Setting and participants.

Two North Italian specialist palliative care units were involved in this study. One unit has a clinical palliative care ward in an acute oncology hospital (IRCCS Reggio Emilia). The second has a palliative home care service and two hospice wards (FARO Foundation, Torino). Both teams are multiprofessional, include cancer and non-cancer patients and are involved in clinical activities (inpatient units, outpatient consultations, home care services and primary care support) in research and education (Quality of life and PROMs development, palliative care for non cancer patients and symptoms control).

To obtain validation of the patients’ version, semi-structured individual interviews with patients receiving specialist palliative care were conducted. For the staff version one focus group in each of the two clinical centres were run. Participants were professionals working in the two multidisciplinary palliative care teams (MPCT).

For the individual interviews to patients a purposive sampling technique was chosen, with the aim to select participants from different clinical settings (home care, hospice or hospital wards) and geographical settings, with different primary diagnosis who could provide in-depth insight about the tool structure. For this reason, we excluded patients whose clinicians deemed to be too ill or distressed.

For the focus groups we included professionals who work daily in the field of palliative care, with experience in the use of PROMs such as IPOS or POS.

Inclusion criteria for the patient interviews were: age > 18, Italian as first language, no relevant cognitive impairment and capacity to give consent. For the focus groups participants had to be professionals expert in palliative care (at least 3 years of clinical practise in the field), fluent Italian speaking and actually working in one of the two MPCT. Exclusion criteria for the patients were being too debilitated or cognitively impaired to sustain an interview or not willing to participate.

Patients were selected through a first identification from the clinical nurse who was in charge of their care and knew the inclusion-exclusion criteria, followed by a discussion with the research team. Researcher then got in touch with the potential participants describing the study and asking their consent to participate.

Professionals were selected by the research team according to their previous experience and interest on the use of IPOS and PROMs, and then asked to provide consent and to participate to the study.

Interviews were conducted where the patients were receiving care, in one of the 3 described settings. Focus groups (FG) were run in the two working settings.

The interviewers and focus groups conductors were expert researchers working in the MPCT previously trained according to the “Topic Guide for Integrated Palliative Care Outcome Scale (IPOS) Pilot survey (Topic Guide)” provided by the IPOS translation team. This tool prompts towards the general aspects of the test, comprehension, retrieval, judgement and response process used by the interviewees.

The trained interviewers were one nurse working in the palliative care ward and 3 psychologists working in the hospice wards and home care team. All of them knew the interviewed patients and participated to their preselection.

The two PI supervised the data collection, provided education training to the interviewers and took part to some interviews and focus groups.

Study procedures.

The overall study design is shown in the flow chart exposed in Fig 2. For the purpose of this study the 7 days versions of IPOS were adopted. After the translation process the cognitive phase was performed, representing the phase 5 of “The Palliative care Outcome Scale (POS) Manual for cross-cultural adaptation and psychometric validation”. (S2 Appendix)

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Stages of the translation and cognitive validation process of IPOS_ITA.

https://doi.org/10.1371/journal.pone.0208536.g002

In the individual interviews participants filled the IPOS and were hinted to read and answer the questions one at a time, then stop and talk about each question before moving onto the next. Interviewees were invited to ‘think out loud’, explaining how they arrived at the answers, rather than the result of the answer itself. They were also asked to report any thoughts or views they might have had about the questions. According to the Topic guide manual, conductors explored general views on the tool as well as comprehension, retrieval, judgement and other potential factors impacting the cognitive evaluation.

All the selected patients invited to the study consented to participate. The interviews took on average 45 minutes (range 30–60).

It was not possible to calculate the time of IPOS_ITA completion since the interview focused on each single item, with a thorough discussion of the comprehension and acceptability of the wording. All the interviews were concluded without evidence of distress of the participants, none of them asked to interrupt for excessive burden and no questions were declined or deemed invasive.

The two focus groups involved 12 professionals with expertise in palliative care. The events were run separately and conducted by the principal investigators of the two study centers. All participants had previous experience of IPOS_ITA use, some conducted interviews with patients, others used it in clinical practise. Both focus groups lasted about 1 hour.

In the focus groups participants were asked to go through the IPOS staff version, with the same approach suggested to the patients. They were also invited to provide reflections on how they thought their patients might have issues in understanding and interpreting the tool’s questions and answers.

Data from the events were registered by written field notes, or audio registrations, based on the researcher’s choice. In all cases comments were recorded. The transcriptions were not translated in English, only the emerged issues were reported in English to the original authors for the following validation audit.

Data analysis.

Data analysis was individually performed by the two PI of the study, through reading and re-reading of the transcriptions, focusing on each item. All the IPOS_ITA fields, questions and answers, were checked in terms of wording, acceptability, ambiguous meaning, as well as format and layout issues. Using the Topic Guide as a reference framework, a thematic analysis [7] allowed us to focus on the most relevant issues.

The thematic analysis was performed using an informal approach based on questions and answers. Any difficulties emerged during the filling of IPOS or in the discussion with participants were considered as a possible issue and coded to be reviewed. The categories (see S1 Appendix) used to perform the analysis were: comprehension (what does the respondent believe the question to be asking), retrieval (Could they recall the information required by the question? Was the time frame suitable?), judgement (Is the respondent able to make an evaluation based on the information recalled?), response (Is the respondent able to map their internally generated answer to a response option?) and others, (including additional comments, the questionnaire overall impression, questions to be removed, layout issues).

The emerged themes were compared and aggregated for each question and each category in both centres, through consensus between the PI. The aim was to highlight the problems or perplexities emerging from the tool, in order to refine and edit it in a new output (IPOS_ITA_AFTER COGNITIVE) to be sent to the English author for proof reading, comments and remarks.

A final validation audit, involving the PI, the main English author (IH) and the author of POS Italian versions (MC), was set up to merge all the issues and arrange the final IPOS_ITA tool to be published on the official website.

The audit process aimed to check any single change, throughout the match and comparison among all the items of the other following POS family versions: original English IPOS, Italian POS and the new and final IPOS_ITA versions.

Ethics.

The study protocol was approved by the Ethical Committees of the two research centres (CE Interaziendale A.O.U. Città della Salute e della Scienza di Torino, reference number CS\599 protocol 0070308, and CE Provinciale Arcispedale S. Maria Nuova di Reggio Emilia, protocol 2015/0021569).

Confidentiality of the data is guaranteed by anonymity of the participants through a coding process. Sensible data are securely stored by the PI. To preserve patients and professionals by possible psychological distress they could stop the interviews at any time or decline invasive questions.

Written consent to the study participation was obtained by all participants.

Layout and presentation.

In the interviews and focus groups some difficulties in reading and filling the forms were complained about by participants, therefore some layout changes were made to make the tool more easy to complete, for example increasing the types font and reprinting with wider interline space.

Between Q5 and Q6 the order of the possible answers reverses because of the nature of the questions (Q5, asks about depression, a negative symptom, Q6 is about being at peace, a positive feeling of wellbeing). Thus, Q3-Q5 are rated Not at all (0) → to Always (4), Q6-Q8 is the opposite, Always (0) → Not at all (4). This caused difficulties since some patients did not notice the change reporting a wrong answer.

“The answers are not well done because I get confused with the following question” a female (aged 50) patient from the hospital ward.

“I want to underline the risk of wrong answers due to the positioning in the same column of answers with opposite meaning” (Professional working in the home palliative care team)

A change in the output was made in order to separate the different blocks of questions and reduce the possible errors.

Content, acceptability, comprehension and meaning.

Many positive comments were made about the choice of prioritizing open questions rather than starting with a problem checklist.

“I was positively surprised when you asked me about my problems before asking me if I were in pain” (a patient in the hospice ward)

The 7 day version of the IPOS adopted in the study required that the patients had to rate symptoms intensity that could have had fluctuations and different impact on their experience over 7 days. This caused problems of retrieval and difficulties in choosing an answer.

“I had shortness of breath every day in the last week, but pain for example varied. So I am not sure what to choose” (a patient with IPF, home care)

Professionals reported difficulties in estimating the degree which patients were affected by symptoms explored in Q2. They did not know if they had to report the highest intensity or report a mean value.

“When they asked me about what to choose I was not sure what to say. In general in palliative care we tend to consider the highest degree of a problem, but here I just didn’t know. Maybe they had an awful symptom for few hours in a week, should they call it unbearable? Or because it went away quickly and didn’t come back it was rather mild or moderate?” (Professional working in the home palliative care team)

For this reason further explanation in brackets was added, suggesting that in the case of fluctuating symptoms a mean value could be entered.

Questions about the symptoms were found very clear and well-constructed. The proposed answers did not create confusion and participants could easily reflect on their experiences. The only exception was the item on sore or dry mouth.

“If I have a slight sore mouth and a severe dry mouth, which one I sign?” (female patient in the hospice ward)

Being a double choice of two very common and unpleasant symptoms, some patients debated if they had to choose between the 2 symptoms or rate the worst, or deny both since one was present but the other not.

One patient suggested dividing the question in two:

“Dry mouth and pain in the mouth are 2 completely different things and should be divided” (Patient in the hospital ward)

In order to ask a single question comprehensive of both symptoms, a new formulation was proposed, according to the NEURO_IPOS symptom checklist recently published [8].

Other changes were applied to the possible answers in Q3-Q8 questions: the word “occasionally–occasionalmente” was changed in “rarely- raramente”, since in Italian this makes more sense. This is consistent with the Italian version of POS [4]. Furthermore in Italian it is hard to difference between “occasionally” and “sometimes”, whereas “rarely” is more distinguishable.

“I was undecided and doubtful between occasionally and sometimes since I did not recall the exact difference” patient, home care)

Some difficulties emerged in the second part of the tool, regarding the psychosocial, existential and social issues.

Question 3 asked about “feeling anxious or worried about the illness or treatment”. This generated hesitation due to the multiple choices exploring anxiety, worries and being directed towards the illness and the treatments. Several patients and professionals reflected on these issues from different perspectives. Patients were more concerned about the difference between being worried about the illness or the treatments: some said they were worried about the disease and not the therapies or vice versa and did not know how to answer. Professionals perceived the word anxiety as a diagnostic category, see for example DSM-5 manual [9], that might cause misunderstanding.

“I fear that the term Anxiety refers to a pathological condition, I wouldn’t like that they felt more ill than they already are. Being worried can be more acceptable for our frail patients” (Psychologist working in hospice)

In a focus group a new formulation was proposed: “have you been worried for your health?” that was felt to be more clear and acceptable and including both options. During the validation process audit this rewording was rejected, because the original IPOS included all the different choices, explored in multiple needs assessment of palliative care patients and to maintain congruence with the fully validated POS_ITA version.

Question 4 explores how much family or friends have been anxious or worried for the patient. According to the previous considerations about difficulties related to multiple choices, a proposal of change was “Have your dear ones been worried for you?”

“I think that we should either split the question for family and friends or keep it together focusing on the people important for the patient” (participant of a Focus Group)

“What if my family are ok but my friends are extremely suffering for my condition? My girlfriend seems well adapted to my condition, but my friends and one of my son who lives in South America are very concerned” (Patient, hospital ward)

For the same thoughts of Q3 the double answer “anxious or worried” was maintained, but a new simplified and inclusive wording was adopted to substitute “family and friends”. It seems that “dear ones” in the Italian setting is more clear, inclusive and acceptable.

Question 5, exploring low mood, was maintained unchanged using “have you been feeling depressed?” even though some issues, as for anxiety, were raised. Depression can be seen as a stigma in Italian culture as well as in Germany [5],professionals tend to see it as a diagnostic category sometimes covering a physiological decrease of the mood state. However it was highlighted how this question was shown useful in the screening for potentially clinical significant depression that could lead to specific therapeutic approach [10].

Question 6 explores the existential-spiritual domain of being at peace. Some patient reported the question as too generic or non-specific:

“I would reformulate the question avoiding the word “at peace” that is unusual. I’d rather use–serene- or -feel well-“(Patient, hospice ward)

“What does it mean being at peace? Only the dead are at peace”. (patient, home care).

“Some patients think this concept of peace is for religious people, for other peace is just the opposite of war, above all those elderly who lived the war.” (Professional in a focus group).

For others this term was “too much religious”. Participants suggested “serene” which is very used in Italian. It is not “being calm” which is seen as the opposite of “worried”, therefore explored in Q3. A new version “have you been at peace within yourself” was adopted since this is coherent with POS_ITA.

Question 7 explores how much patients were able to share their feelings with family and friends. According to the variation proposed in Q4 “your loved ones” was proposed. This new formulation encountered the authors’ favourable opinion because it is inclusive of any person loved by the patient independently from the boundaries or formal labels.

No particular issues came out for the information item except for a minor word change exposed in the table of the results (S3 Appendix). The last question exploring social, practical, financial and personal problems caused some misunderstanding:

“The question is vague. What kind of problems are these? Of health care? I would distinguish personal and financial”. (Patient, hospital ward). In Italian the words “financial or personal” were felt confusing for both participants and professionals. Economical seems more clear in our cultural setting than the word financial, which is linked to stock exchange meaning. Social seems to explore directly the need, personal seems not to reach the point. Many patients asked what was meant with personal, probably thinking about personality-behaviour concerns.

According to the these findings, new IPOS_ITA versions (IPOS_ITA after cognitive) were made and sent to the original authors for revision and approval.

To obtain a final revised version that include the findings from the cognitive phase of IPOS_ITA, maintain the IPOS_English structure as much as possible and integrate the POS_ITA items whenever possible, a proof reading audit (phase 6 of the Manual, S1 appendix) among the researchers and the original authors was established. Some suggested changes proposed after the cognitive phase were accepted and maintained. Other issues could not be applied and therefore a common agreement was achieved. At the end of this process a final IPOS_ITA version was produced, including two patient versions and two staff versions (3 and 7 days of recall).