Domains for inclusion in a novel Treatment Response Measure for Systemic Lupus Erythematosus (TRM-SLE): results of a modified Delphi study

Abstract

Objectives

To achieve consensus on domains of active disease for inclusion in a novel outcome measure for SLE randomised controlled trials (RCTs), the Treatment Response Measure for SLE (TRM-SLE).

Methods

Domains nominated by TRM-SLE Taskforce members were rated in a two-stage modified Delphi study. Each stage comprised two online survey rounds separated by a structured discussion meeting. In Stage 1, expert lupus clinicians and patient representatives rated domain ‘importance’ (impact on symptoms, function or survival). In Stage 2, clinicians rated ‘important’ domains on three characteristics relevant to RCT utility: ‘appropriateness’ for evaluating change in disease activity, ‘representation’ in patients with active SLE and ‘measurability’ in an RCT context. Consensus for domain inclusion was prespecified as all four characteristics achieving a rating ≥7 on a 1–9 scale by ≥70% of participants.

Results

Domain nominations from 36/59 (61%) TRM-SLE Taskforce members yielded 34 potential domains which were rated in the modified Delphi study. At least one Delphi round was completed by 87 clinicians and 13 patient representatives. In Stage 1, 14 domains met consensus on ‘importance’ in both clinician and patient groups, and 11 domains met consensus among patients only. After Stage 2, eight of these domains also reached consensus on ‘appropriateness’, ‘representation’ and ‘measurability’: alopecia, arthritis, haemolytic anaemia, nephritis, mucosal ulcers, rash, serositis and thrombocytopenia.

Conclusions

Considering patient and clinician perspectives, we reached consensus to include eight disease activity domains for future development into the novel TRM-SLE clinical trial outcome measure, aiming to improve trial interpretability and success.

Correspondence to Professor Eric Morand; [email protected]

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Current outcome measures used to define response in SLE clinical trial endpoints have performed inconsistently and are not fit-for-purpose.

• The Treatment Response Measure for SLE (TRM-SLE) project aims to develop a novel instrument to measure meaningful improvement in disease activity, specifically for SLE clinical trials.

WHAT THIS STUDY ADDS

• Using Delphi methods, we achieved consensus to include eight domains of active disease to define treatment response in TRM-SLE, considering patient and clinician perspectives on domain importance, and factors influencing domain utility in the specific context of a clinical trial.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Treatment response defined by our novel outcome assessment will involve measurement of clinically meaningful improvement in key SLE disease activity domains; this may improve trial interpretability and support the approval of novel therapies.

Introduction

SLE is a complex disease associated with substantial individual and public health burden.^1–3 Despite many promising treatments being studied, only three drugs have been approved for SLE or lupus nephritis in the last 60 years.^4–8 A major barrier faced in randomised controlled trials (RCTs) is the complexity associated with measuring disease activity changes in response to treatment, due to the multisystem involvement and clinically heterogeneous nature of SLE.⁹ Clinical outcome assessments (COAs) currently used to measure treatment response in SLE trial endpoints have repeatedly produced inconsistent outcomes, compromised the interpretability of trial results and encumbered regulatory approval.^10–14 New ways to define treatment response in SLE clinical trials, for both adult and childhood-onset patients with SLE, are urgently needed.¹⁵

To support regulatory approval of new treatments, contemporary guidance requires that RCTs demonstrate treatment benefit relevant to how patients ‘feel, function or survive’, using validated COAs specific for the intended ‘context of use’ (the target patient population, study design and setting).^{16 17} Therefore, COAs used to measure treatment efficacy in SLE RCTs should measure concepts that are meaningful to patients with SLE and additionally must consider factors specific to the RCT setting and patient population. Current primary efficacy endpoints used in the vast majority of SLE RCTs, the SLE Responder Index (SRI)¹⁸ and British Isles Lupus Assessment Group (BILAG)-based Combined Lupus Assessment (BICLA),¹⁹ define responder status based on improvement in legacy SLE disease activity instruments. These instruments were not developed for RCTs or to specifically quantify treatment responses and did not include patient input in their design.¹⁰ Therefore, the concepts included in current endpoints and scoring of responses were not specifically developed to capture clinically meaningful treatment benefits in an RCT context.

When assessing treatment response, COAs used in current trial endpoints evaluate virtually all manifestations of active disease. Although comprehensive in breadth, detailed assessment of individual manifestations is precluded by the vast number of items that require evaluation. Understanding responses to treatment and the efficacy of novel therapies at a more granular organ-specific level is therefore limited with current endpoints. We know from both cohort studies and a recent review of SLE RCTs that many SLE features occur very infrequently.^{20 21} Furthermore, some SLE manifestations are currently very difficult to characterise and measure; the inclusion of such features when defining responders in an RCT may compromise the reliability, validity and interpretability of trial results.

The Treatment Response Measure for SLE (TRM-SLE) project aims to develop a novel COA designed specifically to measure clinically meaningful treatment effects in an RCT endpoint that can support the regulatory approval of novel therapeutic agents in SLE. The project is led by the TRM-SLE Taskforce consisting of adult and paediatric clinicians, patient partners, industry, trial and regulatory experts, and methodologists. As outlined in the published study protocol, a series of five aims comprise a literature review, establishment of the measurement goals and context of use, achieving consensus on which domains to measure, determining how to measure response in each domain and finally compiling the domain measures into an integrated multidomain instrument (figure 1).¹¹ We previously reported the first outcomes of the project, in which consensus was achieved on the measurement goals and context of use.¹¹ In brief, the novel COA will measure ‘active immune-mediated disease manifestations that impact on the patient and are modifiable by therapy to reduce or control disease activity’ in the specific context of an SLE RCT that may include adult and/or adolescent patients with SLE.¹¹ The aim of the current study, as the next step in developing this COA, was to achieve consensus on the domains of active disease that will be measured to capture treatment response in the novel instrument, prioritising domains based on importance from both patient and clinician perspectives, and characteristics relevant to utility in an RCT context. Consensus on domain inclusion was achieved via a domain generation survey followed by a two-stage modified Delphi study supported by focused literature reviews.

Figure 1. Overview of the domain selection process, in the context of the five stages of TRM-SLE instrument development. TRM-SLE, Treatment Response Measure for SLE.

Methods

Domain generation survey

TRM-SLE Taskforce members, who are drawn from all continents, were invited to nominate domains deemed potentially suitable for inclusion in the TRM-SLE instrument. Using the Qualtrics online survey platform, participants could nominate domains representing active SLE based on those included in the SLE Disease Activity Index 2000²² and/or suggest domains in a free-text field. Domains nominated by only a single respondent in the initial survey were presented in a second survey, to allow supporting nominations from other participants. Only domains receiving more than one vote across the two surveys were considered further. Synonymous or related nominations were grouped, and a core list of domains generated to proceed to the subsequent modified Delphi study.

Modified Delphi study for domain selection

Consensus on domain inclusion in our novel COA was established via a two-stage modified Delphi study, with each stage comprising two online survey rounds separated by a virtual discussion meeting (figure 1). Delphi methodology is a widely accepted structured process for reaching expert consensus via survey rounds in multiple iterations,²³ and a commonly used technique in the development of outcome measures in rheumatology.²⁴ Online survey rounds were conducted using DelphiManager software, and discussion meetings were held via Zoom video communications. In each Delphi survey round, participants rated each domain on the prespecified characteristic(s) of interest (importance, appropriateness, representation or measurability, as defined in figure 1) on a numerical scale of 1 (lowest rating) to 9 (highest rating). Participants were also provided with an ‘unable to rate’ option and free-text fields. Following each survey round or discussion meeting, participants were provided with the results of the preceding phase to inform their subsequent participation. In the second survey round of each Delphi stage, this included each participant’s own ratings alongside group-level data, as well as access to a recording and written summary of the discussion meeting. Only participants who completed the first round of each Delphi stage were invited to participate in the subsequent discussion meeting and survey round(s).

Modified Delphi study: stage 1

In Stage 1, clinicians and patient partners rated each domain on ‘importance’, defined as the extent to which domain activity impacts aspects of health that are meaningful to patients with SLE, that is, how they ‘feel, function or survive’. All lupus clinicians and patient partner members of the TRM-SLE Taskforce were invited to participate, in addition to external clinician experts nominated by Taskforce members, based on known expertise in lupus clinical care and/or trials. Participants rated domain ‘importance’ from 1 (not important) to 9 (critically important). The prespecified consensus threshold for domain ‘importance’ was a rating of ≥7 by at least 70% of participants in both clinician and patient groups. However, after review of the Stage 1 results by the TRM-SLE Taskforce, including patients, it was agreed to revise the consensus criteria to additionally retain domains rated as ‘important’ by either clinician or patient groups in Stage 2.

Stage 1 was supported by a literature review identifying associations of each domain with how a patient with SLE ‘feels, functions or survives’.^{16 17} First, we performed a systematic search of three databases (MEDLINE, CINAHL and PsycInfo) to identify patient experience data reporting on symptoms and impacts of SLE deemed important by patients. Then, targeted literature searches of MEDLINE were performed to identify studies reporting associations between domains and key SLE outcomes, including the symptoms and impacts highlighted by the previously identified patient experience data, as well as quality of life, damage and mortality. Statements summarising literature results were provided to Delphi participants as ‘help text’ in the online Delphi survey platform (see online supplemental material).

Modified Delphi study: stage 2

Domains meeting consensus for ‘importance’ in clinician and/or patient groups in Stage 1 proceeded to a second Delphi stage. Here, clinicians who participated in Stage 1 were invited to rate domains on three additional characteristics relevant to their utility in assessing treatment responses in an SLE RCT: ‘appropriateness’ based on the measurement goals for TRM-SLE (whether a domain is an active immune-mediated disease manifestation that impacts the patient and is modifiable by therapy to reduce or control disease activity¹¹), ‘representation’ in patients with active SLE (whether the domain occurs with sufficient frequency in patients with SLE with active disease to warrant inclusion) and ‘measurability’ (whether change in domain activity in response to treatment can be quantified in the context of a SLE RCT). Participant ratings of ‘representation’ were supported by a literature review of clinical trial and cohort data summarising the reported frequency of domain activity in patients with active SLE (see online supplemental material). Patient partners, who did not complete the Delphi surveys in Stage 2, were nonetheless invited to participate in the discussion meeting and encouraged to ask questions and comment on clinician ratings. Consensus was predefined as at least 70% of participants rating a domain ≥7 (1–9) on all three characteristics rated in Stage 2.

Statistical analysis

Statistical analyses were performed using Stata SE V.16.0 (StataCorp). Study participant characteristics were summarised descriptively. Delphi survey responses were reported as median and IQR and the proportion of participants rating a domain across three predefined rating categories (1–3=unsuitable to include, 4–6=uncertain suitability and 7–9=suitable to include, based on the respective characteristic being rated), including stratification by clinician/patient subgroup where relevant. The disagreement index (DI)²⁵ was used to quantify the level of agreement for each domain rating, with high agreement indicated by a DI <1. Qualitative data (from survey comment fields and discussion meetings) were grouped and reported thematically. No imputations were done for missing data.

Patient and public involvement

The TRM-SLE Taskforce includes a patient advisory panel which provides patient-centred input and strategic guidance throughout the instrument development process. The patient advisory panel consists of patients with SLE, caregivers and lupus patient organisation representatives, with membership across Europe, North America, Latin America and Asia. In this study, the patient advisory panel contributed to study design, in particular leading decisions around the nature and scope of patient involvement. They were also invited to contribute as study participants and to review and provide a patient perspective on outcomes.

Results

Domain generation

The domain generation survey was completed by 36/59 (61%) TRM-SLE Taskforce members. There were 64 unique nominations, of which 47 were nominated by multiple participants. A second survey limited to domains nominated by only a single respondent (n=17) was completed by 29/63 (46%) Taskforce members, resulting in exclusion of two domains which did not receive further supporting votes. The 62 remaining nominations were consolidated by grouping related suggestions into a core list of 34 domains, which were rated in the subsequent modified Delphi study. This process is summarised in figure 2.

Figure 2. Domain generation process and participants. SLEDAI, SLE Disease Activity Index; TRM-SLE, Treatment Response Measure for SLE.

Modified Delphi study: participant characteristics

Of 118 individuals invited to participate, 100 (85%) completed at least one Delphi survey round. Table 1 describes participant characteristics. Expert SLE clinicians (n=87) included rheumatologists (90%) and other specialists who provide care to both adult (92%) and paediatric (22%) patients. 56 (64%) clinicians had not participated in previous phases of the TRM-SLE project. Patient partners (n=13) were members of the TRM-SLE Patient Advisory Panel, of whom 10 had diagnoses of SLE. Both clinician and patient groups included representatives from multiple geographical regions.

Table 1

Characteristics of Delphi study participants

Participant characteristics	Median (IQR) or n (%)
Clinician experts (n=87)
Age (years)	52 (43–63)
Gender
Female	36 (41)
Male	51 (59)
Region
Africa	4 (5)
Asia	11 (13)
Australia/New Zealand	17 (20)
Europe	24 (28)
Latin America	7 (8)
North America	24 (28)
TRM-SLE Taskforce member	31 (36)
Medical specialty*
Rheumatology	78 (90)
Dermatology	2 (2)
Immunology	3 (3)
Internal medicine	12 (14)
Nephrology	4 (5)
Other	1 (1)
Years of experience managing SLE	21 (11–30)
Manages adult patients with SLE	80 (92)
Manages paediatric patients with SLE	19 (22)
Patient partners (n=13)
Age (years)	52 (43–63)
Gender
Female	11 (85)
Male	2 (15)
Geographical region
Africa	0 (0)
Asia	1 (8)
Australia/New Zealand	0 (0)
Europe	6 (46)
Latin America	1 (8)
USA/Canada	5 (38)
Diagnosed with SLE	10 (77)
Disease duration	19 (16–33)
Age of diagnosis	25 (19–32)

*Multiple responses possible.

TRM-SLE, Treatment Response Measure for SLE.

Modified Delphi study: stage 1

Stage 1 results are summarised in table 2; the full set of results is available in online supplemental material. The first Delphi survey round was completed by 100/118 (85%) invitees, of whom 87 were clinicians and 13 were patient partners. Of these, 73/100 (73%) attended the discussion meeting and 85/100 (85%) completed the second Delphi survey round.

Table 2

Proportion of participants rating domains highly (7–9 on a 9-point scale) after two Delphi voting rounds based on ‘importance’, defined as being associated with how a patient feels, functions or survives, when assessing treatment effect in an SLE clinical trial

Consensus on importance	Candidate domain	Rated highly important (7–9)
		Clinicians	Patient partners
Consensus (≥70% participants scoring ≥7) in both clinician and patient groups	Nephritis*	86/87 (99%)	13/13 (100%)
	Neuropsychiatric lupus*	85/87 (98%)	12/13 (92%)
	Rash	71/74 (96%)	8/11 (73%)
	Haemolytic anaemia	69/74 (93%)	9/11 (82%)
	Arthritis*	78/87 (90%)	11/13 (85%)
	Thrombocytopenia	65/74 (88%)	10/11 (91%)
	Ophthalmic lupus	65/74 (88%)	9/11 (82%)
	Myositis	64/74 (86%)	10/11 (91%)
	Vasculitis*	71/86 (83%)	9/12 (75%)
	Fatigue	56/74 (76%)	11/11 (100%)
	Serositis*	67/87 (77%)	9/12 (75%)
	Brain fog	54/74 (73%)	11/11 (100%)
	Antiphospholipid syndrome	55/74 (74%)	10/11 (91%)
	Pulmonary lupus*	62/87 (71%)	11/12 (92%)
Consensus (≥70% participants scoring ≥7) achieved in patient group only	Constitutional symptoms	51/74 (69%)	9/11 (82%)
	Mucosal ulcers	49/74 (66%)	9/11 (82%)
	Alopecia	47/74 (64%)	8/11 (73%)
	Gastrointestinal lupus	47/74 (64%)	8/11 (73%)
	Depression/anxiety	35/74 (47%)	9/11 (82%)
	Osteoporosis	26/74 (35%)	9/11 (82%)
	Fibromyalgia	30/74 (41%)	10/11 (91%)
	Hypertension	18/74 (24%)	9/11 (82%)
	Raynaud’s/poor circulation	18/74 (24%)	9/11 (82%)
	Sjogren’s syndrome/dryness	13/74 (18%)	10/11 (91%)
	Overlap autoimmune disease	6/74 (8%)	8/10 (80%)
Consensus not achieved in either group	Complement	40/74 (54%)	4/11 (36%)
	Anti-dsDNA antibodies	35/74 (47%)	0/11 (0%)
	Leucopenia	30/74 (41%)	4/11 (36%)
	Anaemia of chronic disease	20/74 (27%)	4/11 (36%)
	Lymphadenopathy	20/74 (27%)	0/11 (0%)
	Erythrocyte sedimentation rate	11/74 (15%)	2/11 (20%)
	Anti-Smith antibodies	5/74 (7%)	3/11 (38%)
	Costochondritis	1/74 (1%)	1/11 (10%)
	Serum iron	0/74 (0%)	0/11 (0%)

*Met consensus after one round of voting, therefore not rerated in the second Delphi round.

anti-dsDNA, anti-double stranded DNA.

Of the 34 domains rated, 6 met consensus on ‘importance’ in both clinician and patient groups after the first Delphi survey round, increasing to 14 domains following the discussion meeting and second voting round. Discrepancies between clinician and patient ratings were noted across a number of domains, including 11 domains that met consensus on ‘importance’ in the patient group only after two Delphi voting rounds. Patients were more likely to rate symptom-based domains (eg, fatigue) and domains less clearly attributable to active inflammation (eg, depression/anxiety) as being of higher ‘importance’. Both clinician and patient groups rated domains representing major organ involvement highly and placed lower importance on laboratory measures. Key findings arising from the Delphi survey that were reviewed during the discussion meeting included clarifying the definition of ‘importance’ and ensuring this was the focus of Stage 1 ratings. Participants were encouraged to rate ‘importance’ in isolation, with other relevant considerations (eg, clinical trial feasibility) being addressed in separate phases of domain selection and instrument development. Each individual domain rating was discussed, with a focus on the distribution of ratings and level of agreement, discordance between clinician and patient ratings, and discordance between participant ratings and findings of the literature review. Following the second voting round, the 25 domains which met consensus on ‘importance’ in either clinician and/or patient groups advanced to Stage 2.

Modified Delphi study: stage 2

Of the 87 clinicians who participated in Stage 1, 74 (85%) completed the first survey round of Stage 2. The discussion meeting was attended by 46/74 (62%) and the second survey round completed by 70/74 (95%) of first round participants. Results of Stage 2 are summarised in table 3, and the full set of all domain rating results is available in the online supplemental material.

Table 3

Proportion of participants rating domains highly (7–9 on a 9-point scale) after two Delphi voting rounds on ‘appropriateness’, ‘representation’ and ‘measurability’, when assessing treatment effect in an SLE clinical trial

Candidate domain	Rated highly (7–9)
	‘Appropriate’	‘Representative’	‘Measurable’
Alopecia	55/70 (79%)	62/70 (89%)	53/70 (76%)
Antiphospholipid syndrome/thrombosis	20/70 (29%)	31/70 (44%)	32/70 (46%)
Arthritis*	72/74 (97%)	73/74 (99%)	73/74 (99%)
Brain fog/cognitive impairment	14/70 (19%)	23/70 (31%)	10/70 (14%)
Constitutional symptoms	43/70 (58%)	42/70 (57%)	34/70 (46%)
Depression/anxiety	7/70 (10%)	14/70 (19%)	19/70 (26%)
Fatigue	23/70 (31%)	38/70 (51%)	34/70 (46%)
Fibromyalgia	2/69 (3%)	9/69 (12%)	6/69 (8%)
Gastrointestinal lupus	47/69 (64%)	23/69 (32%)	28/69 (38%)
Haemolytic anaemia*	73/74 (99%)	53/73 (73%)	72/73 (99%)
Hypertension	8/70 (11%)	13/70 (18%)	36/70 (49%)
Mucosal ulcers*	58/74 (78%)	57/74 (77%)	54/74 (73%)
Myositis	64/70 (87%)	35/70 (47%)	61/70 (82%)
Nephritis*	74/74 (100%)	74/74 (100%)	74/74 (100%)
Neuropsychiatric lupus	63/70 (85%)	48/70 (65%)	41/70 (55%)
Ophthalmic lupus	56/70 (76%)	24/70 (32%)	34/70 (46%)
Osteoporosis	3/70 (4%)	10/70 (14%)	19/70 (26%)
Pulmonary lupus	61/70 (82%)	36/70 (49%)	48/70 (65%)
Rash*	74/74 (100%)	74/74 (100%)	72/74 (97%)
Raynaud’s/poor circulation	18/70 (24%)	18/70 (24%)	21/70 (28%)
Overlap autoimmune disease	8/70 (11%)	9/70 (12%)	12/70 (16%)
Serositis*	71/74 (96%)	66/74 (89%)	57/74 (77%)
Sjögren’s syndrome/dryness	13/70 (18%)	18/70 (24%)	19/70 (26%)
Thrombocytopenia*	71/74 (96%)	62/74 (84%)	71/74 (96%)
Vasculitis	69/70 (95%)	45/70 (62%)	49/70 (66%)

*Met consensus after one round of voting, therefore not rerated in the second Delphi round.

After the first round of voting in Stage 2, seven domains met consensus for inclusion in our novel COA: arthritis, haemolytic anaemia, mucosal ulcers, nephritis, rash, serositis and thrombocytopenia. Key points addressed in the Stage 2 discussion meeting included feasibility considerations specific to the clinical trial setting, the relevance of considering rare manifestations in a clinical trial responder definition, potential mechanisms to handle excluded domains and challenges in the attribution of some manifestations to active disease. Following the discussion meeting and a second Delphi survey round, one additional domain (alopecia) met consensus for inclusion.

Figure 3 provides detailed domain rating results across the two Delphi stages for the eight domains that ultimately met consensus for inclusion in the novel COA. Figure 4 offers illustrative examples of domain rating results across the two Delphi stages for domains that did not reach consensus for inclusion after Stage 2. Results in this style were provided for all domains to Delphi participants between rating rounds and during discussion meetings (available in online supplemental material). Figure 5 summarises the overall outcomes of the modified Delphi study and includes categorisation of the reasons why domains failed to reach consensus for inclusion.

Figure 3. Detailed results from the two-stage Delphi study for the eight domains meeting consensus on all four characteristics: ‘importance’, ‘appropriateness’, ‘representation’ and ‘measurability’. Full results for all domains are provided in the online supplemental material . Red <=30% agreement, yellow >30-<70% agreement, green >=70% agreement. DI, disagreement index (<1 indicates high agreement).

Figure 4. Detailed results from the two-stage Delphi study for example domains that did not meet consensus for inclusion in our novel instrument. Full results for all domains are provided in the online supplemental material . Red <=30% agreement, yellow >30-<70% agreement, green >=70% agreement. DI, disagreement index (<1 indicates high agreement).

Figure 5. Summary of domain selection outcomes based on ratings of ‘importance’, ‘appropriateness’, ‘representation’ and ‘measurability’ and potential implications for domain measurement in clinical trials.

Discussion

New approaches to measuring clinically meaningful treatment effects on disease activity in SLE RCTs are urgently needed to improve our ability to understand the efficacy of new treatments and support their regulatory approval, in both adult and childhood-onset SLE. Two recent Phase 3 trial programmes for SLE therapies have had discordant findings, where parallel trials yielded divergent outcomes, making interpretation of results very challenging.^10–14 The endpoints currently used to determine treatment efficacy in SLE RCTs are one major addressable contributor to these challenges facing SLE trial programmes. The TRM-SLE project aims to develop a new COA designed specifically for use in an SLE RCT endpoint that focuses on the detailed measurement of domains of active disease that are meaningful to patients and most suited to determining treatment response in an RCT setting.¹¹ Via a multistage domain selection process that used Delphi consensus methodology and incorporated input from all major stakeholder groups, we achieved consensus on eight domains that will determine treatment response in the novel COA: alopecia, arthritis, haemolytic anaemia, mucosal ulcers, nephritis, rash, serositis and thrombocytopenia.

The first Delphi stage achieved consensus in both SLE clinician and patient groups on the importance of 14 domains for measuring treatment effect in an SLE RCT. We also identified a number of domains where discrepancies existed between clinician and patient perspectives: 11 domains rated highly ‘important’ by patients did not meet consensus among clinicians. After reviewing these discordant results, the TRM-SLE Taskforce, with input from participating patient partners, decided to revise the predefined consensus definition and allow these domains with discrepant ratings to progress through to the second Delphi stage. Notably, two of these domains (alopecia and mucosal ulcers) ultimately met consensus for inclusion in the novel COA, highlighting the vital contribution of the patient perspective, particularly for domains that have an impact on quality of life, but may be less prioritised by clinicians relative to domains that are organ-threatening or life-threatening. Divergence of priorities between physicians and patients is well documented in SLE.²⁶ Other domains rated highly by patients but not clinicians were generally regarded to have important patient impact, but not fall within the specific measurement goals of the TRM-SLE instrument (eg, domains representing damage or comorbidities, rather than active inflammation in SLE).

It was notable that some domains included in disease activity instruments used in current clinical trial endpoints (eg, leucopenia, complement, anti-double stranded DNA antibodies) were not rated highly on ‘importance’ in our study. We defined ‘importance’ as being associated with how a patient with SLE ‘feels, functions or survives’ based on contemporary guidance emphasising that COAs used to support regulatory approval need to demonstrate treatment benefit based on this paradigm.^{16 17} Therefore, biomarkers that lack empirical evidence of being associated with meaningful aspects of patient health should be omitted from the assessment of treatment response, although measurement may still be relevant for other purposes. In general, participant ratings of ‘importance’ in our Delphi study aligned with available evidence from our literature review. The well-documented association of serological activity with flare risk will likely require ongoing measurement of serological activity in trials, even if it is not part of the primary efficacy endpoint.

In the second Delphi stage, domains meeting consensus on ‘importance’ in either clinician and/or patient groups were rated on ‘appropriateness’ as an active immune-mediated disease manifestation, ‘representation’ in patients with active SLE and ‘measurability’ of change in disease activity in an RCT setting. With the approval of the patient advisory panel, only the clinician expert group voted in the second Delphi stage, due to the technical nature of the study questions posed. However, patients were invited to be active participants in the discussion meeting and encouraged to voice their opinions on each item discussed. Several domains did not meet the consensus threshold for inclusion based predominantly on the ‘appropriateness’ filter (figure 5); this included domains felt to represent damage or comorbidity rather than activity (eg, osteoporosis), or manifestations where the likelihood of response to an immune-directed therapy was uncertain (eg, fatigue). Patient-reported symptoms such as fatigue and brain fog were a topic of several discussions; while the importance of such symptoms to patients was agreed on, several participants raised concerns regarding the ability to accurately attribute these symptoms to underlying immunopathology amenable to treatment and the ability to measure change over the course of a clinical trial. Notably, this was a view also raised by patient representatives, who expressed their concerns about the potential discordance of fatigue with clinically meaningful improvements in disease activity, such as the achievement of remission, and its associated long-term prognostic benefit. Future studies will need to determine how best to allow for manifestations considered important by patients and clinicians alike, such as fatigue, which did not reach consensus for inclusion in this set of domains due to other aspects. For example, fatigue and other important patient-reported outcomes could be used as anchors in validation studies of the TRM-SLE endpoint.

Another group of domains was deemed both ‘important’ and ‘appropriate’ for TRM-SLE, but was excluded based on the filters of ‘representation’ and/or ‘measurability’ (figure 5). Several domains were excluded based on their infrequent occurrence in cohort studies and RCT populations (eg, ophthalmic, gastrointestinal, pulmonary, myositis), a conclusion which was supported by findings of our literature search. Some of these domains (eg, ophthalmic, gastrointestinal) were also deemed to have poor measurability characteristics for a SLE RCT. Other domains (eg, neuropsychiatric lupus, vasculitis) stimulated discussion on points including the exclusion of organ or life-threatening manifestations from many SLE RCTs and concerns about whether the need for advanced imaging or other specialised investigations limited the feasibility of their inclusion in a multidomain SLE trial. These concerns were balanced against the recognition of these domains as being highly consequential to patients and representing important areas of unmet therapeutic need.

The approach to domain inclusion in TRM-SLE is based on facilitating detailed measurement of domains that are important to patients and also high value for determining treatment response in the clinical trial setting, from among the vast number of heterogeneous SLE manifestations. This approach represents a change in perspective from how activity and response are traditionally measured by disease activity instruments used in SLE RCT endpoints, which has focused on capturing breadth of disease activity, rather than quantifying changes in the activity of individual manifestations in more detailed fashion. The latter, however, is necessary to achieve a nuanced interpretation of treatment efficacy for specific disease features as part of a responder definition. In addition, there is considerable evidence that the range of frequently observed manifestations of SLE in cohort studies and clinical trial populations is much narrower than the sum of possible manifestations of SLE.²⁰ These observations have led to recent SLE RCTs incorporating more detailed organ-specific COAs, assessing common manifestations, as secondary endpoints, such as the Cutaneous Lupus Erythematosus Disease Area and Severity Index for mucocutaneous manifestations,²⁷ or joint counts in patients with active arthritis. The endpoint being developed by the TRM-SLE Taskforce will define response based on detailed assessment of a core set of key lupus domains active at baseline, focusing on the manifestations of active disease most relevant to the clinical trial context and associated patient population. This includes consideration of the relevance of domains for patients with childhood-onset disease, as it is intended that the TRM-SLE measure can be applied to trials involving adolescents, as well as adults, with SLE. It is likely that a consequence of measuring response using TRM-SLE will also enable revision of trial eligibility criteria away from legacy disease activity measures, potentially allowing a wider range of patients to be eligible for clinical trials. Setting clinically meaningful disease activity thresholds will be an important aspect of the next stage of this project.

Although our novel COA focuses on a limited set of domains, this does not preclude other domains or criteria from being incorporated within the overall definition of response reflected in the trial endpoint. For example, although excluded domains will not contribute to the improvement measure in TRM-SLE, they may still be measured to ensure no overall worsening of activity, or flare in other organ systems. The specific mechanisms for capturing activity in domains that did not meet consensus for inclusion in TRM-SLE are a designated focus of a subsequent instrument development stage.¹¹ The exclusion of some domains in our study also emphasises the need for research into improving the measurability of domains that are highly important, but poorly defined or quantifiable, such as neuropsychiatric lupus. We hope that future iterations of the instrument being developed by the TRM-SLE Taskforce may be able to be inclusive of domains that are ‘important’ and ‘appropriate’, as more feasible measurement methods evolve.

Limitations of this study include the small number of patient participants relative to clinician numbers, which restricted the potential breadth and diversity of patient representation. Nonetheless, patients with experience of a range of disease manifestations were included, and both adult and childhood-onset SLE were represented. While consideration was given to including a wider group of patients, in consultation with the TRM-SLE Patient Advisory Panel, it was felt important to ensure patient partners had a technical understanding of the goals and context for TRM-SLE, and ideally possessed experience as patient with SLE advocates to allow them to represent the views and experiences of the SLE patient population more broadly. This approach was supported by the strong patient partner contributions made during the domain selection process, which on multiple occasions influenced the study outcomes. In addition, the number of clinicians who participated in the three stages of consensus varied, with a maximum of 87 and a minimum of 46. It is unclear whether having larger numbers of clinicians would have altered the outcomes.

In conclusion, through formal Delphi methodology, consensus was achieved to include eight domains of active SLE (alopecia, arthritis, haemolytic anaemia, nephritis, mucosal ulcers, rash, serositis and thrombocytopenia) for defining meaningful treatment response in a novel COA for SLE clinical trials. Our domain selection process considered both clinician and patient perspectives on domain importance, and key factors relevant to the suitability and utility of domains in the intended context of use. The next step in instrument development will determine the detailed measurement of each included domain, including scoring methods and thresholds defining clinically meaningful response. This will be followed by the determination of the overall responder definition and endpoint, and prospective instrument validation including in RCTs. We hope our novel measure will find application in future trials of novel therapies for SLE, with the goal of improving the efficiency of RCTs in SLE and bringing improved therapies, and outcomes, to patients.

We would like to acknowledge all members of the TRM-SLE Taskforce and those who generously contributed their time and expertise to our Delphi study.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Data relevant to this study are available upon reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

The study was approved by the Monash University Human Research Ethics Committee. Institute name: Monash University. Reference number 25666.

Footnote

X @edvital, @gponsestel, @hermineibrunne1, @n/a

KC and RK contributed equally.

Collaborators TRM-SLE Taskforce Members: Alain Cornet, Alan Friedman, Alessandro Sorrentino, Anca Askanase, Ann Eldred, Anna Stevens, Bianca Fedriz, Blanca Rubio Hernández, Cailin Sibley, Catharina Lindholm, Catherine (Cachou) Barbey, Charlotte Lupton, Chris Brooks, Christian Stach, Claudia Pena Rossi, Cristina Vazquez-Mateo, Cynthia Aranow, Dalila Tremarias, Dalilah Kalla, Darshini Ayton, Dennis Grasela, Dzifa Dey, Ed Vital, Elaine Karis, Eloisa Bonfa, Eric Morand, Eric Zollars, Erik Thomas, Erika Noss, Eve Smith, Flavie Moreau, George Stojan, Gonzalo Tobar Carrizo, Guillermo Pons-Estel, Heath Guay, Hermine Brunner, Hussein Al-Mossawi, Imasha Adisa, Inmaculada De La Torre, Jeanette Andersen, Jeffrey Lockman, Joan Merrill, Jorge Ross Terres, Joseph F. Merola, Joy Buie, Justine Maller, Karen Costenbader, Kate Gregory, Kathryn Connelly, Kenneth Kalunian, Khadija Dantata, Laura Eades, Laurent Arnaud, Laurie Burke, Lee Simon, Maja Hojnik, Maria Dall’Era, Maria Juarez, Maria Silk, Marta Mosca, Narae Yun, Nicki Bush, Nikolay Delev, Nitin Kumar, Oliver Guenther, Patrick Marquis, Qing Zuraw, Rachel Koelmeyer, Rangi Kandane-Rathnayake, Richard Furie, Robert Townsend, Ronald van Vollenhoven, Samantha Pomponi, Sandra Garces, Sanjeev Roy, Shelly Kafka, Shiori Nagamori, Sibongile Komati, Stephanie Scoggins, Subhashis Banerjee, Susanne Udengaard Gydesen, Terence Rooney, Thierry Sornasse, Thomas Lehman, Thomas Morel, Tim Coulom, Toni Grimes, Vera Golder, Veronica Vargasgill, Victoria Werth, Vinita Haroun, Ying Sun, Yoshiya Tanaka, Youmna Lahoud, Yulia Pincus, Zoe Karakikla-MitsakouDelphi Participants: Aaron Alejandro Barrera Rodriguez, Alain Cornet, Alan Friedman, Alexandre Voskuyl, Alfred Kim, Ali Duarte, Anca Askanase, Andrea Doria, Anisur Rahman, Benjamin Chong, Bernardo A. Pons-Estel, Blanca Rubio Hernandez, Bonnie Bermas, Caroline Gordon, Chanakya Sharma, Cherica Tee, Chi Chiu Mok, Chris Wincup, Christiaan Scott, Christophe Richez, Cindy Flower, Claire Barrett, CS Lau, Cynthia Aranow, Dalila Tremarias, David D'Cruz, David Jayne, Deborah M Levy, Dzifa Dey, Dafna Gladman, Dwarakanathan Ranganathan, Ed Vital, Eloisa Bonfa, Eric Morand, Eve Smith, Fiona Goldblatt, Fiona Mackie, Gonzalo Tobar Carrizo, Graciela S. Alarcon, Guillermo Pons-Estel, Guillermo Ruiz-Irastorza, Hermine Isabella Brunner, Hussein Al-Mossawi, Jeanette Andersen, Jiacai Cho, Jinoos Yazdany, Johanna Mucke, Jonathan Akikusa, Jorge A Ross Terres, Joy Buie, Justine Maller, Karen Costenbader, Kate Gregory, Kathryn Connelly, Kavita Makan, Kenji Oku, Kenneth Kalunian, Khadija Dantata, Laura Eades, Laurent Arnaud, Leila Khalili, Mahmood MTM Ally, Mandana Nikpour, Manuel Ugarte-Gil, Maria Dall’Era, Marta Mosca, Martin Aringer, Matthias Schneider, Maureen Rischmueller, Michael Beresford, Michael Tee, Michel Tsang-A-Sjoe, Michelle Kahlenberg, Michelle Petri, Murray Urowitz, Nathalie Costedoat-Chalumeau, Odirlei Andre Monticielo, Pravin Hissaria, Ronald Van Vollenhoven, Ricard Cervera, Richard Furie, Roberto Caricchio, Sandra Garces, Sandra V. Navarra, Sang-Cheol Bae, Sanjeev Roy, Savino Sciascia, Sean O’Neill, Shereen Oon, Shiori Nagamori, Stephanie Scoggins, Susanne Udengaard Gydesen, Ted Tsai, Toni Grimes, Vera Golder, Victoria Werth, Vinita Haroun, Worawit Louthrenoo, Yasuhiro Katsumata, Yoshiya Tanaka, Youmna Lahoud, Zoe Karakikla-Mitsakou.

Contributors All authors contributed to generation of data, writing or editing of the manuscript, and approval of the manuscript. All authors contributed to the design of the study and interpretation of the data. KC and RK analysed the study data. KC and EM drafted the manuscript and all authors reviewed it critically for important intellectual content. All authors gave approval for the final version of the manuscript and agree to be accountable for all aspects of the work. EM is guarantor.

Funding The work reported in this manuscript was directly funded via a multi-institution agreement between Monash University and AbbVie, Amgen, AstraZeneca, Biogen, Bristol Myers Squibb, Eli Lilly, EMD Serono, Genentech, Janssen and UCB. Since the completion of this work, the TRM-SLE Taskforce has also received funding from GSK, Novartis and Takeda. EM reports funding from the National Health and Medical Research Council of Australia.

Competing interests RK reports grants/research support from GSK and Novartis. AF reports shares in AbbVie and consulting fees from Eli Lilly, Merck, Sun Pharmaceuticals and Tenet, and previous employment by AbbVie. AA reports consulting fees from AbbVie, Amgen, AstraZeneca, Aurinia, BMS, Celgene, Eli Lilly, Idorsia, Janssen, Genentech, GSK, Mallinckrodt, Pfizer and UCB. CA reports consulting fees from Synthekine, Bristol Myers Squibb, Kezar, AstraZeneca, GSK and received grants/research support from GSK. EV reports speaker fees from AstraZeneca, UCB, Novartis; consulting fees from Roche, Pfizer, UCB, AstraZeneca, Novartis, AbbVie, Merck, Lilly, Otsuka and grants/research support from AstraZeneca and Sandoz. GP-E reports grants, consulting fees and participation as a speaker and/or advisor and/or steering committee for: Alumis, AstraZeneca, Boehringer Ingelheim, GSK, Janssen, Novartis, Pfizer, RemeGen, Sanofi and Werfen Diagnostics. HB reports consulting fees from AstraZeneca, BI, Novartis, Pfizer, UCB, BMS, Takeda, Eli Lilly, J&J, Biogen and received grants/research support from Pfizer. KK reports consulting fees from BMS, GSK, Artiva, Cabaletta Bio, Roche/Genentech, AstraZeneca, Artiva, Eli Lilly, Kezar, Aurinia and AbbVie. LA reports consulting fees from Alexion, Amgen, AstraZeneca, AbbVie, Alpine Pharmaceuticals, Biogen, BMS, Boehringer Ingelheim, Chugai, GSK, Grifols, Janssen, Kezar, LFB, Lilly, Menarini France, Medac, Novartis, Pfizer, Roche, UCB. LB reports consulting fees from multiple companies across many different therapeutic areas; with no active consulting in SLE. SG reports shares in Amgen and employment by Amgen. QZ reports shares in Remegen Biosciences and employment by Remegen Biosciences, and previous employment by Janssen. VPW reports consulting fees from Pfizer, Janssen, Neovacs, Idera, Octapharma, CSL Behring, Corbus, Rome, AstraZeneca, Biogen, Celgene, Gilead, Lilly, BMS, Nektar, AbbVie, Akira, Viela, GSK, EMD Serona, Sanofi, Xencor, ONO, Cabaletta and received grants/research support from Pfizer, Corbus, Amgen, Janssen, Biogen, Gilead, Viela, Ventus, Regeneron, Argenx. YBS reports employment at Merck KGaA. YT reports speaker fees from Eli Lilly, AstraZeneca, AbbVie, Gilead, Chugai, Boehringer Ingelheim, GlaxoSmithKline, Eisai, Taisho, Bristol Myers, Pfizer, Taiho and grants/research support from Mitsubishi Tanabe, Eisai, Chugai, Taisho. YL reports shares in Biogen and is an emplyoee of Biogen. ASo reports shares in Galapagos and is an employee of Sanofi and previous employee of AstraZeneca. AR reports speaker fees from Elli Lilly. ASt reports consulting fees from AbbVie, Amgen, AstraZeneca, Aurinia, BMS, Celgene, Eli Lilly, Idorsia, Janssen, Genentech, GSK, Mallinckrodt, Pfizer and UCB. CB reports shares in Biogen and is an emplyoee of Biogen. IDD reports speaker fees from Pfizer, Roche. EK reports shares in Amgen and is an emplyoee of Amgen. EN reports shares in Johnson and Johnson and is an employee of Janssen Research and Development, a Johnson and Johnson company. GS reports shares in UCB. JFM reports consulting fees from Amgen, AstraZeneca, Boehringer Ingelheim, Bristol Myers Squibb, AbbVie, Dermavant, Eli Lilly, Incyte, Moonlake, Novartis, Janssen, UCB, Sanofi-Regeneron, Sun Pharma, Biogen, Pfizer and Leo Pharma. JART reports shares in Genentech/Roche and is an emplyoee of Genentech/Roche. JB reports consulting fees from Novartis and grants/research support from Aurinia. JMal reports shares in Genentech/Roche and is an emplyoee of Genentech/Roche. MMo reports speaker fees for GSK, AstraZeneca, Janssen; consulting fees from AstraZeneca, AbbVie, UCB, GSK, Otsuka and grants/research support from GSK. MH reports shares in Eli Lilly. MD’E reports consulting fees from Genentech, Aurinia, GSK, AstraZeneca, Janssen. RAF reports speaker fees from AZ, GSK; consulting fees from BMS, GSK, Genentech, AstraZeneca and Biogen, and received grants/research support from GMS, GSK, Genentech, Biogen, Kyverna. RFvV reports speaker fees from AbbVie, AstraZeneca, BMS, Galapagos, GSK, Janssen, Pfizer, UCB; consulting fees from AbbVie, AstraZeneca, Biogen, BMS, Galapagos, GSK, Janssen, Pfizer, RemeGen, UCB and; grants/research support from AstraZeneca, BMS, Galapagos, MSD, Novartis, Pfizer, Roche, Sanofi, UCB. SB reports shares in BMS and is an emplyoee of BMS. EM reports speaker fees from AstraZeneca, Merck, Gilead, Roche; consulting fees from EMD Serono, AstraZeneca, Biogen, Bristol Myers Squibb, Eli Lilly, Genentech, GlaxoSmithKline, Janssen, Novartis, AbbVie, Galapagos, IgM and grants/research support from AbbVie, Amgen, AstraZeneca, Biogen, BMS, Eli Lilly, EMD Serono, Genentech, GSK, Janssen, UCB (all institutional). The remaining authors have no disclosures.

Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ Cornet A, Andersen J, Myllys K, et al. Living with systemic lupus erythematosus in 2020: a European patient survey. Lupus Sci Med 2021; 8: e000469. doi:10.1136/lupus-2020-000469

² Meacock R, Dale N, Harrison MJ. The humanistic and economic burden of systemic lupus erythematosus : a systematic review. Pharmacoeconomics 2013; 31: 49–61. doi:10.1007/s40273-012-0007-4

³ Barber MRW, Drenkard C, Falasinnu T, et al. Global epidemiology of systemic lupus erythematosus. Nat Rev Rheumatol 2021; 17: 515–32. doi:10.1038/s41584-021-00668-1

⁴ Furie R, Petri M, Zamani O, et al. A phase III, randomized, placebo-controlled study of belimumab, a monoclonal antibody that inhibits B lymphocyte stimulator, in patients with systemic lupus erythematosus. Arthritis Rheum 2011; 63: 3918–30. doi:10.1002/art.30613

⁵ Navarra SV, Guzmán RM, Gallacher AE, et al. Efficacy and safety of belimumab in patients with active systemic lupus erythematosus: a randomised, placebo-controlled, phase 3 trial. Lancet 2011; 377: 721–31. doi:10.1016/S0140-6736(10)61354-2

⁶ Furie RA, Morand EF, Bruce IN, et al. Type I interferon inhibitor anifrolumab in active systemic lupus erythematosus (TULIP-1): a randomised, controlled, phase 3 trial. Lancet Rheumatol 2019; 1: e208–19. doi:10.1016/S2665-9913(19)30076-1

⁷ Morand EF, Furie R, Tanaka Y, et al. Trial of Anifrolumab in Active Systemic Lupus Erythematosus. N Engl J Med 2020; 382: 211–21. doi:10.1056/NEJMoa1912196

⁸ Rovin BH, Teng YKO, Ginzler EM, et al. Efficacy and safety of voclosporin versus placebo for lupus nephritis (AURORA 1): a double-blind, randomised, multicentre, placebo-controlled, phase 3 trial. Lancet 2021; 397: 2070–80. doi:10.1016/S0140-6736(21)00578-X

⁹ Morand EF, Fernandez-Ruiz R, Blazer A, et al. Advances in the management of systemic lupus erythematosus. BMJ 2023; 383: e073980. doi:10.1136/bmj-2022-073980

¹⁰ Connelly K, Golder V, Kandane-Rathnayake R, et al. Clinician-reported outcome measures in lupus trials: a problem worth solving. Lancet Rheumatol 2021; 3: e595–603. doi:10.1016/S2665-9913(21)00119-3

¹¹ Connelly K, Eades LE, Koelmeyer R, et al. Towards a novel clinical outcome assessment for systemic lupus erythematosus: first outcomes of an international taskforce. Nat Rev Rheumatol 2023; 19: 592–602. doi:10.1038/s41584-023-00993-7

¹² Dolgin E. Lupus in crisis: as failures pile up, clinicians call for new tools. Nat Biotechnol 2019; 37: 7–8. doi:10.1038/nbt0119-7

¹³ Morand EF, Vital EM, Petri M, et al. Baricitinib for systemic lupus erythematosus: a double-blind, randomised, placebo-controlled, phase 3 trial (SLE-BRAVE-I). Lancet 2023; 401: 1001–10. doi:10.1016/S0140-6736(22)02607-1

¹⁴ Petri M, Bruce IN, Dörner T, et al. Baricitinib for systemic lupus erythematosus: a double-blind, randomised, placebo-controlled, phase 3 trial (SLE-BRAVE-II). Lancet 2023; 401: 1011–9. doi:10.1016/S0140-6736(22)02546-6

¹⁵ Tse K, Sangodkar S, Bloch L, et al. The ALPHA Project: Establishing consensus and prioritisation of global community recommendations to address major challenges in lupus diagnosis, care, treatment and research. Lupus Sci Med 2021; 8: e000433. doi:10.1136/lupus-2020-000433

¹⁶ US Department of Health and Human Services Food and Drug Administration. Patient-focused drug development guidance series for enhancing the incorporation of the patient’s voice in medical product development and regulatory decision making. 2022. Available: https://www.fda.gov/media/159500/download

¹⁷ Powers JH, Patrick DL, Walton MK, et al. Clinician-Reported Outcome Assessments of Treatment Benefit: Report of the ISPOR Clinical Outcome Assessment Emerging Good Practices Task Force. Value Health 2017; 20: 2–14. doi:10.1016/j.jval.2016.11.005

¹⁸ Furie RA, Petri MA, Wallace DJ, et al. Novel evidence-based systemic lupus erythematosus responder index. Arthritis Rheum 2009; 61: 1143–51. doi:10.1002/art.24698

¹⁹ Wallace DJ, Kalunian K, Petri MA, et al. Efficacy and safety of epratuzumab in patients with moderate/severe active systemic lupus erythematosus: results from EMBLEM, a phase IIb, randomised, double-blind, placebo-controlled, multicentre study. Ann Rheum Dis 2014; 73: 183–90. doi:10.1136/annrheumdis-2012-202760

²⁰ Connelly K, Vettivel J, Golder V, et al. Measurement of specific organ domains in lupus randomized controlled trials: a scoping review. Rheumatology (Oxford) 2022; 61: 1341–53. doi:10.1093/rheumatology/keab777

²¹ Carter LM, Gordon C, Yee C-S, et al. Easy-BILAG: a new tool for simplified recording of SLE disease activity using BILAG-2004 index. Rheumatology (Oxford) 2022; 61: 4006–15. doi:10.1093/rheumatology/keab883

²² Gladman DD, Ibañez D, Urowitz MB. Systemic lupus erythematosus disease activity index 2000. J Rheumatol 2002; 29: 288–91.

²³ Hsu C, Sanford B. The Delphi technique: making sense of consensus. Practical Assessment, Research and Evaluation 2007; 12.

²⁴ Humphrey-Murto S, Crew R, Shea B, et al. Consensus Building in OMERACT: Recommendations for Use of the Delphi for Core Outcome Set Development. J Rheumatol 2019; 46: 1041–6. doi:10.3899/jrheum.181094

²⁵ Fitch K, Berry SJ, Aguilar MD, et al. The RAND/UCLA appropriateness method user’s manual. Arlington, USA: RAND, 2001.

²⁶ Golder V, Ooi JJY, Antony AS, et al. Discordance of patient and physician health status concerns in systemic lupus erythematosus. Lupus (Los Angel) 2018; 27: 501–6. doi:10.1177/0961203317722412

²⁷ Morand E, Pike M, Merrill JT, et al. Deucravacitinib, a Tyrosine Kinase 2 Inhibitor, in Systemic Lupus Erythematosus: A Phase II, Randomized, Double-Blind, Placebo-Controlled Trial. Arthritis Rheumatol 2023; 75: 242–52. doi:10.1002/art.42391