This study is a prospective, randomized, open-label, multicenter clinical trial. The study was conducted at 8 hospitals located in Tokyo and Kanagawa, Japan (Textbox 1). We aimed to evaluate the efficacy of DialBetesPlus on microalbuminuria in T2DM patients. Figure 1 shows a flowchart of this trial. The intervention group used DialBetesPlus and conventional therapy for 12 months, while the control group was treated with conventional therapy alone. The primary outcome was a change in albuminuria levels in a first-morning void after 12 months. The final follow-up was in 18 months, 6 months after the 12-month intervention period.

We recruited patients who passed all the inclusion and exclusion criteria before randomization (Textbox 2). The criteria were designed to include T2DM patients experiencing the early stage of DKD without any restrictions on physical activity. Eligible participants had T2DM with microalbuminuria (30-299 mg/g creatinine), hemoglobin A_1c (HbA_1c) of 6.5% or more, and an estimated glomerular filtration rate (eGFR) of 45mL/min/1.73m² or more. We obtained written, informed consent from all the patients participating in this study. Prior to the randomization, all enrolled patients were asked to use DialBetesPlus for 2 weeks to confirm their ability to use the system and devices properly. Eligible participants were those able to use DialBetesPlus for over 7 days during the initial 2-week confirmation period. The participants were randomized into either the intervention or control group in a one-to-one fashion based on albuminuria levels, gender, and age. The research team consisted of diabetologists, nephrologists, pharmacists, dieticians, nurses, a laboratory technician, and technical support experts. While the team handled technical troubles and monitored usage of the system, the patients continued to consult their attending doctors about their general health status. If 7 days passed without any input into the system, an alert was sent to encourage the patient to resume providing input. If 3 weeks passed without a patient inputting any data, we defined the patient as a dropout. Participants with whom the study team has lost contact were also treated as dropouts. The criteria for study discontinuance were a serious adverse event, patient request for discontinuance, pregnancy, or the judgment of a lead physician.

The details of the DialBetesPlus system are shown in Figure 2. Patients measured blood glucose, blood pressure (BP), body weight, and pedometer counts at home. The data were transferred from each device to a smartphone by either Near Field Communication (NFC) or Bluetooth, and then immediately sent to a server, where they were automatically evaluated following the Japan Diabetes Society (JDS) guideline’s target values [19]. Optimal values include blood glucose below 110 mg/dl before breakfast and 140 mg/dl at bedtime, BP below 125/75 mmHg, and pedometer counts above 8000 steps per day. DialBetesPlus determined if each reading satisfies the JDS guideline requirements and immediately sent the results to the patient’s smartphone. Patients also entered the contents and quantity of their meals and the type and duration of exercise not counted by a pedometer. Then, their intake and consumed calories were automatically calculated and sent to the smartphone along with specific advice regarding lifestyle modifications based on JDS guidelines. Critical values with blood glucose levels above 400 mg/dl or below 70 mg/dl, systolic BP above 220 mmHg, or diastolic BP above 110mg were automatically reported to the research team, and the team informed attending doctors when necessary.

The patients allocated to the intervention group used DialBetesPlus for 12 months. Patients in the intervention group receive an NFC-enabled glucometer (MS-FR201B; Terumo), a Bluetooth-enabled BP monitor (HEM-7271T; Omron), a pedometer (MT-KT02DZ; Terumo), and a scale (HBF-255T; Omron). These devices were all paired with a single provided smartphone (Arrows F-02H; Fujitsu or Galaxy Note3 SC-01F; Samsung) that transmits readings to the DialBetesPlus server via a wireless network. In addition, as a part of standard therapy, the participants in the control group were provided with sphygmomanometers to measure BP at home.

In this study, we divided the participants into 2 models (a hospital-led model and a pharmacy-led model) depending on the location of study enrollment. In the hospital-led model (the University of Tokyo Hospital, Yokohama City University Hospital, Saiseikai Yokohamashi Nanbu Hospital, and Fujisawa City Hospital), the instructions included guidance on DialBetesPlus use provided by health care providers at the patient’s hospital. In a pharmacy-led model (run by the other 4 hospitals), instructions on using DialBetesPlus were provided by pharmacists supporting the study. The pharmacists are Nihon Chouzai Co, Ltd employees, with whom we signed a business cooperation contract to conduct the pharmacy-led model portion of the study.

The primary outcome of this study is a change in albuminuria levels at 12 months. We estimated that baseline first-morning-void albuminuria level would be 200 plus or minus 200 mg/g creatinine and hypothesized the difference in the change in albuminuria between the intervention group and the control group at 12 months would be 100 plus or minus 200 mg/g creatinine. Based on previous reports [20,21], we calculated 63 patients were required in each group to achieve a 2-sided significance level of .05 and a statistical power of 80%. Factoring ineligibility and dropouts, we calculated a final target number of 80 patients per group.

We defined the primary outcome as a change in albuminuria levels in the first-morning void over 12 months. A first-morning void is less influenced by hydration and physical activity than a spot urine sample [22-25]. As the participants in this study are restricted to patients with microalbuminuria but without macroalbuminuria, we validated albuminuria in a first-morning void.

The secondary outcomes were changes in physical parameters (BMI and BP), blood tests, lifestyle habits, self-management, medication therapy, adherence to the diabetes medication regimen, and quality of life (QoL). Blood test parameters included renal function (eGFR), glycemic control (HbA_1c and fasting blood glucose), and lipid metabolism (high-density lipoprotein cholesterol [HDL-C], low-density lipoprotein cholesterol [LDL], and triglycerides [TG]). We also assessed all-cause deaths, composite cardiovascular outcomes, composite renal endpoints, safety, and the usage frequency of DialBetesPlus.

Changes in lifestyle habits were assessed via food-log diaries covering 3 days. Dietitians calculated caloric intake and nutrient balance. The Summary of Diabetes Self-Care Activities Measure (SDSCA)—which evaluates 7 aspects of the diabetes regimen: general diet, specific diet, exercise, medication taking, blood-glucose testing, foot care, and smoking—is a widely used self-reporting tool for patients with diabetes in the United States [26]. We evaluated changes in self-management using the Japanese version of the SDSCA, the J-SDSCA. Change in medication therapy was assessed by evaluating prescriptions. Diabetes medication adherence was monitored using the 8-item Morisky Adherence Scale (MMAS-8). The MMAS-8 is reliable and valid in measuring the adherence of patients with multiple chronic conditions worldwide, including T2DM [27-29]. We measured patients’ QoL using the Japanese version of the Audit of Diabetes-Dependent Quality of Life (JP-ADDQoL). The ADDQoL is a diabetes-specific QoL measurement scale producing reliable and valid scores [30,31]. JP-ADDQoL also showed adequate reliability and acceptable validity [32].

Composite cardiovascular outcomes included the first occurrence and recurrence of myocardial infarction and stroke, the first occurrence of percutaneous coronary intervention and coronary artery bypass, hospitalization for unstable angina and heart failure, and cardiovascular death. The composite renal endpoint was defined as ESRD and more than a 2-fold increase in serum creatinine.

To ensure safety, we monitored the number of hypoglycemic events, other adverse events, and issues with the DialBetesPlus system throughout the study. The results of the hospital-led model and the pharmacy-led model will also be compared.

At registration, we collected information on patients’ background, albuminuria level in a first-morning void, physical parameters, medications, blood tests, food-log diaries, J-SDSCA, MMAS-8, and JP-ADDQoL. Follow-up visits were scheduled at 2 months (plus or minus 4 weeks), 6 months (plus or minus 6 weeks), 12 months (plus or minus 6 weeks), and 18 months (plus or minus 6 weeks). We collected and recorded information according to the schedule in Table 1. Even if the interventional treatment was discontinued, we collected as much information as possible with the consent of participants.

Due to the COVID-19 pandemic, we made some adjustments to ensure complete data collection if participants canceled their hospital visits to avoid infection risk. The Research Ethics Committee of The University of Tokyo Graduate School of Medicine and its affiliated institutions formally approved adjustments to the trial protocol. First, we adopted a self-administered blood collection kit to measure blood glucose, HbA_1c, eGFR, HDL-C, LDL-C, and TG levels when participants canceled in-person hospital visits. Albuminuria levels were also measured using the first-morning void mailed with the self-collected blood sample. We can conduct a blood examination by collecting 65 µl of blood from a fingertip with the self-administered blood collection kit, Ouchide-doc (Halmek Ventures, Inc). It is reported that the assay results with this blood collection method are quite comparable to the conventional methods used in hospitals [33]. Second, to minimize the face-to-face contact for data collection, we sent questionnaires to participants' homes in advance of their hospital visits.

The study is being carried out in compliance with the Declaration of Helsinki. This protocol and informed consent forms were approved by the Research Ethics Committee of The University of Tokyo Graduate School of Medicine and affiliated institutions. This study was registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN000033261) [34].

All participants are included after providing their signed and informed consent to participate in the trial. The participants are also informed of their right to withdraw from the study at any time. After the study concludes, data will be accessible by study groups for analysis and dissemination. All results of any analyses will be presented at major national and international scientific conferences and submitted for peer-reviewed journals of international repute and visibility.

Data regarding patients’ characteristics are presented as mean (SD) or median (IQR). We will compare changes in albuminuria levels, physical parameters, blood tests, and nutritional intake between the intervention and control groups. These will be analyzed using the 2-tailed t test or Mann-Whitney U test, as appropriate. Changes in J-SDSCA, MMAS-8, and JP-ADDQoL scores will also be analyzed using the Mann-Whitney U test. We will compare the proportion of hypoglycemia during the study in the intervention group to the proportion in the control group using Fisher's exact test. P values <.05 will be considered statistically significant. Statistical analyses will be performed using SAS (version 9.4; SAS Institute Inc).