Evaluation of a Tiered Potassium Replacement Protocol in Post-Operative Patients Admitted to a Pediatric Cardiac Intensive Care Unit

Introduction

Potassium is an intracellular cation that plays a major role in many physiologic functions, including the regulation of cardiovascular electrical activity.^1,2 A disruption in the balance between intracellular and extracellular (3.5–5.5 mEq/L) concentrations can have significant and life threatening consequences.¹

Hypokalemia is considered one of the most common electrolyte disturbances in critically ill patients and can cause disruptions in the electrical activity of cardiac, skeletal, and smooth muscles. These effects may lead to arrhythmias, cardiac arrest, respiratory failure, and muscular paralysis.^2,3 Pediatric patients undergoing cardiovascular surgery are highly susceptible to hypokalemia in the post-operative period. During this time, patients often require vasopressors, inotropes, and high-dose diuretics, which can contribute to hypokalemia by shifting potassium intracellularly or disturbing the renal mechanisms for regulation. In addition, the clinical picture is often further clouded by the dilutional effect seen during procedures requiring cardiopulmonary bypass (CPB) as well as the need for intravenous (IV) fluid management until enteral feeds begin.^1,3,4

Pediatric patients undergoing cardiovascular surgery are also highly susceptible to hyperkalemia in the post-operative period due to the use of blood products, potential for over-supplementation, and the risk of transient acute kidney injury after CPB.³ Hyperkalemia can be further classified into mild (serum potassium 5.5–6.5 mEq/L), moderate (serum potassium 6.6–7.5 mEq/L) and severe (serum potassium greater than 7.5 mEq/L).¹ In most cases, the cause of hyperkalemia can be attributed to either an impaired elimination of potassium or a shift of potassium extracellularly.¹

A retrospective cohort review published in 2020 by Amirnovin et al³ compared the safety and efficacy of 3 separate potassium replacement protocols in pediatric patients who were admitted to the cardiac intensive care unit post-operatively. The authors assessed a low-dose protocol (using 0.5 mEq/kg of IV potassium chloride for potassium concentrations less than or equal to 3.2 mEq/L), a high-dose protocol (using 1 mEq/kg of IV potassium chloride for potassium concentrations less than or equal to 3.2 mEq/L), and a tiered-dose protocol (using 0.5 mEq/kg of IV potassium chloride for potassium concentrations 2.9 to 3.2 mEq/L and 1 mEq/kg of IV potassium chloride for potassium concentrations less than or equal to 2.8 mEq/L) with a maximum dose of 10 mEq in all 3 protocols. The authors concluded that the tiered-dose protocol improved the efficacy of the low-dose protocol while maintaining its safety profile.³

In 2019, our institution implemented a tiered-dose potassium replacement protocol, presented in Table 1, utilizing 0.5 mEq/kg (maximum of 20 mEq/dose) of IV potassium chloride for serum potassium concentrations of 3.3 to 3.7 mEq/L (Tier 1) and 1 mEq/kg (maximum of 40 mEq/dose) of IV potassium chloride for serum potassium concentrations of less than or equal to 3.2 mEq/L (Tier 2). The replacement dose is administered over 2 hours and a repeat serum potassium concentration is drawn 1 hour after the end of infusion. This protocol is ordered for all patients post-operatively, unless omitted by the physician due to patient characteristics warranting against use, such as patients with significant renal impairment or requiring renal replacement therapy. Routine potassium concentrations are obtained per provider order and if hypokalemia is present, the nurse requests the appropriate protocol replacement dose. This protocol was implemented due to an increased incidence of junctional ectopic tachycardia observed in this population in an effort to be more proactive and have a higher sensitivity of capturing hypokalemia. Our protocol is also entirely nurse-driven with no maximum number of doses. This design allows the nurse to continue giving replacement doses as needed to resolve the hypokalemia without additional provider orders, to help eliminate any potential delay in serum electrolyte correction. In contrast to the protocol evaluated by Amirnovin et al,³ this protocol utilizes a more aggressive dosing strategy requiring supplementation for potassium concentrations less than 3.7 mEq/L and allowing up to 40 mEq per dose. The purpose of this study was to evaluate the efficacy and safety of our potassium replacement protocol in pediatric patients admitted post-operatively to the pediatric cardiac intensive care unit.

Table 1.Institutional Intravenous Potassium Replacement Protocol

View Fullscreen

Materials and Methods

Results

Study Population.

Demographic data and clinical characteristics for patients are presented in Table 2. Of the 40 patients screened for inclusion, 17 patients were excluded. The reasons for exclusion included: age greater than or equal to 18 years of age (6 patients) and protocol not followed (11 patients). Patients classified as not following the protocol included those with dose rounding greater than 10% (3 patients), those who did not receive a potassium replacement dose within the specified time frame (5 patients), and those in which no episodes of hypokalemia were treated following the protocol (3 patients).

Table 2.Baseline Patient Characteristics

View Fullscreen

In the 23 patients who met inclusion, a total of 153 hypokalemia episodes were identified with 95 episodes meeting inclusion criteria. The primary reason for episode exclusion was due to not following the protocol (43 episodes), followed by repeat serum potassium concentration being drawn too early (11 episodes) and no serum potassium concentration being drawn within 6 hours of administering the replacement dose (4 episodes). Episodes grouped as not following the protocol were most commonly excluded due to no protocol dose being administered for a hypokalemia concentration or the administered replacement dose did not match the protocol (see Supplemental Figure). Concomitant medications that are known to cause hypokalemia or hyperkalemia for each episode are presented in Table 3. Doses of diuretics utilized during episodes of hypokalemia are presented in Table 4.

Table 3.Important Concomitant Medications Associated With Hypokalemic Episodes

View Fullscreen

Table 4.Diuretic Doses Utilized During Hypokalemic Episodes

View Fullscreen

Efficacy.

Of the 95 episodes of hypokalemia included, 76 (80%) episodes had an initial serum potassium concentration between 3.3 and 3.7 mEq/L (Tier 1) and 19 (20%) episodes had an initial serum potassium concentration that was less or equal to than 3.2 mEq/L (Tier 2) (Table 5).

Table 5.Primary Outcome

View Fullscreen

For Tier 1 episodes, the average number of doses needed to resolve hypokalemia was 1.3 (SD 0.36) with a median of 1. The average dose, in mEq/kg, to resolve hypokalemia was 0.68 (SD 0.41) with a median of 0.5.

For Tier 2 episodes, the average number of doses needed to resolve hypokalemia was 2.05 (SD 2.44) with a median of 1. The average dose, in mEq/kg, to resolve hypokalemia was 1.92 (SD 2.28) with a median of 1. The average values for this group were largely influenced by 3 outlier episodes. One of these outlier episodes required a total of 4 doses (4 mEq/kg), 1 of these outlier episodes required a total of 5 doses (4 mEq/kg) and 1 of these outlier episodes required a total of 11 doses (10.5 mEq/kg) to correct the hypokalemia.

Safety.

A total of 9 incidences of hyperkalemia were identified (Table 6). Five incidences of hyperkalemia were determined to be false with proven or suspected hemolysis as a repeat concentration was performed within 10 minutes of the elevated result and returned within normal limits. One incidence of hyperkalemia was noted to be drawn during the infusion of a potassium replacement dose and another was drawn during administration of massive transfusion protocol. Due to this, both incidences were excluded. The remaining 2 incidences (2.1% of the total episodes) were suspected or proven to be true hyperkalemia, but neither was associated with ECG abnormalities or required treatment. Of the 2 incidences of hyperkalemia, only 1 was associated with a high-dose (Tier 2) replacement.

Table 6.Secondary Outcomes

View Fullscreen

Discussion

We used a serum potassium greater than 3.7 mEq/L after replacement as a marker of efficacy. Sixty episodes (78.9%) with an initial serum potassium concentration of 3.3 to 3.7 mEq/L (Tier 1) were corrected after 1 dose. Thirteen episodes (68.4%) with an initial serum potassium concentration less than 3.2 mEq/L (Tier 2) were corrected after 1 dose. For both tiers of our potassium replacement protocol, the median number of IV doses to correct hypokalemia was 1, which indicates that our protocol is effective. The average dose to resolve hypokalemia in Tier 2 episodes was largely influenced by 3 outliers. For each of these episodes, patients were receiving diuretics at doses higher than the mean for the population, which may have impacted the need for additional potassium replacements. The types of diuretics and mean doses observed in this study are consistent with our unit practice.

The study conducted by Amirnovin et al³ found a hyperkalemia rate, which they defined as a serum potassium concentration greater than or equal to 4.8 mEq/L, of 0.5% (associated with 3 doses) with the tiered-dose protocol. They utilized a high-normal potassium concentration as a marker for safety to increase sensitivity. We found a hyperkalemia rate of 2.1% (associated with 2 doses) with our tiered-dose protocol, which included no incidences of severe hyperkalemia. Our protocol, although more aggressive than that studied by Amirnovin et al,³ does not appear to lead to clinically significant higher rates of hyperkalemia for our population.

There was significant variability in time to administration of an IV potassium replacement dose given inconsistencies in our unit practice. Our protocol is entirely nurse driven and requires the nurse to request a replacement dose to be made and sent from the pharmacy. Some nurses proactively request replacements to have available on the unit when needed, but this is not standard practice nor part of our protocol, making it entirely dependent on the nurse and the patient’s presumed level of acuity. This also varies with time, as nurses are more likely to maintain replacement doses on the unit for the first 24 to 48 hours after surgery compared with post-operative day 5. There was less variability in time to repeat serum potassium concentration. Our protocol defaults to a 2-hour infusion time for each dose with a minimum infusion time of 1 hour. Nurses are prompted to repeat a serum potassium concentration 1 hour after the replacement dose completes. Given this, an average time to repeat concentration of 3.18 hours from the start of infusion aligns with our protocol.

This study adds to the limited existing literature for potassium replacement in this population. However, given that this is a retrospective and observational study, the ability to capture safety parameters, such as ECG abnormalities, is restricted, and our study lacked a comparison group. This study was also performed after a recent change in our institution’s electronic health record which limited the time frame for data collection. This study followed patients through post-operative day 5 and did not evaluate efficacy or safety in non-surgical pediatric patients admitted to the pediatric cardiac intensive care unit. The median age of 3.5 months in this study may limit the generalizability to other pediatric cardiac intensive care units. Limitations also exist in the ability to accurately measure serum potassium and the level of hemolysis present, as our unit commonly uses point of care testing via samples obtained for blood gas analysis for repeat serum potassium concentrations. Additionally, we were unable to quantify the impact of potassium administration outside of the protocol, such as through intravenous fluids or parenteral nutrition, as these varied throughout the study period and were impacted by many other factors. Our unit largely relies on nurses and pharmacists to monitor and dispense potassium replacement doses, which could be an area of opportunity to improve timely potassium replacement. At the time this study was conducted, to our knowledge, there was no published literature evaluating the higher parameters used for our potassium replacement protocol or a completely nurse-driven protocol. Future prospective studies with a larger sample size may be warranted.

Conclusion

The presence of hypokalemia is common in pediatric patients following cardiac surgery. This study aimed to evaluate the efficacy and safety of our tiered IV potassium infusion replacement protocol in pediatric patients admitted post-operatively to the pediatric cardiac intensive care unit. This protocol utilizes higher serum potassium concentrations for replacement and higher maximum dosing than published protocols. We found our protocol to be effective at resolving hypokalemia, as defined by our protocol, with minimal hyperkalemia that was comparable to prior studies.