Page 57 - Exploring the Potential of Self-Monitoring Kidney Function After Transplantation - Céline van Lint

Patient Acceptance of a Self-Management Support System 55
while. These five patients indicated a variety of reasons for this: variety in self-measured creatinine values (n = 3), stress caused by self-monitoring (n = 1), and too little benefit (n = 1). The profile of the participants who responded to T0 and T1 questionnaire is shown in Table 1. In both cases, 46 patients completed the questionnaires. Although these populations were not made up of the exact same responding patients, no significant differences in profile were found between the populations who responded at T0 and T1.
Data preparation
Not Applicable and Missing Data
A distinction was made between situations where participant specifically indicated that a question was not applicable (NA) for them, or when they had left the question unanswered, i.e. missing values. The relative NA percentage, i.e., the number of NA/(the number of participants - the number of missing values) × 100% for each item was calculated. The majority of questionnaire items (77.03%) had less than 5% of the participants rated the question as NA. However, items with a relative NA percentage above 1.5 × interquartile range (4.88%) + 3rd quartile (4.88%) = 12.20% were regarded as outliers [50] as apparently an unusual number of patients considered them as not applicable to their situation and were therefore not appropriate items to capture the underlying constructs across the patient sample. Twelve items (18%) turned out to be outliers and were therefore removed from the analysis, leading to the removal of the social influence construct all together and facilitating condition item 3 and 4 (all at T0 and T1, Additional file 1). For the remaining items, ‘not applicable’ was treated as missing.
There were 394 (12.71%) values missing in total. Fifteen out of fifty (30%) participants answered all the questionnaire items, and none of the items was answered by all participants. To avoid exclusion of participants and thereby biasing the analysis [51], Maximum Likelihood methods using the expectation– maximization (EM) algorithm was applied to substitute missing data of the RTPTA questionnaire items. This method produces unbiased parameter estimates with missing (completely) at random data [52]. Patients’ age, gender, type of donor, and pre-transplant status were used as predictors.
3