Implementation of the Lupus Low Disease Activity State in Pediatric Rheumatology Care: The Role of the Visual Analog Scale

Abstract

Objective

We compared the measurement properties of a traditional physician global assessment of disease activity (PhGA) 10‐cm visual analog scale (PhGA_0–10) with that of the three‐point numeric scale (PhGA_0–3) in childhood‐onset systemic lupus erythematosus (cSLE) as part of the childhood Lupus Low Disease Activity State (cLLDAS).

Methods

We used a secondary data analysis from a convenience sample of 100 patients with cSLE followed every three months for up to seven visits. Ratings of PhGA_0–10, PhGA_0–3, parent assessment of patient well‐being (ParGA) (range: 0= very poorly, 10 = very well), disease activity as measured by the SLE disease activity index 2000 (SLEDAI‐2k), Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA) SLEDAI, and the British Isles International Lupus Activity Group index (BILAG; A = 9, B = 3, C = 1, D/E = 0) were compared. After linear transformation of PhGA_0–10 to a 0 to 3 range (tPhGA_0–10), the frequency of PhGA_0–3 ≤1 was assessed to estimate the impact of scale type on the scoring of the cLLDAS.

Results

In 600 visits, the median (range) scores of PhGA_0–10, PhGA_0–3, SLEDAI‐2k, SELENA‐SLEDAI, and BILAG were 2 (0–10), 1(0–3), 4 (0–28), 4 (0–32), and 2 (0–28), respectively. PhGA_0–10 and PhGA_0–3 ratings were strong to moderately correlated with (r = 0.73; P < 0.0001) and with more variability for PhGA_0–3 ≥2. SELENA‐SLEDAI and SLEDAI‐2k scores were moderately correlated with PhGA_0–10 (r = 0.56/0.54; P < 0.0001). ParGA values were weakly correlated with all other measures considered (all r = −0.19 to −0.34). There were 490 of 600 visits with PhGA₀₋₃ ≤1 and 497 of 600 visits with tPhGA₀₋₁₀ ≤1 (κ (SE) =0.59 (0.04), McNemar P = 0.4).

Conclusion

PhGA_0–3 and PhGA_0–10 have comparable measurement properties and yield almost identical cLLDAS rates when used in cSLE.

INTRODUCTION

Systemic lupus erythematosus (SLE) is a chronic multisystem inflammatory disease that continues to result in considerable morbidity and mortality. Patients with childhood-onset SLE (cSLE) have been found to have more multiorgan involvement, including lupus nephritis, and persistently active disease when compared to adult-onset SLE.¹

---

SIGNIFICANCE & INNOVATIONS

• The physician global assessment of disease activity (PhGA) 10-cm visual analog scale (PhGA_0–10) can be reliably used for childhood Lupus Low Disease Activity State (cLLDAS) assessment in childhood-onset systemic lupus erythematosus (cSLE).
• Use of the PhGA_0–10 and PhGA with a three-point numeric scale (PhGA_0–3) scales in cLLDAS assessment in cSLE showed comparable measurement properties.
• The PhGA_0–10 scale is routinely assessed in pediatric rheumatology practices and as a measure in the Provisional Pediatric Rheumatology International Trials Organization/American College of Rheumatology (ACR) criteria for treatment response, the Preliminary ACR Criteria for Flare and the Provisional Criteria of Minimally important improvement of cSLE, hence, potentially enabling easier widespread adoption of the PhGA_0–10 in cLLDAS assessment in cSLE.

---

In recent years, treat-to-target strategies for managing SLE and cSLE have been formulated to improve disease outcomes and avoid damage development.² To operationalize the above concept, the Lupus Low Disease Activity State (LLDAS) was developed and proposed as a treat-to-target goal of adults and children with SLE.^2,3 LLDAS determination requires patients to tolerate immunosuppressive medication, use no more than low dose glucocorticosteroids (GCs), and to maintain a low level of SLE activity.³ Indeed, achievement and maintenance of LLDAS in adults is associated with reduced damage accrual and improved survival in both adults and children.^4,5 Recently, the childhood LLDAS (cLLDAS) was derived by consensus to facilitate life course studies. This is a modification of the adult derived LLDAS and includes similar parameters such as a physician global assessment (PhGA) scale, but with steroid dosing based on body weight.⁶

Visual analog scales (VASs) of disease activity are widely used in rheumatology and allow for continuous scaling of disease activity, that is, allow physicians the consideration of disease activity features of a patient that are not considered in traditional disease activity measures. Furthermore, VASs provide an opportunity for studies to determine clinically significant changes, rather than relying on predetermined glossary-based definitions as indicators of disease severity. Unfortunately, past studies of VASs in SLE have yielded inconsistent results, likely because of the potential variations in how clinicians interpret these scales.⁷

Specifically in pediatric rheumatology, the 10-cm VAS to score PhGA of disease activity (PhGA_0–10) is widely used for assessing disease activity in juvenile idiopathic arthritis, juvenile dermatomyositis, and cSLE.^8–10 Conversely, the cLLDAS repurposes a three-point numeric Likert scale (PhGA_0–3) that is part of the Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA) version of the SLE disease activity index (SLEDAI) and its flare tool.^6,11 Therefore, comparison between the measurement properties of the PhGA_0–10 and PhGA_0–3 scales is needed because the completion of the PhGA_0–3 is currently not routine in pediatric rheumatology practice and to enhance the integration of the cLLDAS as a treat-to-target goal in cSLE. Thus, we aimed to evaluate the impact of using the PhGA_0–10 as opposed to the PhGA_0–3 as part of the cLLDAS. We hypothesized that the associations between various measures of disease activity used in cSLE with the PhGA_0–3 and the PhGA_0–10 are comparable.

METHODS

Patients and data

An exploratory analysis of an available dataset from aNational Institutes of Health–supported study in cSLE was performed. The original dataset was assembled to assess the intrarater reliability, construct validity, and responsiveness of the disease indices in cSLE.⁸ There were a total of 600 visits from a total of 100 patients with cSLE fulfilling the 1997 revised American College of Rheumatology classification criteria for SLE before the age 18 years.¹² For these patients, serial assessments were done about every three months to document the course of cSLE. Collected data included the measurement of disease damage, medication usage, including the dose of prednisone equivalent, and patient characteristics. The study used fully registry data from an institutional review board–approved protocol. All patients and/or caretakers signed assent and consent forms approved by the Cincinnati Children's Institutional Review Board (IRB 2008-0635). The datasets generated and/or analyzed during the current study are not publicly available due to individual data privacy but may be available from the corresponding author on reasonable request.

Outcomes

Patient assessments included measurement of disease activity, patient well-being, and the physician assessment of global disease activity. The following disease activity indices were evaluated: British Isles International Lupus Activity Group index (BILAG; numeric conversion of the domain ratings [A–E]: A = 9, B = 3, C = 1, D/E = 0),¹³ the SLEDAI (version 2000; range 0–105),¹⁴ and the SELENA-SLEDAI.¹¹ The latter features a three-point numeric Likert scale of disease activity (PhGA_0–3), which provides distinct markers when assessing disease activity. These are inactive for a value of 0; mild for a value of 1, moderate or severe for values of 2 and 3, respectively. The rater can assign any value on the PhGA_0–3 between 0 and 3 in response to the question “How do you assess your patient's current disease activity?” The treating physician also rated patient global disease activity on a 10-cm VAS with the anchors “inactive” for a value of 0 and “very active” for a value of 10, in response to the question: “Considering the findings at today's visit, the overall disease activity of the patient is.”

The patient global assessment of well-being (Pat-well) was also rated on a traditional 10-cm VAS with the anchors “very poor” for a value of 0 and “very well” corresponding to a value of 10, in response to the question: “When considering your medications, school, work, life at home, doctor visits, pains, and feelings, how would you rate your overall well-being?”

From the available data, cLLDAS was present if all of the following were met: SLEDAI-2k score of ≤4 without activity in major organ systems, no new features of disease activity compared to the previous assessment, physician global assessment (PhGA_0–3) of ≤1 (scale 0–3), daily prednisone equivalent dose of ≤7.5 mg (or a maximum of 0.15 mg/kg body weight per day for children under 50 kg), and well-tolerated standard maintenance doses of immunosuppressives.⁶

Statistical analysis

Descriptive analysis was done for the baseline status of the patients at the time of the enrollment into the study. Numeric values were summarized as arithmetic mean (SD) and median (range) and ordinal and nominal variables as frequencies, respectively. We also determined mean (SD) and median (range) of Pat-well and the disease activity measures (SLEDAI-2k, SELENA-SLEDAI, BILAG, PhGA_0–10, PhGA_0–3) for all study visits.

To study the concurrent validity of the disease activity measures (SLEDAI-2k, SELENA-SLEDAI, BILAG, PhGA_0–10, PhGA_0–3) and the Pat-well values, Spearman correlation coefficients (r) were calculated, which can be interpreted as follows: values of r < 0.2 were considered not correlated. Weak correlation is defined by r ≥ 0.2–0.4; moderate correlation by r ≥ 0.4–0.6; strong to moderate correlation by r ≥ 0.6–0.8; and strong correlation by r ≥ 0.8. Smooth plots were generated using locally estimated scatterplot smoothing to further evaluate the relationship between PhGA_0–3 and PhGA_0–10. To test the effect of the use of the PhGA_0–10 for determining cLLDAS, we first performed a linear transformation of the physician-provided ratings on the PhGA_0–10 to yield values between 0 and 3 (tPhGA_0–10). McNemar test and kappa (SE) statistics were then used to compare cLLDAS frequencies during the study period when using the tPhGA_0–10 or the traditional PhGA_0–3 scale in the cLLDAS algorithm. Agreement, based on kappa values can be interpreted as κ ≤ 0, no agreement; κ = 0.01–0.20, no more than slight agreement; κ = 0.21–0.40, fair agreement; κ = 0.41– 0.60, moderate agreement; κ = 0.61–0.80, substantial agreement; and κ = 0.81–1.00, almost perfect agreement. McNamar agreement test P > 0.05 indicate that there is no statistically significant disagreement between measures.

RESULTS

Patients and visits

There were 600 visits for 100 patients available for this analysis. Complete PhGA_0–3 and PhGA_0–10 data were missing for 2 of the 100 enrolled patients from sites in the United States and Canada (Table 1).

Table 1 Demographics and SLE features at baseline*

Parameter	N	Mean ± SD
Age, y	98	15.3 ± 2.85
Disease duration, y	98	1.5 ± 2.0
Current medications
Prednisone, mg/day	82	15.1 ± 1.8
Azathioprine, mycophenolate mofetil, methotrexate	47
Cyclophosphamide	6
Hydroxychloroquine	73
Nonsteroidal anti-inflammatory drugs	24
At least one antihypertensive medication	38
Biopsy-proven lupus nephritisa	39
Disease damage, SDI score (0 = no damage)	98	0.42 ± 0.1
Physician assessment of overall disease activity (PhGA0–10)	98	2.5 ± 1.93
Physician assessment of overall disease activity (PhGA0–3)	99	0.85 ± 0.59
Disease activity score	98
SLEDAI-2k		4.93 ± 4.25
SELENA-SLEDAI		4.82 ± 4.04
BILAG (0–9)		5.31 ± 5.44

Rating and association of outcome measures during the study

Across all 600 visits mean (SD) and median (range) of the SLEDAI-2k were 4.63 (4.14) and 4 (0–28) and of the SELENA-SLEDAI were 4.51 (4.1) and 4 (0–32). Mean (SD) and median (range) of the BILAG was 2.82 (3.34) and 2 (0–28) and that of Pat-well was 7.74 (1.77) and 8 (1–11). Table 2 shows the correlations of the PhGA_0–3 and the PhGA_0–10 to Pat-well and disease activity index scores to be almost identical.

Table 2 Associations of well-being with disease activity and PhGA measures*

	PhGA0–10	PhGA0–3	SLEDAI-2k	SELENA-SLEDAI	BILAG
Par-well	−0.34 (<0.0001)	−0.30 (<0.0001)	−0.19 (<0.0001)	−0.20 (<0.0001)	−0.28 (<0.0001)
SLEDAI-2k	0.54 (<0.0001)	0.53 (<0.0001)	1.00 (NA)	0.97 (<0.0001)	0.58 (<0.0001)
SELENA-SLEDAI	0.52 (<0.0001)	0.52 (<0.0001)	0.97 (<0.0001)	1.00	0.57

Comparison of ratings of PhGA of disease activity

Across all study visits, the mean (SD) and median (range) ratings of the PhGA_0–10 were 2.13 (1.87) and 2 (0–10) and of the PhGA_0–3 were 0.79 (0.64) and 1 (0–3). Figure 1A depicts the relationship between the rating on these two scales. Notably, 95% confidence intervals increase for PhGA_0–3 >2.

[IMAGE OMITTED. SEE PDF]

Impact of the use of the PhGA₀₋₃ versus the PhGA₀₋₁₀ on the scoring of the cLLDAS

Scoring of the cLLDAS using complete data was available for 600 visits. Figure 1B depicts the comparison of the two PhGA scales in measurement of the cLLDAS in cSLE. When using the PhGA_0–3, there were 456 visits with values of ≤1 compared to 489 visits when considering the tPhGA_0–1 (κ [SE] = 0.59 [0.04]; McNemar P = 0.4). Misclassification rates were ≤6.7%. There was a larger variability between the PhGA_0–10 and PhGA_0–3 ≥2 (unpooled t-test, t = −11.93; Satterthwaite P ≤ 0.0001).

DISCUSSION

Routine standardized assessment of disease activity in pediatric rheumatic diseases is advocated. Thus, the VAS of physician assessment of disease activity is relevant and commonly performed. In the past, the PhGA₀₋₁₀ was almost universally used for measuring flare and treatment response in juvenile idiopathic arthritis and cSLE. More recently, the cLLDAS has been proposed as a treat-to-target for cSLE. This is because patients with cSLE in LLDAS experienced fewer disease flares and limited the accrual of disease damage in cSLE compared to patients not reaching LLDAS.⁴⁻⁶ In our study, we show that correlations with validated disease activity indices using the two types of VAS (PhGA₀₋₃ and PhGA₀₋₁₀) are almost identical. Further, rates of cLLDAS using either PhGA₀₋₃ or PhGA₀₋₁₀ are comparable, with misclassifications rates of 5% to 7%. This can be interpreted as that, in 93% of the patient assessments, the choice of the VAS type had no impact on whether the patient with cSLE was deemed inactive. We noted that there is larger variability (≥2) between the PhGA₀₋₁₀ and PhGA₀₋₃. This indicates a potential ceiling effect of the PhGA₀₋₃ and suggests that the PhGA₀₋₁₀ is preferable because it would allow a wider numeric range to better discriminate the various levels of disease activity in patients with cSLE.

An assessment of cLLDAS in clinical practice for the purpose of treat-to-target strategies requires information from patient medical records, which are more often in electronic format in pediatric rheumatology centers in North America and several countries around the world. Arguably, to measure the LLDAS in adult SLE cohorts, it is sensible to use PhGA₀₋₃ because this is the predominant type of VAS to measure disease activity in clinical practice and research. Conversely, considering the widespread use of the PhGA₀₋₁₀ in pediatric rheumatology, the authors opine that the new introduction of this PhGA₀₋₃ scale may add to the documentation burden of physicians and impede the ease of adaptation of treat-to-target strategies for cSLE in routine clinical practice.

Our study provides quantitative assessment of the measurement properties of both PhGA₀₋₁₀ and PhGA₀₋₃ in cSLE. Based on the results of this study, an alternate approach for the adaptation of the cLLDAS^5,6 could be to incorporate the PhGA₀₋₁₀ in addition to the already modified low steroid criterion that adjusts the ≤7.5 mg of prednisone equivalent for pediatric patients with smaller body weights. Notably, the PhGA₀₋₁₀ scale is already used as a measure in the Provisional Pediatric Rheumatology International Trials Organization/ACR criteria for treatment response,¹⁰ the Preliminary ACR Criteria for Flare,¹⁵ and the Provisional Criteria of Minimally important improvement of cSLE.⁹

In conclusion, the PhGA₀₋₃ and PhGA₀₋₁₀ have comparable measurement properties and yield almost identical cLLDAS rates when used in cSLE. This may suggest that both scales can be used to score the cLLDAS in cSLE. Further studies are needed to understand the measurement properties of both PhGA₀₋₃ and PhGA₀₋₁₀ scales in the Provisional Pediatric Rheumatology International Trials Organization/ACR criteria for treatment response, the Preliminary ACR Criteria for Flare, and the Provisional Criteria of Minimally important improvement of cSLE.

ACKNOWLEDGMENTS

We are indebted to the clinical research staff Catherine Robben, Megan Quinlan Waters, and Angela Merritt for their research support.

AUTHOR CONTRIBUTIONS

All authors contributed to at least one of the following manuscript preparation roles: conceptualization AND/OR methodology, software, investigation, formal analysis, data curation, visualization, and validation AND drafting or reviewing/editing the final draft. As corresponding author, Dr Brunner confirms that all authors have provided the final approval of the version to be published, and takes responsibility for the affirmations regarding article submission (eg, not under consideration by another journal), the integrity of the data presented, and the statements regarding compliance with institutional review board/Helsinki Declaration requirements.

References

Mina R, Brunner HI. Pediatric lupus–are there differences in presentation, genetics, response to therapy, and damage accrual compared with adult lupus? Rheum Dis Clin North Am 2010;36(1):53–80, vii–viii.

Smith EMD, Aggarwal A, Ainsworth J, et al. Towards development of treat to target (T2T) in childhood‐onset systemic lupus erythematosus: PReS‐endorsed overarching principles and points‐to‐consider from an international task force. Ann Rheum Dis 2023;82(6):788–798.

Franklyn K, Lau CS, Navarra SV, et al. Definition and initial validation of a Lupus Low Disease Activity State (LLDAS). Ann Rheum Dis 2016;75(9):1615–1621.

Smith EMD, Tharmaratnam K, Al‐Abadi E, et al. Attainment of low disease activity and remission targets reduces the risk of severe flare and new damage in childhood lupus. Rheumatology (Oxford) 2022;61(8):3378–3389.

Cody EM, Wilson BE, Ogbu EA, et al. Usefulness of the LLDAS as a treatment target in childhood‐onset SLE. Lupus Sci Med 2023;10(1):e000884.

Smith EMD, Aggarwal A, Ainsworth J, et al. PReS‐endorsed international childhood lupus T2T task force definition of childhood LLDAS (cLLDAS). Clin Immunol 2023;250:109296.

Isenberg DA, Allen E, Farewell V, et al. An assessment of disease flare in patients with systemic lupus erythematosus: a comparison of BILAG 2004 and the flare version of SELENA. Ann Rheum Dis 2011;70(1):54–59.

Brunner HI, Higgins GC, Wiers K, et al. Prospective validation of the provisional criteria for the evaluation of response to therapy in childhood‐onset systemic lupus erythematosus. Arthritis Care Res (Hoboken) 2010;62(3):335–344.

Brunner HI, Holland MJ, Beresford MW, et al; Paediatric Rheumatology International Trial Organisation and Pediatric Rheumatology Collaborative Study Group. American College of Rheumatology Provisional Criteria for Clinically Relevant Improvement in Children and Adolescents With Childhood‐Onset Systemic Lupus Erythematosus. Arthritis Care Res (Hoboken). 2019;71(5):579–590.

Ruperto N, Ravelli A, Oliveira S, et al; Pediatric Rheumatology International Trials Organization. The Pediatric Rheumatology International Trials Organization/American College of Rheumatology provisional criteria for the evaluation of response to therapy in juvenile systemic lupus erythematosus: prospective validation of the definition of improvement. Arthritis Rheum. 2006;55(3):355–363.

Petri M, Kim MY, Kalunian KC, et al; OC‐SELENA Trial. Combined oral contraceptives in women with systemic lupus erythematosus. N Engl J Med. 2005;353(24):2550–2558.

Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997;40(9):1725.

Gordon C, Sutcliffe N, Skan J, et al. Definition and treatment of lupus flares measured by the BILAG index. Rheumatology (Oxford) 2003;42(11):1372–1379.

Touma Z, Urowitz MB, Gladman DD. SLEDAI‐2K for a 30‐day window. Lupus 2010;19(1):49–51.

Brunner HI, Mina R, Pilkington C, et al. Preliminary criteria for global flares in childhood‐onset systemic lupus erythematosus. Arthritis Care Res (Hoboken) 2011;63(9):1213–223.