Deep learning algorithms for predicting renal replacement therapy initiation in CKD patients: a retrospective cohort study.

BACKGROUND: Chronic kidney disease (CKD) requires accurate prediction of renal replacement therapy (RRT) initiation risk. This study developed deep learning algorithms (DLAs) to predict RRT risk in CKD patients by incorporating medical history and prescriptions in addition to biochemical investigations. METHODS: A multi-centre retrospective cohort study was conducted in three major hospitals in Hong Kong. CKD patients with an eGFR < 30ml/min/1.73m2 were included. DLAs of various structures were created and trained using patient data. Using a test set, the DLAs' predictive performance was compared to Kidney Failure Risk Equation (KFRE). RESULTS: DLAs outperformed KFRE in predicting RRT initiation risk (CNN + LSTM + ANN layers ROC-AUC = 0.90; CNN ROC-AUC = 0.91; 4-variable KFRE: ROC-AUC = 0.84; 8-variable KFRE: ROC-AUC = 0.84). DLAs accurately predicted uncoded renal transplants and patients requiring dialysis after 5 years, demonstrating their ability to capture non-linear relationships. CONCLUSIONS: DLAs provide accurate predictions of RRT risk in CKD patients, surpassing traditional methods like KFRE. Incorporating medical history and prescriptions improves prediction performance. While our findings suggest that DLAs hold promise for improving patient care and resource allocation in CKD management, further prospective observational studies and randomized controlled trials are necessary to fully understand their impact, particularly regarding DLA interpretability, bias minimization, and overfitting reduction. Overall, our research underscores the emerging role of DLAs as potentially valuable tools in advancing the management of CKD and predicting RRT initiation risk.

Use of intraperitoneal urokinase for resistant bacterial peritonitis in continuous ambulatory peritoneal dialysis.

Intraperitoneal (IP) urokinase is a fibrinolytic agent that has been used in the adjunctive treatment of continuous ambulatory peritoneal dialysis (CAPD) and resistant and relapsing peritonitis. However, its efficacy and role in treating resistant CAPD bacterial peritonitis remain unclear and results from previous prospective studies have been conflicting. We prospectively randomized 88 CAPD patients with bacterial peritonitis resistant to initial empirical IP antibiotics into two groups: IP urokinase 60,000 IU and a placebo group. Patients were treated concomitantly with susceptible antibiotics according to culture results. Peritoneal dialysate grew pseudomonas aeruginosa in 13 patients (14.8%), non-pseudomonas bacteria in 63 patients (71.6%) and negative cultures in 12 patients (13.6%). For the clinical outcomes, there were no significant differences in the primary response rates (61.4 vs. 50%), relapse rates (9.1 vs. 13.6%), Tenckhoff catheter removal rates (22.7 vs. 29.5%) and mortality rates (6.8 and 6.8%) between the urokinase group and the controls (p=ns). Subgroup analysis of culture negative patients (n=12) also demonstrated no sgnificant benefit for urokinase treatment. No significant adverse effects were encountered with the IP urokinase instillation. Total median peritonitis-related length of hospitalization for the urokinase group and controls were 7 and 11 days, respectively (p=0.32). We concluded that IP urokinse plays no significant role as an adjuvant therapy in the treatment of bacterial CAPD peritonitis resistant to initial IP antibiotic therapy.

The use of corticosteroid as treatment in SARS was associated with adverse outcomes: a retrospective cohort study.

OBJECTIVE: To study the effect of corticosteroids in the treatment of severe acute respiratory syndrome (SARS). METHODS: A retrospective cohort of 78 consecutive adult SARS patients admitted to a regional hospital in Hong Kong between March and May 2003 was analysed to study the effectiveness of corticosteroid. They were categorized according to whether or not corticosteroid therapy was given, and compared in terms of demographic characteristics, comorbidities, peak lactate dehydrogenase (LDH) levels and clinical outcomes. Established adverse prognostic factors including old age, comorbidities and high LDH levels were used as covariates in multiple logistic regressions to adjust for their confounding effect on adverse outcomes. RESULTS: Among 78 patients, 66 patients (84.6%) received corticosteroid. The LDH level was similar in both groups. The corticosteroid group had more adverse outcomes (37.9% vs. 16.7%) despite younger age and less comorbidity. In multivariate analysis, corticosteroid treatment was associated with a 20.7-fold increase in risk of either ICU admission or mortality, independent of age and disease severity. CONCLUSION: Despite more favourable baseline characteristics and similar peak LDH levels, SARS patients given corticosteroid had more adverse outcomes.