively completed by the nurse. Healthcare utilization outside the hospital, informal care and absenteeism from paid work will be assessed using an adapted version of the TiC-P questionnaire [50]. The adapted cost questionnaire will be administered at T10, T24, and T48 with a recall period of 3 months.

As we assume that the oncological treatment in both the experimental and control group of patients with mCRC is equal, the oncological treatment costs will not be included in the cost measurements. Absenteeism from paid work will be valued according to the friction cost method [47] using mean age- and gender-specific income of the Dutch population.

Absenteeism from unpaid work and informal care is valued using the shadow price method. The shadow price of voluntary work and informal care is assumed to be equal to the tariff for cleaning work.

Quality of life will be measured by the EuroQol (EQ-5D). Quality adjusted life years (QALY’s) will be calculated by multiplying the utility of a health state by the time spent in this health state. The Dutch valuation tariff will be used to value health states [48].

Primary outcome is the difference in treatment effect over time in psychological distress, as assessed with the HADS, between the TES program and usual care. Secondary outcomes include quality of life, patient evaluation of care, recognition and management of psychological distress, and costs. Data are analyzed according to the intention to treat principle, using (mixed model) analysis of covariance. The exact interval between assessments is modelled in the analyses. Patients entering stepped care at S0, S10, or S18 are accounted for via a covariate in the analyses. Response to treatment (progression or not) is accounted for via a covariate as well.

In the intention to treat analysis mixed effects models for longitudinal data are used to deal with missing data (attrition). Alternatively, depending on the missing data pattern, Multiple Imputation (MI) using the Multivariate Imputation by Chained Equations (MICE) is used to impute missing effect and cost data [51-54]. The number of imputed data sets to be created is determined based on the fraction of missing information[55]. All datasets are analyzed separately and the results of the separate analyses are pooled using Rubin’s rules [56].

A complete data analysis and subgroup analysis based on patient and treatment characteristics are performed as secondary analyses. Additionally, during the study, data on the pre-study number of referrals to psychological or psychiatric services is gathered at each department of medical oncology (retrieved from medical records). If substantial differences exist, pre-study number of referrals is used as a covariate in the analyses, to correct for these differences.

Both cost-effectiveness analyses (using the HADS) and cost-utility analyses (using QALYs) will be performed. Missing cost and effect data are imputed using multiple imputation [54], as previously described. Differences in costs and effects are analyzed using linear multilevel regression analyses including clustering at the levels of hospitals and/or nurses implementing TES. Adjustment for confounding and/or effect modification is done if necessary. Incremental cost effectiveness ratios (ICERs) are calculated by dividing the differences in mean total costs between both treatment groups, by the difference in mean effects between both treatment groups. To account for the typically skewed distribution of costs, bias-corrected and accelerated bootstrapping (5000 replications) is used to estimate the 95% confidence intervals around the mean cost differences and the uncertainty surrounding the ICERs. To account for the clustering of data, bootstrap replications are stratified for hospital [57]. The bootstrapped ICERs are graphically presented in cost-effectiveness planes. Cost-effectiveness acceptability curves are estimated [58,59] to show the probability of the intervention program to be cost-effective in comparison with usual care for a range of different ceiling ratios, thereby showing decision uncertainty. In order to evaluate robustness of the results, sensitivity analysis on the most important cost drivers is performed.

The study aims to evaluate whether implementation of screening and subsequent treatment is effective. The power analysis takes into account that only a subset of patients (i.e. 33%, see below) is actually treated for psychological distress; the treatment effect obtained in this subset is ‘diluted’ by patients who are not treated for psychological distress. The power analysis addresses the question on the effectiveness of screening and treatment in all patients: the overall expected effect size is calculated in all patients, including patients who are actually treated and patients who are not treated [6]. The overall expected effect size of the implementation of screening and subsequent treatment is calculated as follows:

Setting the within subject correlation rho = 0.3, the overall expected effect size d = 0.18, alpha = 0.05 (two tail), beta = 0.80, the required sample size is 604 patients (302 per group).

Since this is a cluster randomized trial, patients within a cluster (i.e. hospital) cannot be assumed to be independent. To correct for the clustering effect, the obtained sample size should be inflated [61]. Assuming an intra cluster correlation of 0.005 [62] and a cluster size of 38 (604 divided by 16 hospitals), the obtained sample size is multiplied by 1 + 0.005 (38–1) = 1.18 [62]. This results in 604 x 1.18 = 715 patients.

Setting the exclusion rate at 40% (primarily due to lack of informed consent), 1192 eligible patients are required. Patients are recruited during 28 months in 16 oncology departments, which on average see 35 eligible patients per year. This results in an estimate of 1300 patients. Thus, recruitment is feasible, even taking into account a high exclusion rate. In previously published intervention studies on quality of life aspects, however, about 70% of eligible patients agreed to participate [24,63,64], and an exclusion rate of 40% seems rather pessimistic.