Validation and important differences for the Sarcoidosis Assessment Tool. A new patient-reported outcome measure.

AJRCCM Articles in Press. Published on 16-January-2015 as 10.1164/rccm.201410-1785OC

Page 1 of 28

Validation and important differences for The Sarcoidosis Assessment Tool: A new patient reported outcome measure Marc A. Judson1, Michael Mack2, Jennifer L. Beaumont3, Rosemary Watt2, Elliot S. Barnathan2, David E. Victorson3

1

Department of Medicine, Division of Pulmonary and Critical Care Medicine, Albany Medical College, Albany, NY 2

Immunology, Janssen Research & Development, LLC, Spring House, PA

3

Department of Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, Illinois

Address Correspondence and reprint requests to: Marc A. Judson, M.D. Professor of Medicine Chief, Division of Pulmonary and Critical Care Medicine Department of Medicine, MC-91 Albany Medical College 47 New Scotland Avenue Albany, New York 12208 USA 1-518-262-5196 FAX: 1-518-262-6472 [email protected]

Copyright © 2015 by the American Thoracic Society


Running head: Validation of the Sarcoidosis Assessment Tool Descriptor #: 9.38 - Sarcoidosis: Clinical Evaluation and Treatment Funding Support: Janssen Research & Development, LLC Word count: 2448/3500

At a Glance Commentary Scientific Knowledge on the Subject: The study describes the performance of the Sarcoidosis Assessment Tool (SAT), a sarcoidosis specific patient reported outcome measure (PRO) in a randomized clinical trial. The SAT was found to have excellent reliability and consistency. What This Study Adds to the Field: Clinically important differences for this instrument and its subscales were determined, making it useful for clinical application. Several of the subscales were organ specific for the lung or skin, and these subscales were validated against appropriate anchors. We believe that these findings suggest that the SAT will be a useful clinical tool to measure the sarcoidosis patient’s assessment of the impact of their disease and response to therapy. Conflicts of Interest: M.A. Judson is a consultant for Questcor, Janssen, Mistubishi-Tanabe, Celgene, and Novartis. Drs. Barnathan, Watt and Mack are or were employees of Janssen Research & Development, LLC at the time of the study and own(ed) stock in Johnson & Johnson.

Conception and design: MAJ, RW, ESB, MM, DEV Acquisition of the data: MAJ, RW, ESB, MM Analysis and interpretation: MAJ, RW, ESB, MM, JLB, DEV Drafting the manuscript and revising it for important intellectual content: MAJ, RW, ESB, MM, JLB, DEV


Page 2 of 28


Page 3 of 28

Abstract Rationale: Patient reported outcome (PRO) measures have been developed to measure symptoms and other aspects of health related quality of life. Objective: The Sarcoidosis Assessment Tool (SAT), a sarcoidosis specific PRO, was administered in a lung and skin sarcoidosis treatment trial. We explored SAT performance characteristics and correlation with standard clinical measurements in order to validate it as a useful clinical sarcoidosis-specific PRO.

Methods: The SAT analyses focused on baseline and week 16 assessments. Besides the SAT, participants underwent clinical and physician assessments plus additional PROs that were used as anchor variables and were compared to the SAT. Reliability was evaluated by using Cronbach’s alpha coefficient. Spearman correlation coefficients were used to evaluate the association between SAT scores with clinical and other PRO measures. Changes between assessments in the clinical and PRO “anchor” variables were classified as improved, stable, or worsened. Mean differences between adjacent categories of the known groups and mean changes from the ability to detect change analyses were reviewed for appropriate clinically important difference estimates.

Measurements/main results: Results from 173 patients were analyzed. Each SAT module reflected appropriate anchor variables both at baseline and in terms of change. The Cronbach’s alpha for each of these modules was at least 0.87. In addition, we successfully established a clinically important difference range for each SAT module.

Conclusions: We demonstrated that the SAT is a reliable and consistent sarcoidosis-specific PRO. It has excellent internal consistency and reliability. A range of clinically important differences has been established for the SAT modules.



Abstract word count: 250/250 Key Terms: sarcoidosis, assessment tool, PRO


Page 4 of 28


Page 5 of 28

Introduction The clinical manifestations of sarcoidosis are typically assessed and monitored using objective clinical outcome measures including radiographic findings, pulmonary function testing, skin lesion status, and ophthalmologic findings. However, it has been demonstrated that physician assessment of sarcoidosis correlates poorly with patient self-assessment.1 Furthermore, the patient is most concerned with the status of his/her health related quality of life (HRQoL) rather than the results of objective medical tests.

Patient reported outcome (PRO) measures have been developed to measure disease symptoms, treatment-related side effects and aspects of HRQoL such as emotional health, and social well-being. A PRO is a measurement of any aspect of a patient’s health status that comes directly from the patient (i.e., without the interpretation of the patient’s responses by a physician or anyone else).2 PRO’s have the advantage of eliciting the symptoms and concerns directly from the patient without interpretation by another individual. Furthermore, these tools may be scaled and validated so that these data may reliably reflect the perceived HRQoL experienced by patients themselves.3

We conducted a three-arm trial of golimumab, ustekinumab, and placebo for the treatment of lung and skin sarcoidosis.4 As part of that trial we administered the Sarcoidosis Assessment Tool, a sarcoidosis-specific HRQoL PRO that had previously been constructed.5,6 We now report, for the first time, an analysis of SAT results from an interventional trial and explore its



performance characteristics and correlation with standard clinical measurements in order to validate it as a useful clinical PRO in patients with sarcoidosis.

Methods Protocol 1275148SCD2001 was a phase II, multicenter, randomized, double-blind parallelgroup, placebo-controlled study evaluating the safety and efficacy of treatment with ustekinumab and golimumab in patients with chronic sarcoidosis who had sarcoidosis involvement of the lung, skin, or both organs. One hundred and seventy three patients were enrolled in this trial, the results of which have been published elsewhere.4 Study drug was administered through week 24. The primary endpoint was assessed at week 16, and after that, those who were eligible entered a steroid taper phase through week 28.

The Sarcoidosis Assessment Tool (SAT) is a new PRO tool developed for use in sarcoidosis clinical trials and clinical practice. 5,6 The SAT is comprised of select measures from the Patient Reported Outcomes Measurement Information System (PROMIS),3,7-9 as well as newly developed sarcoidosis specific patient-reported measures. Through a multistep process of selecting the highest performing PROMIS items, balanced with important clinical relevance in sarcoidosis, several modified PROMIS short forms were assembled for use in the SAT, including items from: 1) Physical Function, 2) Satisfaction with Roles and Activities, 3) Fatigue, 4) Pain Interference and 5) Sleep Disturbance. In addition, several new sarcoidosis specific PRO


Page 6 of 28


Page 7 of 28

measures were developed using item-response theory (IRT) approaches. These measures represented the most commonly affected areas and issues in sarcoidosis, including: 1) Skin Concerns, 2) Skin Stigma, 3) Lung Concerns, and 4) Eye Concerns (NOTE: because the drug trial concerned only subjects with lung and/or skin sarcoidosis, the Eye Concerns short form was not administered in trial 1275148SCD2001 and is thus not included in this report). The following specific domains of the SAT were used in this trial: daily activities, satisfaction with life activities, fatigue, pain, sleep disturbance, lung concerns, skin concerns, and embarrassment over skin appearance. Using the trial data we assessed the SAT’s reliability, construct validity (convergent, discriminant, known-groups), its ability to detect change, and estimated the magnitude of clinically meaningful change. The SAT was administered at baseline, week 8, week 16, week 28, and week 44. The validation analyses focused on the baseline and week 16 assessments only. Week 16 was selected instead of week 8 because a greater number of validation measures were administered at the week 16 assessment and because the primary endpoint of the trial was assessed at Week 16 before the steroid taper phase of the trial. In addition to the SAT, participants also underwent clinical, physiologic, physician, and patient reported outcome assessments described in Table 1 that were used as anchor variables against which the SAT was compared.

Statistical analysis The treatment arms were combined for all analyses, because there were no significant differences between treatment groups in any of the SAT scores at Week 16 (p > 0.10) (data not shown).



Analyses were conducted on the full sample except where indicated that only the patients with lung involvement or with skin involvement were used. No adjustment for multiple comparisons was made. All analyses were based on the available data; there was no imputation of missing data, as approximately 5% to 6% of the SAT scores were missing at week 16. All PRO instruments were scored according to developers’ instructions. Reliability was evaluated by using Cronbach’s alpha coefficient to measure internal consistency. Spearman correlation coefficients were used to evaluate the association between SAT scores and the clinical and other patient reported measures. We hypothesized moderate to high correlation coefficients (r>0.3)10 between variables measuring the same or related constructs [e.g., SAT SarcoidosisLung Concerns with Medical Research Council (MRC) Dyspnea, Physician Global Assessment (PGA), SGRQ total score, chest radiographic scoring, lung Physician Organ Assessment, percent predicted forced vital capacity (FVC), diffusing capacity for carbon monoxide (DLCO), 6-minute walk distance (6MWD), and Patient Global Assessment (PGA)]. Mean scores were compared across categories of relevant anchor variables at each assessment. Analysis of variance (ANOVA) was used to test the differences among categories. Effect sizes (mean difference / pooled standard deviation) were calculated for the differences between adjacent groups to allow for comparisons with other instruments. For example, patients were grouped according to MRC Dyspnea category and the SAT Physical Functioning and Sarcoidosis-Lung Concerns scores were compared between categories.

An important aspect of validation is determining the extent to which important changes in the anchor variables are captured by changes in the instrument score. Change scores were


Page 8 of 28


Page 9 of 28

calculated for the SAT scores and for the other patient-reported and clinical measures by subtracting the baseline score from the Week 16 score. Changes between assessments in the clinical and patient reported “anchor” variables were classified as improved, stable, or worsened as defined in Table 1. Paired t-tests were used to evaluate whether mean SAT change scores within those groups defined by the anchor variable changes were significantly different from zero. The standardized response mean (mean SAT change in anchor group / SD of change in that group) was calculated. SAT change scores were similarly assessed within groups defined by changes in other anchor variables (Table 1):

Distribution- and anchor-based methods were used to identify clinically important differences for the SAT scores. Distribution-based methods included 1/3 standard deviation (SD), 1/2 SD and one standard error of measurement (SEM). The SEM is computed as follows: SEM = σ x 1 − rxx where: rxx

=

the reliability of the measure (i.e., the Cronbach’s alpha)

σx

=

the standard deviation of the measure at baseline.

Anchor-based methods used the known-groups and change analyses described above. Mean differences between adjacent categories of the known groups and mean changes from the ability to detect change analyses were reviewed for appropriate clinically important difference (formally known as “minimally important difference”) estimates. We applied the same three a priori criteria for useable anchor-based estimates as used by Yost and colleagues.11 The first is a



correlation between an anchor and SAT score of at least 0.3. The second criterion requires a sample size of at least 10 in the clinically distinct group (e.g., MRC Dyspnea 0, 1, 2, or 3) or change score group (e.g., improved, worsened). The final criterion is that the anchor-based estimate has a corresponding standardized mean difference within a plausible range of 0.2-0.8; that is, standardized estimates 0.8 standard deviation units are unlikely to be "minimal." If an anchor-based estimate did not meet all of these criteria, it was not considered in the final clinically important difference determination. We have summarized the results for each score using nonparametric statistics, namely medians and interquartile ranges.

Results Subjects randomized to one of 3 treatment groups (placebo, golimumab, ustekinumab) were generally well balanced in terms of demographic and clinical variables, as well as by smoking history, pulmonary function, dyspnea, and quality of life measures (Table 2). At study entry, similar numbers of patients across treatment groups were receiving similar daily doses of oral corticosteroids (median dose of 10 mg) and other immunomodulators for sarcoidosis such as methotrexate, azathioprine, and mycophenolate mofetil, although use of these latter drugs were infrequent across treatment groups.

Descriptive statistics for the SAT component scales at baseline and week 16 are presented in Table 3. Reliability was very good for all scales, with Cronbach’s alpha coefficients of 0.86 or greater (Table 3).


Page 10 of 28

Page 11 of 28


Most of the SAT measures correlated moderately well with appropriate PRO and other anchor variables (Table 4). One exception was the SAT skin measures (skin concerns and skin stigma) that did not correlate with many of the anchor skin measures. However, SAT Sarcoidosis - Skin Concerns and Skin Stigma were highly correlated with DLQI and moderately correlated with patient global assessment. It should also be noted that almost all correlations with patient global assessment were at least moderate whereas all correlations with physician global assessments were small to negligible.

We next compared SAT Module scores between anchor groups at week 16 (Table 5). Almost all of the SAT module scores differed significantly between groups defined by appropriate anchor variables. Most effect sizes were in the 0.30 to 1.0 range.

We then examined the change over 16 weeks in the anchor variables as compared to the change in the SAT module scores (Table 6). Changes in almost all of the anchor variables were in the appropriate direction and at least moderately correlated with changes in the appropriate SAT module score. Change in SAT Sarcoidosis - Skin Concern was moderately correlated with several measures of skin involvement while SAT Skin Stigma change scores were moderately correlated with change in skin global assessment only.



We next evaluated the change in SAT module scores from baseline to week 16 compared to the change in the anchor variables partitioned into three groups of differing amounts of change (Table 7). In general, the changes in SAT module scores were in the appropriate direction as the three groups for each anchor variable. Most of the effect sizes were moderate.

Finally, we calculated the distribution-based estimates of clinically important differences (summarized in Table 8). For the cross-sectional and longitudinal anchor-based analyses, only those estimates that met the criteria outlined in the methods are listed in Table 8. The estimated clinically important differences for the scales varied but were generally in the range of 2-5 points.

Discussion We found that in our population of chronic lung and skin sarcoidosis patients, the SAT, a sarcoidosis-specific HRQoL PRO, was consistent with other anchor indicators of disease in sarcoidosis. The SAT is composed of modules reflecting various organ involvement and other clinical aspects of sarcoidosis. Each of these modules reflected appropriate anchor variables both at baseline and in terms of change. These results suggest that the SAT is a HRQoL PRO that has good construct validity and is a reliable reflection of sarcoidosis patients’ health concerns. The Cronbach’s alpha for each of these modules was at least 0.87, suggesting that each module had excellent reliability. It had previously been shown that the SAT fatigue module, that was used as the fatigue PRO in Patient Reported Outcomes Measurement Information Systems


Page 12 of 28


Page 13 of 28

(PROMIS), funded by the National Institutes of Health (NIH),7,12 had superior reliability to the Fatigue Assessment Scale,13,14 a PRO commonly used to assess fatigue in sarcoidosis.

In addition, we were able to establish an estimated clinically important difference range for each of the modules of the SAT. The determination of a clinically important difference for each SAT module allows for clinically meaningful changes in sarcoidosis patient self-assessment of their HRQoL to be measured. Since sarcoidosis is a disease with disparate manifestations, specific modules of the SAT can be used to determine the status of several specific manifestations.

The SAT was developed based on a conceptual model of health-related quality of life in sarcoidosis.5 This PRO was developed and tested using item response theory, where items were scaled and short forms for each module were constructed using standard methodology.15 Because item banks for each SAT module were created, additional short forms could be constructed to examine patients with particularly mild or severe disease (overcome ceiling and floor effects).16 In addition, computer adapted testing (CAT) could be constructed so that reliable measurements of each PRO module could be made using a small number of items if administered to patients via a computer.17 The SAT requires approximately 5 to 10 minutes to complete so that it is appropriate for use in clinical settings.



One potential limitation of our analysis is that we pooled the SAT results across three treatment groups in this study. However, this study was a negative trial both in terms of the primary endpoint (change in the percent predicted FVC at 16 weeks in the lung sarcoidosis subjects) and all major secondary endpoints that included the change from baseline to week 28 in 6MWD, SGRQ total score, % predicted FVC in the lung group, and the proportion of SPGA responders (score of ≤1) in the skin group.4 In addition, our analysis showed that there were no significant treatment effects of golimumab, ustekinumab, and placebo on the SAT modules. Therefore, we believe that pooling of the three treatment groups to analyze the SAT data is appropriate. Although the trial was a negative study, and, therefore, none of the three study treatment groups demonstrated a statistically significant improvement, we were able to adequately determine the characteristics of the SAT instrument by examining study subjects who had a significant change in anchor variables over the 16 weeks. We acknowledge that some of the subgroups of anchor change were very small, and this may limit the validation of SAT change across the cohort. However, almost all the anchor change subgroups had changes in the SAT score in a similar direction; furthermore, all of the anchor change subgroups had changes in the SAT score in a similar direction when the SAT changes were statistically significant.

Correlations with other sarcoidosis-specific HRQoL PROs such as the Sarcoidosis Health Questionnaire (SHQ)18 and the King’s Sarcoidosis Questionnaire (KSQ)19 were not made. Although the SHQ has been well studied, a clinically important difference has not been established. In addition, the SHQ is not divided into domains/modules so that various aspects of sarcoidosis are lumped together into a total score. Therefore, the SHQ may be insensitive to


Page 14 of 28

Page 15 of 28


changes in specific aspects of sarcoidosis-related HRQoL such as fatigue or skin changes. The KSQ PRO was constructed using a similar methodology to the SAT and also contains modules reflecting specific aspects of sarcoidosis. However, the KSQ has fewer modules than the SAT. Furthermore, because of the SAT’s foundation in IRT-based item banks from NIH PROMIS and newly created organ-specific banks, its ability to measure the full spectrum of a given latent trait appears to be more flexible and potentially profound. Finally, a clinically important difference has not been established for the KSQ, making the clinical importance problematic to interpret from this PRO. It will be useful to compare all of these specific sarcoidosis specific HRQoL PROs in the future.

It is interesting, albeit not that surprising, that while almost all correlations with patient global assessment were at least moderately good, the correlations with physician global assessment were not. This emphasizes the importance of capturing the patient’s perspective using validated instruments. It has been previously shown that the correlation between the patient and physician assessment of the health impact of sarcoidosis may be very discordant.1

In summary, we have demonstrated that the SAT is a reliable and consistent sarcoidosis-specific PRO that accurately reflects various disparate aspects of the disease. It appears to have excellent internal consistency and reliability. A clinically important difference has been established for the SAT modules. We believe that the SAT should be considered for clinical and research use in sarcoidosis.



Table. 1

Anchor variables

Ref #

Clinical Assessments - Sarcoidosis Activity and Severity Index (SASI), an assessment of the extent and severity of the involvement of the face - Chest radiograph scoring, grading reticulo-nodular opacities (R-score) based on extent and profusion - Skin Physician’s Global Assessment (SPGA), a physician assessment of the status of the patient’s skin involvement, based on overall induration and erythema of the skin lesions - Skin lesion Percent Body Surface Area (BSA), a calculation to estimate the percentage of sarcoidosis involvement over the patient’s entire body - Target lesion assessment, an assessment of a prospectively identified target lesion for erythema and induration.

Physiologic Assessments - Spirometry: - Percent predicted FVC

Type of scoring

anchor variable versus variable compared to SAT scores

20

0-48

Comparison to SAT

21

0-4 scale

Anchor variable

0-4

Anchor variable

decrease by at least one category, no change, increase by at least one category decrease by at least one category, no change, increase by at least one category

Comparison to SAT

20

0-4 scale for erythema and for induration

Comparison to SAT

22

Comparison to SAT

- Diffusing Capacity of the Lung for Carbon Monoxide (DLCO)

23

Anchor variable

- 6-Minute Walk Test

24

Anchor variable

Physician Assessments - Physician Organ Assessment (POA) for skin and lung, determine degree of

Changes in anchor variables partitioned into 3 groups assessed by SAT

25

7 point scoring

Anchor variable


greater than 7% predicted decline, -7% predicted to +7 % predicted change, greater than 7% predicted improvement greater than 20% decline, -20% to +20% change, greater than 20% improvement greater than 30 m decline, -30 m to 30 m change, greater than 30 m improvement

decrease by at least one category, no

Page 16 of 28


Page 17 of 28

organ involvement, based on all information available to the physician (eg laboratory analyses, physical examination findings, and other diagnostic tests) at the time of the evaluation - Physician Global Assessment (PGA), a VAS scale adapted from the American College of Rheumatology (ACR) core set of assessments, to measure the patient’s overall well being from the physician’s perspective. PRO's - St. George’s Respiratory Questionnaire (SGRQ) - 36-item Short Form health survey (SF36) Physical Functioning, Social Functioning, Bodily Pain, Vitality scales - Medical Research Council (MRC) Dyspnea Score

scale

change, increase by at least one category

VAS scale

Comparison to SAT

decrease by at least one category, no change, increase by at least one category

26

0-100

Anchor variable

+8

27

0-100 (higher score better) 1-5

Anchor variable

+4

Anchor variable

decrease by at least one category, no change, increase by at least one category +4

28

29

- Fatigue Assessment Scale a 10-item 0-50 Anchor variable instrument assessing physical and mental fatigue 30 - Dermatology Life Quality Index 0-30 Anchor variable +3.2 assess the impact of the skin disorder on the patient’s daily living. 31 - Patient Global Assessment (PGA), a VAS scale Comparison to greater than 10% VAS scale adapted from the ACR core SAT decline, -10% to +10% set of assessments, to measure the change, greater than patient’s overall wellbeing from the 10% improvement patient’s perspective Ref#: reference number; SAT: Sarcoidosis Assessment Tool; FVC: forced vital capacity; VAS: visual analog scale; PROs: patient reported outcome measures

Table 2: Subject demographics and clinical characteristics Placebo

Golimumab

Ustekinumab

Total

58

55

60

173

Mean (SD) Median

49.5 (9.5) 49

50.0 (9.4) 51

49.8 (10.2) 50

49.8 (9.7) 50

Male Female

29 (50.0%) 29 (50.0%)

28 (50.9%) 27 (49.1%)

31 (51.7%) 29 (48.3%)

88 (50.9%) 85 (49.1%)

N Age (years)

Sex



Page 18 of 28

Race White Black Other

32 (55%) 23 (40%) 3 (5%)

36 (66%) 16 (29%) 3 (5%)

38 (63%) 19 (32%) 3 (5%)

106 (61%) 58 (34%) 9 (5%)

%predicted FVC Mean (SD)

71.6 (12.5)

73.2 (15.4)

69.9 (17.0)

71.5 (15.1)

%predicted FEV1 Mean (SD)

61.8 (18.2)

63.4 (17.8)

62.0 (20.7)

62.4 (18.9)

DLCO (mL/min/mmHg) Mean (SD)

13.4 (6.8)

11.4 (6.3)

13.6 (7.7)

12.8 (7.0)

6MWD (Meters) Mean (SD)

394 (97.4)

405 (99.3)

401 (92.1)

400 (95.7)

SGRQ total score Mean (SD)

47.9 (19.9)

46.1 (21.9)

45.9 (20.7)

46.6 (20.7)

MRC dyspnea score 3.00 (1.06) 2.95 (0.87) 2.95 (0.96) 2.97 (0.96) Mean (SD) Smoking status Mean (SD) total pack year 8.1 (5.2) 9.7 (3.7) 8.9 (8.1) 8.8 (5.7) Never Smoker 26 (59%) 26 (62%) 33 (72%) 85 (64%) Current Smoker 4 (9%) 1 (2%) 1 (2%) 6 (5%) Former Smoker 14 (32%) 15 (36%) 12 (26%) 41 (31%) SD: standard deviation; FVC: forced vital capacity; FEV1: forced expiratory volume in one second; DLCO: single breath diffusing capacity for carbon monoxide; 6MWD: 6-minute walk distance; MRC: Medical Research Council



Page 19 of 28

Table 3. The SAT module scales: descriptive statistics

N

Mean (SD)

Median (range)

Cronbach’s coefficient alpha

Baseline Week 16

173 163

39.5 (7.2) 41.0 (7.5)

38.5 (23 – 57) 40.9 (26 – 57)

0.90 0.90

Baseline Week 16

173 162

45.2 (8.5) 47.6 (9.0)

45.2 (31 – 67) 48.0 (31 – 67)

0.94 0.96

Baseline Week 16

173 162

56.9 (9.3) 54.1 (9.8)

58.6 (42 – 77) 55.9 (42 – 74)

0.96 0.97

Baseline Week 16

173 163

53.3 (9.5) 51.1 (9.9)

54.5 (34 – 76) 51.0 (34 – 76)

0.92 0.94

173 163

57.6 (8.6) 54.7 (9.3)

58.2 (34 – 77) 54.3 (34 – 77)

0.92 0.94

173 163

47.0 (7.2) 44.3 (7.6)

47.7 (29 – 63) 45.5 (29 – 58)

0.89 0.91

132 124

48.2 (6.4) 45.1 (6.9)

48.8 (33 - 63) 46.0 (29 - 60)

0.89 0.91

162 158

49.1 (10.8) 46.1 (10.2)

48.6 (33 – 78) 47.0 (33 – 69)

0.90 0.87

58 55

56.8 (9.2) 53.8 (8.9)

56.4 (39 - 78) 53.5 (33 - 69)

0.90 0.87

162 158

50.8 (10.2) 50.0 (10.2)

49.5 (41 – 72) 48.8 (41 – 72)

0.92 0.92

58 55

60.7 (7.8) 59.7 (7.7)

60.8 (41 - 72) 58.2 (41 - 72)

0.92 0.92

SAT Module Physical Functioning

Satisfaction with Roles and Activities

Pain Interference

Sleep Disturbance

Fatigue Baseline Week 16 Sarcoidosis – Lung Concerns entire cohort Baseline Week 16 Sarcoidosis – Lung Concerns lung cohort only Baseline Week 16 Sarcoidosis – Skin Concerns entire cohort Baseline Week 16 Sarcoidosis – Skin Concerns skin cohort only Baseline Week 16 Sarcoidosis – Skin Stigma entire cohort Baseline Week 16 Sarcoidosis – Skin Stigma skin cohort only Baseline Week 16 SAT: Sarcoidosis Assessment Tool



Page 20 of 28

Table 4. Spearman correlation coefficients at week 16 Between SAT and comparison measures Comparison measures

Physical Functioning [+]

Satisfaction with Roles and Activities [+]

Pain Interference [-]

Sleep Disturbance [-]

Fatigue [-]

Sarcoidosis – Lung Concerns* [-]

Sardoidosis – Skin Concerns† [-]

Sarcoidosis – Skin Stigma† [-]

SF-36 Physical Functioning 0.73 [+] SF-36 Social Functioning [+] 0.66 SF-36 Bodily Pain [+] -0.83 SF-36 Vitality [+] -0.46 -0.78 FAS [-] -0.58 0.47 0.84 SGRQ total score [-] -0.81 -0.66 0.80 MRC Dyspnea [-] -0.64 0.43 0.38 6MWD [+] 0.44 -0.36 -0.38 Percent predicted FVC [+] 0.34 -0.15 -0.23 DLQI [-] -0.22 0.66 0.69 SASI [-] 0.34 0.06 0.10 Skin Global Assessment [-] 0.16 0.09 0.17 Skin lesion BSA [-] -0.04 0.07 0.18 Target lesion induration [-] 0.08 0.04 0.08 Target lesion erythema [-] 0.24 -0.14 0.05 Physician Global Assessment 0.12 0.26 0.05 0.13 -0.18 -0.16 0.10 0.12 [+] Patient Global Assessment [+] 0.55 0.57 -0.44 -0.36 -0.53 -0.52 -0.31 -0.26 SAT: Sarcoidosis Assessment Tool; *Population with lung involvement; †Population with skin involvement; SF-36: Sghort Form 36; FAS: Fatigue Assessment Scale; SGRQ: St. George Respiratory Questionnaire; MRC: Medical Research Council; 6MWD: 6-minute walk distance; FVC: forced vital capacity; DLQI: Dermatology Life Quality Index; SASI: Sarcoidosis Activity and Severity Index; BSA: body surface area Empty cells indicate that a moderate correlation was not hypothesized a priori [+]: higher score indicates better HRQoL or functioning; [-]: higher score indicates worse HRQoL or functioning



Page 21 of 28

Table 5. SAT Module scores within anchor groups at week 16

DLCO 0% - 75% (n=11)

Mean difference between adjacent groups Physical Functioning 37.9 (6.1) 3.9 41.8 (7.2) 0.3 42.1 (8.0) 2.7 44.8 (8.5) --

FAS

The Elbow Self-Assessment Score (ESAS): development and validation of a new patient-reported outcome measurement tool for elbow disorders.

The Spine Functional Index: development and clinimetric validation of a new whole-spine functional outcome measure.

The burden of osteoarthritis: development and validation of a new assessment tool (BONe'S).

Content Validity for the VVSymQ® Instrument: A New Patient-Reported Outcome Measure for the Assessment of Varicose Veins Symptoms.

The Munich Wrist Questionnaire (MWQ) - development and validation of a new patient-reported outcome measurement tool for wrist disorders.

Patient-important outcome for the assessment of fracture repair.

The Munich Knee Questionnaire: Development and Validation of a New Patient-Reported Outcome Measurement Tool for Knee Disorders.

The obesity quotient: a new tool to measure childhood obesity.

Development and validation of the Surgical Outcome Risk Tool (SORT).

Standardized patient and standardized interdisciplinary team meeting: validation of a new performance-based assessment tool.

The Performance Scales disability measure for multiple sclerosis: use and sensitivity to clinically important differences.

Cognitive Aids in Medicine Assessment Tool (CMAT): preliminary validation of a novel tool for the assessment of emergency cognitive aids.

Validation of a new tool for the assessment of study quality and reporting in exercise training studies: TESTEX.

Important differences between a triage pedigree and risk assessment.

Development and Validation of the Adolescent Assessment of Preparation for Transition: A Novel Patient Experience Measure.

The VVSymQ® instrument: Use of a new patient-reported outcome measure for assessment of varicose vein symptoms.

Development and validation of the Tibetan primary care assessment tool.

Development and validation of a patient-reported outcome measure for stroke patients.

NPH Log: Validation of a New Assessment Tool Leading to Earlier Diagnosis of Normal Pressure Hydrocephalus.

Validation of the flemish CARES, a quality of life and needs assessment tool for cancer care.

Reliability assessment and validation of the postacne hyperpigmentation index (PAHPI), a new instrument to measure postinflammatory hyperpigmentation from acne vulgaris.

Stability of an ERP-based measure of brain network activation (BNA) in athletes: A new electrophysiological assessment tool for concussion.

Validation of the Patient-Reported Outcome Mortality Prediction Tool (PROMPT).

Development and validation of the Diabetic Peripheral Neuropathic Pain Impact (DPNPI) measure, a patient-reported outcome measure.