Testing a pain self-management intervention by exploring reduction of analgesics’ side effects in cancer outpatients and the involvement of family caregivers: a study protocol (PEINCA-FAM)

Pain is one of cancer patients’ most frequent and distressing symptoms. However, analgesics’ side effects often increase symptom burden [1, 2]. The International Association for the Study of Pain defines pain as “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage” [3]. World Health Organization (WHO) guidelines note that the main focus of pain management is analgesics. As depicted in Table 1, these guidelines focus on a 3-step progression of analgesics from non-opioids to strong opioids, which are associated with burdensome side effects [4, 5]. For opioids, the most common of these side effects are constipation, nausea, emesis, and concentration difficulties; the most feared, though very rare, are respiratory depression and death. However slight the risk, fear may still lead to early discontinuation, underdosing, or otherwise inadequate analgesia [6, 7].

Although pain is a treatable symptom, up to 40% of cancer patients experience inadequate pain management. System-level barriers include limited access to pain specialists or opioids [8] and misconceptions of healthcare professionals (e.g., misconceptions and knowledge gaps concerning pain management [9]). Patient-level barriers can be divided into four groups: cognitive (e.g., lack of information), affective (e.g., stress, anxiety, depression), sensory (e.g., analgesics’ side effects) and practical (e.g., implementing pain management in everyday life) [10]. Family caregivers (FCs) directly involved in patients’ pain management commonly experience the same barriers as the patient. FCs are defined as relatives, partners, friends, or neighbours who provide care to patients [11].

With the home becoming the primary setting for many aspects of cancer care, a key goal is to improve patients’ pain and side effect self-management [12]. Self-management consists largely of strategies for solving problems, making decisions, taking action, utilizing resources, forming patient/healthcare provider partnerships, and dealing with physical and psychological issues to avoid or delay deterioration [13, 14]. FCs commonly play central roles in patients’ pain self-management. However, while FCs may administer medications, implement pain relief strategies, and provide emotional support [10, 15], they often assume their tasks with little or no preparation, knowledge, resources, or skills [16]. The psychological and physical costs include a high prevalence of anxiety and depression, social isolation, sleep disturbances, fatigue, and even an elevated risk of coronary heart disease [17‐19]. The responsibility for pain management without corresponding skills often results in FCs’ loss of self-efficacy [20]. Self-efficacy is defined as the confidence in one’s ability to perform the behaviours necessary to achieve a target outcome [21]. Recent studies on cancer pain management reveal that high self-efficacy improves both patients’ well-being and FCs’ mood [22, 23]. However, little research has focused on FCs’ influence and self-efficacy vis à vis supporting self-management of pain and analgesics’ side effects in cancer outpatients.

Three recent meta-analyses focusing on the effects of psychosocial interventions to support cancer pain management from 1983 to 2012 found that, while gaining attention, the psychosocial approach to pain management remains understudied [24‐26]. The two most common intervention types were skill training and education [25]. Regarding the main outcome measures, i.e., effects on pain intensity, pain interference in daily activities, knowledge, and attitudes towards cancer pain and analgesics, these meta-analyses revealed various significant, though moderate overall improvements. Further, a systematic evaluation of interventions’ content, structure, and efficacy to improve patients’ self-management of cancer pain revealed no discernible patterns with respect to components, duration, or delivery type or mode [27]. We found no studies that focused on the side effects of analgesics.

To date, little research has focused on interventions supporting FCs in patients’ cancer pain self-management. First, Keefe et al. [28] tested the feasibility of a partner-guided cancer pain management training program with 78 end-of-life patients and their FCs. Intervention group FCs showed significantly higher levels of self-efficacy regarding support for the control of pain and other symptoms. Second, in a randomized controlled trial (RCT) of caregiver assisted coping skill training for cancer patients and their FCs, Porter et al. [29] reported significant results compared with an education program. Additionally, enhanced knowledge of cancer pain management led to both decreased anxiety symptoms in both groups and increased symptom management self-efficacy in FCs. However, as this study lacked a standard care control condition, these results should be interpreted with caution. Third, Hendrix et al. [23] tested an individualized training intervention on self-efficacy in 120 patient-caregiver dyads. FCs in the intervention group demonstrated significant increases in self-efficacy regarding pain control, prevention of infections, maintenance of nutrition, and practical home care issues.

However, no studies have evaluated the effectiveness of interventions focusing on analgesic side effect management to improve self-management in cancer patients and their FCs. Similarly, no study has yet investigated correlations between FCs’ cancer pain management knowledge and self-efficacy related to pain management and patients’ pain intensity.

One psychoeducational intervention included in the above-noted meta-analyses, the PRO-SELF © Pain Control Program (PCP), an intervention to support self-management of pain in adult oncology outpatients and their FCs, showed statistically significant and clinically meaningful reductions compared to standard care in both average and worst-pain intensity in a large RCT (N = 174) [30]. In further research, the intervention was expanded regarding its duration (from 6 to 10 weeks) resulting in the PRO–SELF © Plus PCP [31, 32]. Koller, Miaskowski, De Geest, Opitz, and Spichiger [33] translated and adapted this version for use in German-speaking populations. A pilot RCT using 39 oncology outpatients yielded a statistically significant increase in test subjects’ knowledge and demonstrated this version’s feasibility. Based on the pilot RCT’s findings, the intervention was adapted and is currently being tested within the multi-centre “Mixed methods study to test the efficacy of the adapted German PRO–SELF © Plus PCP, an intervention directed at outpatients with cancer and their family caregivers to reduce pain and related symptoms” (PEINCA) in Switzerland. The overall aim of the multi-centre mixed methods PEINCA study is to evaluate the efficacy of the adapted German PRO-SELF© Plus PCP, which was designed to improve outpatients’ and their FCs’ management of pain and pain intensity. The purpose of this sub-study of PEINCA is to test the efficacy of the adapted German PRO-SELF © Plus PCP at both reducing analgesic side effects and enhancing cancer patients’ and FCs’ knowledge of cancer pain as well as to explore their learning processes. In addition, FCs’ involvement in patients’ pain self-management and association between their self-efficacy and knowledge of cancer pain with the cancer patients’ pain intensity will be investigated (PEINCA-FAM).

The overall framework of the PRO-SELF © Plus PCP was the Theory of Symptom Management (TSM) [34]. A further development of the TSM, the Symptom Management Model (SMM, see Fig. 1), which depicts interacting concepts, serves as the theoretical framework for this study [35]. Precipitating factors (antecedents) such as demographic, sociocultural and psychological characteristics, can influence perceived symptoms at baseline (intercepts) and over a given trajectory (slope). By charting the trajectories of symptoms in relation to interventions (symptom management strategies) and patient/FC/nurse interactions, we can illustrate these interactions’ continued influence over patients’ and their FCs’ outcomes (consequences). In this study, consistent with the SMM, cancer pain is viewed as a complex, multidimensional experience frequently accompanied by analgesics’ side effects. Precipitating factors, e.g., social environments and lack of knowledge, influence symptom trajectories. Within the framework’s symptom management strategy, our intervention focuses on supporting cancer patients’ pain and analgesic side effect self-management and the involvement of FCs. Patients as well as FCs are seen as key players who interact with intervention nurses. Alongside improved knowledge of cancer pain and self-efficacy in FCs, reduction of analgesics’ side effects is a key intervention outcome.

In addition, Bandura’s Social Cognitive Theory (SCT), an adult learning theory, provided the supplementary conceptual framework for the intervention [36]. SCT acknowledges that learning is a cognitive process that takes place in a social context and can occur through processes of observation, with each individual possessing a self-regulating system that affects motivation and learner differentiation. As part of this system, Bandura introduced the concept of self-efficacy [37], along with major influencing factors: mastery experiences, live modelling, performance exposure and positive appraisal [21]. Through our intervention to educate cancer patients and their FCs, we aim to develop their knowledge regarding pain management, their self-efficacy regarding pain management, and their symptom self-management skills. This entails keeping a diary, setting goals and establishing a management plan to reach those goals.

PEINCA-FAM aims to test the adapted German PRO-SELF © Plus PCP pain self-management intervention to reduce cancer outpatients’ analgesics’ side effects, to explore patients’ and FCs’ knowledge of cancer pain and their learning processes, and to investigate FCs involvement in patients’ pain self-management and associations between their self-efficacy, knowledge of cancer pain and patients’ pain intensity. Specific aims of this study are: (1) to test the efficacy of the adapted German PRO-SELF © Plus PCP to reduce analgesics’ side effects by focusing on constipation, nausea, emesis and concentration difficulties; (2) to test the efficacy of the adapted German PRO-SELF © Plus PCP to improve knowledge of cancer pain management in patients and FCs and to qualitatively explore patients’ and FCs’ learning process during the intervention; and (3) to explore the association between FCs’ self-efficacy and knowledge of cancer pain with patients’ pain intensity and to qualitatively explore FCs’ influence on patients’ pain self-management.

Embedded in the ongoing PEINCA study, the PEINCA-FAM sub-study applies a mixed method approach. In a concurrent embedded strategy, an RCT is combined with qualitative parts [38, 39]. All data are generated within PEINCA. The following describes the PEINCA study’s essential methods as they apply to PEINCA-FAM.

Patients with cancer pain are recruited from oncology outpatient clinics at three university hospitals in the German speaking part of Switzerland. These individuals are included if they (1) are aged ≥18 years; (2) have experienced any type of recurrent cancer pain, i.e., rated ≥3 on a 0–10 Numeric Rating Scale (NRS, 0 = no pain, 10 = worst imaginable) over the past week; (3) have an estimated life expectancy of > 6 months; (4) are able to understand, read and write German; and (5) have access to a telephone. Participants are excluded if they (1) have cognitive dysfunction or hearing impairment; (2) are hospitalized > 2 weeks during the study; and (3) are suffering solely from neuropathic pain. To be eligible, FCs must be (1) aged ≥18 years; (2) able to understand, read and write German; and (3) willing to participate in all intervention sessions.

The results of the above-mentioned pilot study (Koller et al. [33]) included a maximum mean alleviation of 0.99 NRS (0–10) for average pain and a mean alleviation of 1.27 NRS (0–10) for the worst pain compared to the control group in week nine. For these two endpoints, based on average pain improvement data, an alpha error of 0.05 (95%) and a beta error of 0.2 (80%) would require a 136-patient sample to achieve sufficient statistical power. Estimating an attrition rate of 35% over the 6-week study period, it will be necessary to recruit a total sample of 210 patients. We expect to recruit 50% of patients along with FCs. This number will be sufficient. Based on Wells, Hepworth, Murphy, Wujcik, and Johnson [40], who assessed initial effects of an intervention to increase knowledge and positive beliefs about cancer pain management in patients using the Family Pain Questionnaire (FPQ), an improvement in knowledge and beliefs were found [F (1,62) = 18.2, p < 0.001]. At an alpha error level of 0.05, a sample size of 32 FCs should be sufficient to achieve a statistical power of 80%.

With respect to the sample size for the qualitative part, a sample of 42–60 patients and FCs should supply sufficiently redundant data to discover patterns, commonalities and differences among the groups.

Specifically trained nurses from the outpatient clinics work as research assistants (RAs) for the study. All RAs are experienced registered nurses or graduate students in nursing who have received study-specific training by the study coordinator (HR). RAs screen potential participants and verify inclusion criteria via patient files and discussions (regarding life expectancy, etc.) with their treating physicians. If the initial inclusion criteria are met, the RAs contact the patients directly during their outpatient appointment to check whether they have experienced repeated pain that they would rate on an NRS as ≥3 over the last week, are able to understand, read and write German, and have access to a telephone. The RAs then inform patients about the study and invite them to participate. Each patient is also asked if an FC is involved in their daily pain self-management. If yes, the FC is also invited to participate in the study. Written information and consent forms (including those for FCs, if necessary) are then given to the patient. After at least 24 h, an RA contacts each patient (or patient/FC dyad) again to provide verbal information to the FC as needed, ask about their decision regarding participation, and collect the signed written consent form. In all settings, these written informed consent forms are stored safely in the Investigator Site File (ISF) by the RAs. Only afterwards are participants randomized to the IG or CG. Patients and FCs have the right to withdraw from the study at any time without consequences.

Concerning the recruitment of participants for qualitative data collection, all patients (and FCs) are asked at the beginning of the main study whether they are willing to participate in an interview following completion of the study’s 6 weeks of procedures. At the end of the final home visit, the intervention nurse (IN) or an RA asks selected patients (and FCs) whether they are still willing to participate in the interview. They then provide verbal and written information and obtain written consent from those who agree. Purposive sampling is applied to ensure approximately equal sample sizes per study site, as well as variation regarding pain intensity, intervention adherence, age, gender, education, living situation, and tumour entity.

As shown in Fig. 2, patients and FCs who have provided written consent are stratified by site and randomized 1:1 either to a six-week intervention group (IG) or to a usual care group (CG). A permuted block procedure, with blocks of 2, 4 and 6, is used to create a computer generated randomization list. This procedure should ensure approximately equal distribution of patients per group and per centre. The Clinical Trial Unit (CTU) of one of the included university hospitals provides the list, which is accessible from all settings.

Blinding is not possible because the INs use collected data directly for the intervention (pain diary, Patient Pain Questionnaire (PPQ)). And while treating physicians are not informed, they may still become aware of group allocation if participants ask questions or take their diary to an appointment.

Both quantitative and qualitative data will be collected from March 2016 until December 2018. Patients in both groups complete a daily pain and symptom diary; patients and FCs fill in questionnaires. The same 6-week data collection protocol is applied via in-home visits at baseline and during weeks 1 and 6. In the IG, data are collected by the IN. In order to minimize diffusion of the intervention to CG participants, an RA collects data in the control group. Between in-home visits at baseline and weeks 1 and 6, patients/FCs receive brief telephone calls every second week, during which the RA ensures that the pain diary is being completed on a daily basis.

Table 2 provides an overview of the study variables and data collection points for patients and FCs.

Questionnaires employed in this study are applied in their German versions, for which validity and reliability have been established. The primary patient outcomes are selected analgesics’ side effects (constipation, nausea, emesis, and concentration difficulties), self-efficacy, and knowledge of cancer pain management. Primary FC outcomes are self-efficacy and knowledge of cancer pain management. The secondary outcomes are patients’ average and worst-pain intensity. Anxiety, depression, functional status, and cancer-related symptoms are measured as covariates. Our analytical models will also control for age, education, time since diagnosis and current disease status.

Quantitative data will be compiled and analysed using SPSS 24.0. Data will be retrieved from the SecuTrial® database. Data will be systematically examined for out of range values and data inconsistencies. As appropriate, descriptive statistics will be calculated, including means and standard deviations for interval variables and frequencies, as well as percentages for categorical variables. An intent-to-treat analysis will be applied, using a significance level of .05 [55]. Although premature withdrawal is expected to be a random process, participants who completed the study will be compared with those who did not based on demographics, treatment group and other salient variables.

Data are handled confidentially and stored in a locked cabinet for 10 years. Data are anonymized by applying numerical IDs. The master file with patient names and associated ID numbers is kept secure. Only the study team has access to the written and electronic data. With regard to the final trial data set, only the primary investigator, the study coordinator, the PhD student and the person who enters the data will have access. In order to communicate trial results to participants, health care professionals and the public, we plan to publish results of primary and secondary outcomes. Furthermore, the study team and nurse managers of participating university hospitals will be informed about study results via scientific presentations in various settings.