Abstract
Background:
Enteral phosphate replacement offers multiple advantages over intravenous phosphate replacement, but optimal dosing and administration are unclear.
Objective:
This study aimed to compare 2 different dosing strategies (PRE vs POST) of enteral phosphate replacement for mild hypophosphatemia (serum phosphorus = 2.1-2.5 mg/dL) and to evaluate our dosing strategy for moderate hypophosphatemia (serum phosphorus = 1.5-2 mg/dL) in critically ill patients.
Methods:
This single-center quasiexperimental, institutional review board–approved study was conducted in adult patients admitted to an intensive care unit (ICU) who received enteral phosphate replacement with a follow-up serum phosphorus level obtained within 24 hours of replacement. Patients were excluded if they had diabetic ketoacidosis, acute kidney injury, or received total parenteral nutrition or targeted temperature management. The mild PRE and POST groups received 48 and 72 mmol of enteral phosphate, respectively. The moderate group received 72 mmol of enteral phosphate. Effectiveness outcomes included normalization of serum phosphorus, defined as 2.6 to 4.5 mg/dL, and median change in phosphorus concentration. Safety outcomes included the incidence of new onset diarrhea and hyperphosphatemia (serum phosphorus >4.5 mg/dL) within 24 hours after phosphate replacement.
Results:
Of the 138 patients included, the median age was 59 (44-70) years and 69.6% were admitted to the surgical trauma ICU. Normalization of serum phosphorus occurred in 55 (58.2%) patients in mild PRE and 44 (72.7%) patients in mild POST (P = 0.13). Change in serum phosphorus was 0.4 (0.1-0.9) mg/dL in mild PRE vs 0.7 (0.3-1.2) mg/dL in mild POST (P = 0.06). Normalization occurred in 39 (66.7%) patients in the moderate group with a change in serum phosphorus of 1.0 (0.6-1.6) mg/dL. Rates of new onset diarrhea and hyperphosphatemia were not significantly different between groups.
Conclusion and Relevance:
Enteral phosphate replacement appears safe and effective in critically ill patients with mild-to-moderate hypophosphatemia and may serve as an alternative to intravenous phosphate.
Introduction
Hypophosphatemia, or low serum phosphorus, is frequently encountered in critically ill patients, occurring in 28% to 100% of medical and surgical intensive care unit (ICU) patients.1-4 A wide range of factors may affect serum phosphorus concentration in the ICU, such as malnutrition, refeeding syndrome, medications, and gastrointestinal tract losses.4-7
The clinical importance of hypophosphatemia has been debated. There is no high-quality evidence indicating the correction of hypophosphatemia leads to improved outcomes. Most patients with hypophosphatemia are asymptomatic, although complications such as ventilatory muscle weakness, decreased myocardial contractility, arrhythmias, hemolysis, rhabdomyolysis, and seizures have been reported.4,8-10 Serum phosphorus levels less than 2.5 mg/dL have been associated with a greater risk of failure to wean mechanical ventilation. 11 Correction of hypophosphatemia has been shown to attenuate these complications.3,9,12 Despite the lack of standardized guidelines to treat hypophosphatemia in the ICU, phosphate replacement remains common practice. 4
Hypophosphatemia is often corrected using intravenous formulations in clinical practice, presumably because previous studies have primarily evaluated intravenous phosphate formulations and intravenous administration ensures 100% bioavailability.13-15 Intravenous formulations, however, are more costly than enteral formulations and require a large infusion volume (eg, 250-500 mL), which can contribute to fluid overload. 16 In addition, intravenous phosphate requires a prolonged duration of administration, introducing issues with line access and compatibility.7,16,17 Recent studies have suggested that enteral phosphate replacement may be an alternative to intravenous phosphate. However, a protocol-driven approach to enteral phosphate replacement has not been well established.17-19
The use of electrolyte replacement protocols, including phosphorus, in intensive care settings improves timeliness and effectiveness of replacement without increasing the occurrence of adverse events. 13 Beginning in 2018, a critical care electrolyte replacement protocol (CCERP) was implemented at our institution to address potassium, magnesium, and phosphate replacement (Table 1). The protocol was built into a nursing-driven order set, which is initiated with a provider order and allows nurses to replace electrolytes as needed according to lab results. The order set remains active as long as the patient does not develop contraindications to CCERP use or once the provider order for order set use is discontinued (eg, on ICU discharge, contraindication arises). The CCERP incorporated the use of enteral phosphate supplementation (potassium and sodium phosphate powder, or Phos-NaK) for hypophosphatemia. We previously examined our dosing strategy in which we found that only 56% of patients with a serum phosphorus of 2.1 to 2.5 mg/dL receiving enteral phosphate replacement achieved normalization of their serum phosphorus. 20 We subsequently increased the enteral phosphate dosing in our protocol, and we aim to evaluate the effectiveness and safety of this modified enteral phosphate dosing strategies to correct mild-to-moderate hypophosphatemia (serum phosphorus level = 1.5-2.5 mg/dL) at our institution.
Institutional Adult Critical Care Electrolyte Replacement Protocol (CCERP) for Enteral and Intravenous Sodium Phosphate Replacement.
CCERP Exclusions:
Acute kidney injury (urine output less than 0.5 mL/kg/h for past 6 hours) or creatinine clearance < 30 mL/minute
Renal replacement therapy (hemodialysis/continuous renal replacement therapy/peritoneal dialysis)
Pediatrics (age < 18 years)
Diabetic ketoacidosis
Targeted temperature management
CCERP Monitoring:
If serum phosphorus concentration < 1.5, recheck 2 hours after intravenous infusion completion
If serum phosphorus concentration ≥ 1.5, recheck with morning labs the following day
Materials and Methods
A single-center quasiexperimental study was performed to assess consecutive adult patients (18 years or older) admitted to an ICU (surgical trauma, medical, neurological, or cardiac) who received Phos-NaK supplementation utilizing the CCERP order set and had a follow-up serum phosphorus level obtained within 24 hours of replacement. The electronic medical records (EPIC Systems, Verona, WI) of all patients who met these criteria were evaluated for inclusion. Our institutional review board approved this retrospective review, and the need for informed consent was waived.
Patients were excluded if they received total parenteral nutrition, targeted temperature management, renal replacement therapy, intravenous phosphate replacement within 24 hours of enteral phosphate replacement or were diagnosed with diabetic ketoacidosis or acute kidney injury, defined as urine output less than 0.5 mL/kg/h for 6 consecutive hours, or creatinine clearance less than 30 mL/min calculated via Cockcroft-Gault equation. Patients were also excluded if they received less than 3 doses of Phos-NaK ordered using the CCERP order set. For patients who received multiple treatment courses of enteral phosphate within the same admission, only the earliest course was included for analysis to minimize the potential contribution of phosphate replacement from an earlier course.
Phos-NaK (Cypress Pharmaceutical Inc., Madison, MS 39110) was initiated for the management of hypophosphatemia in patients with oral or gastric/small bowel access and a serum phosphorus concentration between 1.5 and 2.5 mg/dL per our institutional CCERP. Each packet contains 8 mmol of elemental phosphate, 7.1 mEq of potassium and 6.9 mEq of sodium. 21 Patients from two 12-month periods who received 2 or 3 packets of Phos-NaK every 4 hours for at least 3 doses were evaluated. The mild PRE group represents the original CCERP and included patients admitted between January 1 and December 31, 2019 with mild hypophosphatemia (defined as serum phosphorus concentration 2.1-2.5 mg/dL) and received 48 mmol of elemental phosphate (Phos-NaK 2 packets every 4 hours for 3 doses). In May 2021, the CCERP was updated to treat patients with mild hypophosphatemia with 72 mmol of elemental phosphate (Phos-NaK 3 packets every 4 hours for 3 doses). The mild POST group represents the modified CCERP and included patients admitted between June 1, 2021 and May 31, 2022. The moderate group included patients from both time periods with moderate hypophosphatemia (defined as serum phosphorus concentration 1.5-2.0 mg/dL) who received 72 mmol of elemental phosphate. Enteral phosphate was administered in divided doses to limit gastrointestinal intolerance. A follow-up serum phosphorus concentration was automatically ordered with routine morning labs following enteral phosphate replacement per our institutional protocol.
A case report form was created a priori and was not modified between the 2 timeframes. The PRE cohort was collected by 2 investigators (JTL and KES) and the POST cohort was collected by 2 investigators (KMW and KES) with KES ensuring consistency in data collection across the 2 timeframes. A third party (CW) was available for both cohorts to adjudicate discrepancies. Certain data points were secured via a report from electronic health record (eg, age, hospital and ICU lengths of stay, and Phos-NaK administration), the rest were obtained via manual chart review. Baseline demographic and clinical characteristics included age, sex, race, body mass index, admitting ICU service, nutritional intake, mechanical ventilation, hospital and ICU lengths of stay, ICU mortality, serum creatinine on the day of phosphate replacement, medications known to lower serum phosphorus concentration, and risk factors for hypophosphatemia. Medications included those that cause intracellular shifts of phosphorus (catecholamines and insulin), those that increase urinary excretion of phosphorus (loop diuretics, thiazide diuretics, acetazolamide, and mannitol), and phosphate binders (sucralfate and calcium carbonate). 22 Risk factors collected were pancreatitis, sepsis, malnutrition/refeeding syndrome, alcohol use disorder, burn injury, or trauma/traumatic brain injury and were identified based on documentation in the medical record.4-7
The primary effectiveness outcomes of this study were the normalization of serum phosphorus following enteral phosphate replacement, defined as 2.6 to 4.5 mg/dL by our institutional laboratory, and median change in serum phosphorus concentrations. Patients who did not achieve normalization of serum phosphorus were designated as treatment failures. Treatment failure was further subcategorized as no change in hypophosphatemia severity, the severity of hypophosphatemia worsened, or additional phosphate supplementation was required within 48 hours of initial replacement. Safety outcomes included the incidence of hyperphosphatemia, defined as a serum phosphorus concentration greater than 4.5 g/dL, and the incidence of new onset diarrhea, defined as 3 or more occurrences of Bristol type 7 stools or 250 mL or more of stool per day if a fecal management system was used, 23 as documented in the nursing flowsheet and collected 24 hours before and 24 hours after phosphate replacement.
Data are summarized using descriptive statistics, both overall and after stratification by study subgroups. Continuous data are expressed as median (interquartile range). Categorical data are summarized as count (percentage). Data analysis was performed using R version 4.4.1 (2024-06-14 ucrt). Comparison of the mild PRE and mild POST groups was conducted using the Pearson chi-square with 2-sided tail test for the incidence of normalization and the Wilcoxon rank-sum test for change in serum phosphorus. Fisher’s exact test was used for any data point with an n < 5. A P value of <0.05 was considered statistically significant.
Results
Baseline Characteristics
A total of 176 patients were assessed for eligibility in the two 12-month periods after implementation of the original and modified CCERPs, respectively. A total of 38 patients were excluded, leaving 138 patients who met study criteria to be eligible in the final analysis (Figure 1). Our cohorts consisted of a predominantly white, male, mixed ICU population with the majority being surgical/trauma patients (70%) (Table 2). One-hundred eighteen (85.5%) patients had at least one risk factor for hypophosphatemia. Concomitant medications known to cause hypophosphatemia were administered to 81 (58.7%) patients. The majority of patients received enteral nutrition (88.1%) with most receiving greater than or equal to 50% of their goal rate at the time of enteral phosphate replacement.
Flow diagram of patient enrollment.
Baseline Characteristics.
Abbreviation: TBI, traumatic brain injury.
Patients may have more than one risk factor or concomitant medication that increases the risk of hypophosphatemia.
The patient ICU characteristics of the mild PRE and POST groups were similar overall, aside from longer median hospital length of stay (24 [13-29] days vs 30 [15-47] days; P = 0.04) and median duration of invasive mechanical ventilation (8.3 [4.6-13.4] days vs 13.6 [8.5-30] days; P = 0.01) in the mild POST group. The median ICU length of stay was 11 (6-14) days for mild PRE and 13 (6-26) days for mild POST (P = 0.15) groups. Thirty-four (61.8%) patients in mild PRE vs 29 (65.9%) patients in mild POST required invasive mechanical ventilation (P = 0.67). ICU mortality occurred in 6 (10.9%) patients in mild PRE and 7 (15.9%) patients in mild POST (P = 0.51).
Patients in the moderate group had median ICU and hospital lengths of stay of 15 (8-22) days and 25 (15-51) days, respectively. ICU mortality occurred in 11 (28.2%) patients. Twenty-nine (74.4%) of patients required invasive mechanical ventilation with a duration of 12 (5.9-19.2) days.
Effectiveness Outcomes
The median time from first dose of Phos-NaK to follow-up serum phosphorus level in the overall cohort was 18.9 [15.9-21.8] hours. Normalization of serum phosphorus occurred in 32 (58.2%) of the mild PRE compared with 32 (72.7%) in the mild POST (P = 0.13, RR = 1.3 (95% confidence interval [CI] 0.9-1.7); Figure 2) group. The median change in serum phosphorus was 0.4 (0.1-0.9) mg/dL in the mild PRE and 0.7 (0.2-1.3) mg/dL in the mild POST (P = 0.06) groups (Figure 3). Treatment failure occurred in 23 (41.2%) of mild PRE compared with 12 (27.3%) in mild POST (P = 0.13) group (Table 3).
Normalization of serum phosphorus within 24 hours.
Change in serum phosphorus.
Treatment Failure.
Data are presented as n (%).
Twenty-six (66.7%) patients had normalization of serum phosphorus in the moderate cohort. Median change in serum phosphorus was 1 (0.5-1.6) mg/dL. Treatment failure occurred in 13 (33.3%) of patients.
Safety Outcomes
New onset diarrhea occurred in 5 (9.1%) patients in the mild PRE group compared with 9 (20.5%) in the mild POST group (P = 0.15). Five (12.8%) patients in the moderate cohort developed new onset diarrhea. One patient in the mild POST and one patient in the moderate group developed hyperphosphatemia following replacement (serum phosphorus level = 4.6 and 4.8 mg/dL, respectively).
Discussion
Our study demonstrated that protocolized enteral phosphate replacement at a dose of 72 mmol was safe and effective; this resulted in overall normalization of serum phosphorus in 70% of patients with mild or moderate hypophosphatemia without increasing the risk of diarrhea or hyperphosphatemia.
The results of this study are comparable with existing literature evaluating intravenous and enteral phosphate replacement in critically ill patients. Normalization of serum phosphorus in the studies assessing intravenous replacement have been variable, ranging from 30% to 98%.7,12,24-27 Brown and colleagues included 79 adult trauma ICU patients who received intravenous phosphate replacement per a graduated dosing scheme and observed a 50% normalization rate in patients with serum phosphorus concentration between 1.6 and 2.2 mg/dL. 24 Two other studies dosing intravenous phosphate according to serum phosphorus and/or patient weight reported normalization rates of 78% and 81.5% in patients with serum phosphorus concentrations ranging from 1.3 to 2.5 mg/dL.25,26 We observed similar serum phosphorus normalization (58.2%-72.7%) in patients with mild and moderate hypophosphatemia within 24 hours of enteral replacement.
Our institution utilizes Phos-NaK packets for enteral replacement, however, there is a lack of published or internal manufacturer data regarding the bioavailability of this specific formulation; it is presumably 1% to 20% as reported in other enteral phosphate products. 21 Two previous studies utilizing intravenous phosphate dosing of 15 to 20 mmoL found a mean increase in serum phosphorus concentration of 0.8 mg/dL.25,26 Another study noted a mean increase in serum phosphorus concentration of 1.1 mg/dL following an intravenous phosphate dose ranging from 20 to 40 mmol. 1 In comparison, we observed a median increase in serum phosphorus concentration of 0.4 and 0.7 to 1.0 mg/dL after receiving 48 and 72 mmol of enteral phosphate, respectively. Thus, we propose that an enteral phosphate dose of 72 mmol with 20% bioavailability is roughly equivalent to 15 mmol of intravenous phosphate.1,25,26 Our findings of a 0.7 to 1.0 mg/dL increase with the modified enteral phosphate dosing (72 mmol for both mild and moderate hypophosphatemia) are consistent with those reported with intravenous dosing.
Lemon and colleagues were the first to demonstrate that enteral phosphate supplementation can be used to treat mild or moderate hypophosphatemia in the ICU and found no significant effects on average serum phosphorus concentration and duration of mechanical ventilation after switching from intravenous to enteral phosphate replacement. The study, however, did not specify the dosing of enteral phosphate administered or the magnitude of serum phosphorus increase postsupplementation. 18 A more recent retrospective chart review assessed intragastric phosphate replacement mixed with saline enema solution in critically ill patients requiring enteral nutrition. 19 The study used a dose of 34 mmol of elemental phosphate with the option to re-dose if needed for a total of 68 mmol of elemental phosphate. Although not significant, those who received 2 doses had improved serum phosphorus concentration compared with one dose. The study included patients with baseline serum phosphorus of less than 3.0 mg/dL, whereas our study included patients with baseline serum phosphorus of less than 2.6 mg/dL. They similarly did not find an increased incidence of diarrhea with enteral phosphate dosing. 19 Furthermore, a recent study by Nguyen and colleagues found that enteral phosphate was noninferior to intravenous phosphate replacement in a critically ill population and substantially reduced cost and waste of resources. 16 The study was comparable with ours as it included a similar number of patients with mild and moderate hypophosphatemia, assessed normalization of serum phosphorus concentrations at 24 hours, and included a similar number of patients on mechanical ventilation. 16 Overall, our study adds to the existing efforts of evaluating the effectiveness of enteral phosphate replacement by utilizing a protocolized approach. In addition, we contextualize the normalization and extent of increase in serum phosphorus concentrations after enteral supplementation, echoing the studies by both Harris and colleagues and Nguyen and colleagues.16,19
Practical advantages to correcting mild-to-moderate hypophosphatemia with enteral phosphate formulations versus intravenous formulations in patients with intact gastrointestinal tracts include cost, fluid stewardship, and availability as an effective alternative. Phos-NaK for enteral administration is inexpensive, averaging $0.38 per packet in wholesale price (Cardinal Health; accessed January 2025). In comparison, each single-use 45 mmol/vial of intravenous sodium phosphate or potassium phosphate both cost approximately $13. A dosing regimen of 3 Phos-NaK packets every 4 hours for 3 doses per our CCERP (72 mmol total) yields an acquisition cost of $3.42, which is approximately $10 less expensive than a single dose of intravenous phosphate, not accounting for the costs of diluent bags, sterile preparation by pharmacy, and nursing administration. In addition to cost, there are practical administration considerations; intravenous potassium phosphate requires a long duration of administration, ranging from 4 to 12 hours depending on the dose and line access at our institution. From a fluid stewardship standpoint, intravenous phosphate is commonly diluted in a volume of 250 to 500 mL. Switching to an enteral phosphate formulation limits the contribution of fluids and avoids potential intravenous incompatibility issues in patients with limited intravenous access.7,16,17,26 Our findings together with prior studies may improve provider comfort in utilizing enteral phosphate supplementation for mild-to-moderate hypophosphatemia when feasible, especially if another national intravenous phosphate shortage should occur.
This descriptive study is reflective of a single-center experience, which limits our generalizability. The retrospective design made the study susceptible to potential inaccurate documentation in the electronic medical record. As such, it was difficult to make clear associations between our defined safety outcomes of hyperphosphatemia and diarrhea with enteral phosphate replacement. We did not collect presence or type of ostomy to describe output with enteral phosphate replacement. In addition, we did not assess the impact of phosphate intake from nutrition sources, although we collected enteral nutrition formulas and rates, nor did we assess type of feeding tube used or rate of occlusion with enteral phosphate administration. Potassium and sodium concentrations were not assessed; our institutional CCERP recommends separate potassium replacement in patients with concomitant hypophosphatemia and hypokalemia. We did not have an intravenous comparator group, which limits our ability to validate previous studies’ conclusions that enteral phosphate is an equal alternative. We did not specify a minimum time requirement between the administration of the last dose of Phos-NaK and the follow-up serum phosphorus level, however, most patients received all 3 doses within a reasonable time frame, as the median time from first dose to follow-up serum phosphorus level was 18.9 hours. Our CCERP excluded patients with acute kidney injury, patients requiring targeted temperature management, patients receiving total parenteral nutrition, and patients with diabetic ketoacidosis; therefore, we cannot comment on the safety and effectiveness of enteral phosphate replacement for these patients. We did not collect on conditions that can influence phosphate metabolism and excretion, such as Fanconi’s syndrome, hyperparathyroidism, or chronic kidney disease, although we did exclude patients with creatinine clearance less than 30 mL/min. In addition, we did not collect vitamin D levels or specifically collect on the administration of sodium phosphate enemas or sevelamer use. We did not assess the impact of patient body weight on phosphate dosing. Differences in the patient populations may impact interpretation of the results, including higher incidence of alcohol use disorder in the mild PRE cohort. The data collection timeframe included the COVID-19 pandemic, which may explain the longer hospital length of stay and duration of invasive mechanical ventilation in the mild POST group. Moreover, nearly half of our study population had traumatic or burn injuries which may limit generalizability to other ICU populations. Finally, our study was limited to patients with mild-to-moderate hypophosphatemia; patients with severe hypophosphatemia (serum phosphorus concentration <1.5 mg/dL) were not evaluated due to common practice and reliability of intravenous phosphate to promptly replete these levels.
Conclusion and Relevance
In conclusion, our study suggests enteral phosphate replacement at a dose of 72 mmol may be safe and effective in critically ill patients with mild or moderate hypophosphatemia. Our study corroborates similar studies that investigated the effectiveness of enteral phosphate as an alternative to intravenous phosphate replacement. These promising results warrant validation in prospective studies.
Footnotes
Acknowledgements
None.
Ethical Considerations
This study received ethical approval from the MaineHealth IRB (approval #2195471) on July 17, 2024. This is an IRB-approved retrospective study, all patient information was de-identified and patient consent was waived. Patient data will not be shared with third parties.
Consent to Participate
The requirement for informed consent to participate was waived by MaineHealth Institutional Review Board.
Author Contributions
All authors listed were substantially involved in the project conception, design, collection, analysis, and interpretation of the data; drafting the article or revising it critically for important intellectual content; approval of the version to be published; and agree to be accountable and willing to investigate and resolve all questions pertaining to accuracy and/or integrity of the work.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.