Abstract: FR-PO650
Reflex Ordering of Polystyrene Sulfonate for "Red-Flagged" Mild Hyperkalemia: A Problem of Perception
Session Information
- Fluid and Electrolytes: Clinical - Potassium, Sodium, Water
November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Fluid and Electrolytes
- 902 Fluid and Electrolytes: Clinical
Authors
- Hirsch, Jamie S., Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, United States
- Sakhiya, Vipulbhai, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, United States
- Bock, Kevin R., Northwell Health, New Hyde Park, New York, United States
- Parikh, Rushang, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, United States
- Barnett, Richard L., Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, United States
- Jhaveri, Kenar D., Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, United States
Background
Lab test results, typically reported with reference ranges, are often flagged and red when abnormal. For potassium (K), while treatment recommendations have not changed, reference limits may have. How the reference upper limit (RUL) impacts provider responsiveness to hyperkalemia treatment with polystyrene sulfonate (PS, Kayexalate) is unknown.
Methods
We queried the EHR at 14 hospitals in a large integrated health system, for orders of PS between 2012-2018. The serum K at the time of an order was deemed the result that occasioned treatment. We extracted the RUL for the result, which changed from 5.1 to 5.3 mEq/L in December, 2015. We compared PS ordering based upon RUL using a two-sample Kolmogorov–Smirnov (KS) test. KS is a nonparametric test to determine if two samples are drawn from the same distribution; the null hypothesis is that the distribution of orders by K level remains the same regardless of RUL.
Results
There were 43,497 orders for PS, almost evenly split between RUL 5.1 and 5.3. The two distributions were statistically different (D = 0.17478, p-value < 0.0001). For results with RUL 5.1, the initial peak in PS ordering occurred at K 5.2, whereas it occurred at 5.4 with RUL 5.3. The maximum absolute difference between the two cumulative curves (D statistic), 0.17, occurs at K 5.3, falling within the “high” range for RUL 5.1, but normal for RUL 5.3. For RUL 5.1, approximately 10% of results between 5.2-5.3 led to a PS order, whereas only 2% led to an order for RUL 5.3.
Conclusion
While the physiologic understanding of K and its serum levels have not changed, we found that modifying the RUL led to major changes in PS prescribing for hyperkalemia. This abrupt and sustained shift suggests a reflexive approach to treatment, whereby providers are ordering PS based upon an abnormal flag or red-colored value, rather than clinical indication. Educating providers and using appropriate nudges or defaults can lead to more thoughtful approaches to management of K and other electrolyte disorders.