Page 109 - Exploring the Potential of Self-Monitoring Kidney Function After Transplantation - Céline van Lint

Self-monitoring of renal function: a randomised trial 107
of life. This questionnaire contains 12 multiple-choice items covering Physical Component Summary (PCS), and the Mental Component Summary (MCS). The Partners in Health scale[30] (PiH) was used to measure self-efficacy regarding self-management behaviour. The PIH contains 12 items on a nine-point scale. The Client Satisfaction Questionnaire(CSQ-8) [31] was used to measure satisfaction with care and consists of 8 items that are rated on a four-point scale. All questionnaires were completed at baseline and 12-month follow-up.
The interviews were recorded and transcribed in full. Answers to questions relating to satisfaction with self-monitoring after kidney transplantation (1. If possible, would you like to continue self-monitoring supported by the SMSS? Why (not)?, 2. Would you recommend self-monitoring kidney function supported by the SMSS to other kidney transplant patients? Why (not)?) were extracted from the data and categorized.
2.5. Data analysis
Baseline continuous variables are reported as mean (SD, standard deviation) or median (IQR, interquartile range) in case of skewed data. Categorical variables are reported in percentages. The baseline characteristics of the intervention and control group were compared using independent t- tests for continuous variables and chi-square tests for categorical variables. Linear mixed modelling was used to compare intervention and control patients concerning the change in kidney function (eGFR), blood pressure, QoL (physical and mental component scale), level of satisfaction with care and self-management behaviour over time. Linear mixed modelling is the recommended method for analysing repeated measures, as it accounts for correlation between repeated measurements of the same patient.
To analyse the difference between intervention and control patients for number of ftf and telephonic consults, total number of reimbursable minutes spent over 365 days post-discharge (including ftf visits, telephonic consults and laboratory analysis only) and number of SAEs, univariate linear regression was used.
Besides univariate linear mixed models and linear regression analyses, sensitivity analyses were performed using multivariate models adjusting for significant baseline difference(s) between the two study groups. Missing baseline data were imputed using multiple imputations (n=10). For the variables with a hypothesized non-inferior outcome (eGFR, blood pressure, quality of life, satisfaction, number of SAEs), only the per protocol population was included in the analyses. Including intention-to-treat patients would dilute the potential difference between intervention and control patients, facilitating a non-inferior result. For the variables with an expected difference between the
5