Zsa Zsa Weerts

Smart Data Collection for the Assessment of Treatment Effects in IBS 227 9 electronic CRF to collect additional secondary outcomes. A troublesome issue that occurred was a routing error in one of the questionnaires that was discovered too late and had already led to a high proportion of missing data ( Supplementary Table S9.1 ). This applied to only a single question, but routing errors can have potentially disastrous consequences. As such, investigators and data-managers should take appropriate care and time when testing questionnaires. Data-exports should furthermore be examined in an early testing phase and preferably by more than one investigator and data-manager. Similar to the diary, the web-based questionnaires and the eCRF featured built-in routing of questions, data validation, and response requirements to stimulate data- quality and completeness. Overall, these steps allowed for guaranteed standardized data-collection with a completeness of more than 99% for the web-based questionnaire and eCRF items. Additional advantages of the combined framework for digitalized data-collection are: 1) the ability to monitor patients and their adherence; 2) a reduction in paperwork and physical archiving, e.g. in this study the paperwork was reduced to one single informed consent file; 3) manual data transcription can be omitted as research data enter the database immediately; 4) the possibility to adjust and individualize the smartphone application, eCRF, and web-based questionnaires according to the needs of each particular study and 5) more accurate and standardized data reporting as no error- prone re-entry necessary as compared to paper diaries. The described framework for digitalized data-collection can therefore be employed across different disease entities. Our findings should be interpreted in light of some potential limitations. First, the study was not primarily designed for the analysis of adherence to the digital symptom diary, but for measuring the main clinical outcome. 16 However, since almost 200 patients were included, the sample size is sufficiently large to estimate adherence with enough precision. Second, adherence rate was not assessed within a controlled trial with a more traditional method of data collection ( i.e. paper-and-pencil diaries, IVRS) as a comparison. However, the rapid diffusion toward digitalized approaches in healthcare and clinical research renders such comparison less meaningful from a practical point of view as the use of such techniques become inevitably ubiquitous. In addition, it is unlikely that these traditional approaches to data collection would result in even higher adherence than those observed here. Another limitation was that patient satisfaction with the digital diary or web-based questionnaires was not quantified by means of a questionnaire. In the current study, the feasibility of the used framework was evaluated primarily on the basis of patients’ and investigators’ adherence and the proportion of complete data, whereas quantified patient satisfaction was not taken into account.