Effect of Behavior Modification on Outcome in Early- to Moderate-Stage Chronic Kidney Disease: A Cluster-Randomized Trial

Abstract

Objectives: Owing to recent changes in our understanding of the underlying cause of chronic kidney disease (CKD), the importance of lifestyle modification for preventing the progression of kidney dysfunction and complications has become obvious. In addition, effective cooperation between general physicians (GPs) and nephrologists is essential to ensure a better care system for CKD treatment. In this cluster-randomized study, we studied the effect of behavior modification on the outcome of early- to moderate-stage CKD.

Design: Stratified open cluster-randomized trial.

Setting: A total of 489 GPs belonging to 49 local medical associations (clusters) in Japan.

Participants: A total of 2,379 patients (1,195 in group A (standard intervention) and 1,184 in group B (advanced intervention)) aged between 40 and 74 years, who had CKD and were under consultation with GPs.

Intervention: All patients were managed in accordance with the current CKD guidelines. The group B clusters received three additional interventions: patients received both educational intervention for lifestyle modification and a CKD status letter, attempting to prevent their withdrawal from treatment, and the group B GPs received data sheets to facilitate reducing the gap between target and practice.

Main outcome measure: The primary outcome measures were 1) the non-adherence rate of accepting continuous medical follow-up of the patients, 2) the collaboration rate between GPs and nephrologists, and 3) the progression of CKD.

Results: The rate of discontinuous clinical visits was significantly lower in group B (16.2% in group A vs. 11.5% in group B, p = 0.01). Significantly higher referral and co-treatment rates were observed in group B (p<0.01). The average eGFR deterioration rate tended to be lower in group B (group A: 2.6±5.8 ml/min/1.73 m2/year, group B: 2.4±5.1 ml/min/1.73 m2/year, p = 0.07). A significant difference in eGFR deterioration rate was observed in subjects with Stage 3 CKD (group A: 2.4±5.9 ml/min/1.73 m2/year, group B: 1.9±4.4 ml/min/1.73 m2/year, p = 0.03).

Conclusion: Our care system achieved behavior modification of CKD patients, namely, significantly lower discontinuous clinical visits, and behavior modification of both GPs and nephrologists, namely significantly higher referral and co-treatment rates, resulting in the retardation of CKD progression, especially in patients with proteinuric Stage 3 CKD.

Trial registration: The University Hospital Medical Information Network clinical trials registry UMIN000001159.

Fig 1. Study clusters and patients.

We recruited 49 local medical associations (clusters) in 15 different prefectures, which were classified into four regions (strata) based on the level of increase in the rate of dialysis patients [9]. We recruited 557 GPs and 2,417 patients; 2,379 patients from 489 GPs gave consent. After randomization, 68 patients in group B chose to withdraw, while only 13 patients in group A did so. Most of the patients in group B withdrew just after randomization due to an aversion to the educational intervention. Finally, 1,107 patients in group A and 1,029 patients in group B completed this cluster-randomized trial.

Fig 2. Effects of advanced CKD care system on CKD treatment targets.

Items were BMI changes (A), HbA1c changes (B), systolic (C) and diastolic blood pressure changes (D), non-HDL cholesterol changes (E), and hemoglobin changes in subjects with Stage 3 CKD or later (F) during the study period. Patients in group B had greater improvements than those in group A in terms of the CKD practice guide targets, except for hemoglobin level. In particular, BMI, HbA1c, and blood pressure differences between group A and group B gradually increased over the time course of the interventions. Average BMI was significantly reduced in group B patients 2 years after starting the advanced intervention, and HbA1c was also significantly reduced in group B patients 2.5 years after starting the advanced intervention. * indicates p<0.05 between group A and group B by Student’s T test.

Fig 3. CKD treatment control achievements during the study period.

CKD treatment cumulative achievements for BMI <25 (A), HbA1c <6.9% (B), controlled blood pressure (C), smoking cessation (D), non-HDL-C <150 mg/dl (E), and proportion of hemoglobin concentration controlled to 10–12 g/dl among CKD stages 3, 4, and 5 (F). There was no significant difference between group A and group B in CKD treatment control achievements by generalized linear mixed model. Patients in group B had greater cumulative incidences of achieving the targets of BMI <25, glycated hemoglobin <6.9%, and other measured risk factors, except for non-HDL cholesterol level and smoking cessation rate. In particular, cumulative differences in the rates of achieving BMI, HbA1c, and blood pressure targets between group A and group B gradually widened over the course of the interventions.

Fig 4. Renal function outcomes.

Both the number of subjects with a doubling of serum creatinine and the number with a 50% reduction in eGFR significantly decreased in group B. The cumulative incidence of a doubling of serum creatinine or a 50% reduction in eGFR also exhibited a gradually increasing difference between groups A and B over the course of the study. A significant difference was shown between Group A and Group B by generalized linear mixed model.