Global endpoints in functional gastrointestinal disease: Authors’ reply

Abstract

Sirs, We appreciate the interest of Drs Mangel and Fehnel from RTI Health Solutions in our paper and completely agree with them that the selection of endpoints is critical for successful clinical trials. There is no doubt that binary endpoints (e.g. dead or alive) are much easier to handle in clinical trials. Thus, global response measures (e.g. complete vs. incomplete control of symptoms) have been used in a number of clinical trials. On the other hand, this approach does not allow determining specific symptoms that respond to therapy. As a consequence, questionnaires that assess responses of various items and domains provide important information during the drug development. Drs Mangel and Fehnel address a number of methodological issues that in principal require consideration during the development of the symptom score. In this context, we appreciate that the authors state that the development of the GIS followed the established approaches. Indeed most methodological issues raised already have been carefully addressed in our paper. In order to avoid repetition, we would like to refer with regard to most issues raised to our paper and would like to specifically address only two points namely the potential response bias and the issue of outweighed scoring. It might well be that during a face-to-face interview some bias responses occur. However, this bias similarly influences patients on placebo and patients on active treatment. Thus, it is unlikely that the overall difference between placebo and active medication is influenced. On the other hand, in the context of usually 8–12 week trials, self-administered questionnaires have a number of limitations that range from the reliability of the timing of completion (if classic paper-based instruments are being used) to random responses of patients who are annoyed by the repeated administration of questionnaires or bothering symptom assessments with electronic devices. These limitations can only partly be overcome by phone or palm-based assessments. Regarding the weighing of individual symptoms which in principle was considered during the design of the GIS, it is important to recognize that the GIS was developed as a universal tool for the routine clinical setting to properly assess symptoms as well as an outcome measure for clinical trials. Thus, we had to balance the complexity of the scoring procedure against the potential benefit of a more sophisticated weighing of individual items. As it was felt that the additional complexity does not yield relevant improvements of the scores, the approach without adding additional weigh to specific items was favoured. Indeed even without using different weights for the various items, the GIS fulfil all criteria with regard to reliability, validity and responsiveness. Nevertheless, it is obvious that Drs Mangel and Fehnel are trying to make a case for a global response measure. However, the evidence to favour a global response over a valid symptom score is insufficient, vice versa. Most likely the combination of both approaches will yield most information. Further clinical trials will address this issue and we are grateful for the comments from Drs Mangel and Fehnel that have allowed us to clarify this issue. We are in agreement with the authors of the letter that the development of instruments for the assessment of symptoms is an ongoing process. The same holds true for the design of trials in functional gastrointestinal disorders. In this context, the GIS is a tool that allows quantifying treatment responses. We look forward to trials that prospectively compare the responsiveness of the GIS and global response measures. Most likely, both approaches complement each other and it might be advisable to include both in future clinical trials. We have concluded our manuscript with the statement: ‘…In summary, based upon our data, the GIS is a valid and reliable instrument for the assessment of dyspeptic symptoms that is suitable for the quantification of symptom severity in clinical trials, as well for the standardized assessment of symptoms in the clinical setting…’. The discussion raised by Mangel and Fehnel provides further support for this statement and thus we are thankful for their comments.