Background
The primary goal of palliative care is to enhance the quality of life (QOL) of people who have life-threatening illness and their families through the prevention and relief of suffering. This requires early identification of their concerns [
1]. Hence there is a need to systematically assess both patients’ and family carers’ QOL [
2]. Self-reported measurement instruments are increasingly used to identify physical, psychological, social and spiritual care needs, assess changes and evaluate interventions [
3]. QOL measures developed for use in palliative care irrespective of the underlying diagnosis are few [
4]. Although several disease specific instruments are available [
2], there is a need for measures for use in palliative care irrespective of the disease.
The McGill Quality of life Questionnaire (MQOL) is a measure that was originally developed in Canada (in English and French) for patients with life-threatening illness (irrespective of specific disease) [
5,
6]. It is now recommended and used internationally for palliative care irrespective of diagnosis [
2,
4]. Developed for the specific situation when facing the end of life, MQOL incorporates assessment of existential well-being and includes not only negative but also positive contributors to QOL, while also considering the length of the instrument [
5,
6]. This is important as measures should not be too burdening even though they need to comprise several aspects of QOL [
7]. The increasing interest in self-reported measurement instruments for palliative care has resulted in the original MQOL being translated into about 20 languages. The measure has been recently revised and initially validated (MQOL-Revised; MQOL-R) [
8]. The MQOL-R measures four domains: physical, psychological, existential, and social; it also includes an item assessing overall QOL. As the situation can change quickly at the end of life, it has a timeframe of 2 days. In addition, an expanded version has been developed, MQOL-Expanded (MQOL-E) [
9], to include a wider variety of domains that people with life-limiting illness indicate are important to their QOL [
10,
11]. MQOL-E incorporates MQOL-R and additional dimensions of QOL: feeling of being a burden, environment, cognition and quality of health care. It consists altogether of 21 items including the single item about overall QOL [
9]. All items have a numeric response scale from 0 to 10, anchored with verbal responses at the ends/extremes. After reversed items have been rescored, zero describes the worst situation. An example of an item is “Over the past two days (48 hours) I felt: Physically terrible (0) vs. Physically well (10)” [
12]. A summary of the item content for MQOL-E is presented in Table
1.
Table 1Description of the McGill quality of life questionnaire - expanded (MQOL-E)
Single item | Overall QoL | Overall QoL |
1 | Physical | Problems with physical symptoms |
2 | Physical state |
3 | Problems due to physical functioning |
4 | Psychological | Being depressed |
5 | Being nervous or worried |
6 | Feeling sad |
7 | Fear of future |
8 | Existential | Meaning in life |
9 | Achievement of life goals |
10 | Control over life |
11 | Feeling about oneself |
12 | Social | Communication with people I care about |
13 | Relationships with people I care about |
14 | Feeling supported |
15 | Burden | Feeling about how one’s situation affects people I care about |
16 | Environment | Physical surroundings |
17 | Cognition | Clarity of thought |
18 | Memory function |
19 | Health care | Access to information |
20 | Availability of health carea |
21 | Quality of care |
The family carers’ situation is interwoven with the situation of the patient approaching the end of life [
13,
14], and their QOL is affected as well. Hence the QOL of family carers should also be assessed. Furthermore, family carers are critical for care of the patient and home palliative care organization and they may need support [
15]. The Quality of Life in Life-Threatening Illness - Family carer version (QOLLTI-F) is a companion instrument to MQOL-E and has been psychometrically validated with carers of cancer patients [
16‐
18]. The QOLLTI-F originates from qualitative interviews exploring what family carers of people at the end of life indicate is important for their own QOL. The interview results identified not only burdens but also positive experiences in the caregiver’s situation which are included in the questionnaire. The QOLLTI-F includes seven subscales assessing different domains: environment, patient condition, the carer’s own state, carer’s outlook, quality of care, relationships and financial worries. The QOLLTI-F v2 consists of 17 items and, as in MQOL-E, a single item about overall QOL. Unique to QOLLTI-F is an item asking family carers about the patient’s condition, which is very important to their own QOL [
17]. Like MQOL-E, each item has a numeric response scale ranging between 0 and 10, anchored with verbal responses at the ends/extremes and after reversed items have been rescored, zero describes the worst situation [
17]. An example of an item is “Over the past two days (48 hours) I had time to take care of myself: Never (0) vs. Always (10)”. A summary of the item content for QOLLTI-F v2 is presented in Table
2. The QOLLTI-F has been translated into about 10 languages.
Table 2Description of the quality of life in life-threatening illness - family carer/caregiver version (QOLLTI-F v2)
Single item | Overall QoL | Overall QoL |
1 | Environment | Satisfaction with place of care |
2 | Privacy |
3 | Patient condition | Distress related to patient condition |
4 | Carers own state | Control over life |
5 | Time to take care of oneself |
6 | Clarity of thought |
7 | Physical state |
8 | Emotional state |
9 | Carers outlook | Feeling about caring for the family member (patient) |
10 | Comfort from outlook, faith or spirituality |
11 | Meaning in life |
12 | Quality of care | Agreement with decision making process for patient |
13 | Availability of health carea |
14 | Quality of care |
15 | Relationships | Interaction with patient |
16 | Interaction with the other important people |
17 | Financial worries | Stress due to financial situation |
The availability of the instruments in different languages facilitates international comparisons. However, the validity of such comparisons requires that people are interpreting and responding to items of translated instruments in the same way (i.e., the translated items have the same meaning). To address this, contemporary measurement validity theory emphasizes examining response processes as a central aspect of measurement validity. Response processes describe “what people do, think, or feel when interacting with, and responding to, the item” [
19]. Investigation of response processes and understanding how people interpret and respond to items is particularly important for determining whether the items and scores can be interpreted in the same way when measurement instruments are translated and adapted for use in different cultures. Evidence pertaining to content validity and response processes is essential to ensuring linguistic equivalence and cultural appropriateness of the items and validity of the overall measurement instrument [
20]. Therefore, the aim of this study was to translate and examine content validity and response processes during completion of MQOL-E and QOLLTI-F v2 among Swedish patients with life-threatening illness and family carers.
Discussion
This study describes the translation into Swedish and subsequent initial validation of the content of two self-report companion instruments to measure QOL, MQOL-E for patients and QOLLTI-F v2 for family carers. Through the cognitive interviews, we examined response processes, i.e. how patients with different life-threatening illnesses and family carers interpret and respond to MQOL-E or QOLLTI-F v2 items while completing the translated instruments. Response processes are important as validity evidence to support the intended use of the measurement instruments and hence the validity of inferences in a new cultural context [
19]. This is crucial as how people think and feel while responding to items may vary between different cultures, which may consequently influence the meaning of the scores.
The results of response processes pertaining to the Swedish translations of both MQOL-E and QOLLTI-F v2 contribute evidence regarding content validity, linguistic equivalence and cultural appropriateness of the translated instruments. After some changes to the translation based on the initial cognitive interviews, the items were understood and interpreted as intended and the instruments were appreciated as relevant and important to the target groups. The items triggered reflections in areas of importance and participants emphasized that responding was meaningful to them.
The results from the cognitive interviews provided important insights on the translation, which made it possible to correct and improve the clarity of the Swedish versions. The analysis showed that the issues were questions of choosing the optimal word for clarity in the translation rather than cultural differences. The results also show that the content was generally considered relevant for both MQOL-E and QOLLTI-F v2. This may be related to the fact that qualitative interviews were the foundation for both instruments, which were developed from patients’ or family carers’ views on what is important to them towards end of life [
6,
17].
Two patients who at the time were in a more stable stage perceived that the MQOL-E would have increased relevance later in their illness trajectory. QOLLTI-F v2 was also proposed to be particularly relevant for carers of severely ill patients in homecare. This suggests that the instruments may be more relevant in late rather than early palliative care, but results also indicate that respondents felt that for understanding changes in their QOL over time (or as the disease progresses), it is important to include these items at earlier as well as later stages.
The patients’ suggestions for additional items needed to capture QOL in MQOL-E concerned different aspects of other persons’ influence on QOL. This highlights the significance of people in the environment and social interaction towards end of life. The original MQOL included an item on “the world has been: an impersonal, unfeeling place vs. caring and responsive to my needs” but was taken out in favor of an item referring to “people I care about …” because “the world” was unclear to some people.
Some family carers suggested that QOLLTI-F v2 is missing items on some aspects of being a family carer. One suggested a specific question on information from health care professionals to the family. This is in line with studies showing that to know what to expect in the future is a major need in family carers [
25‐
27].
Both for MQOL-E and QOLLTI-F v2 the suggested additional items illuminate that QOL is an individual construct and that different aspects vary in importance [
28]. However, in general the participants found the instruments relevant and the content important and engaging. In instrument development there are always decisions on what is most important versus the number of items and respondent burden [
29]. In clinical care, one way to handle items suggested as additions without expanding the instruments might be to make place for the possibility to comment in free text on the content of the instrument [
30].
Some family carers found that the timeframe of 48 h in QOLLTI-F v2 was problematic. Interestingly, these were in different ways i.e. both as too short and framing and as too long since some items vary over hours. This highlights the instability of some aspects of QOL in life-threatening illness and is important to consider when assessing it. This variability towards the end of life was also the reason why the original developer decided on a 48-h timeframe for the responses in MQOL and QOLLTI-F, rather than the more common timeframes of a week or a month [
6,
17]. However, it is important to bear in mind that the last 48 h may have been extreme in some sense and scores should not be interpreted as representative of QOL over a longer time.
There were some comments that it would be easier if the response scale was narrowed to 0–5. However, in development of these measures it was found that a scale of 0–10 was accepted as intuitively easy by most participants (Cohen, personal communication) which was also the case in the present study. There were also some comments that the opposing end anchors varied in strength. However, this was purposefully done to get a decent distribution for some items, as in the development of both the MQOL and the QOLLTI-F, no one selected responses near the extreme negative end anchor (Cohen, personal communication).
The present results show that some patients occasionally responded reverse to what they intended in the items where the response scale changes the direction. This was also pointed out by family carers as tricky and demanded attention before responding. This issue is however complex and not easy to avoid as people have different ways of thinking of “10”: either as “the most” (whether or a positive or negative thing) or as “the best”. The consequence of changing the positive or negative direction in response scales has been studied before but it was found that there was not enough reason to change this as several factors influence answers [
31]. Therefore, given the intent to retain comparability with the original instrument, the directions of the response scales remained unaltered in the Swedish translated versions.
When reflecting over items several patients and some family carers showed emotional reactions. These results show that the contents of MQOL-E and QOLLTI-F v2 are important and central to respondents and that responding may be an opportunity to reflect over important issues/questions and to express feelings as sadness but also joy. These results also show that MQOL-E and QOLLTI-F v2 may support meaningful conversations on aspects of QOL between patients and/or family carers and health care professionals. The use of QOLLTI-F v2 could be a way for family carers to communicate their own situation and be acknowledged by the health care professionals, which are aspects fundamental in palliative care. Importantly, studies have also shown that vulnerable groups such as patients at end of life [
32] and their family members [
33,
34] may find it valuable and important to participate in research.
Methodological considerations
Patients had different life-threatening illnesses such as advanced heart failure, advanced chronic obstructive pulmonary disease, end-stage kidney disease or advanced cancer and both patients and family caregivers were recruited from different settings (inpatient, outpatient and home care) which add to the strength of the study. In studies using cognitive interviews the number of interviews needed varies and depends on the objective and the data quality [
23]. Purposive sampling was used to achieve variation in age, gender, education and relationships, which added richness to the data. Furthermore, data in interviews was repetitive. Hence, the number of participants was considered sufficient to investigate content validity and response processes. Nevertheless, it is always possible that eligible participants that declined participation may have had different perspectives from the ones included. The description of background characteristics of the actual sample and the findings should facilitate judgement of transferability of findings to other contexts.
The patients were in both more stable phases and at the end of life. The patients that were interviewed about the MQOL-E were however not in the very end of life (as within a few days or weeks). This may influence the results as their views on QOL items may be related to the stage of their illness. While participants thought that the measures would capture change over time, we do not know in fact that they do in Sweden. Further validity evidence is needed to confirm psychometric equivalence, i.e. measurement invariance, of the translated and original versions of the instruments, and to ascertain sensitivity in measuring change over time [
35].
In the present study we translated and validated an interim version of MQOL-E and the version 2 of QOLLTI-F. During the study a validated version of MQOL-E has been published and item 20 (Availability of health care) has been excluded. No other changes were made [
9]. Furthermore, a version 3 of QOLLTI-F has been developed. The only change that has been made from QOLLTI-F v2 is that item 13 (Availability of health care) has been excluded. The present validation is therefore significant also for the latest versions of the instruments
1.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.