Validation of a 7-Point Global Overall Symptom Scale to Measure the Severity of Dyspepsia Symptoms in Clinical Trials

S. J. O. Veldhuyzen Van Zanten; N. Chiba; D. Armstrong; A. N. Barkun; A. B. R. Thomson; V. Mann; , S. Escobedo; B. Chakraborty; K. Nevin

Aliment Pharmacol Ther. 2006;23(4):521-529. 

Summary and Introduction

Summary

Background: Currently there is no consensus on the optimal method to measure the severity of dyspepsia symptoms in clinical trials.
Aim: To validate the 7-point Global Overall Symptom scale.
Methods: The Global Overall Symptom scale uses a 7-point Likert scale ranging from 1 = no problem to 7 = a very severe problem. Validation was performed in two randomized-controlled trials (n = 1121 and 512). Construct validity: Global Overall Symptom was compared with the Quality of Life in Reflux And Dyspepsia, Gastrointestinal Symptom Rating Scale, Reflux Disease Questionnaire and 10 specific symptoms using Spearman correlation coefficients. Test–retest reliability: The Intraclass Correlation Coefficient was calculated for patients with stable dyspepsia defined by no change in Overall Treatment Effect score over two visits. Responsiveness: effect size and standardized response mean were also calculated.
Results: Construct validity: Change in Global Overall Symptom score correlated significantly with Quality of Life for Reflux And Dyspepsia, Gastrointestinal Symptom Rating Scale, Reflux Disease Questionnaire and specific symptoms (all P < 0.0002). Reliability: The Intraclass Correlation Coefficient was 0.62 (n = 205) and 0.42 (n = 270). Responsiveness: There was a positive correlation between change in Global Overall Symptom and change in symptom severity. The effect size and standardized response mean were 1.1 and 2.1, respectively.
Conclusion The Global Overall Symptom scale is a simple, valid outcome measure for dyspepsia treatment trials.

Introduction

There has been much discussion on how to most appropriately measure severity of dyspepsia symptoms in clinical trials.[1-3] Dyspepsia is a term describing a symptom complex thought to originate in the upper gastrointestinal (GI) tract. Epigastric pain or discomfort are the cardinal symptoms and may be associated with other symptoms such as heartburn, acid regurgitation, excessive burping and belching, upper abdominal bloating, early satiety and nausea.[4, 5] There is disagreement as to whether heartburn and acid regurgitation are symptoms within the symptom complex of dyspepsia.[4, 5] Clinical trial patients presenting with dyspepsia who do not undergo endoscopy are classified as having uninvestigated dyspepsia. A diagnosis of functional (non-ulcer) dyspepsia is made when endoscopy does not detect any structural abnormalities that may explain the dyspepsia symptoms.

A methodological challenge in clinical trials of patients with uninvestigated dyspepsia or functional dyspepsia is the lack of objective measures, such as an ulcer or oesophagitis, to which assessments of symptom severity can be correlated. In such patients one must rely on the subjective reporting of symptoms and the impact that these symptoms have on quality of life to decide whether a treatment is providing a benefit. A 4- or 5-point ordinal (Likert) scale has been frequently used to measure severity of individual dyspepsia symptoms,[6] but 7-point scales have also been utilized.[7, 8] The advantage of a 7-point scale over a 4- or 5-point scale is the ability to detect smaller differences.[1, 9, 10] Here, we report the validation of a 7-point Likert scale to measure overall severity of dyspepsia symptoms by documenting construct validity, test–retest reliability and responsiveness (ability to detect change), to determine the validity of the scale for use as an outcome measure in clinical trials and practice.

Methods

The Global Overall Symptom (GOS) scale is a 7-point scale, adapted from a previously validated 5-point scale.[11] The GOS score is self-reported by the patient after the context of the scale, contained in , is either read to them, or alternatively, it is read by the patients themselves. Patients are asked to grade the overall severity of their dyspepsia symptoms defined as upper abdominal symptoms (located centrally between the breastbone and umbilicus) over a certain retrospective period of time, either 28 days (GOS28) or 2 days (GOS2).

Table 1.  The Global Overall Symptom (GOS) scale

Please take a few moments to think about the symptoms you have had due to your overall stomach problems in the (specified time period). It is very important for us to get this information. We will be using it as one of the ways to find out if the treatment is helping your stomach problems throughout the study. Overall, how have your stomach problems been over the (specific time period)? According to the scale below, please indicate the severity of your overall symptoms over the (specified time period)
1. No problem
2. Minimal problem (can be easily ignored without effort)
3. Mild problem (can be ignored with effort)
4. Moderate problem (cannot be ignored but does not influence my daily activities)
5. Moderately severe problem (cannot be ignored and occasionally limits my daily activities)
6. Severe problem (cannot be ignored and often limits my concentration on daily activities)
7. Very severe problem (cannot be ignored and markedly limits my daily activities and often requires rest)

Patients were also asked to rate the severity of 10 specific upper GI symptoms [specific symptom subtypes (SSS)] using the same 7-point Likert scale as for the GOS scale: epigastric pain (pain in the middle of the upper abdomen just below the breastbone), epigastric discomfort (epigastric pain and discomfort are both assessed as some patients perceive these differently), heartburn (a burning sensation felt under the lower part of the centre of the chest which rises towards or into the neck), acid regurgitation (a backward flow of sour or bitter fluid from the stomach into the food pipe), upper abdominal bloating, excessive belching, nausea, early satiety (a feeling the stomach is full soon after starting to eat), postprandial fullness (an unpleasant sensation of persistent fullness in the stomach after a meal) and other epigastric symptoms.

Validity Tools

The following outcome measures were used to assess the validity of the GOS: the Overall Treatment Effect (OTE),[7] the Quality of Life in Reflux And Dyspepsia (QoLRAD) questionnaire,[12] the Gastrointestinal Symptom Rating Scale (GSRS)[13] and the Reflux Disease Questionnaire (RDQ).[14]

The OTE is a validated scale that rates the change in symptoms on a 15-point scale (−7 to −1 = worse; 0 = no change and 1 to 7 = better).[7] It then categorizes the change with a second question asking how patients perceived the importance of the change on a 7-point scale from: (1) not important, (2) slightly important, (3) somewhat important, (4) moderately important, (5) important, (6) very important, (7) extremely important.

The QoLRAD is a validated disease-specific questionnaire that measures quality of life in patients with dyspepsia and gastro-oesophageal reflux disease (GERD).[12] It consists of 25 questions covering five domains: emotional distress, sleep disturbance, food/drink problems, physical/social functioning and vitality. This instrument uses 7-point Likert scales with higher scores indicating better QoL. Results are reported as a mean item score for each of the five dimensions.

The GSRS questionnaire is a validated, self-administered questionnaire that includes 15 questions, which assess severity of GI symptoms using a 7-point Likert scale in five domains: indigestion, diarrhoea, constipation, abdominal pain and reflux.[13] The severity of symptoms reported in the GSRS increases with increasing score.

The RDQ is a self-administered questionnaire, in which subjects are asked to report on the frequency and severity of the following symptoms − heartburn, regurgitation, epigastric pain and epigastric burning, over the past 4 weeks.[14] Symptom frequency is ranked according to six choices (did not have, <1 day a week, 1 day a week, 2−3 days a week, 4−6 days a week, daily), as is symptom severity (did not have, very mild, mild, moderate, moderately severe, severe).

Patient Population

Validation of the GOS was tested within the Confirmatory Acid Suppression Test (CAST) study[15] and the Canadian Adult Dyspepsia Empirical Treatment − Heartburn Negative (CADET-HN) study.[16] In both studies, the Canadian Dyspepsia (CanDys) definition of dyspepsia was used to categorize patients.[5] This is a broad definition of dyspepsia, which includes the symptoms of heartburn and regurgitation.

CAST Study. In the CAST study, 1121 patients with moderate to severe dyspepsia (GOS: ≥4) were recruited at 93 Canadian family practitioner centres.[15] The key inclusion criterion required patients present to their family doctors with uninvestigated dyspepsia of at least 1-month duration. The severity of dyspepsia in the 2-days prior to enrolment had to be of at least moderate severity (GOS score: ≥4). Use of proton-pump inhibitors, prokinetic agents, antibiotics or bismuth compounds was not permitted 4 weeks prior and use of H2-receptor antagonists not permitted 2 weeks prior to initial study evaluation.

Eligible patients were randomized to receive esomeprazole, either 40 mg once (in the morning with matching placebo tablets in the evening) or twice daily for 7 days. Following this, all subjects received open-label 40 mg esomeprazole tablets once daily for 3 weeks. The main objective of the study was to assess the positive and negative predictive values of a 7-day treatment with esomeprazole 40 mg once or twice daily used as a diagnostic test to determine whether symptoms were acid-related. This was assessed by comparing the 7-day therapy results to those after 4 weeks of treatment.

CADET-HN Study. In the CADET-HN study, 512 patients with moderate to severe uninvestigated dyspepsia (GOS score: ≥4) were recruited at 35 Canadian family practitioner centres.[16] Patients with heartburn and/or regurgitation alone were considered to have a diagnosis of GERD and were excluded. Patients had to be negative for Helicobacter pylori infection. The severity of dyspepsia in the month prior to enrolment had to be of at least moderate severity (GOS score: ≥4).

Eligible patients were randomized to one of four treatments for 4 weeks: omeprazole 20 mg daily, ranitidine 150 mg twice daily, cisapride 20 mg twice daily or placebo.

The main objectives of the study were to measure the proportion of patients that had minimal symptoms after 1 month of daily treatment and then after 5 months of on-demand treatment (i.e. based on symptoms) with the same medication.

Data Analysis

Both studies used the GOS score as their main outcome measure. To be counted as a treatment success (patient responder, R) the patient's GOS score had to decrease to either no or minimal severity, i.e. a GOS score of either 1 or 2 after 4 weeks of treatment. For non-responders (NR), the GOS score remained >2 after 4 weeks treatment. In the CAST study patients were asked to rate the severity of their symptoms over the previous 2 days (GOS2), and in the CADET-HN study, over the last 28 days (GOS28). The reason was differences in study objectives and study duration.

Test−Retest Reliability

Test−retest reliability was measured by calculating the Intraclass Correlation Coefficient (ICC) between the GOS scores reported at two separate visits for patients with stable symptoms. A patient was considered stable if there was no change in their OTE score (−1, 0, +1) between two specified visits. In the CAST study, the ICC was calculated between 21 and 28 days of treatment and for the CADET-HN study, between 28 and 84 days of treatment. The ICC ranges from 0, no correlation to 1.0 perfect correlation.

Responsiveness

Responsiveness implies the ability of the instrument to detect important differences over time. This property is also referred to as the sensitivity of the instrument to change and is a critical property of evaluative instruments.[9] Two distinct methods have been used to estimate the response to change. The first method uses anova, a technique used to estimate sources of variation in a data set.[17] The second method assesses effect size (ES), which estimates the change in GOS relative to the variability in the GOS at baseline, and is calculated by dividing the change in GOS by the standard deviation at baseline (ES = ΔGOS/s.d. at baseline).[18] A scale with an ES of 0.8 or larger has been defined as being very responsive to change.[19] As the variability between patients is, in general, larger than the variability within individual patients, the standardized response mean (SRM) has also been proposed as a measure of responsiveness and is calculated by dividing the change in GOS by the standard deviation of the change (SRM = ΔGOS/s.d. of the change).[9, 18]

Additionally, to determine responsiveness, the change in GOS scores were compared descriptively with OTE scores for change and perceived importance of that change at 4 weeks in each study. The OTE scores for change were aggregated as no (0, ±1), small (±2, ±3), moderate (±4, ±5) and large (±6, ±7). For the importance rating of change, the response categories were aggregated as: none (not important), slight (slightly important, somewhat important), important (moderately important, important) and very (very important and extremely important).

Construct Validity

Construct validity was determined, using Spearman correlation coefficients, by correlating changes in the GOS to changes in severity of individual and the mean score for SSS, QoLRAD overall score and dimensions in both studies; RDQ overall score and dimensions in the CAST study and GSRS overall score and dimensions in the CADET-HN study. We considered that correlations >0.4 would be clinically relevant. We also predicted that correlations would be higher for the abdominal pain and reflux domains than constipation and diarrhoea domains.

Results

The results of the CAST study have been reported elsewhere.[15] In total 1121 patients were randomized and included in the intention-to-treat (ITT) analysis. The mean age of the enrolled patients was 46 years (range: 18–90) and 50% were male. The proportion of responders (R = GOS: ≤2) at 4 weeks for the esomeprazole 40 mg once daily group was 66% (367 of 556) and for the twice daily group 64% (362 of 565).

The results of the CADET-HN study have been reported elsewhere.[16] In total 512 patients were randomized and included in the ITT analysis. The mean age of enrolled patients was 40 years (range: 18−78) and 47% were male. The average duration of dyspepsia symptoms varied from 8 to 10 years.

The proportion of responders (R = GOS: ≤2) at 4 weeks for each treatment group was omeprazole 51% (69 of 135), ranitidine 36% (50 of 139), cisapride 31% (32 of 105) and placebo 23% (31 of 133). The differences between omeprazole and all other three treatments were statistically significant (P = 0.012, 0.001 and <0.001 respectively), as was the difference between ranitidine and placebo (P = 0.023).

Test−Retest Reliability

The ICC for stable patients [unchanged OTE score (−1, 0, +1) over two visits] were calculated as 0.62 (n = 205) for GOS2 and 0.42 (n = 270) for GOS28.

Responsiveness

The mean change in GOS scores for each patient group, as defined by their OTE score, is presented in . For the change rating for both GOS2 and GOS28, a trend is evident displaying how the mean change in GOS score matches changes in symptoms reported by patients on the OTE scale. Further, responders (R) when compared with NR reported a progressively larger magnitude of change in their symptoms on the OTE scale, which correlated with a similar progressive change in their mean GOS score. The magnitude of change was much larger for patients who were responders than NRs. The data for the importance rating showed similar results as for the change rating: those patients that reported greater importance in their change in symptoms (using the OTE scale) had a greater change in the GOS score, with responders placing greater importance on the change in GOS score than NRs.

Table 2.  Responsiveness: the mean change in GOS score (s.d.) for both change and importance rating for all enrolled patients who were R and NR

Patients
GOS2 data – CAST study
Change None (74) Small (105) Moderate (225) Large (601)
  All (995) 0.89 (1.2) 1.4 (1.1) 2.2 (1.2) 3.3 (1.1)
  R (717) 2.4 (0.5) 2.4 (0.9) 2.9 (0.9) 3.4 (1.0)
  NR (288) 0.5 (1.0) 1.1 (0.9) 1.4 (1.1) 1.3 (1.1)
Importance None (17) Slight (71) Important (181) Very (631)
  All (900) 2.6 (1.6) 1.8 (1.3) 2.3 (1.4) 3.1 (1.2)
  R (678) 3.1 (1.1) 2.8 (1.1) 3.1 (1.0) 3.4 (1.0)
  NR (222) 0.0 (0.0) 1.1 (0.9) 1.1 (1.0) 1.5 (1.1)
GOS28 data – CADET-HN study
Change None (129) Small (101) Moderate (140) Large (112)
  All (482) 0.71 (1.1) 1.2 (1.2) 1.5 (1.3) 2.7 (1.2)
  R (181) 2.6 (0.76) 2.6 (0.86) 2.6 (0.72) 2.9 (0.87)
  NR (301) 0.49 (0.87) 0.82 (0.93) 0.97 (1.2) 1.3 (1.5)
Importance None (19) Slight (119) Important (74) Very (138)
  All (350) 1.21 (1.23) 1.4 (1.3) 1.8 (1.3) 2.2 (1.4)
  R (159) 2.5 (0.84) 2.6 (0.77) 2.8 (0.94) 2.9 (0.83)
  NR (191) 0.62 (0.87) 0.80 (0.88) 0.90 (1.1) 1.0 (1.3)

 

For all values, reductions in GOS score means improvement in symptoms.
Values are expressed as mean (s.d.).
All: all included patients; R, responders, defined as GOS ≤ 2 after treatment; NR, non-responders, defined as GOS > 2 after treatment; GOS, Global Overall Symptom; CAST, Confirmatory Acid Suppression Test; CADET-HN, Canadian Adult Dyspepsia Empirical Treatment – Heartburn Negative.

The results of the ES and SRM analysis are reported in . Both show large changes for responders after treatment indicating the GOS is responsive.

Table 3.  Responsiveness: the effect size and standardized response mean values for responsiveness

Patients (nfor CAST/ CADET-HN) GOS2 data (CAST study) GOS28 data (CADET-HN study)
Effect size Standard response, mean Effect size Standard response, mean
All (931/491) 3.0 1.9 1.1 2.1
R (700/182) 3.8 3.3 3.3 3.7
NR (231/309) 1.2 1.0 0.69 0.86

 

All values were reductions meaning improvement in symptoms.
All, all patients; R, responders; NR, non-responders; effect size, ΔGOS/s.d. at baseline; standardized response mean, ΔGOS/s.d. of the change; CAST, Confirmatory Acid Suppression Test; CADET-HN, Canadian Adult Dyspepsia Empirical Treatment - Heartburn Negative; GOS, Global Overall Symptom.

Construct Validity

The GOS results for responders and NRs were compared with QoLRAD, SSS and RDQ (CAST study only) or GSRS (CADET-HN study only). As presented in  , there is a trend where a worsening GOS score (increasing score from 1 to 7) is associated with a worsening QoLRAD scores (decreasing mean overall scores) and worsening GSRS, RDQ and SSS scores (increasing mean overall scores). This assessment is supported by moderate to high Spearman correlation coefficient values (0.41−0.80) achieved for each of the outcome measures displaying good correlation with the GOS scale.

Table 4.  Construct validity: correlation of GOS scores with other outcome measures

  GOS score (n) r
Responders (722) Non-responders (295)
1 (481) 2 (241) 3 (148) 4 (102) 5 (31) 6 (11) 7 (3)
GOS2 data - CAST study
  QoLRAD 6.6 5.9 5.1 4.7 4.0 4.1 4.0 −0.69
  SSS 1.1 1.6 2.2 2.5 3.2 3.0 4.2 0.80
  RDQ 3.8 9.6 16.5 19.3 24.4 18.8 25.7 0.64
  GOS score (n)  
Responders (182) Non-responders (308)  
1 (60) 2 (122) 3 (112) 4 (124) 5 (57) 6 (12) 7 (3)  
GOS28 - CADET-HN study
  QoLRAD 6.8 6.3 5.9 5.4 4.6 4.3 3.1 −0.62
  SSS 1.3 1.6 2.0 2.5 3.0 3.5 5.4 0.41
  GSRS 1.5 1.9 2.2 2.6 3.1 2.8 3.4 0.54

 

Mean overall questionnaire scores (QoLRAD, GSRS and RDQ) and overall mean of individuals symptoms (SSS) are presented, followed by Spearman correlation coefficients (r).
Responder, GOS ≤ 2; non-responder, GOS > 2.
GOS, Global Overall Symptom; CAST, Confirmatory Acid Suppression Test; CADET-HN, Canadian Adult Dyspepsia Empirical Treatment - Heartburn Negative; QoLRAD, Quality of Life in Reflux And Dyspepsia; SSS, specific symptom subtype; RDQ, Reflux Disease Questionnaire; GSRS, Gastrointestinal Symptom Rating Scale.

The correlation coefficients for each domain of the outcome measures used in this study, the QoLRAD, GSRS, RDQ and for each symptom of the SSS were also calculated. As can be seen in , correlation coefficients varied from 0.26 to 0.72. Correlations for the domains indigestion 0.47, reflux 0.54 and abdominal pain 0.56 of the GSRS were similar to the overall mean of 0.54 and higher than for diarrhoea 0.26 and constipation 0.25. Correlations for all the RDQ domain scores, heartburn 0.47, regurgitation 0.46, GERD 0.55, epigastric pain 0.59 were somewhat lower than the overall mean correlation 0.64. For QoLRAD, all correlations were >0.4 as were most of the specific symptoms, except early satiety. Correlations for the QoLRAD domains taken from the CADET-HN study (GOS28) ranged from 0.47 for sleep disturbance to 0.62 for food/drink problems and vitality compared with overall score 0.62; and from the CAST study (GOS2) ranged from 0.58 for physical/social functioning to 0.66 for food/drink problems and vitality compared with overall score 0.69. The correlation for the SSS symptoms was highest for epigastric pain, 0.65 (GOS28) and 0.63 (GOS2) and epigastric discomfort, 0.68 (GOS28) and 0.73 (GOS2). All correlations were in the expected directions.

Table 5.  Construct validity: Spearman correlation coefficient showing correlation between changes in GOS with changes in individual domains of other outcome measures for the two studies: QoLRAD, RDQ, GSRS and individual specific symptoms

Outcome measure: specific domain GOS2 GOS28
n r n r
QoLRAD: emotional distress 1011 −0.65 489 −0.60
QoLRAD: sleep disturbance 1010 −0.59 489 −0.47
QoLRAD: food/drink problem 1010 −0.66 489 −0.62
QoLRAD: physical/social 1010 −0.58 489 −0.49
QoLRAD: vitality 1010 −0.66 489 −0.61
RDQ: heartburn 1011 0.47 - -
RDQ: regurgitation 1011 0.46 - -
RDQ: epigastric pain 1011 0.59 - -
RDQ: GERD 1011 0.55 - -
GSRS: diarrhoea - - 488 0.26
GSRS: indigestion - - 488 0.47
GSRS: constipation - - 488 0.25
GSRS: abdominal pain - - 487 0.56
GSRS: reflux - - 488 0.54
SSS: epigastric pain 1015 0.63 489 0.65
SSS: epigastric discomfort 1015 0.68 489 0.72
SSS: heartburn 1015 0.48 489 0.47
SSS: regurgitation 1015 0.45 489 0.40
SSS: bloating 1015 0.61 489 0.40
SSS: belching 1015 0.50 489 0.48
SSS: nausea 1015 0.45 489 0.41
SSS: early satiety 1015 0.50 489 0.30
SSS: postprandial symptoms 1015 0.57 489 0.41

 

GOS, Global Overall Symptom; GERD, gastro-oesophageal reflux disease; QoLRAD, Quality of Life in Reflux And Dyspepsia; SSS, specific symptom subtype; RDQ, Reflux Disease Questionnaire; GSRS, Gastrointestinal Symptom Rating Scale.

Discussion

Currently there is no consensus on the optimal method to measure the severity of dyspepsia symptoms in clinical trials of either uninvestigated or functional (non-ulcer) dyspepsia patients.[1-3] The 1999 Rome Working Party on Design of Clinical Trials recommended use of a global outcome measure, which integrates the heterogeneous symptoms that form dyspepsia or alternatively a validated quality of life questionnaire.[1] This report also discusses the advantages and disadvantages of 5- vs. 7-point scales, the main advantage of a 7-point scale being the ability to detect smaller differences. The GOS was developed as a valid overall outcome measure to assess responsiveness (change over time) as a result of therapeutic interventions. Similar 7-point scales have been used as the main outcome measure to evaluate the severity of epigastric pain in other functional dyspepsia studies.[7, 8] A 7-point scale has also been validated for use in diary cards in dyspepsia trials.[20]

Our results demonstrate that assessment of the severity of dyspepsia symptoms using the 7-point GOS score is a valid outcome measure. The face validity of this outcome measure is high and the improvement in the GOS correlates highly with other outcome measures including severity of individual symptoms (SSS), OTE and validated quality of life outcome measures such as QoLRAD, GSRS and RDQ. It should be noted that GSRS is not a true quality of life instrument as it mainly assesses GI symptom severity in several domains. We considered correlations ≥0.40 important, and all domains of QoLRAD and RDQ showed higher correlations, with individual symptoms (SSS) displaying high correlations ( ). For the GSRS, we expected the correlations for the abdominal pain and reflux domains to be higher as these symptoms are more pertinent to dyspepsia, than the constipation and diarrhoea domains, which are not. The likely explanation for the higher correlations of GOS2 than GOS28 is the shorter time frame over which symptoms are assessed and longer time in between assessments.

Table 5.  Construct validity: Spearman correlation coefficient showing correlation between changes in GOS with changes in individual domains of other outcome measures for the two studies: QoLRAD, RDQ, GSRS and individual specific symptoms

Outcome measure: specific domain GOS2 GOS28
n r n r
QoLRAD: emotional distress 1011 −0.65 489 −0.60
QoLRAD: sleep disturbance 1010 −0.59 489 −0.47
QoLRAD: food/drink problem 1010 −0.66 489 −0.62
QoLRAD: physical/social 1010 −0.58 489 −0.49
QoLRAD: vitality 1010 −0.66 489 −0.61
RDQ: heartburn 1011 0.47 - -
RDQ: regurgitation 1011 0.46 - -
RDQ: epigastric pain 1011 0.59 - -
RDQ: GERD 1011 0.55 - -
GSRS: diarrhoea - - 488 0.26
GSRS: indigestion - - 488 0.47
GSRS: constipation - - 488 0.25
GSRS: abdominal pain - - 487 0.56
GSRS: reflux - - 488 0.54
SSS: epigastric pain 1015 0.63 489 0.65
SSS: epigastric discomfort 1015 0.68 489 0.72
SSS: heartburn 1015 0.48 489 0.47
SSS: regurgitation 1015 0.45 489 0.40
SSS: bloating 1015 0.61 489 0.40
SSS: belching 1015 0.50 489 0.48
SSS: nausea 1015 0.45 489 0.41
SSS: early satiety 1015 0.50 489 0.30
SSS: postprandial symptoms 1015 0.57 489 0.41

 

GOS, Global Overall Symptom; GERD, gastro-oesophageal reflux disease; QoLRAD, Quality of Life in Reflux And Dyspepsia; SSS, specific symptom subtype; RDQ, Reflux Disease Questionnaire; GSRS, Gastrointestinal Symptom Rating Scale.

The ES and SRM analysis demonstrated that the scale was very responsive. The change in GOS was significantly higher in responders than NRs, and was also rated as being important by the patients ( ). The test–retest reliability as assessed by the ICC varied from 0.42 to 0.62. Others have also shown some variability in other chronic GI diseases, such as in patients with inflammatory bowel disease.[21] Therefore, the GOS is useful in patients, in whom there is no change in health status.

Table 2.  Responsiveness: the mean change in GOS score (s.d.) for both change and importance rating for all enrolled patients who were R and NR

Patients
GOS2 data – CAST study
Change None (74) Small (105) Moderate (225) Large (601)
  All (995) 0.89 (1.2) 1.4 (1.1) 2.2 (1.2) 3.3 (1.1)
  R (717) 2.4 (0.5) 2.4 (0.9) 2.9 (0.9) 3.4 (1.0)
  NR (288) 0.5 (1.0) 1.1 (0.9) 1.4 (1.1) 1.3 (1.1)
Importance None (17) Slight (71) Important (181) Very (631)
  All (900) 2.6 (1.6) 1.8 (1.3) 2.3 (1.4) 3.1 (1.2)
  R (678) 3.1 (1.1) 2.8 (1.1) 3.1 (1.0) 3.4 (1.0)
  NR (222) 0.0 (0.0) 1.1 (0.9) 1.1 (1.0) 1.5 (1.1)
GOS28 data – CADET-HN study
Change None (129) Small (101) Moderate (140) Large (112)
  All (482) 0.71 (1.1) 1.2 (1.2) 1.5 (1.3) 2.7 (1.2)
  R (181) 2.6 (0.76) 2.6 (0.86) 2.6 (0.72) 2.9 (0.87)
  NR (301) 0.49 (0.87) 0.82 (0.93) 0.97 (1.2) 1.3 (1.5)
Importance None (19) Slight (119) Important (74) Very (138)
  All (350) 1.21 (1.23) 1.4 (1.3) 1.8 (1.3) 2.2 (1.4)
  R (159) 2.5 (0.84) 2.6 (0.77) 2.8 (0.94) 2.9 (0.83)
  NR (191) 0.62 (0.87) 0.80 (0.88) 0.90 (1.1) 1.0 (1.3)

 

For all values, reductions in GOS score means improvement in symptoms.
Values are expressed as mean (s.d.).
All: all included patients; R, responders, defined as GOS ≤ 2 after treatment; NR, non-responders, defined as GOS > 2 after treatment; GOS, Global Overall Symptom; CAST, Confirmatory Acid Suppression Test; CADET-HN, Canadian Adult Dyspepsia Empirical Treatment – Heartburn Negative.

We believe the way the GOS question is phrased allows the scale to capture the heterogeneity of the dyspepsia symptom complex and is in keeping with the recommendation of the Rome Working Party.[1] Furthermore, it is preferred that the patient performs the assessment of the severity of their symptoms rather than the doctor or research personnel conducting the study. Other validated outcome measures to rate dyspepsia have been reported.[22] These include, among others, the Glasgow Dyspepsia Severity Scale (GDSS),[23] Leeds Help Questionnaire,[24] Severity of Dyspepsia Assessment (SODA)[25] and Nepean Dyspepsia Index.[26] It will be interesting to compare the performance of these instruments with the GOS scale. Depending on the study design, such as duration of treatment, use of intermittent or on-demand therapy or length of follow-up, the duration over which symptoms are assessed will vary. Given the validity of the GOS scale and good correlation with other outcome measures when measured over the previous 2 days or the preceding 4 weeks, we believe the GOS scale will perform well when administered over different time frames.

An important aspect of the design of the CAST and the CADET-HN studies was the use of a strict definition of what constitutes a treatment responder. Patients needed to have at least baseline moderate severity (GOS score: ≥4) to participate in the studies. An improvement to either no or minimal symptoms (GOS score: 1 or 2) was considered as clinically meaningful. In support of meaningful change, improvements in the GOS score correlated with improvements in individual symptoms and with the overall scores and domain scores of quality of life measures. Furthermore, there was a difference in the degree of improvement seen in the GOS and other outcome measures in those patients who were responders when compared with NRs. Our analysis of ratings of the importance of change for the patient supported that a change from GOS ≥4 to ≤2 is clinically meaningful and as such believe this is a minimal clinically important difference (MCID).[9] One could also define a treatment responder differently by, for example, accepting a change of at least two points (ΔGOS: ≥2). However, in doing so the placebo response rate will increase. There is evidence in functional dyspepsia[27] and in irritable bowel syndrome (IBS)[28] that the placebo response is lowest when absence of symptoms is the outcome measure.

In conclusion, we provide evidence that the GOS, a 7-point Likert scale assessing the overall severity of dyspepsia symptoms is a valid and reliable outcome measure that is also sensitive to change and thus can be used in clinical trials evaluating dyspepsia treatments.

References

  1. Veldhuyzen van Zanten SJO, Talley NJ, Bytzer P, Klein KB, Whorwell PJ, Zinsmeister AR. Design of treatment trials for functional gastrointestinal disorders. Gut 1999; 45 (Suppl. II): II69–77.

  2. Veldhuyzen van Zanten SJO. Assessment of outcome in dyspepsia: has progress been made? Gut 2002; 50 (Suppl. IV): 23-5.

  3. Talley NJ. A critique of therapeutic trials in Helicobacter pylori-positive functional dyspepsia. Gastroenterology 1994; 106: 1174-83.

  4. Talley NJ, Stanghellini V, Heading RC, Koch KL, Malagelada JR, Tytgat GN. Functional gastroduodenal disorders. Gut 1999; 45 (Suppl. 2): II37-42.

  5. Veldhuyzen van Zanten SJ, Flook N, Chiba N, et al. An evidence-based approach to the management of uninvestigated dyspepsia in the era of Helicobacter pylori. Canadian Dyspepsia Working Group. CMAJ 2000; 162 (Suppl. 12): S3-23.

  6. Veldhuyzen van Zanten SJO, Cleary C, Talley NJ, et al. Drug treatment of functional dyspepsia: a systematic analysis of trial methodology with recommendations for design of future trials. Am J Gastroenterol 1996; 91: 660-73.

  7. Blum AL, Talley NJ, O'Morain C, et al. Lack of effect of treating Helicobacter pylori infection in patients with non-ulcer dyspepsia. N Engl J Med 1998; 339: 1875-81.

  8. Talley NJ, Janssens J, Lauritsen K, Racz I, Bolling-Sternevald E. Eradication of Helicobacter pylori in functional dyspepsia: randomised double blind placebo controlled trial with 12 months' follow up. The Optimal Regimen Cures Helicobacter Induced Dyspepsia (ORCHID) Study Group. BMJ 1999; 318: 833-7.

  9. Jaeschke R, Singer J, Guyatt GH. Measurement of health status. Ascertaining the minimal clinically important difference. Control Clin Trials 1989; 10: 407-15.

  10. Veldhuyzen van Zanten S, Talley N, Bytzer P, Klein K, Whorwell P, Zinsmeister A. Design of treatment trials for the functional gastrointestinal disorders. Rome II: the Functional Gastrointestinal Disorders 1999, Chapter 11. MacLean, VA, USA: Degnon Associates, 2000: 577-622.

  11. Veldhuyzen van Zanten SJO, Tytgat KMAJ, Pollak PT, et al. Can severity of symptoms be used as an outcome measure in trials of non-ulcer dyspepsia and Helicobacter pylori associated gastritis? J Clin Epidemiol 1993; 46: 273-9.

  12. Wiklund IK, Junghard O, Grace E, et al. Quality of life in reflux and dyspepsia patients. Psychometric documentation of a new disease-specific questionnaire (QOLRAD). Eur J Surg 1998; 583: 41-9.

  13. Svedlund J, Sjodin I, Dotevall G. GSRS − a clinical rating scale for gastrointestinal symptoms in patients with irritable bowel syndrome and peptic ulcer disease. Dig Dis Sci 1988; 33: 129-34.

  14. Shaw MJ, Talley NJ, Beebe TJ, et al. Initial validation of a diagnostic questionnaire for gastroesophageal reflux disease. Am J Gastroenterol 2001; 96: 52-7.

  15. Armstrong D, Veldhuyzen van Zanten SJO, Barkun AN, et al. Symptom response at 1-week predicts responders at 4-week with esomeprazole (40 mg bid and od) in patients with uninvestigated dyspepsia. Am J Gastroenterol 2003; 98 (9 Suppl 1): S48-9 (Abstract 139).

  16. Veldhuyzen van Zanten SJO, Chiba N, Armstrong D, et al. A randomized trial comparing omeprazole, ranitidine, cisapride or placebo in Helicobacter pylori negative, primary care patients with dyspepsia: The CADET-HN study. Am J Gastroenterol 2005; 100: 1477-88.

  17. Norman G. Issues in the use of change scores in randomized trials. J Clin Epidemiol 1989; 42: 1097-105.

  18. Deyo RA, Diehr P, Patrick DL. Reproducibility and responsiveness of health status measures. Statistics and strategies for evaluation. Control Clin Trials 1991; 12 (Suppl. 4): 142S-58S.

  19. Talley N, Fullerton S, Junghard O, Wiklund I. Quality of life in patients with endoscopy-negative heartburn: reliability and sensitivity of disease-specific instruments. Am J Gastroenterol 2001; 96: 1998-2004.

  20. Junghard O, Lauritsen K, Talley NJ, Wiklund IK. Validation of 7-graded diary cards for severity of dyspeptic symptoms in patients with non-ulcer dyspepsia. Eur J Surg 1998; 583 (Suppl. ): 106-11.

  21. Guyatt G, Mitchell A, Irvine EJ, et al. A new measure of health status for clinical trials in inflammatory bowel disease. Gastroenterology 1989; 96: 804-10.

  22. Fraser A, Delaney B, Moayyedi P. Symptom-based outcome measures for dyspepsia and GERD trials: a systematic review. Am J Gastroenterol 2005; 100: 442-52.

  23. El-Omar EM, Banerjee S, Wirz A, McColl KE. The Glasgow Dyspepsia Severity Score – a tool for the global measurement of dyspepsia. Eur J Gastroenterol Hepatol 1996; 8: 967-71.

  24. Moayyedi P, Duffett S, Braunholtz D, et al. The Leeds Dyspepsia Questionnaire: a valid tool for measuring the presence and severity of dyspepsia. Aliment Pharmacol Ther 1998; 12: 1257-62.

  25. Rabeneck L, Wristers K, Goldstein JL, Eisen G, Dedhiya SD, Burke TA. Reliability, validity, and responsiveness of Severity of Dyspepsia Assessment (SODA) in a randomized clinical trial of a COX-2-specific inhibitor and traditional NSAID therapy. Am J Gastroenterol 2002; 97: 32-9.

  26. Talley NJ, Verlinden M, Jones M. Validity of a new quality of life scale for functional dyspepsia: a United States multicenter trial of the Nepean Dyspepsia Index. Am J Gastroenterol 1999; 94: 2390-7.

  27. Moayyedi P, Delaney BC, Vakil N, Forman D, Talley NJ. The efficacy of proton pump inhibitors in nonulcer dyspepsia: a systematic review and economic analysis. Gastroenterology 2004; 127: 1329-37 (Review).

  28. Atkinson W, Sheldon TA, Shaath N, Whorwell PJ. Food elimination based on IgG antibodies in irritable bowel syndrome: a randomised controlled trial. Gut 2004; 53: 1459-64.