Potassium Reduction Associated with Lactulose and Sodium Polystyrene Sulfonate Use in Hospitalized Patients with Hyperkalemia: A Real-World Observational Study

Introduction

Hyperkalemia is a common electrolyte disorder in hospitalized patients and is associated with life-threatening cardiac arrhythmias and respiratory muscle weakness.¹ A systematic review reported an overall global prevalence of hyperkalemia of 6.3% and an incidence of 2.8 per 100 patient-years.² Our center reported a prevalence of 3.7% in adults over 1 year.³

The management of hyperkalemia necessitates prompt intervention to stabilize the cardiac cell membrane, redistribute potassium intracellularly, enhance its elimination from the body, and eliminate dietary sources of potassium and drugs that interfere with renal elimination.⁴ Potassium-binding resins are a class of drugs used to increase gastrointestinal potassium excretion. These include sodium- or calcium polystyrene sulfonate (SPS/CPS), patiromer⁵ and sodium zirconium cyclosilicate (SZC);⁶ however, the latter two resins are not currently available in South Africa. Therefore, SPS is the only prescribed resin in our setting.

SPS was first approved for clinical use by the Food and Drug Administration (FDA) in 1958.⁷ In recent years, the efficacy of SPS has been questioned as most studies have been non-randomized controlled trials (RCTs). A systematic review that included 30 studies could only identify four RCTs, while 70% were observational studies.⁸ A total of 23 studies reported on the reduction in the serum potassium concentration ([K+]). Within the first 72 h of therapy, the [K+] decreased by -0.55 mmol/L to -1.5 mmol/L, while the concentration decreased by -0.2 mmol/L to -1.8 mmol/L after 72 h. The study concluded that there was a paucity of high-quality data for the utilization of SPS in the treatment of hyperkalemia. Although the evidence was more robust for the treatment of chronic hyperkalemia, the evidence for the treatment of acute hyperkalemia was weak.

Laxatives are frequently used to prevent and treat constipation caused by SPS. Previously, a suspension of SPS with 70% sorbitol was available, but following concerns of an increased risk of colonic necrosis, the FDA added a black box warning and recommended against the co-administration of sorbitol. Recent studies suggest that laxatives may also aid in reducing the [K+] by enhancing potassium secretion via BK channels located in the apical membrane of cells lining the colon;⁹ however, there is limited data comparing the efficacy of SPS to laxatives alone in reducing [K+] in hospitalized patients.

The objective of this study was to describe and compare the potassium reduction associated with lactulose, SPS alone, and combined lactulose-SPS use in hospitalized patients with hyperkalemia.

Materials and Methods

This study was a post-hoc analysis of a previous retrospective cohort study of adult hospitalized patients diagnosed with hyperkalemia, defined as a [K+] ≥5.5 mmol/L, performed at our centre.³ The study population included adult patients (≥18 years old) hospitalized at Tygerberg Hospital in Cape Town, South Africa, between 01 January and 31 December 2019.

We compared the efficacy of potassium-lowering interventions for the treatment of hyperkalemia across three treatment groups, assessed at two time points: within the first 24 h and again at 2-7 days. At our institution, the only available laxative for the treatment of hyperkalemia was the osmotic laxative, lactulose, while the only potassium-binding resin available was SPS (Kexelate). Therefore, we divided the cohort into three treatment groups based on the prescription of lactulose and/or SPS as follows: lactulose alone (L-group), SPS alone (S-group), or a combination of lactulose and SPS (LS-group).

Each group was prescribed standard treatments in varying frequencies, comprising insulin with dextrose, salbutamol nebulization, and intravenous sodium bicarbonate. Patients who were undergoing kidney replacement therapy, including hemodialysis or peritoneal dialysis, were excluded.

Additional laboratory data, including follow-up serum potassium measurements, were obtained from the online TrakCare system of the National Health Laboratory Service (NHLS) database. Demographic and clinical data were extracted, including age, sex, and comorbidities such as chronic kidney disease (CKD), hypertension, diabetes mellitus, heart disease, and HIV infection. Data on additional hyperkalemia treatments, including insulin therapy, salbutamol nebulization, and intravenous sodium bicarbonate, were also collected. Laboratory data included [K+] at baseline, 24 h, and 2-7 days post-treatment. Additional laboratory parameters included serum creatinine and estimated glomerular filtration rate (eGFR), using the CKD-EPI equation, for patients with CKD. CKD was defined as an eGFR <60 mL/min per 1.73 m² for three months or more.

The outcome of the study was the difference in [K+] at 24 h and at 2-7 days among the three treatment groups.

Results

Baseline characteristics

A total of 134 hospitalized patients with hyperkalemia were included in the analysis, of whom 40 (30%) were in the L-group, 71 (53%) in the S-group, and 23 (17%) in the LS-group. There were no statistically significant differences in age, sex, hypertension, diabetes mellitus, heart disease, or HIV status (p >0.05 for all) across groups. All patients in the LS-group had kidney disease compared to the S-group (93%) and L-group (70%) (p <0.001), and CKD was more common in the LS-group (56.5%) than in the S-group (32.4%) or L-group (22.5%) (p = 0.022) [Table 1].

Table 1: A comparison of baseline characteristics of patients who had hyperkalemia and received treatment with either lactulose only, SPS only, or combined lactulose and SPS

	Lactulose only N = 40	SPS only N = 71	Combined N = 23	p value
Baseline characteristics
Age, median (IQR)	56 (40-72)	50 (37-62)	57 (38-66)	0.335
Male sex	19 (47.5)	38 (53.5)	15 (65.0)	0.397
Hypertension	28 (70)	42 (59)	16 (70)	0.436
Diabetes mellitus	16 (40)	19 (27)	9 (41)	0.252
Heart disease	7 (17.5)	15 (21)	3 (14)	0.831
HIV	8 (20)	13 (18.3)	4 (18.2)	0.202
Kidney disease	28 (70)	66 (93)	23 (100)	<0.001
KDIGO AKI stage	19 (47.5)	43 (61)	10 (43.5)	0.231
Stage 1	4 (21)	7 (16)	1 (10)	0.520
Stage 2	5 (26)	5 (12)	1 (10)
Stage 3	10 (53)	31 (72)	8 (80)
CKD	9 (22.5)	23 (32.4)	13 (56.5)	0.022
Stage 3	2 (22)	5 (22)	0	0.219
Stage 4	1 (11)	5 (22)	1 (8)
Stage 5	6 (67)	13 (56)	12 (92)
Laboratory results
Baseline [K+], mmol/L, median (IQR)	6.0 (5.6-6.9)	6.4 (6.0-7.0)	7.2 (6.1-7.6)	0.003
Mild (5.5–5.9)	21 (52.5)	17 (24)	2 (9)	0.001
Moderate (6.0–6.9)	11 (27.5)	33 (46)	9 (39)
Severe (≥7.0)	8 (20)	21 (30)	12 (52)
Creatinine (AKI), µmol/L, median (IQR)	488 (167-883)	586 (313-1323)	681 (441-1000)	0.310
eGFR (CKD), ml/min per 1.73 m2, median (IQR)	11 (4.2-29.3)	9 (4.2-29.7)	4.4 (3.1-10.3)	0.066
Other emergency therapies
Insulin and dextrose	21 (52.5)	50 (70.4)	21 (91.3)	0.005
Salbutamol nebulization	6 (15)	12 (17)	3 (13)	1.000
Intravenous sodium bicarbonate	2 (5)	10 (14.1)	3 (13)	0.340

IQR: Interquartile range, SPS: Sodium polystyrene sulfonate, HIV: Human immunodeficiency virus, KDIGO: Kidney disease improving global outcomes, AKI: Acute kidney injury, [K+]: Potassium concentration, eGFR: Estimated glomerular filtration rate, CKD: CKD.

Baseline [K+] differed significantly between groups, being highest in the LS-group (median [K+]: L: 6.0 mmol/L [IQR 5.6-6.9], S: 6.4 mmol/L [6.0-7.0], LS: 7.2 mmol/L [6.1-7.6]; p = 0.003) [Figure 1]. A significantly greater proportion of patients in the LS-group presented with severe hyperkalemia (≥7.0 mmol/L) compared to the L- and S-groups (52% vs. 20% and 30%, respectively; p = 0.001). Use of other potassium-lowering therapies was highest in the LS-group, with 91.3% receiving insulin and dextrose (p = 0.005), while use of salbutamol and sodium bicarbonate was similar across all groups [Table 1].

Close

Figure 1:

A comparison of the serum potassium concentrations at baseline, at 24 h, and days 2-7 in patients who had hyperkalemia and received treatment with either lactulose only, SPS only, or combined lactulose and SPS. Abbreviation: [K+], serum potassium concentration, SPS, sodium polystyrene sulfonate.

Export to PPT

Baseline demographic and comorbidity characteristics were similar across treatment groups in the full cohort and in sensitivity analyses restricted to patients with available potassium measurements at 24 h and days 2-7 [Table S1]. Differences in kidney disease burden across treatment groups were present at baseline and persisted in analytic subsets. CKD was both more prevalent and more advanced in the combination group. This suggests confounding by indication rather than bias due to missing data, as patients in this group were more likely to receive combination therapy. Overall, these findings indicate that missing data were unlikely to have introduced substantial bias into the complete-case analyses.

Supplementary Table 1

Download

Serum potassium response at 24 h

The median change in [K+] at 24 h did not differ significantly between treatment groups (p = 0.146). The median [K⁺] reduction was -0.9 mmol/L in the L group, -1.1 mmol/L in the S-group, and -1.3 mmol/L in the LS-group [Table 2]. In a multilinear regression model (n = 63), adjusted for age, sex, kidney disease including AKI and CKD, baseline [K⁺], insulin, salbutamol and sodium bicarbonate therapies, only baseline [K⁺] was independently associated with 24-hour potassium reduction, with each 1 mmol/L increase in baseline potassium corresponding to a 0.55 mmol/L greater decline over 24 h (β -0.55; 95% CI -0.91 to -0.19; p = 0.004) [Table 3 and Figure 2].

Table 2: A comparison of the serum potassium response at 24-h and days 2-7 of patients who had hyperkalemia and received treatment with either lactulose only, SPS only, or combined lactulose and SPS

Follow-up [K+]	Lactulose only N = 40	SPS N = 71	Combined N = 23	p value
[K+] at 24-h, mmol/L, median (IQR)	5.8 (5.4–6.1)	6.5 (5.9–7.1)	6.1 (5.5–6.9)	0.146
[K+] difference at 24-h, mmol/L, mean ±SD	-0.36 ±0.82	-0.40 ±0.99	-0.63 ±1.32	0.737
[K+] at 2–7 days, mmol/L, median (IQR)	5.1 (4.6–5.7)	5.0 (4.7–5.5)	5.0 (4.7–5.9)	0.610
[K+] difference at 2-7 days, mmol/L, mean ±SD	-0.90 ±0.92	-1.36 ±1.29	-1.75 ±1.42	0.038
Number of days from ≥2-days, mean ±SD	4.8 ±2.00	4.0 ±1.80	4.5 ±1.97	0.166

Abbreviations: [K+]: Serum potassium concentration, SPS: Sodium polystyrene sulfonate.

Table 3: Multilinear regression analyses of predictors of the absolute serum potassium reduction at 24 h and days 2-7

Absolute [K+] reduction at 24-h	b coefficient	p value	95% confidence interval
Age, years	0.01	0.217	-0.01 to 0.02
Female sex	-0.17	0.555	-0.74 to 0.40
Baseline [K+], mmol/L	-0.55	0.004	-0.91 to -0.19
Kidney disease	0.45	0.299	-0.41 to 1.31
Insulin	0.15	0.662	-0.52 to 0.82
Salbutamol	0.10	0.774	-0.60 to 0.80
Sodium bicarbonate	-0.20	0.623	-1.03 to 0.62
Treatment groups
S group	0.22	0.458	-0.38 to 0.82
LS group	-0.06	0.875	-0.82 to 4.51
Absolute [K+] reduction at days 2–7
Age, years	-0.01	0.133	-0.02 to 0.002
Female sex	-0.28	0.169	-0.68 to 0.21
Baseline [K+], mmol/L	-0.87	<0.001	-1.15 to -0.59
Kidney disease	0.34	0.328	-0.35 to 1.03
Insulin	0.32	0.187	-0.16 to 0.80
Salbutamol	-0.34	0.285	-0.96 to 0.28
Sodium bicarbonate	-0.64	0.053	-1.28 to 0.01
Treatment groups
S group	-0.37	0.116	-0.83 to 0.09
LS group	-0.49	0.106	-1.09 to 0.11

Abbreviations: [K+]: Serum potassium concentration: S group: Sodium polystyrene sulfonate only, LS group: Lactulose plus sodium polystyrene sulfonate only, R-squared (24-h): 24.5%; R-squared (2-7 days): 43.9%.

Close

Figure 2:

A comparison of the differences in serum potassium concentrations at 24 h and days 2-7 in patients who had hyperkalemia and received treatment with either lactulose only, SPS only, or combined lactulose and SPS. Abbreviations: [K+], serum potassium concentration, SPS, sodium polystyrene sulfonate.

Export to PPT

Discussion

In this post-hoc analysis of hospitalized patients with hyperkalemia, we compared the potassium reduction associated with lactulose only, SPS only, and their combination at two time intervals: within the first 24 h and between days 2-7. While prior studies have described the potassium-lowering effect of SPS, data on lactulose as monotherapy are limited, and studies directly comparing these interventions are even more scarce.

At 24 h, we found that the median reductions in [K⁺] were not significantly different between the groups. This finding is consistent with previous literature suggesting that both SPS and laxatives exert a modest and delayed effect in the early phase of treatment. Hunt et al.¹⁰ evaluated the effectiveness of a single 30 g oral dose of SPS suspended in 33% sorbitol; a formulation historically used to reduce the risk of constipation. In their retrospective analysis of patients with CKD Stage 4 or 5, the [K⁺] was reassessed within 24 h. They reported a median potassium reduction of 0.8 mmol/L (IQR 0.4-1.1 mmol/L).¹⁰ Similarly, Hagan et al.¹¹ found that a single 30 g dose of SPS resulted in a statistically significant but modest decline in potassium within 24 h (0.93 mEq/L).

The effects of laxatives alone have been less extensively studied. Sumida et al.,¹² who conducted a large cohort study involving over 36,000 United States veterans with advanced CKD transitioning to dialysis, demonstrated that time-varying laxative use was associated with a 21% lower risk of hyperkalemia (odds ratio 0.79, 95% CI: 0.76-0.84), suggesting a possible role for gastrointestinal potassium loss. This supports a mechanism of enhanced colonic potassium secretion via upregulation of apical ‘big potassium’ or BK channels.⁹ In our study, the use of lactulose alone was associated with a median [K⁺] reduction of 0.9 mmol/L by days 2-7. This aligns with the findings reported by Abdulameer et al.¹³ where a daily dose of lactulose administration led to a notable decrease in [K⁺] by day 5 after all other measures had failed. In our study, although lactulose was less effective than combination therapy, it is possible that the dosing and frequency were suboptimal, as more than half of the patients in the L-group had only mild hyperkalemia; however, in the absence of dosing data, this remains speculation.

A systematic review by Wong et al.,⁸ evaluated SPS and CPS use across 30 studies. Among the 25 studies reporting serum potassium outcomes, 88% showed a reduction of ≥0.5 mmol/L, with high-quality studies demonstrating a range of 0.14-1.04 mmol/L. Notably, only 10 studies reported reductions within 24 h, highlighting the delayed onset of action associated with these agents. The most significant effects occurred after 48-72 h.

Importantly, the greatest [K⁺] reduction in our cohort was observed in the combination therapy group by days 2-7; however, since patients in the LS-group had a higher baseline [K⁺], the greater efficacy may reflect regression to the mean. In support of this, sensitivity analyses demonstrated that while combination therapy was associated with greater unadjusted potassium reductions, these differences were attenuated when analyses were restricted by kidney disease status or stratified by baseline hyperkalemia severity. This was further supported by the multivariable linear regression model for potassium reduction at days 2-7, in which baseline serum potassium concentration emerged as the strongest predictor of potassium reduction. This suggests that baseline potassium concentration was the primary determinant of the observed potassium reductions.

Conversely, more appropriate dosing of lactulose and SPS, combined with consistent administration, may have led to improved efficacy. As early as 1961, Flinn et al.¹⁴ reported that administering SPS with 70% sorbitol to oliguric patients enhanced faecal potassium elimination and helped mitigate the constipating effects of the resin. This study laid the foundation for the long-standing clinical practice of co-administering SPS with sorbitol or other laxatives to optimize gastrointestinal potassium clearance. Emmett et al.¹⁵ further substantiated this approach. In a controlled study of healthy volunteers, the combination of SPS with phenolphthalein-docusate significantly increased faecal potassium excretion compared to either agent alone. Approximately 24% of the total faecal potassium loss was directly attributed to SPS binding, while the remainder was enhanced by the laxative-induced acceleration of intestinal transit and colonic secretion.¹⁵

Our findings support the argument that while SPS and lactulose may not offer sufficient efficacy within the initial 24 h, they may be suitable for subacute or chronic potassium control in non-dialysis CKD patients. Newer agents such as patiromer and SZC have demonstrated more predictable effects and improved tolerability, though their onset is similarly delayed to 24-48 h.^5,6 Nevertheless, their high cost limits prescription, and they are currently not accessible in South Africa. In low-resource settings with limited access to newer potassium binders and delayed availability of dialysis, the combination of SPS and lactulose may be a pragmatic option to consider for the management of hyperkalemia.

Concerns around the gastrointestinal safety of SPS persist, particularly regarding its use with sorbitol; however, current evidence suggests that the absolute risk is very low and largely restricted to specific high-risk populations. In a large analysis, Watson et al.¹⁶ emphasized that colonic necrosis is an exceedingly rare event, estimated at 0.2-0.3%, and mostly occurring in patients with recent abdominal surgery, bowel ischaemia, or pre-existing bowel dysfunction.

Although this was a small, retrospective, single-center study, to the best of our knowledge, it is one of the largest to compare the effect of lactulose only, SPS only, or their combination in reducing the [K⁺] to date. Although missing data may have influenced outcomes, sensitivity analysis indicated that missingness was unlikely to be strongly differential by treatment group, supporting a missing-at-random (MAR) assumption. We did not collect data on dosing or frequency of administration of therapies, which may have affected the potassium reduction within each group.

To conclude, among hospitalized patients with hyperkalemia, the combination of lactulose and SPS was associated with larger reductions in serum potassium at days 2-7, although causal inference cannot be made. A randomized controlled trial is needed to evaluate the efficacy and safety of combined lactulose and SPS therapy in reducing serum potassium and to better define clinical outcomes, dosing strategies, and adverse effects relative to standard or alternative therapies.