Hans van den Heuvel

Participant interaction and compliance During the 15 weekday study period, the 13 participants who completed the study sent in their BP and symptom checklist on average, respectively, 14.0 and 12.8 times. The total compliance rate, as shown in Table 4, was 93% for blood pressure and 85% for symptom checklist uploads. Sevenparticipants (54%) provideduswithmoredata than requested, resulting in compliance rates (theoretically) above 100%. One of them also uploaded data during Saturday and Sunday, despite our study guidelines only to measure during weekdays. Two participants (S06 and S09) showed a clearly lower compliance rate (less than 55%) compared to the rest of the group. One participant noted she did not feel the urge to measure on a daily basis, as she was not experiencing symptoms or high blood pressure during the study period. The other participant did not mention a reason for lower compliance. The accuracy of the automatic alert system In the BP alarm system 7 alerts occurred (3.8% of all 182 BP measurements). In 4 of these 7 alarms, the upper threshold for either systolic or diastolic BP was exceeded. These 4 alerts were all sent by the one participant (S10) who was subsequently diagnosed with chronic hypertension. The other three alerts appeared because of an increase of more than 20 mmHg in either systolic or diastolic pressure compared to the previous measurement. These accidental rises of blood pressure could be handled with expectant management after reviewing the BP trend and/or the absence of preeclampsia symptoms, as could be directly reviewed in the adjacent checklist. After reviewing all variables of the seven BP alerts, no false positive alerts were found. Subsequent manual comparison of all other received values with our set thresholds detected no missing or incorrect alerts (Table 5). In 15 days, 167 symptom checklists were uploaded, of which 73 (43.7%) alarmed because of present symptoms. Frequently, participants reported multiple symptoms in the same checklist. This resulted in a total of 111 symptoms in 73 separate alarms for the checklists, reporting at least one symptom at a time. Braxton-Hicks contractions (42x, 37.8%) and peripheral edema (29x, 26.1%) were the most common reported symptoms (Figure S2). Out of all 167 received symptom checklists, 5 (3.0%) resulted in the advice to consult a health care provider. Reasons for referral included an episode of vaginal blood loss or a combination of hypertension and symptoms. Combinations of a positive symptom checklist in normotensive participants were handled with expectant management, as per protocol. CHAPTER 4 60