Introduction: National drug shortages result in numerous challenges for clinicians, including the need to make practice changes to limit or conserve supply. With the ongoing shortage of all intravenous phosphate products, our institution had to change practice by revising the CRRT phosphorus replacement protocol. Once sodium phosphate became unavailable, potassium phosphate was used for replacement and the serum phosphorus level threshold for when phosphate is replaced was changed. Previously, the replacement threshold was levels less than 2.9 mg/dL. The revised protocol has a replacement threshold for levels less than 1.6 mg/dL (severe hypophosphatemia). Methods: Single center, retrospective chart review in which phosphate replacement was analyzed in 26 patients, age 18 years and older, who received CRRT during a six week period. Data collection included patient demographics, serum potassium and phosphorus levels before and after phosphate replacement, adverse effects from hypophosphatemia, and all sources of potassium and phosphorus administered. Results: Of the 26 patients on CRRT, 6 (31%) required phosphorus replacement for severe hypophosphatemia. A total of 156 episodes of hypophosphatemia, defined as serum phosphorus less than 2.9 mg/dL, noted in all CRRT patients. Of these 156 episodes, 21 were severe (less than 1.6 mg/dL) meeting the criteria for replacement with the revised protocol. Average increase in serum phosphorous level was 0.67 mg/dL after replacement, with 100% of levels correcting to a moderate hypophosphatemia (1.6-2.2mg/dL). None of the patients experienced hyperkalemia after receiving a replacement dose. There were no documented adverse events in this patient population. Conclusions: Potassium phosphate was able to effectively replace sodium phosphate for use in our CRRT population without causing any incidents of hyperkalemia. It was also determined that supplies of intravenous phosphate could safely be conserved by lowering the threshold for treatment of hypophosphatemia to 1.6 mg/dL. We also determined that changing the threshold for replacement did not result in any adverse events.