Abstract
INTRODUCTION: Patient adherence is important for long-term outcomes of peritoneal dialysis (PD). Artificial intelligence is a good tool to manage patients. However, there are limited data regarding its impact on the patient adherence and the effect of patient adherence on serum phosphate, calcium, and intact parathyroid hormone (iPTH) control in PD patients. METHODS: This was a single-center, prospective cohort study including PD patients in Guangdong Provincial People's Hospital. Adult patients (age ≥18 years) who were included in the smart PD care program from September 1, 2020, to April 31, 2023, were enrolled. Patient adherence was assessed using the patient-reported daily PD prescription data and calculated as the total days with PD ultrafiltration reported divided by the total days of follow-up. Good adherence was defined as the reporting rate ≥80%. The primary outcome was serum phosphate, calcium, and iPTH values achieved the treatment targets at 12-month follow-up. Unadjusted and adjusted generalized estimating equations were used to evaluate the association of patient adherence with the serum phosphorus, serum calcium, and iPTH control. RESULTS: A total of 267 patients were included in this study. The mean age of the whole cohort was 43.3 ± 12.8 years, 130 (48.7%) were females, and 52 (19.5%) had diabetes. Patient adherence improved after being included in the smart PD care program and the overall patient adherence during 12-month follow-up was 77.1% ± 26.4%, 93.0% ± 7.4%, and 50.9% ± 25.6% for the entire cohort, patients with good adherence, and those with poor adherence, respectively. Compared to patients with poor adherence, those with good adherence were associated with a better serum calcium (adjusted OR: 3.76; 95% CI: 2.67-5.30; p < 0.001) and iPTH control (adjusted OR: 2.20; 95% CI: 1.56-3.11; p < 0.001) but not for serum phosphorus control (adjusted OR: 1.31; 95% CI: 0.89-1.91; p = 0.17) after being adjusted for potential confounders. Results were similar when assessing the relationship between patient adherence and the longitudinal changes of serum calcium, iPTH, and phosphorus during follow-up. CONCLUSIONS: Smart PD care program was effective in improving patient adherence. Good patient adherence was associated with better serum calcium and iPTH control but not for phosphorus control in PD patients. Further studies should be done to evaluate the effect of the smart PD care program on long-term patient outcomes.