Conclusions
Better information recall, a trusting relationship with healthcare professionals, and a time out period for the consideration of treatment options are important determinants that influence patients in their treatment decisions and their preferred participation role during shared decision-making.
Keywords : Shared decision-making, determinants, option talk, decision talk, pancreatic neoplasm
Introduction
Patients with pancreatic cancer need to make difficult decisions about intensive medical treatment weighed against their quality of life in a short period of time. Pancreatic cancer carries a poor prognosis with an overall survival rate of 2–7% and has a great impact on a patient’s life.1,2 Over the period 1990-2017, the incidence of patients with pancreatic cancer in the Netherlands increased by 2-3 times due to aging, overweight and diabetes.1 The only curative option is resection of (a part of) the pancreas, often combined with (neo)adjuvant chemotherapy. Resection of (a part of) the pancreas is a major operation with the risk of severe complications, such as sepsis, bleeding or wound infection.3 Long-term consequences such as diabetes and digestive problems can also occur.4 In addition, in almost 50% of patients the tumour relapses within 2 years after surgery.3 This information, together with the up-to-date short-term outcome results of these operations, should be shared with patients and considered in light of patients’ expectations regarding quality of life.
Shared decision-making (SDM) is an approach whereby clinicians and patients make decisions together using the best available evidence. Patients are encouraged to think about the available treatment options and the likely benefits and harms of each, so that they can communicate their preferences and help select the best treatment for them.5
While the most common cancers studied in relation to SDM are breast and prostate, little is known about SDM in patients with pancreatic cancer.6 Since 2017, healthcare professionals working in an University Medical Center for pancreatic cancer in the Netherlands have used SDM according to the revised three-talk model of Elwyn et al..7 This SDM model is characterised by an option talk, a time out period, and a decision talk. In the option talk, the surgeon and case manager discuss the treatment options in terms of the potential harms and benefits of each treatment. After this, the patient is given a time out period of four days (reflection period). During the time out, the patient considers the treatment options together with their general practitioner (GP) or social network. In the decision talk, the surgeon and case manager guide the patient to make a preference-based treatment decision.
In the period of 2017–18, in the run up for the present study, a cohort of 30 patients, characterised by (borderline) resectable pancreatic cancer, an older age (over 75 years) and multiple comorbidities, was followed at the outpatient clinic at the Radboud University Medical Center. After starting with SDM as a communication model, almost 30% of these patients did not opt for surgery in their decision talks. Before 2017, treatment options and best supportive care were not discussed according to the SDM model.5
The observation that 30% of these patients did not opt for surgery after the implementation of the SDM model was the reason for initiating the present study. We aimed to identify the determinants that influence the treatment decisions of patients with pancreatic cancer during SDM.
Methods
Study design and setting
We performed a qualitative study consisting of a literature review on SDM determinants, followed by interviews and observations to substantiate the findings of the review with practical experiences. The study was carried out at the outpatient clinic of the pancreas center in the Department of Surgery at the Radboud University Medical Center in the Netherlands. Approximately 350 patients are discussed annually by the pancreatic tumour board (a multidisciplinary team of clinicians) at this institution. Of these patients, approximately 200 have their option and decision talks at the outpatient clinic.
The study was approved by the local medical ethical committee (registration number: 2021/7316). There was no pre-existing relationship between the researcher (LD) and the participants in the observations and interviews. LD is an experienced nurse and trained in conducting observations and interviews.
Literature review
We reviewed the literature on the determinants that influence patients with pancreatic cancer during SDM. Search terms were developed to capture three concepts: pancreatic neoplasms, patient participation and patient decision-making. Appendix 1 provides a detailed list of the search strings. Full-text studies published between 2000 and 2019 on PubMed (including MEDLINE), CINAHL and PsycINFO were included. The references of the included studies were manually checked, and the authors’ personal files and the bibliographies of previously published related reviews were searched to identify additional relevant studies (snowballing). Only English-language studies were considered for this review. The inclusion criteria for the studies were 1) the determinants affecting SDM were described, 2) the population consisted of patients with pancreatic cancer, and 3) the studied patients were over 18 years old. Studies were excluded if only medical treatments were reported. The titles and abstracts of the included studies were initially screened by one researcher (LD). Subsequently, the full text of the selected studies was screened by LD in consultation with a second researcher (MH-S).
Quality assessment of the selected literature
The methodological quality of the selected studies was assessed by LD in consultation with a second researcher (MH-S). For the selected qualitative studies, the Criteria for Reporting Qualitative Research guidelines (COREQ)8 were used for evaluation. For observational (cohort) studies, the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)8 was used. For systematic reviews, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) were used.8 For rating and standardising the methodological assessments of the studies, we developed a four-point scoring form, with studies rated from ‘Very good (++): Fully met the standards of COREQ, STROBE, PRISMA’ to ‘Poor (- -): Assessment was not possible, exclusion of the study ’ (e.g., the study design does not match with the research data analysis).
Observations of patients
Patients who were referred for surgery at the outpatient clinic of the Radboud University Medical Center were purposive sampled to represent the patient population of this outpatient clinic in terms of the (no) present of comorbidities and chemotherapy treatments. The included patients had an option or decision talk.
Patients gave verbal informed consent to the researcher (LD) before their option or decision talk. They gave permission for the observations to be written up and published (including quotes). The researcher (LD) observed the conversation and used a standardised observation list based on the determinants found in the literature review (Appendix 2). LD made field notes about the patients’ context (e.g., social support, disease stage) during the observations for correct interpretation of the data. The observations took place in a regular consultation room in the period September–December 2019.
Interviews with professionals and patients
A purposive sampling strategy was used to ensure a representative sample of healthcare professionals in terms of positions (clinician, case manager and physician assistant) and characteristics, such as their role within SDM. LD approached patients with pancreatic cancer for an interview after their option talk at the pancreatic outpatient clinic. Patients were recruited based on their age (< 65 and ≥ 65 years old) and gender. The patients were informed about the study by e-mail and provided verbal informed consent at the beginning of the interview. Their verbal informed consent was audio recorded. The interviews with the healthcare professionals took place at the hospital. Patients were interviewed at their homes or at a hospital location with sufficient privacy, depending on their preference. LD conducted the semi-structured face-to-face interviews (45 min each) from September 2019 to January 2020. The interviews were guided by an interview topic guide (see Appendix 3), audio recorded and then transcribed ad verbatim by LD. The transcripts were returned to the interviewees for comment and correction.
Data analysis
The study characteristics and outcomes, such as study design, sample, population and methodology, were tabulated (see the data extraction form in Appendix 4). The data from the literature review, observations and interviews were analysed manually. We used Bandura’s Social Cognitive theory (SCT) to identify the determinants of SDM, since this theory has shown to be helpful in explaining behaviour in relation to health outcomes.9 Bandura emphasised that behaviour is the result of continuous interactions between aspects of the social environment, the person, and the person’s behaviour.9
The determinants of SDM in the included studies were identified and analysed by LD in consultation with two researchers (MH-S and AJ). The identified determinants of SDM were divided into the three categories related to behaviour, the physical and socio-cultural environment and the person9, and further classified into subcategories.
The interview transcripts and observation lists were systematically analysed according to directed content analysis.10Relevant data were identified and structured by initial codes based on Bandura’s SCT9, and open coding to create new codes (see Appendix 5). The coding took place under the supervision of MH-S and AJ. Issues that occurred during the analysis were resolved by consensus meetings.
The categories, subcategories and determinants from the literature review, interviews and observations were merged, restructured and tabulated. The determinants were substantiated with illustrative quotes. We used the COREQ guideline to report the qualitative data8 (see Appendix 6 for the completed checklist for this study).
Results
Literature search
Our initial search for studies on the determinants influencing SDM resulted in 141 studies (Figure 1). Additionally, 22 studies were identified manually (snowballing). After scanning the titles, abstracts and full texts, the methodologies were assessed. The final set consisted of 13 studies: 6 qualitative studies, 4 quantitative studies and 3 systematic reviews (Appendix 7).