Remote Symptom Monitoring Using Patient-Reported Outcomesin Patients With Chronic Kidney Disease:Process Evaluation ofa Randomized Controlled Trial

Background: In Denmark, outpatient follow-up for patients with chronic kidney disease (CKD) is changing from in-hospitalvisits toward more remote health care delivery. The nonuse of remote patient-reported outcomes (PROs) is a well-known challenge,and it can be difficult to explain which mechanisms of interventions influence the outcome. Process evaluation may, therefore,be used to answer important questions on how and why interventions work, aiming to enhance the implications for clinicalpractice. Objective: This study aimed to provide insight into the intervention process by evaluating (1) the representativity of the studypopulation, (2) patient and physician use patterns, (3) patient adherence to the intervention, and (4) clinical engagement. Methods: A process evaluation determining the reach, dose, fidelity, and clinical engagement was carried out, alongside amulticenter randomized controlled trial (RCT). We developed and implemented an intervention using PRO measures to monitoroutpatients remotely. Data were collected for the PRO intervention arms in the RCT from 4 sources: (1) PRO data from theparticipants to determine personal factors, (2) the web-based PRO system to identify key usage intervention patterns, (3) medicalrecords to identify clinical factors relating to the use of the intervention, and (4) semistructured interviews conducted with involvedphysicians.Results: Of the 320 patients invited, 152 (47.5%) accepted to participate. The study population reflected the target population.The mean adherence rate to the PRO intervention arms was 82% (95% CI 76-87). The questionnaire response rate was 539/544(99.1%). A minority of 13 (12.9%) of 101 patients needed assistance to complete study procedures. Physicians assessed 477/539(88.5%) of the questionnaires. Contact was established in 417/539 (77.4%) of the cases, and 122/539 (22.6%) of the patients didnot have contact. Physicians initiated 288/417 (69.1%) and patients requested 129/417 (30.9%) of all the contacts. The primarycauses of contact were clinical data (242/417, 58%), PRO data (92/417, 22.1%), and medication concerns and precautionaryreasons (83/417, 19.9%). Physicians found the use of PRO measures in remote follow-up beneficial for assessing the patient'shealth. The inclusion of self-reported clinical data in the questionnaire motivated physicians to assess patient responses. However,some barriers were emphasized, such as loss of a personal relationship with the patient and the risk of missing important symptomsin the absence of a face-to-face assessment. Conclusions: This study demonstrates the importance and practical use of remote monitoring among patients with CKD. Overall,the intervention was implemented as intended. We observed high patient adherence rates, and the physicians managed most questionnaires. Some physicians worried that distance from the patients made it unfeasible to use their "clinical glance," posinga potential risk of overlooking crucial patients' symptoms. These findings underscore key considerations for the implementationof remote follow-up. Introducing a hybrid approach combining remote and face-to-face consultations may address these concerns