Page 134 - Exploring the Potential of Self-Monitoring Kidney Function After Transplantation - Céline van Lint

132
Chapter 6
registered and a mean level of creatinine for the values that were measured on days without any
registration. A paired t-test was performed to compare these means.
2. Selectionofmeasurements,whichisthecasewhenthenumberofperformedmeasurementsthat
is stored in the device is higher than the number of registered measurements on a given date. Therefore only days with a difference between number of measured and number of registered creatinine values were selected (e.g. three measurements stored in the device and one value registered in the SMSS). We then calculated per patient the mean of all values stored in the device and registered in the SMSS and the mean of all values stored in the device, but not registered in the SMSS. A paired t-test was performed to compare these means.
Adherence to feedback
After registration of a creatinine value in the SMSS, patients received an automatically generated advice on the necessary action to take (see figure 1). To investigate level of adherence to the advice generated by the SMSS, we separated between adherence to action feedback (supplied after the registration of a first measurement when further action was required) and adherence to the day conclusion (supplied when no further actions were required after the first registration of a day or when a second and final measurement was registered on the same day).
Action feedback could only appear in case the newly registered creatinine value was higher than the previous ones and required an additional measurement to confirm the first. In these cases, the feedback system of the SMSS showed an orange or red traffic light with the corresponding advice to repeat the measurement. From the SMSS logged data, we selected those cases where a second measurement was requested and checked whether the concerning patients indeed measured and registered a second creatinine value on the same day.
To study adherence to the day conclusion, we only considered the cases in which patients again were confronted with an orange or red traffic light. In case of a request to perform another measurement tomorrow (orange traffic light), the SMSS logged data was checked to see whether the requested action was indeed performed. In case of a request to contact the hospital (red traffic light), patient hospital records were searched for telephonic and/or outpatient contacts on dates following the concerning feedback.
Sensitivity analyses
Two sensitivity analyses were performed to control for potential bias. First, as being hospitalized limits the possibility to keep up with requested measurement and registration frequency, level of adherence