FormalPara Key Summary Points

Physical and mental fatigue can be assessed by the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-Fatigue) scale. Currently, there is limited evidence on how valid the content of this questionnaire is to assess fatigue experienced by patients with systemic lupus erythematosus (SLE).

This study investigated patients’ experience with fatigue related to SLE, and whether the FACIT-Fatigue is an appropriate questionnaire to evaluate fatigue in patients with this disease.

This study highlighted several fatigue-related themes that are important to patients with SLE, and its findings support the use of the FACIT Fatigue as an easily understood method of assessing the burden from fatigue in patients with SLE.

Digital Features

This article is published with digital features, including a summary slide, to facilitate understanding of the article. To view digital features for this article go to https://doi.org/10.6084/m9.figshare.13713916.

Introduction

Systemic lupus erythematosus (SLE) is a complex autoimmune disease with a range of clinical manifestations [1] and periods of relative remission of disease activity followed by exacerbations or flares (increases in symptoms or severity of symptoms) [2]. The estimated incidence of SLE in the United States is 5.1/100,000 per year, and the prevalence is 52.2/100,000; females and those of non-white race are more likely to be affected [3]. SLE can be life-threatening when major organs are impacted. More frequently, it presents with a number of chronic and debilitating non-specific symptoms [1].

Fatigue is one of the most common symptoms and is experienced by the majority of patients with SLE [4, 5]. It has been linked to worse quality of life, depression, increased work disability, and increased health care costs [6,7,8]. As fatigue imposes a high burden in patients with SLE, it is important that it can be accurately assessed to ensure that patients can be effectively monitored and managed. Moreover, both the United States Food and Drug Administration (FDA) and the European Medicines Agency recommend measuring fatigue during clinical trials of treatments for SLE [9, 10].

Despite the importance of measuring fatigue in SLE, there is no consensus on what instruments should be used, and no instrument has been developed specifically for patients with SLE [9]. The Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-Fatigue) is a 13-item patient-reported outcome (PRO) instrument designed to capture multiple aspects of fatigue [11]. Since its development in 1997 for use in patients with cancer, its appropriateness for use in other diseases has been evaluated, including autoimmune conditions like rheumatoid arthritis; and it has been used in numerous SLE clinical trials [12]. Systematic reviews of PRO measures indicated that the FACIT-Fatigue has good face and psychometric validity, and contains concepts appropriate for measuring fatigue in patients with SLE [13, 14]. Content validity of the FACIT-Fatigue has been previously established through evidence reported in studies of patients with SLE [13,14,15], as well as evidence collected through focus groups with patients with SLE [16]. The current study expands on existing evidence of content validity of the FACIT-Fatigue by collecting data directly from patients through individual one-on-one qualitative interviews. This approach allowed more in-depth information to be gathered on patients’ experience of fatigue to assess the appropriateness of the FACIT-Fatigue for use in measuring fatigue in an SLE population.

Objectives

The purpose of the present qualitative study was to assess the content validity of the FACIT-Fatigue using a hybrid approach of concept elicitation (CE) and cognitive debriefing (CD) in individual interviews. The study also aimed to gain a better understanding of patients’ experiences of living with the condition and managing the symptoms of SLE.

Methods

Study Design

This qualitative study (GSK study 209226) was designed to evaluate the content validity of the FACIT-Fatigue for use within the SLE population. The design is in accordance with the FDA Guidance for Industry: Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims and the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Clinical Outcome Assessment Emerging Good Practices Task Force Guidance on Assessing Respondent Understanding of PRO Instruments [17, 18]. The study approach used hybrid CE and CD interviews of patients diagnosed with SLE. The CE component aimed to gain insight into the symptoms experienced by patients living with SLE, and how these symptoms impact daily life including experience and impacts of fatigue. This information was used to better understand whether the concepts covered by the FACIT-Fatigue are relevant and appropriate to measure fatigue in those with SLE. The CD component aimed to determine whether the FACIT-Fatigue is relevant and easily understood by participants. Interviews were conducted by telephone in July 2018.

Study Participants

The sample size of 15 participants was determined based on industry best practice regarding the number of interviews needed to reach saturation of a concept, i.e., the point at which additional interviews are not expected to yield additional information [18,19,20]. Participants were screened and recruited from across the USA via an existing panel provided by a global research company specializing in healthcare data collection. The following inclusion criteria were applied: adults aged ≥ 18 years with a self-reported doctor’s diagnosis of SLE for ≥ 6 months and who self-reported to testing positive for antinuclear antibodies or anti–double-stranded deoxyribonucleic acid, who reported persistent SLE symptoms in the past 6 months or self-reported at least one SLE flare in the past 12 months despite treatment with steroids and/or an immunosuppressant, and who had the ability to complete the interview in fluent English. Participants were selected, based on self-reported screening questions, to ensure a diverse range of disease severity, number of flares, and types of symptoms commonly experienced in SLE. Ethics approval was granted by the New England Independent Review Board (120180137) and participants provided written informed consent prior to the interview, with consent confirmed verbally at the start of each interview. The study was performed in accordance with the Declaration of Helsinki 1964 and its later amendments.

Interview Procedure

Individual interviews (90 min, by phone) were conducted by an experienced qualitative scientist with a semi-structured interview guide. Feedback from early interviews was incorporated into subsequent interviews to verify early emerging themes.

During the CE interview component, participants were asked open-ended questions to investigate their daily experience of SLE-related fatigue and its impact on their daily life. Following the CE component, participants completed the FACIT-Fatigue using a think-aloud CD approach, in which participants were asked to verbalize their thoughts as they worked through the FACIT-Fatigue, articulating how they made sense of instructions, items, and response choices [18, 21]. Participants were also asked to discuss what would be a meaningful change in their symptoms of fatigue as a result of treatment for SLE. The CD interview component was designed to assess the FACIT-Fatigue for relevance, appropriateness, comprehensiveness, and the ability to be easily understood and completed by participants with SLE. Participants were also asked to note any items or aspects of the FACIT-Fatigue that they found confusing, and these were discussed with the interviewer retrospectively upon completion. Following the think-aloud, participants were asked structured questions on the relevance, comprehensiveness, and clarity of the FACIT-Fatigue and were prompted to comment on the format, instructions, items, recall period, and response choices.

Data Analysis

Interviews were audio-recorded and transcribed, and final transcripts were anonymized to ensure any personal health information was excluded. Data were then coded using a grounded theory approach [22]. The grounded theory approach allows for themes to emerge from the data rather than imposing a priori hypotheses [22]. Content coding was completed by two qualitative analysts and a qualitative research scientist. The coding tasks for the CE and CD sections of the interviews were each completed by one qualitative analyst and reviewed by another, to support consistency. The qualitative research scientist reviewed the final coding structure and all codes. Any identified discrepancies between coders were reviewed, discussed, and resolved by the research team to help ensure the reliability of the coding.

All coding and data analysis of the CE component were carried out using NVivo, version 11.0 [23]. The first batch of five interviews was coded by both analysts to establish a preliminary coding structure; these five interviews were then recoded using the agreed set of codes. An analysis of saturation was conducted with the data from the CE component of the interview to provide evidence that enough interviews were completed to fully understand concepts important to participants. The coding process for saturation analysis involved developing a preliminary set of first-level codes by reviewing transcripts 1–5. This review facilitated an understanding of the data and helped determine which themes were emerging as most common across participants. After discussion of the first batch of transcripts, the study coding team (qualitative analysts and qualitative research scientist) reached consensus on the final set of codes with which to proceed. Concepts produced spontaneously were given preference in theme identification. Next, interviews 1–5 were recoded for second-level codes and then coded again using the entire list of new codes. Interviews 6–10 were then coded using the final list of codes. Finally, interviews 11–15 were coded to confirm saturation, ensuring that no additional, relevant concepts emerged. Following completion of the coding and data analysis, item mapping was conducted to demonstrate whether there is correspondence between the fatigue-related concepts identified in the CE interviews and the FACIT-Fatigue scale.

Narrative data from the CD portions of interviews were coded in Excel using a series of ratings to summarize feedback. Each transcript was reviewed and coded for content related to instructions, recall period, the relevance and comprehension of each item and the item response options of the FACIT-Fatigue. Items were assigned two codes: one for whether a problem was reported and a second for whether it was a spontaneous or prompted remark. A pre-specified threshold of ≥ 4 participants having a consistent comment on a particular item was set to guide conclusions, recommendations, and determinations regarding item relevance and comprehension and determining whether a reported problem warranted a change. Every reported problem, however, was evaluated by the research team to determine whether the requested change was an individual preference or style-related request or a true problem with comprehension or responding to an item.

Results

Participant Characteristics

A total of 15 participants with SLE were interviewed. The mean (standard deviation [SD]) age was 52.1 (13.1) years and 13 participants (86.7%) were female (Table 1). Almost half (7/15, 46.7%) of the participants interviewed self-reported moderate SLE, and 4/15 (26.7%) self-reported severe SLE. All participants self-reported at least one flare in the previous 12 months, with 9/15 (60.0%) reporting ≥ 4 flares in this period (Table 1).

Table 1 Patient demographics and characteristics
Full size table

Concept Elicitation

Saturation analysis showed that 90% of the high-frequency themes emerged by the tenth interview and retained their prevalence over the course of analysis, suggesting that additional interviews would not have resulted in further themes or sub-themes. Three late-emerging themes related to emotional impacts were reviewed by the research team to determine whether new information was introduced that warranted additional data collection. It was determined that these individuals had described emotional impacts that were similar to those of other participants using different words, were minor, an outlier, or technically unrelated, and did not warrant further data collection.

Overall, 27 signs and symptoms associated with SLE were reported, affecting a range of bodily systems and functions. Fatigue, pain, sleep disturbance, and cognitive issues were reported by all participants (Fig. 1). Almost all participants (14/15, 93.3%) reported that their symptoms could be brought on or intensified by triggers. Participants most commonly noted exposure to the sun/photosensitivity (12/15, 80.0%) and physical exertion (10/15, 66.7%) as triggers (Table 2). The majority of participants (9/15, 60.0%) reported that the severity of their symptoms and frequency and intensity of flares varied over time.

Fig. 1
figure 1

Commonly reported SLE symptoms*. *Only symptoms reported by at least five participants are shown. SLE systemic lupus erythematosus

Full size image
Table 2 Symptom triggers and most bothersome symptoms
Full size table

Fatigue was reported as the most bothersome symptom by 11/15 (73.3%) participants (Table 2). All reported regularly experiencing fatigue, with most feeling fatigued every day (12/15, 80.0%) and others a few days a week (3/15, 20.0%). The majority (12/15, 80.0%) also reported that their fatigue often fluctuates over the course of the day, or from day to day. Of 13 participants who described experiencing flares, ten (76.9%) reported intense fatigue associated with their flares.

Participants were asked to describe how fatigue impacted their quality of life (Table 3). Physical functioning was reported to be impacted by 9/15 (60.0%) participants. They emphasized that physical difficulties affected their employment and physical inactivity impacted their ability to maintain a healthy weight. All participants reported emotional impacts of fatigue, where the most common was frustration/stress due to being unable to lead a “normal” life and accomplish tasks. Almost all participants (14/15, 93.3%) reported ways in which fatigue had profoundly impacted their social lives. The most frequently reported social functioning difficulty was being unable to participate in social activities. Most participants (12/15, 80.0%) reported that fatigue interfered with their ability to fulfil work- or school-related roles. Additionally, activities of daily living were limited in all participants, who reported difficulty starting/finishing tasks. The majority (14/15, 93.3%) were unable to do much other than rest or sleep due to fatigue.

Table 3 Representative quotes from participants asked to describe impacts of fatigue
Full size table

All participants (14/14 [100%]) who answered the question on what a meaningful change in their fatigue would be reported that the ability to be more active and have a more normal lifestyle would have the most meaningful and important impact on their lives.

Concept mapping showed that all 13 items of the FACIT-Fatigue mapped directly onto concepts spontaneously mentioned by participants during the interviews without major gaps (Table 4).

Table 4 Concept mapping of the FACIT-Fatigue
Full size table

Cognitive Debriefing

Overall, the FACIT-Fatigue was easy to complete and understand by participants and was comprehensive for capturing the symptoms and impacts of fatigue related to SLE. The think-aloud CD method revealed that 13 participants found the instructions easy to understand and 11 endorsed the recall period (7 days) as appropriate. FACIT-Fatigue items were generally found to be clear and relevant to patients’ experiences with fatigue. No items received four or more common concerns from study participants (Table 5) and follow-up queries ensured that any concern reported by participants did not interfere with their ability to understand or respond to an item.

Table 5 Patient comments per item of the FACIT-Fatigue from cognitive debriefing
Full size table

Due to the complexity of SLE and its unique presentation, two participants wanted the response options to capture more detail, such as adding comment boxes to enable elaboration with open-ended responses. However, most participants (11/15, 73.3%) reported that the response options were adequate to capture their experience of fatigue for each item.

Discussion

Participants in this study described fatigue as the most debilitating symptom of their condition. These findings confirm fatigue as an important concept to patients with SLE and suggest the value of including it as a secondary endpoint in randomized controlled trials and longitudinal observational studies to measure efficacy of SLE treatments [14]. Participants also confirmed the importance, relevance, and comprehensiveness of concepts covered in the FACIT-Fatigue. They reported that the instrument accurately captures the type of fatigue they experience as part of their condition, including the range of physical, emotional, and social impacts of SLE-related fatigue in daily life. These results build on previous studies and provide additional qualitative evidence of content validity of the FACIT-Fatigue for use in patients with SLE.

Although previous studies have investigated the content validity of the FACIT-Fatigue in SLE [13, 15, 16], the current study was the first investigation to assess content validity using individual in-depth qualitative interviews. Findings from the current study parallel evidence gathered in a previous study that used focus groups, lending further support to the FACIT-Fatigue as relevant for measuring fatigue in patients with SLE [16]. This study also expanded on the prior work by Kosinski et al., who reported that patients found two items of the FACIT-Fatigue (“I’m too tired to eat” and “I have energy”) less relevant than other items. The time spent with each individual in our study allowed for deeper probing into the meaning of these items and the relevance of the concepts to the participants.

While Kosinski et al. found that only a small number of patients were physically too tired to eat due to the effort required for meal preparation [16], nearly all participants in the current study (14/15, 93.3%) reported that the item “I’m too tired to eat” covered a relevant and important aspect of SLE-related fatigue. In-depth probing on these items revealed that most participants (12/15, 80.0%), interpreted the item as intended, namely as the act of eating rather than food preparation. A subset of these participants (5/12, 41.7%) also indicated that preparing food would be a separate item for which they might provide a different response, further demonstrating their interpretation of the original item as the physical act of eating. Many participants also reported difficulty maintaining a healthy diet because SLE-related fatigue interfered with both eating and meal preparation.

The item “I have energy” was also reported in Kosinski et al. to be less relevant to patients with SLE, as most patients were continually tired and without energy [16]. Individual interviews in the current study allowed for additional inquiry of those participants who reported “0 – not at all” to this item during the cognitive debriefing exercise. Most participants (14/15, 93.3%) described this item as relevant and easy to understand, and all reported experiencing times when they did have energy. Those who had responded “0 – not at all” on this item also reported experiencing moments when their energy levels fluctuated and were closer to what they considered normal. Nine of these participants described energy levels that fluctuated over a week or a day, and two described their energy as self-motivated. Thus, despite findings of previous studies to the contrary, this study supports the relevance of these two items (eating and energy) as important and relevant to patients experiencing SLE-related fatigue.

Overall, cognitive interviews in the current study demonstrated that participants had no problem understanding the instructions, items, or response choices of the FACIT-Fatigue, and found the recall period to be acceptable. Although some difficulties were reported in interpreting items, no consistent problems or suggestions were reported by participants that interfered with their ability to understand or respond to an item. Of the minimal problems reported, many were from an 85-year-old participant whose SLE-related symptoms may have also been confounded by other age-related symptoms.

Like most qualitative research in rare diseases, conclusions drawn from the current study are based on a small sample size. Additionally, given the higher prevalence of SLE in females [24, 25] there were difficulties in recruiting male participants. However, while formal analyses were not conducted, the experience of fatigue and the responses from the two male participants were not noticeably different from those of the female participants. Another limitation is that the diagnosis of SLE, the number of flares patients experienced in the past 12 months, and severity were all self-reported by participants during screening rather than derived through disease activity indices or clinician confirmation. Potential misclassification relating to reported diagnosis was minimized by requesting information known only to those with SLE. Furthermore, almost all participants (14/15) indicated during screening that they had received a positive antinuclear antibody test. Both the range of severity levels reported during screening and the wide array of symptoms and impacts revealed during interviews confirmed the inclusion of multiple levels of severity in this sample. While our aim was to ensure the FACIT-Fatigue was appropriate to capture fatigue across different levels of severity, as experienced by the patient, we also recognized that the absence of a validated tool in this study suggests that reported levels may not parallel physician assessment of patient disease activity. Discrepancies between patient and physician assessment of SLE disease activity are well documented [26]; and thus, participants in this study could have overestimated the level of severity. Furthermore, four out of 15 participants reported that they did not know their level of severity. Other clinical characteristics such as medication use and existing comorbidities were not collected. Therefore, any potential impact of medication use and conditions comorbid to SLE on reported levels of fatigue is unknown and should be acknowledged as a limitation. In order to test the English version of the FACIT-Fatigue, the sample was limited to those who read and spoke fluent English. Further research would therefore be necessary to confirm that the items would be as well understood in a less-fluent English-speaking or non-English-speaking population. Additionally, information on socio-economic status was not collected for this sample. It is possible that facets of socio-economic status, including income and education level, could impact understanding of the FACIT-Fatigue items.

Despite these limitations, this study used a sound and rigorous methodological approach to explore the primary research questions. Qualitative interviews were conducted on an individual level, which allowed for detailed data to be collected without responses being influenced by other participants. It also allowed for in-depth probing to expand on and clarify findings from previous studies. Additionally, the exploratory nature of the CE component facilitated the uncovering of the most important concepts related to participants’ experience of fatigue in SLE; and the think-aloud approach of the CD component confirmed relevance of the concepts included in the FACIT-Fatigue and participants’ understanding of the instrument items.

Conclusions

The results of this study provide qualitative evidence to support the FACIT-Fatigue as an appropriate measure for assessing fatigue and its impacts on health-related quality of life in patients with SLE. The instrument measures fatigue-related concepts relevant to SLE, using items that are easily understood. Fatigue has been shown to be the most commonly reported symptom of SLE, greatly impacting how patients are able to function in their day-to-day life, warranting more standardized use of measures that are able to capture both the essence of this fatigue and its impacts. Furthermore, the results support the use of the FACIT-Fatigue in SLE clinical trials and other research exploring fatigue in this population.