Abstract
Purpose
To refine the previously developed scleroderma gastrointestinal tract (SSC-GIT 1.0) instrument.
Methods
We administered the SSC-GIT 1.0 and SF-36 to 152 patients with SSc; 1 item was added to SSC-GIT 1.0 to assess rectal incontinence. In addition, subjects completed a rating of the severity of their GIT involvement (from very mild to very severe). Evaluation of psychometric properties included internal consistency reliability, test-retest reliability (1.1 week mean time interval), and multitrait scaling analysis.
Results
Study participants were female (84%) and Caucasian (81%); 55% had diffuse SSc. Self-rated severity of GIT involvement ranged from no symptoms to very mild (39%), to mild (21%), to moderate (31%), to severe/ very severe (9%). Of an initial 53 items in SSC-GIT 1.0, 19 items were excluded, leaving a 34-item revised instrument (UCLA Scleroderma Clinical Trial Consortium GIT 2.0 [UCLA SCTC 2.0]). Analyses supported 7 multi-item scales: reflux, distention/bloating, diarrhea, fecal soilage, constipation, emotional well-being, and social functioning. Test-retest reliability estimates were ≥ 0.68 and coefficient alphas ≥0.67. Participants who rated their GIT disease as mild had lower scores on a 0-3 scale on all 7 scales. Also symptom scales were able to discriminate subjects with corresponding clinical GIT diagnoses. Total GIT Score, developed by averaging 6 of 7 scales (excluding constipation), was reliable and provided greater discrimination between mild, moderate, and severe self-rated GIT involvement than individual scales.
Conclusion
This study provides support for the reliability and validity of the UCLA-SCTC GIT 2.0, an improvement over SSC-GIT 1.0 and supports a Total GIT Score in SSc-GIT.
Gastrointestinal tract (GIT) involvement occurs in approximately 90% of patients with systemic sclerosis (SSc)(1;2) and has a negative impact on health-related quality of life (HRQOL)(3;4). Support for the reliability and validity of the Scleroderma Gastrointestinal Tract Involvement 1.0 (SSC-GIT 1.0) instrument was previously reported (5). However, the 52-item SSC-GIT 1.0 is fairly time consuming to administer, especially when administered in combination with other patient-reported measures. The goal of this study was: 1) to refine the SSC-GIT 1.0 and develop a shorter but still reliable and valid instrument; 2) to differentiate reflux symptoms from the symptoms of distention/ bloating; 3) to add a scale to evaluate rectal incontinence because of its high prevalence in SSc(6); and 4) develop a composite score that captures overall GIT burden association with SSc. We followed the Food and Drug Administration draft guidance on the development/ modification of a patient-reported outcome measure(7) and developed a 34-item revised instrument (UCLA Scleroderma Clinical Trial Consortium GIT 2.0 [UCLA SCTC 2.0]).
Materials and Methods
Development and evaluation of SSC-GIT 1.0
We searched the Pub Med electronic database for existing GIT questionnaires, convened an expert panel of rheumatologists, gastroenterologists, and psychometricians, pilot tested SSC-GIT 1.0(5) in focus groups comprised of patients with SSc, and field tested SSC-GIT 1.0 in 88 patients with SSc and GIT involvement. Analyses of SSC-GIT 1.0 supported 6 multi-item HRQOL scales: reflux/indigestion, diarrhea, constipation, pain, emotional well-being, and social functioning. SSC-GIT 1.0 was found to be feasible, reliable, and valid.
Field-testing of the UCLA SCTC-GIT 2.0
Patients with SSc and GIT involvement were invited to participate at the following 3 Scleroderma centers in the United States: UCLA, Los Angeles, CA; University of Michigan, Ann Arbor, MI; and University of Texas at Houston, Houston, TX. The protocol was approved by the IRB at each institution and each subject signed a consent form prior to completing the questionnaires. In addition to completing the paper-and-pencil SSC-GIT 1.0 questionnaire(5), all patients provided sociodemographic information (age, sex, ethnicity, and level of education), completed the SF-36 version 2(8), a global self-rating of GIT severity (In the past 1 week, how severe were your gastrointestinal (gut, GI) symptoms overall?: none, very mild, mild, moderate, severe, very severe), and a self-rating of upper and lower GIT severity: “In the past 1 week, how severe were your upper/ lower gastrointestinal (gut, GI) symptoms?” (none, very mild, mild, moderate, severe, very severe).
Each physician did a physical examination to determine the type of SSc (limited or diffuse) and provide their GIT diagnoses.
Patient- reported measures
The SF-36 version 2 is a generic health status measure consisting of 36 items assessing 8 domains(8;9):physical functioning [10 items], bodily pain [2 items], role limitations due to physical health perceptions [4 items], general health perceptions [5 items], mental health [5 items], role limitations due to emotional problems [3 items], vitality [4 items], and social functioning [2 items]. The 8 SF-36 scales are summarized into Physical Component Summary (PCS) and Mental Component Summary (MCS) scores. The summary scores are normalized to the U.S. general population, for whom the mean score is 50 and the standard deviation is 10. We used the acute (1-week) recall period version of the SF-36 v.2 (8).
SSC-GIT 1.0 is a 52-item instrument with 6 multi-item scales: reflux/indigestion, diarrhea, constipation, pain, emotional well-being, and social functioning(5). Each item is scored on a 0-100 possible range where 100 represents better HRQoL. SSC-GIT 1.0 has a 1-week recall period. In this study, we added one item to assess rectal incontinence. From the data derived from the SCTC-GIT 1.0 we developed a 34-item revised instrument (UCLA Scleroderma Clinical Trial Consortium GIT 2.0 [UCLA SCTC 2.0]) using the psychometric analysis as described below.
Psychometric Analysis
The 53 items (52 items from SSC-GIT 1.0 and 1 rectal incontinence item) were grouped into 8 hypothesized scales based on their content: reflux, distention/bloating, diarrhea, fecal soilage (to assess rectal incontinence), constipation, pain, emotional well-being, and social functioning. Since one of our objectives was to discriminate the reflux symptoms from the symptoms of distention/ bloating, we divided the items in the reflux/ indigestion scale into reflux scale and distention/ bloating scale (better wording than ‘indigestion’). Unlike SSC-GIT 1.0, items were scored on a 0-3 possible range, with lower values indicated better HRQOL. A multitrait analysis was used to evaluate the extent to which each item correlated more strongly with its hypothesized scale than with other scales(10). Items identified as having either low item-to-hypothesized scale correlations (r≤0.40, N=1), poor discrimination across scales (N=5), or any subset of items having high correlations among themselves (N=13), were excluded. Based on these analyses, 7 multi-item scales (reflux, distention/bloating, diarrhea, fecal soilage, constipation, emotional well-being, and social functioning) were retained. Based on the multitrait analysis, dysphagia to solid food (Q1 Appendix A) was retained under the reflux scale. In addition, because only 1 item of pain scale was retained (Q34 Appendix A) after multitrait analysis and pertained to the symptoms of constipation, pain item was merged with the constipation scale. and resulted in the final UCLA SCTC -GIT 2.0 instrument. Version 2 of the GIT instrument contains 33 items from SSC-GIT 1.0 and 1 item to assess fecal soilage.
Mean scores, standard deviations (SD), ranges, and percentages of respondents scoring the minimum (floor) and maximum (ceiling) possible scores were calculated to evaluate scale score distributions.
Internal consistency reliability was estimated using Cronbach's alpha(11). For test-retest reliability, intraclass correlations (ICC) over a mean interval of 1.1 weeks in 25 subjects were calculated using a two-way random-effects model(12) described in detail previously(5). Four subjects each developed diarrhea and constipation during test-retest visits and ICC was calculated including and excluding these patients.
We assessed the construct validity by exploring the association between the UCLA SCTC-GIT 2.0 and other HRQOL measures. Correlations ≤0.29 were considered to be small, between 0.30 and 0.49 moderate, and ≥0.50 as large(13). We hypothesized a priori moderate product-moment correlations between the social functioning scale of the UCLA SCTC-GIT 2.0 and social functioning of the SF-36, the emotional well-being scale of the UCLA SCTC-GIT 2.0 and the role-emotional scale of the SF-36 (r's≥0.30). We also hypothesized that the upper GIT anchor will have larger correlations with upper GIT scales than lower GIT scales and the lower GIT anchor will have larger correlations with lower GIT scales than upper GIT scales.
We examined the ability of the UCLA SCTC-GIT 2.0 to differentiate among patients with mild, moderate and severe GIT involvement. We also hypothesized that symptom-specific scale scores (e.g., reflux scale) will be higher (worse HRQOL) in patients with specific GIT clinical diagnosis (e.g., gastro-esophageal reflux disease). The self-rated severity of GIT (external anchor) was collapsed into 4 groups — no symptoms/ very mild, mild, moderate, and severe/ very severe. Tukey's post hoc adjustment was used to test for any significant differences in the ANOVA analyses. Exploratory factor analysis was performed to evaluate the underlying structure of the 7 multi-item scales (14). Criteria used to select the most plausible model included components accounting for at least 5% of the variance and principal component eigenvalues greater than 1(15). Oblique promax rotation was performed to estimate factor correlations (rather than assume they were uncorrelated)(21). An oblique rotation often yields more interpretable factors with a simpler structure where factors are allowed to correlate than obtained with an orthogonal rotation where factors are uncorrelated.
All analyses were performed by using STATA software version 9.2 (College Station, Texas) and p<0.05 was considered statistically significant.
Results
Study Population
The participants were patients with SSc who were evaluated at 3 academic Scleroderma centers. The majority of participants were female (84.1%) and Caucasian (81.4%); 55.3% had diffuse SSc and their mean age was 50.9 years (Table 1). Mean (SD) SF-36 PCS score was 36.7 (9.3) and SF-36 MCS score was 47.1 (12.5). Self-rated severity of GIT ranged from no symptoms/ very mild (39%), to mild (21%), to moderate (31%), to severe/ very severe (9%). The majority of patients had a diagnosis of GERD (91%), followed by small intestinal bowel bacterial overgrowth, gastroparesis, and diarrhea (11% each). Patients with limited SSc had statistically higher (worse HRQoL) scores in the distention/bloating, constipation, social function, and emotional well-being scales compared to diffuse SSc (p< 0.05; data not shown).
Table 1. Baseline Characteristics of Study Participants.
| Variables | Total Sample (n=152) |
|---|---|
| Age, years | 50.9 (12.3) |
| Gender: Female, N (%) | 128 (84.1) |
| Race* | |
| Caucasian, N (%) | 118 (81.4) |
| African-American, N (%) | 11 (7.6) |
| Asian, N (%) | 9 (6.2) |
| Others, N (%) | 7 (4.8) |
| Ethnicity: Hispanic, N (%) | 26 (17.1) |
| Education | |
| Less than or equal to High school graduate, N (%) | 40 (26.3) |
| Some College, N (%) | 44 (29.0) |
| College graduate, N (%) | 37 (24.3) |
| Graduate degree, N (%) | 31 (20.4) |
| Type of SSc, N (%) | |
| Limited | 62 (40.7) |
| Diffuse | 81 (55.3) |
| Overlap | 6 (4.0) |
| Physician's GIT Diagnoses, % | |
| Gastro-esophageal reflux disease | 91 |
| Gastroparesis | 11 |
| Gastric antral venous ectasia | 4 |
| Small bowel bacterial overgrowth | 11 |
| Bowel pseudo-obstruction | 3 |
| Diarrhea | 11 |
| Rectal incontinence | 7 |
| Constipation | 5 |
| HRQOL | |
| SF-36 Physical Component Summary, mean (SD) | 36.7 (9.3) |
| SF-36 Mental Component Summary, mean (SD) | 47.1 (12.5) |
| Self-Rated Gastrointestinal Severity‡ | |
| No symptoms | 29 (19.1) |
| Very Mild | 30 (19.7) |
| Mild | 33 (21.7) |
| Moderate | 47 (30.9) |
| Severe | 9 (5.4) |
| Very Severe | 4 (2.6) |
Data not available for 7 patients;
Data not available for 3 patients VAS= Visual analog scale
Evaluation of the UCLA SCTC-GIT 2.0
Table 2 provides the descriptive statistics and reliability estimates for the UCLA SCTC-GIT 2.0. Multitrait analyses provided support for 7 separate scales: reflux, diarrhea, distention/ bloating, fecal soilage, constipation, emotional well-being, and social functioning (Appendix 1). Mean (SD) scores ranged from 0.26 (0.51) (possible score 0-3) for the for the social functioning scale to 1.07 (0.82) for distension/ bloating scale. Test-retest reliability ICC estimates for the 7 scales over a mean interval of 1.1 weeks ranged from 0.68 (fecal soilage scale) to 0.89 (constipation scale), and coefficient alpha was >0.70 for all scales except the constipation scale (alpha=0.67). Ceiling effects ranged from 11% for the reflux scale to 80% for the fecal soilage scale.
Table 2. Descriptive statistics and internal consistency reliability of UCLA-SCTC GIT 2.0.
| Scales | Sample Size |
Number of Items |
Mean Score |
Standard Deviation |
Median Score |
Minimum Score |
Maximum Score |
% with Floor Effect |
% with Ceiling Effect |
Cronbach's alpha |
ICC |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Reflux | 151 | 8 | 0.69 | 0.53 | 0.63 | 0 | 2.65 | 0.0 | 11.3 | 0.74 | 0.85 |
| Distention/Bloating | 151 | 4 | 1.07 | 0.82 | 1.00 | 0 | 3.00 | 2.6 | 13.8 | 0.77 | 0.83† |
| Diarrhea | 152 | 2 | 0.56 | 0.67 | 0.50 | 0 | 2.00 | 7.2 | 50.0 | 0.74 | 0.84‡ |
| Fecal soilage | 151 | 1 | 0.30 | 0.67 | 0.00 | 0 | 3.00 | 2.7 | 79.5 | NA | 0.68 |
| Constipation | 151 | 4 | 0.43 | 0.50 | 0.25 | 0 | 2.25 | 0.0 | 39.7 | 0.67 | 0.89 |
| Emotional well-being | 151 | 9 | 0.49 | 0.66 | 0.22 | 0 | 2.80 | 0.0 | 38.8 | 0.91 | 0.78 |
| Social functioning | 151 | 6 | 0.26 | 0.51 | 0.00 | 0 | 0.00 | 3.0 | 60.9 | 0.88 | 0.88 |
| Total Score* | 152 | 30 | 0.66 | 0.46 | 0.59 | 0 | 2.10 | 0.0 | 4.0 | 0.71 | 0.81 |
Total score is sum of 6 of 7 scales (excludes Constipation scale). Each scale is scored from 0 (better HRQOL) to 3 (worse HRQOL). ICC is
0.53 and 0.55
when 4 patients who developed diarrhea and constipation during test-retest period were included.
The social functioning and emotional well-being scales of UCLA SCTC-GIT 2.0 correlated (r=0.36 each; p<0.01 for both) with the social functioning and emotional well-being scales of the SF-36, respectively. Hypothesized, the self-rated GIT severity was variably correlated with the 7 scales and ranged from -0.21 for constipation scale to -0.50 for distention/bloating scale. The reflux scale had a higher correlation with upper GIT self-rate severity, distention/ bloating had a similar correlation with upper and lower self-rate severity, and diarrhea, fecal soilage, and constipation scales had a higher correlation with lower GIT self-rate severity (Table 3).
Table 3. Spearman correlation coefficient between 3 self-rated GIT severity scales vs. UCLA SCTC- GIT 2.0 scales.
| UCLA SCTC- GIT 2.0 | Self-rated Total GIT severity | Self-rated Upper GIT severity | Self-rated Lower GIT severity |
|---|---|---|---|
| Reflux | 0.47 | 0.56 | 0.31 |
| Distention/bloating | 0.50 | 0.48 | 0.47 |
| Diarrhea | 0.36 | 0.22 | 0.45 |
| Fecal soilage | 0.21 | 0.13* | 0.29 |
| Const | 0.20 | 0.28 | 0.30 |
| Social | 0.41 | 0.44 | 0.31 |
| Emotional | 0.43 | 0.38 | 0.55 |
| Total Score | 0.60 | 0.52 | 0.60 |
All coefficients are significant at p< 0.05 except *
The SSC-GIT scales discriminated between the self-rated severity of GIT (Table 4). All 7 scales showed the lowest scores (least involvement) in people with very mild to mild disease and the highest scores (worst disease) in participants with severe to very severe disease.
Table 4. Ability of the UCLA SCTC- GIT 2.0 to differentiate between the severity of self-rated GIT involvement.
| UCLA SCTC 2.0 | No symptoms to very Mild (n=59) | Mild (n=33) | Moderate (n=47) | Severe to very severe (n=13) | F test | P value |
|---|---|---|---|---|---|---|
| Mean Score | Mean Score | Mean Score | Mean Score | |||
| Reflux | 0.41† | 0.64‡¶ | 0.91# | 1.23 | 17 | <0.001 |
| Distention/Bloating | 0.59*† | 1.15¶ | 1.44 | 1.75 | 17 | <0.001 |
| Diarrhea | 0.30† | 0.67¶ | 0.65# | 1.19 | 8.8 | <0.001 |
| Fecal soilage | 0.15 | 0.27 | 0.36 | 0.77 | 3.3 | 0.02 |
| Constipation | 0.26 | 0.55 | 0.54 | 0.51 | 2.5 | 0.02 |
| Emotional well-being | 0.25† | 0.33‡¶ | 0.61# | 1.54 | 21 | <0.001 |
| Social functioning | 0.09† | 0.14‡¶ | 0.41# | 0.84 | 12 | <0.001 |
| Total Score | 0.36*† | 0.64‡¶ | 0.86# | 1.31 | 31 | <0.001 |
p< 0.05 between very mild vs. mild disease;
p< 0.05 between very mild and moderate disease;
p < 0.05 between very mild vs. severe disease;
p< 0.05 between mild vs. moderate disease;
p< 0.05 between mild and severe disease;
p< 0.05 moderate and severe disease. Each scale is scored from 0 (better HRQOL) to 3 (worse HRQOL).
In addition, symptom -specific scales were able to discriminate subjects with corresponding clinical GIT diagnosis (Table 5). For example, reflux scale was able to discriminate patients who had clinical diagnosis of GERD vs. those who didn't (p< 0.05).
Table 5. Comparison of mean scores of the UCLA SCTC GIT 2.0 scales in patient with and without clinical GIT diagnosis.
| UCLA SCTC- GIT 2.0 Scale | Clinical Diagnosis | Mean score of pts with diagnosis | Mean score of pts without diagnosis |
|---|---|---|---|
| Reflux | GERD | 0.72 | 0.38* |
| Distention/bloating | Gastroparesis | 1.63 | 1.00* |
| Distention/bloating | Bacterial overgrowth | 1.66 | 1.01* |
| Distention/bloating | Pseudo-obstruction | 2.25 | 1.04* |
| Diarrhea | Diarrhea | 1.00 | 0.51* |
| Fecal soilage | Rectal incontinence | 0.50 | 0.28 |
| Constipation | Constipation | 0.81 | 0.41* |
Each scale is scored from 0 (better HRQOL) to 3 (worse HRQOL).
P< 0.05; GERD: Gastro-esophageal reflux disease
Exploratory factor analysis of the 7 scales suggested 2 underlying factors with eigenvalues greater than 1, explaining over 34% of the variation in the components (Table 6). Four of the 7 scales had noteworthy loadings on the first factor and the second factor was defined primarily by reflux, distention/bloating, and constipation scales. The estimated correlation between the factors was -0.55. As previously observed(5), the underlying factors did not represent anatomic upper and lower GIT.
Table 6. Oblique Promax Rotated Two-Factor Pattern Matrix for UCLA SCTC- GIT 2.0.
| UCLA SCTC- GIT 2.0 Scale | Factor 1 | Factor 2 |
|---|---|---|
| Reflux | 0.24 | 0.42 |
| Distention/Bloating | 0.25 | 0.44 |
| Diarrhea | 0.74 | − 0.18 |
| Fecal soilage | 0.58 | − 0.11 |
| Constipation | − 0.30 | 0.58 |
| Emotional well-being | 0.71 | 0.11 |
| Social functioning | 0.58 | 0.29 |
Because there was a negative correlation between diarrhea and constipation scales (-0.06), and a higher correlation between the diarrhea scale vs. the other 5 remaining scales and self-rated severity of GIT when compared to constipation scale, we created a Total GIT score excluding the constipation scale. The Total score is the average of the remaining 6 scales and is scored from 0 (better HRQoL) to 3 (worse HRQoL).
Total GIT Score= average (Reflux scale+ Distention/ Bloating scale+ Diarrhea scale+ Fecal soilage scale+ Emotional well-being scale+ Social functioning scale)
The mean Total GIT Score was 0.66 (0.46) with no floor and only small ceiling effect (4%). The test-retest reliability ICC estimate was 0.81 and coefficient alpha was 0.71. The Total GIT Score had large correlation coefficients with self-rated GIT severity (0.60), upper (0.52) and lower GIT severity (0.60). Total GIT Score was able to discriminate among each of the 4 self-rated severity groups — no symptoms/ very mild, mild, moderate, and severe/ very severe (F test 31.3, p< 0.001). The F-ratio was the largest for the total GIT score, indicating that it was the most sensitive measure to self-rated GIT disease (Table 4). However, symptom -specific scales were more discriminatory in separating subjects with corresponding clinical GIT diagnosis compared to the Total GIT Score (Table 5).
Discussion
Gastrointestinal tract involvement has a major impact on the HRQOL of SSc patients (3;4). In this study, we refined our previously developed SSC-GIT 1.0. The results of this study provide support for the reliability and validity of the UCLA-SCTC GIT 2.0. The UCLA-SCTC GIT 2.0 captures SSc-related GIT activity and severity and can be used both in clinical trials and day-to-day care.
The reliability of the UCLA-SCTC GIT 2.0, as assessed by test-retest and internal consistency was found to be satisfactory (correlation coefficient ≥ 0.67 for both). The UCLA-SCTC GIT 2.0 has 7 scales— reflux, distention/ bloating, diarrhea, fecal soilage, constipation, emotional well-being, and social functioning.
We made 5 refinements to our previous instrument, the SSC-GIT 1.0. First, we have decreased the number of items from 52 items to 34 items without sacrificing the reliability of the instrument. This should decrease patient burden especially if UCLA-SCTC GIT 2.0 is administered in combination with other patient reported measures.
Second, SSC-GIT 1.0 had one reflux/ indigestion scale. After the development of SSC-GIT 1.0, it was realized that a combined reflux/ indigestion scale might not be able to differentiate diagnoses of GERD, gastroparesis or bacterial overgrowth. Therefore, we decided to separate the scale into reflux and distention/ bloating (more appropriate wording than ‘indigestion’). The distention/ bloating scale is psychometrically sound and is able to differentiate patients who had the diagnosis of bacterial overgrowth syndrome and gastroparesis vs. those who did not (Table 5). Third, we have added a 1-item fecal soilage scale to capture rectal incontinence, as up to one-third of SSc patients complain about soiling (6) and it has negative impact on HRQOL(17). Fourth, we have developed a composite ‘Total GIT Score’ to capture overall burden of SS-associated GIT. The Total GIT Score was found to reliable and valid and more discriminatory in differentiating among self-rated GIT disease than individual scales. Last, we have simplified the scoring algorithm to improve its feasibility in clinical care; the instrument can be scored on a 0-3 scale and does not require conversion to a 0-100 scale.
Similar to SSC-GIT 1.0, we noticed a high ceiling effect ranging from 11.0% in reflux scale to 80% in the fecal soilage scale with UCLA-SCTC GIT 2.0. This ceiling effect may be due to relatively mild-to-moderate disease recruited in our study, reflect variation in symptoms and anatomic GIT involvement from patient to patient(18;19). For example, only 13 (8%) of patients rated their GIT disease as ‘severe’ or ‘very severe.’ However the Total GIT Score was associated with less ceiling effect (4%).
The 7 scales and Total GIT Score showed both convergent and divergent validity; the self-rated GIT severity correlated with the 7 scales and Total GIT score and ranged from 0.21 for the constipation scale to 0.60 for the Total GIT Score. In addition, 7 scales and the Total GIT Score were able to discriminate between the self-rated severity of GIT. All 7 scales and the Total GIT Score showed the lowest scores (indicating better GIT health) in people with very mild/mild disease and the highest scores (indicating worse GIT health) in participants with severe/ very severe disease. The Total GIT Score had the highest F statistics suggesting that it has greater discriminatory ability than individual scales. We suggest that the Total GIT Score be reported to assess overall burden/ severity of GIT involvement. This can be used in both clinical practice and trials.
On the other hand, individual scales were able to discriminate between individual corresponding clinical diagnoses; scores for a specific scale (e.g. reflux scale) were statistically lower in patients with specific clinical diagnosis (e.g. GERD) compared to patients who did not have the diagnosis. Therefore, the scores of each scale can be used separately in clinical care and as outcome measures in RCTs as an activity index. As an outcome measure, we recommend to present separate scores for each 7 scales as it was more discriminatory for clinical GIT diagnoses and likely to change over time or with specific treatment.
Our study has significant strengths. First, this was a prospective, 3-center study with the primary objective to refine the previously developed GIT instrument. Second, we show that symptom -specific scales are able to discriminate subjects with corresponding clinical GIT diagnoses. These characteristics are sufficiently robust for both clinical trials and clinical care. Third, we show that score is useful to assess overall burden/ severity of GIT involvement, Total GIT Score is more useful.
Our study is not without limitations. We didn't assess UCLA-SCTC GIT 2.0 vs. radiological tests measures such as gastro-esophageal endoscopy, esophageal manometry and breath test because the objective of the current study was to refine the previously developed instrument and due to resource constraints. Also, and we didn't assess symptoms of xerostomia. Second,,the Total GIT Score as an unweighted additive score rather than giving different weights to scales. Nevertheless, the Total GIT Score showed no floor and very low ceiling effects compared to individual scales and a greater discriminatory ability than individual scales. One of our goals was to develop a measure that can be easily scored in clinical practice so an unweighted composite score is preferable. Further, other unweighted composite measures in rheumatology have performed well in observational and clinical trials(20-22).
In conclusion, we report the refinement of the previously developed SSC-GIT 1.0 in SSc. Current analysis is ongoing to examine its responsiveness to change and minimally important differences for the UCLA-SCTC GIT 2.0.
Acknowledgments
We acknowledge the support of the clinical coordinators and participants at the 3 US Scleroderma Centers.
The development of the questionnaire was supported by a grant from the Scleroderma Clinical Trial Consortium, the International Scleroderma Network, and unrestricted funds by the Pettit family to UCLA Scleroderma Program and by the Jonathan and Lisa Rye Scleroderma Research Fund at the University of Michigan. Dr. Khanna was supported by a National Institutes of Health Award (NIAMS K23 AR053858-01A1) and the Scleroderma Foundation (New Investigator Award).
Reference List
- 1.Sjogren RW. Gastrointestinal motility disorders in scleroderma. Arthritis Rheum. 1994;37(9):1265–1282. doi: 10.1002/art.1780370902. [DOI] [PubMed] [Google Scholar]
- 2.Lock G, Holstege A, Lang B, Scholmerich J. Gastrointestinal manifestations of progressive systemic sclerosis. Am J Gastroenterol. 1997;92(5):763–771. [PubMed] [Google Scholar]
- 3.Nietert PJ, Mitchell HC, Bolster MB, Curran MY, Tilley BC, Silver RM. Correlates of depression, including overall and gastrointestinal functional status, among patients with systemic sclerosis. J Rheumatol. 2005;32(1):51–57. [PubMed] [Google Scholar]
- 4.Gliddon AE, Dore CJ, Maddison PJ. Influence of clinical features on the health status of patients with limited cutaneous systemic sclerosis. Arthritis Rheum. 2006;55(3):473–479. doi: 10.1002/art.21999. [DOI] [PubMed] [Google Scholar]
- 5.Khanna D, Hays RD, Park GS, Braun-Moscovici Y, Mayes MD, McNearney TA, et al. Development of a preliminary scleroderma gastrointestinal tract 1.0 quality of life instrument. Arthritis Rheum. 2007;57(7):1280–1286. doi: 10.1002/art.22987. [DOI] [PubMed] [Google Scholar]
- 6.Trezza M, Krogh K, Egekvist H, Bjerring P, Laurberg S. Bowel problems in patients with systemic sclerosis. Scand J Gastroenterol. 1999;34(4):409–413. doi: 10.1080/003655299750026434. [DOI] [PubMed] [Google Scholar]
- 7.Guidance for Industry: patient-reported outcome measures: use in medical product development to support labeling claims. 2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ware J, Kosinski M, Dewey J. How to Score Version Two of the SF-36 Health Survey. Lincoln, RI: QualityMetric Incorporated; 2000. [Google Scholar]
- 9.Ware JE, Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30(6):473–483. [PubMed] [Google Scholar]
- 10.Hadorn DC, Hays RD. Multitrait-multimethod analysis of health-related quality-of-life measures. Med Care. 1991;29(9):829–840. doi: 10.1097/00005650-199109000-00004. [DOI] [PubMed] [Google Scholar]
- 11.Cronbach L. Coefficient alpha and the internal structure of tests. Psychometrica. 1951;16:297–334. [Google Scholar]
- 12.Hays RD. Reliability and validity (including responsiveness) In: Fayers P, Hays RD, editors. Assessing quality of life in clinical trials. second. New York: Oxford; 2005. pp. 25–39. [Google Scholar]
- 13.Cohen J, Cohen P, West SG, Aiken LS. Applied multiple regression/correlation analysis for the behavioral sciences. 2nd. Hillsdale, NJ: Lawrence Erlbaum Associates; 2003. [Google Scholar]
- 14.Joliffe IT, Morgan BJ. Principal component analysis and exploratory factor analysis. Stat Methods Med Res. 1992;1(1):69–95. doi: 10.1177/096228029200100105. [DOI] [PubMed] [Google Scholar]
- 15.Shaw M, Talley NJ, Adlis S, Beebe T, Tomshine P, Healey M. Development of a digestive health status instrument: tests of scaling assumptions, structure and reliability in a primary care population. Aliment Pharmacol Ther. 1998;12(11):1067–1078. doi: 10.1046/j.1365-2036.1998.00399.x. [DOI] [PubMed] [Google Scholar]
- 16.Floyd F, Widaman K. Factor Analysis in the Development and Refinement of Clinical Assessment Instruments. Psychological Assessment. 1995;7(3):286–299. [Google Scholar]
- 17.Kenefick NJ, Vaizey CJ, Nicholls RJ, Cohen R, Kamm MA. Sacral nerve stimulation for faecal incontinence due to systemic sclerosis. Gut. 2002;51(6):881–883. doi: 10.1136/gut.51.6.881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sjogren RW. Gastrointestinal features of scleroderma. Curr Opin Rheumatol. 1996;8(6):569–575. doi: 10.1097/00002281-199611000-00012. [DOI] [PubMed] [Google Scholar]
- 19.Khanna D. Gastrointestinal involvement in systemic sclerosis. In: Font J, Ramos-Casals M, Rodes J, editors. Digestive involvement in systemic autoimmune diseases. Elsevier; 2008. pp. 51–69. [Google Scholar]
- 20.Fries JF, Spitz P, Kraines RG, Holman HR. Measurement of patient outcome in arthritis. Arthritis Rheum. 1980;23(2):137–145. doi: 10.1002/art.1780230202. [DOI] [PubMed] [Google Scholar]
- 21.Clements P, Lachenbruch P, Siebold J, White B, Weiner S, Martin R, et al. Inter and intraobserver variability of total skin thickness score (modified Rodnan TSS) in systemic sclerosis. J Rheumatol. 1995;22(7):1281–1285. [PubMed] [Google Scholar]
- 22.Aletaha D, Smolen J. The Simplified Disease Activity Index (SDAI) and the Clinical Disease Activity Index (CDAI): a review of their usefulness and validity in rheumatoid arthritis. Clin Exp Rheumatol. 2005;23(5 Suppl 39):S100–S108. [PubMed] [Google Scholar]