Common use of Methods against bias Clause in Contracts

Methods against bias. This is an open-label study. Blinding is not possible because active refeeding of stool in the intervention group is obvious to any person participating in the medical care of the patient. Randomization will be performed centrally (with variable block length) and stratified by study center, height of stomata [29, 30] and ,weight at birth (<1000g / ≥1000g), as this is an important prognostic factor for primary endpoint. Randomization will take place after enterostomy creation in order to reduce the amount of missing values due to patient exclusion after surgery (e.g. due to unforeseen need for resection of ileocecal valve). The primary analysis will be performed on the ITT population as this is an open study and parents may have preferences not outspoken before randomization. A per protocol analysis will be conducted as a sensitivity analysis. Consistency between the findings in the ITT population and the per protocol population will be examined as it is an important pre-requisite for a successful interpretation of the study. Drop-Outs are not expected because all patients will constantly undergo neonatal intensive care and will therefore not be lost to follow-up. If parents withdraw their infant from study participation they will be asked to allow data collection at a final analysis in order to avoid that information would be wasted. Nonetheless, if missing values should occur (e.g due to death, or parents’ refusal of data collection) observations will be censored at the last timepoint with known enteral feeding status. Since this censoring may be informative, missing values for time to full feeds will be replaced by the worst observation in each group in a sensitivity analysis in order to check how censoring may have influenced the results. If any death should occur before the respective patient reaches full enteral feeds a sensitivity analysis will be performed on all surviving patients.

Methods against bias. This is an open-label study. Blinding is not possible because active refeeding of stool in the intervention group is obvious to any person participating in the medical care of the patient. Randomization will be performed centrally (with variable block length) and stratified by study center, height of stomata [29, 30] center and ,weight at birth (<1000g / ≥1000g), as this is an important prognostic factor for primary endpoint. Randomization will take place after enterostomy creation in order to reduce the amount of missing values due to patient exclusion after surgery (e.g. due to unforeseen need for resection of ileocecal valve). The primary analysis will be performed on the ITT population as this is an open study and parents may have preferences not outspoken before randomization. A per protocol analysis will be conducted as a sensitivity analysis. Consistency between the findings in the ITT population and the per protocol population will be examined as it is an important pre-requisite for a successful interpretation of the study. Drop-Outs are not expected because all patients will constantly undergo neonatal intensive care and will therefore not be lost to follow-up. If parents withdraw their infant from study participation they will be asked to allow data collection at a final analysis in order to avoid that information would be wasted. Nonetheless, if missing values should occur (e.g due to death, or parents’ ' refusal of data collection) observations will be censored at the last timepoint with known enteral feeding status. Since this censoring may be informative, missing values for time to full feeds will be replaced by the worst observation in each group in a sensitivity analysis in order to check how censoring may have influenced the results. If any death should occur before the respective patient reaches full enteral feeds a sensitivity analysis will be performed on all surviving patients.

Methods against bias. This is an open-label study. Blinding is not possible because active refeeding of stool in the intervention group is obvious to any person participating in the medical care of the patient. Randomization will be performed centrally via fax (with variable block length) and stratified by study center, height of stomata (jejunostomy/proximal Ileostomy or terminal ileostomy) [29, 30] ], weight and ,weight mother’s gestation week at birth (<< 1000g and before 37+0 /< 1000g and before 37+0 / ≥1000g≥ 1000g /and at least 37+0), as this is an are important prognostic factor factors for the primary endpoint. Randomization will take place after enterostomy creation in order to reduce the amount of missing values due to patient exclusion after surgery (e.g. due to unforeseen need for resection of ileocecal valve)) and after the treating physician has determined the full feed kcal goal on the basis of the stratification. The primary analysis will be performed on the ITT population as this is an open study and parents may have preferences not outspoken before randomization. A per protocol analysis will be conducted as a sensitivity analysisanalysis in all patients that have no substantial protocol deviations. Consistency between the findings in the ITT population and the per protocol population will be examined as it is an important pre-pre- requisite for a successful interpretation of the study. DropIn general drop-Outs are not expected because all patients will constantly undergo neonatal intensive care and will therefore not be lost to follow-up. If parents withdraw their infant from study participation participation, they will be asked to allow data collection at a final analysis in order to avoid that information would be wasted. NonethelessNevertheless, up to now, we recognized that for some individual patients it is not feasible to abide by the protocol, especially the nutrition protocol after enterostomy closure. Therefore, if missing values should occur (e.g e.g. due to death, or parents’ refusal of data collection) observations will be censored at the last timepoint time point with known enteral feeding status. Since this censoring may be informative, missing values for time to full feeds will be replaced by the worst observation in each group in a sensitivity analysis in order to check how censoring may have influenced the results. If any death should occur before the respective patient reaches full enteral feeds a sensitivity analysis will be performed on all surviving patients. To avoid direct influence on the change of kcal goal of the treating physician during the study, at the end of the study an independent reviewer will assess at least all changes of kcal goals.

Methods against bias. This is an open-label study. Blinding is not possible because active refeeding of stool in the intervention group is obvious to any person participating in the medical care of the patient. Randomization will be performed centrally (with variable block length) and stratified by study center, height of stomata [29, 30] and ,weight at birth (<1000g / ≥1000g), as this is an important prognostic factor for primary endpoint. Randomization will take place after enterostomy creation in order to reduce the amount of missing values due to patient exclusion after surgery (e.g. due to unforeseen need for resection of ileocecal valve). The primary analysis will be performed on the ITT population as this is an open study and parents may have preferences not outspoken before randomization. A per protocol analysis will be conducted as a sensitivity analysis. Consistency between the findings in the ITT population and the per protocol population will be examined as it is an important pre-requisite for a successful interpretation of the study. Drop-Outs are not expected because all patients will constantly undergo neonatal intensive care and will therefore not be lost to follow-up. If parents withdraw their infant from study participation they will be asked to allow data collection at a final analysis in order to avoid that information would be wasted. Nonetheless, if missing values should occur (e.g due to death, or parents’ ' refusal of data collection) observations will be censored at the last timepoint with known enteral feeding status. Since this censoring may be informative, missing values for time to full feeds will be replaced by the worst observation in each group in a sensitivity analysis in order to check how censoring may have influenced the results. If any death should occur before the respective patient reaches full enteral feeds a sensitivity analysis will be performed on all surviving patients.