P0637SAFETY, EFFICACY AND OUTCOMES OF SLOW LOW EFFICIENCY DIALYSIS (SLED) IN CRITICALLY ILL PATIENTS WITH ACUTE KIDNEY INJURY (AKI) IN MEDICAL INTENSIVE CARE UNITS (ICU)- A PROSPECTIVE STUDY IN A TERTIARY CARE HOSPITAL IN INDIA

Background and Aims Slow low efficiency dialysis (SLED) is increasingly being used as a renal replacement therapy in hemodynamically unstable, critically ill patients with acute kidney injury (AKI). SLED may reduce intradialytic hemodynamic instability as compared with conventional intermittent haemodialysis, while reducing resource demands of continuous renal replacement therapies (CRRT). There are very few studies which have evaluated its safety, efficacy and outcome despite its increasing use, especially in Indian subcontinent. Method We conducted a single centre, prospective study to demonstrate safety, efficacy and outcome of SLED.. Net fluid removal and duration of SLED were based on need and hemodynamic status of the individual patient as decided by treating nephrologist. SOFA score was used as severity illness score. Efficacy of SLED was studied in terms of ability to achieve ultrafiltration goal, correction of acidosis and urea reduction ratio. Safety was studied in terms of hemodynamic and cardiovascular stability and complications during and after SLED. Outcomes were noted at time of discharge and six months later in terms of dialysis dependence, renal parameters (if dialysis independent) and mortality (In hospital and 6 months following discharge). We included hemodynamically unstable patients with AKI from medical ICUs with age >18 years of either gender. We excluded patients with AKI who could tolerate conventional haemodialysis or ESRD on maintenance haemodialysis. Results We analysed 228 patients with AKI in medical ICUs who underwent 576 SLED sessions over period of six months. Mean age was 57.48 ± 15.67 years and 74% (n=169) were male. Comorbidities were hypertension (56%), diabetes mellitus (43%), CKD (33%) and cardiovascular diseases (28%). Sepsis (93%) and hypoperfusion (68%) were most common causes for AKI. Refractory fluid overload (91%) and refractory metabolic acidosis (79%) were most common indications for SLED. Mean SOFA score was 12.2 ± 7.75 with 61 % patients had SOFA score more than 11. 66% were on mechanical ventilator. Out of 576 sessions, 555 sessions (96%) completed the planned duration without any adverse event. Planned ultrafiltration goal was achieved in 94%. SLED was able to correct metabolic acidosis in majority (86.1%). 68% SLED sessions required a vasopressor support and 34.8 % of SLED sessions were associated with hemodynamic instability. Total 14(6.2 %) patients died during SLED session. No documented arrhythmias developed after starting SLED. In-hospital mortality occurred in 61% patients. At 6 months follow up, another 13% patients died. In subgroup analysis, mortality was significantly higher in patients with SOFA score more than 11 (P<0.0001). Ventilatory requirement was also significantly high in non-survivors(P<0.0001). Univariate logistic regression analysis showed that inotropic requirement, higher SOFA severity score, acidosis with pH <7.25 and presence of underlying CKD were associated with significant mortality. Conclusion Our study demonstrated efficacy and safety of SLED in critically ill AKI patients in medical ICU. SLED was able to achieve planned ultrafiltration goal and correct metabolic acidosis in majority of patients. SLED had good hemodynamic tolerability. Mortality was noted in 61 % of patients (not attributable to SLED per se). High SOFA score, underlying co morbidities, vasopressor requirement and severe acidosis (pH <7.25) were associated with high mortality. SLED is a reasonable cost-effective option of RRT in hemodynamically unstable patients with AKI especially in developing countries.