Participants
Any individual who recently received a certain or likely diagnosis of pancreatic cancer was eligible for the study.
Procedure
Recruitment took place between April 2013 and May 2014. Patients were recruited through the Internet or at their visit at the hospital. A call with information about the study was placed on the websites of Dutch foundations which aim to address needs of patients with pancreatic cancer or digestive diseases (Lisa Waller Hayes foundation, Dutch Digestive Diseases Foundation, Foundation for Patients with Alimentary Canal Cancer). The call included a link to an online survey as well as contact information to receive the paper version and a postage-paid return envelope. Furthermore, surgeons, gastroenterologists, physician assistants and nurse practitioners from three academic and three non-academic Dutch hospitals recruited patients, at the time of their outpatient appointment or when admitted to the hospital. The health care professionals gave eligible patients a flyer with information about the study including the link to the online survey, again with the option to complete it on paper. The participants provided informed consent before completing the survey in one of two ways. They were either asked to sign an informed consent form if completing the survey on paper or were asked for informed consent on the first screen of the online survey and were only directed to the actual survey if they responded positively. The Leiden University Medical Center Medical ethics committee approved the study.
Survey development
To generate a list of questions which could be important to patients with pancreatic cancer in the first weeks after their diagnosis, nine patients (four male, five female) were interviewed. Eight patients received the diagnosis pancreatic cancer and one patient first was suspected to have pancreatic cancer, and turned out to have bile duct cancer. The interviewer (LRS) asked patients which questions were important to them in the first weeks after diagnosis. To support patients’ thinking and recollection, the interviewer asked them to name any question they could think of in the following categories [5]: disease, treatments, physical, and psychosocial issues. The questions obtained from these nine interviews were included if they seemed general enough to be relevant to more patients as well as specific enough to the disease and treatment options at stake. This was determined by consensus among the authors (LRS, RJS, AP). Question topics most often mentioned included questions about the treatment procedure, treatment harms, caregiver support, and what diet is most suitable for pancreatic cancer patients. The questions obtained from the interviews were combined with questions that could possibly be relevant to patients with pancreatic cancer in the first weeks after diagnosis, based on studies on information needs in prostate cancer patients [6, 13]. This resulted in a total of 84 questions.
Final survey
The final survey consisted of two parts. It started with questions about the participants’ personal situation, including age, gender, level of education, time since diagnosis, and treatment or treatments received or about to be started in the upcoming month.
Next, patients were asked to think back to the first weeks after diagnosis and to indicate how important it was to them, at that time, that certain questions were answered. The participants were asked to rate the importance of each of the 84 questions on a seven-point Likert-type scale, ranging from 1,“totally not important”, through 4 “not unimportant/not important” to 7, “very important”. The seven-point Likert-type scale was chosen to enable patients to distinguish the importance of different questions. Each point on the response scale was labelled. If the participants deemed it appropriate, they could instead tick one of two boxes: “avoid answer”, to indicate that a question should not be answered, or “not applicable”, to indicate that a question is not deemed relevant to pancreatic cancer soon after diagnosis.
To structure the large amount of questions, they were arranged into the following categories [5]: disease, treatments, physical issues, and psychosocial issues. Within the category “disease”, questions about the nature of the disease were followed by questions about disease progress and prognosis. Within the category “treatments”, questions were ordered as following: questions about a) the choice whether or not to treat the disease, b) treatment procedures, c) treatment harms, and d) treatment results.
Statistical analyses
Descriptive statistics were used to present the number of questions that the participants rated as important or very important and the perceived importance of questions. We assessed whether the number of questions a participant rated as important or very important was significantly related to gender, age, education, or number of days since diagnosis using median tests and spearman correlations. We used the mean perceived importance score to rank order the questions in terms of importance. In case of equal mean importance scores, questions with a smaller standard deviation were considered as more important, assuming that there was comparatively more agreement on their importance among the participants. The participants who indicated that a question was not applicable or that it should be avoided, are shown separately. We compared the top 10 of most important questions, again rank ordered according to their mean score and standard deviation, as a function of time using a median split to separate the patients who had completed the questionnaire earlier (i.e., within 46 days of diagnosis) versus later (i.e., after 46 days). SPSS 22 was used to analyse the data, significance was tested two-tailed at α = 0.05.