69 3 Patient satisfaction with stoma care and their expectations on mobile apps for supportive care February and 17th March 2020. Collected data were anonymized by the Dutch Ostomy Association, and subsequently provided for analysis. Statistical analysis Statistical analyses were performed using SPSS (version 27). Descriptive statistics were used to assess baseline characteristics. Missing data were accounted for using imputation by chained equations for variables with missing data. The pooled results of the five multiple iterations were used for the analysis. For further analysis, the frequencies of ten potential physical and nine potential psychological problems were respectively added together, so the total scores ranged 10-50 and 9-45. Multiple linear regression analysis was performed to investigate the association between the patient characteristics and patient satisfaction. As recall bias for patient satisfaction may be strongly present, the analysis was also conducted separately for each time group (< 1 year, 1-3 years, 3-5 years, 5-10 years and > 10 years having a stoma). Multinominal logistic regression analysis was conducted to investigate whether willingness to use an app could be predicted by patient characteristics, satisfaction, or experience with mobile technology. To discriminate between patients’ willingness and unwillingness, we trichotomized the outcomes to ‘willing’ (with choices ‘very willing’ and ‘willing’), ‘neutral’, and ‘unwilling’ (‘unwilling’ and ‘very unwilling’). The reference category for the logistic analysis was the neutral option. For both regression analyses, all determinants were chosen a priori, based on the literature and expectations. Dummy variables were created for nominal and ordinal variables, and measured relative to their default reference categories (the highest frequency in this study population or the most clinically relevant). A stepwise backward selection method was used to correctly select and remove covariates that were not associated with the outcome. Therefore, only the significant variables (p≤ 0.05) remained in the prediction model.
RkJQdWJsaXNoZXIy MTk4NDMw