If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
This is the largest study of its kind to date in the United Kingdom.
Insulin/dextrose (IDex) is transiently effective for hyperkalemia management.
Over 1 in 3 patients require retreatment with IDex for hyperkalemia within 24 hours.
IDex can cause hypoglycemia in up to 1 in 5 patients within 6 hours of treatment.
Frequent, regular glucose monitoring is crucial after IDex to ensure patient safety.
Hyperkalaemia occurs in up to 10% of hospital admissions but its treatment in the emergency setting is inconsistent.
To describe the emergency management of hyperkalaemia in adults with insulin-dextrose (IDex) and to explore clinical outcomes associated with IDex treatment.
Design and setting
Cohort study using comprehensive electronic health records of all emergency admissions to a large university hospital in the United Kingdom between April 2015 and August 2018.
Adult patients aged ≥16 years with at least one emergency admission and one blood potassium result during the study period.
Main outcomes and measures
Emergency hyperkalaemia treatment was evaluated including the requirement for re-treatment with IDex, episodes of glucose dysregulation, intensive care (ICU) admission and length of hospital stay. Associations with hyperkalaemia, adverse events and IDex treatment were explored by logistic regression.
Amongst 211,993 patients attending the Emergency Department (ED) we identified 11,107 hyperkalaemic adult patients, of whom 1,284 were treated with IDex.
Multiple doses were required in 542 patients (42.2%). Hypoglycaemia (plasma glucose < 4 mmol/L) occurred in 249 patients (19.4%) within 6 hours of IDex. Repeated doses were associated with an increased risk of hypoglycaemia (OR 2.94, 95% CI 2.20 to 3.93) compared to patients receiving a single dose, which, after adjustment was also associated with an increased risk of death (OR 1.56, 95% CI 1.16 to 2.09) during the study period.
Patients who received multiple doses of IDex (OR 2.2, 95% CI 1.6-3.1) and those who received a dose of insulin above the guideline recommended limit (OR 5.6 3.1-10.3) were more likely to be admitted to ICU following IDex than those who received a single dose or the guideline recommended dose of insulin.
Conclusions and Relevance
This study provides novel insight into the emergency management of hyperkalaemia in a large population, demonstrates the high risk of hypoglycaemia and highlights the urgent need for an improved, evidence-based approach to the emergency management of hyperkalaemia.
]. In patients receiving treatment with renin-angiotensin-aldosterone system (RAAS) blocking medication for heart failure, hypertension or diabetic nephropathy, the prevalence of hyperkalaemia exceeds 10% and in those with chronic kidney disease (CKD) (stage 4 or 5), this exceeds 20% [
]. Hyperkalaemia can result in fatal cardiac arrhythmias if not treated promptly. It typically requires in-hospital treatment and constitutes a medical emergency.
Despite the high prevalence of hyperkalaemia, its treatment is inconsistent. One multi-centre observational study reported 43 different treatment combinations with an insulin and dextrose infusion used most commonly, either as a single agent or in combination, in 64% of patients [
] through transfer from the extra-cellular fluid to the intra-cellular space though hyperkalaemia can recur as potassium redistributes requiring repeated treatment with IDex. Dextrose is given to mitigate the potential for insulin-mediated hypoglycaemia, but hypoglycaemia may still occur for many hours after administration [
], thus requiring close monitoring in an inpatient setting. Other emergency treatments for hyperkalaemia include intravenous calcium salts to stabilise the cardiac membrane and nebulised salbutamol, intravenous sodium bicarbonate, loop diuretics, potassium binding-resins and renal replacement therapy to help lower serum potassium [
The objective of this study was to describe the acute management of hyperkalaemia with IDex in emergency medical admissions to a large tertiary hospital in the UK and to explore clinical outcomes associated with IDex treatment.
The study objectives were to: 1) compare a large cohort of patients with and without hyperkalaemia and 2) describe the emergency management of hyperkalaemia with IDex. The second objective focussed on clinical outcomes relating to treatment effectiveness, safety and healthcare resource use amongst hyperkalaemic patients treated with IDex (herein referred to as Ins) in comparison with patients with confirmed moderate-severe hyperkalaemia (serum potassium ≥6.0mmol/L) who did not receive IDex (herein referred to as noIns). These outcomes included requirement for re-treatment with insulin, episodes of glucose dysregulation following treatment, admission to Intensive Care following IDex treatment and length of hospital stay.
Patients were included if they were aged >16 years old, had attended the ED during the study period and had at least one recorded blood potassium result during their follow up. Patients with missing potassium data were excluded. Patients with hyperkalaemia were defined and then stratified on the basis of their highest recorded blood potassium (K+) concentration. Those with a K+ concentration of 5.5 mmol/L or greater were identified as hyperkalaemic, and further stratified into mild (5.5 – 5.9 mmol/L), moderate (6.0 – 6.4 mmol/L) or severe hyperkalaemia (≥ 6.5 mmol/L) categories as per the European Resucitation Council and UK Renal Association guidelines on hyperkalaemia [
]. In this study, non-hyperkalaemic patients were defined as having a K+ <5.5 mmol/L. Hypoglycaemia was defined as blood glucose <4 mmol/L and severe hypoglycaemia <2.8 mmol/L.
Data including demographics, admission and discharge dates, medications, medical history and blood test results (including creatinine, glucose, potassium) were abstracted. Diagnoses were identified from International Classification of Diseases 10th revision (ICD-10) codes and were used to calculate an age-adjusted Charlson co-morbidity score [
]. Morbidity code lists are available in the Supplementary Appendix. Patient characteristics including age, sex, ethnicity and presence of chronic co-morbidities, prescribing of medications known to affect potassium homeostasis, average length of hospital stay and laboratory data were summarised. The estimated glomerular filtration rate was calculated from creatinine values using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation but with no adjustment for race [
]. Exposure to IDex was identified from prescribing data that detailed the time and date of IDex administration. ED admissions were selected since acute medical, surgical and non-elective sub-specialty admissions were all hospitalised via this route.
Categorical variables were described as frequency (%), and continuous variables were described as mean ± standard deviation (SD) or median and interquartile range (IQR) as appropriate to their distribution.
Categorical variables were compared between groups using the Chi-squared test. Continuous variables were compared using the Student's t-test or Mann-Whitney U-test. Associations with hyperkalaemia were explored by logistic regression and odds ratios (ORs) with associated 95% confidence intervals (CIs) were calculated. Results reached statistical significance when p≤0.05.
Co-variates included in the regression models included age, sex, chronic comorbidities (including CKD, diabetes mellitus, heart failure, ischaemic heart disease and hypertension) and medications that were considered to have an effect on the development of hyperkalaemia (renin-angiotensin receptor blockers, beta-blockers, and potassium-sparing diuretics (including amiloride, spironolactone and eplerenone)).
All data management and analyses were conducted using Stata IC version 15 (StataCorp, Texas, USA).
Between April 2015 and August 2018, 211,993 patients registered 382,394 ED attendences and 646,054 hospital attendences including haemodialysis, day surgery and chemotherapy sessions in addition to visits to the ED. We excluded all 40,659 paediatric patients, and excluded 45,069 adult patients for whom there was no serum potassium measurement (Supp figure 1). Of the remaining 126,265 patients, 11,107 (8.8%) had registered a potassium measurement in hyperkalaemic range (defined as K+ ≥ 5.5 mmol/L in this study), of whom 5,146 (3.0% overall) registered moderate hyperkalaemia (K+ ≥ 6.0 mmol/L). Severe hyperkalaemia (K+ > 6.5 mmol/L) was present in 2,420 (1.4%). The mean potassium of all adult patients in whom it was measured (n = 126,265) was 4.21±0.71mmol/L. Patient characteristics are shown in Table 1, stratified by the maximum potassium concentration.
Table 1.Characteristics of the study population stratified by serum potassium concentration.
Non-hyperkalaemic patients (Potassium <5.5mmol/L)
Hyperkalaemic patients(Potassium ≥5.5mmol/L)
Age, median (IQR)
Female N (%)
Still alive at end of study N (%)
Charlson Co-Morbidity Index, median (IQR)
Calculated eGFR (mls/min/1.73m2), median (IQR)
Ethnicity N (%)
Co-morbidites N (%)
Chronic Kidney Disease
End-Stage Kidney Disease
Congestive Cardiac Failure
Ischaemic Heart Disease
Angiotensin 2 Receptor Blockers
*Categorical variables were compared between groups using the Chi-squared test. Continuous variables were compared using the Student's t-test or Mann-Whitney U-test as appropriate to their distribution
Comparison between hyperkalaemic (potassium ≥5.5mmol/L) and non-hyperkalaemic patients
Compared to hyperkalaemic patients, normokalaemic patients were younger (median age 52 years (IQR 29-69) vs 72 years (56-84), p<0.001), less likely to have diabetes (6.3% vs 25.3%, p<0.001) or have chronic kidney disease (CKD) (2.4% vs 19.9%, p <0.001) and normokalaemic patients had fewer additional comorbid conditions (Charlson co-morbidity index: 2 (0-4) vs 4 (2-6), p<0.001). Fewer normokalaemic patients had died at the end of follow up (5.9% vs 23.7%, p<0.001) (Table 1) and in a logistic regression model adjusting for age, sex and co-morbidity the risk of death remained higher in the hyperkalaemic group (Odds Ratio (OR) 3.0, 95% Confidence interval (CI) 2.8-3.1).
A lower proportion of normokalaemic patients, compared to hyperkalaemic patients, were treated with ACE Inhibitors (8.3% vs 24.9%, p<0.001), Angiotensin-2-receptor blockers (3.5% vs 9.9%, p<0.001) or potassium-sparing diuretics (1.6% vs 9.8%, p<0.001) (Table 1).
In a logisitic regression model, CKD (OR 4.0, 95% CI 3.7-4.2), diabetes (OR 2.2 95% CI 2.1-2.3) and use of potassium-sparing diuretics (OR 2.5 95% CI 2.3-2.7) were strongly associated with an increased risk of developing hyperkalaemia (Table 2).
Table 2.Adjusted logistic regression model to explore predictors for developing hyperkalaemia.
Adult patients treated with IDex for hyperkalaemia
1,284 adult patients were treated with IDex (soluble Actrapid insulin (mean dose 10 units) with 50mls of 50% dextrose) for hyperkalaemia. The median age of these patients was 72 years (58-83), 804 (62.6%) were male and 77.9% were of White British ethnic origin. At the end of follow up 709 patients (55.2%) remained alive. The median calculated GFR was 43.2mls/min (IQR 27.5-70.9). The characteristics for these patients including co-morbidities and prescribed medications are presented in Supp table 5.
Intravenous calcium salts (gluconate or chloride) were given to 1,087 patients (84.7%).
Characteristics of hyperkalaemic patients who were treated with IDex (Ins) versus those not treated with IDex (NoIns)
Compared to NoIns patients, Ins patients were older (72 years (IQR 58-83) vs 70 years (IQR 53-82), p<0.001), more likely to have chronic kidney disease (CKD) (39.9% vs 30.5%, p <0.001), and have a greater number of additional comorbid conditions (Supp table 5). Renal function was lower for Ins patients versus NoIns (eGFR 43.2 (IQR 27.5-70.9) vs 61.2 (IQR 33.5-92.2) mls/min/1.73m2). A higher proportion of Ins patients were treated with ACE Inhibitors (30.1% vs 26.7%, p=0.07), Angiotensin-2-receptor blockers (12.3% vs 9.2%, p=0.001) or potassium-sparing diuretics (17.1% vs 12.4%, p<0.001). The median index hyperkalaemic value for NoIns patients was 6.3 (6.1-6.8) mmol/L with a median of subsequent potassium blood tests during the same admission of 4.3 (3.8-4.7) mmol/L.
In a logistic regression model (Supp table 4), CKD (OR 1.4, 1.1-1.6), being male (OR 1.4, 1.2-1.7), use of potassium-sparing diuretics (OR 1.4, 95% CI 1.1-1.8) and an existing diagnosis of hypertension (OR 1.2, 95% CI 1.0-1.5) were significantly associated with needing IDex treatment.
At the end of follow up, 575/1,284 patients (44.8%) in Ins vs 383/1,082 patients (35.3%) had died (p<0.001). In a logistic regression model adjusting for age, sex and co-morbidity, the risk of death remained higher in the Ins group (OR 1.5, 1.3-1.8) (Supp table 6). Exact cause of death was not assessed.
Management of hyperkalaemia with IDex
2,541 total doses of IDex were given to 1,284 adult patients (median - 1 dose per patient (IQR 1-2)) between April 2015 and August 2018. The dose of short-acting insulin given was 10 units in 94% and the most common alternative dose was 15 units in 2.5% of cases. 660 doses (26.0%) were given in the Emergency Department with the remaining 1,881 doses (74.0%) given on the inpatient wards, including 268 doses (10.5%) in intensive care. Median time from admission to treatment with IDex was 19.5 (IQR 4.6-176.4) hours and a total of 1,320 doses (51.9 %) within 24 hours of admission.
Mean potassium concentration immediately (≤ 60 min) pre-treatment was 6.34 (SD 1.2) mmol/L and median time from hyperkalaemic result to treatment with IDex was 50 minutes (IQR 21-103). The mean reduction in potassium at 4-hours post treatment was 0.86 (SD 0.92) mmol/L (excluding patients receiving renal replacement therapy) (Figure 1).
Multiple doses were required in 542 patients (42.2%), of whom 209 (16.3%) were retreated within 4 hours of initial treatment and 445 (34.6%) within 24 hours. Patients receiving multiple doses were more likely to have CKD (44.5% vs 36.5, p=0.002) or congestive cardiac failure (22.9% vs 17.4%, p=0.009) and to have been exposed to ACE inhibitors (33.2% vs 27.9%, p=0.02) or mineralocorticoid antagonists (19.4% vs 15.5%, p=0.04) (Supp table 8), although only CKD remained significantly associated with retreatment in a logistic regression model adjusted for age, sex and co-morbidity (OR 1.4, 95% CI 1.1-1.8). (Supp table 1) Receiving more than one dose of insulin was associated with an increased risk of death (OR 1.3, 1.0-1.7) after adjustment for age, sex and co-morbidity. Median length of hospital stay was significantly longer in Ins patients (11.7 days (IQR 4.9-24.7)) compared to NoIns controls (6.0 days (IQR 1.2-17.3) p<0.001).
Dysregulation of glucose metabolism following IDex
Dysregulation of glucose metabolism occurred in 818 patients (64.0%) within 6 hours of receiving insulin. Baseline glucose prior to receiving IDex was 7.5mmol/L (IQR 6.0-10.2). Hypoglycaemia (plasma glucose < 4 mmol/L) occurred in 249 patients (19.4%) within 6 hours of insulin administration, and 72 patients (5.6%) experienced severe hypoglycaemia (plasma glucose < 2.8 mmol/L), of whom 93.1% received 10 units of insulin and 6.9% received 15 untis or more. Hyperglycaemia (plasma glucose >10mmol/L) occurred in 700 (54.5%) insulin-treated patients within 6 hours of IDex.
The likelihood of developing hypoglycaemia was increased by receiving multiple doses of IDex (OR 3.0, 2.2 to 4.0), having a baseline glucose of less than 7mmol/L (OR 3.4, 2.5-4.6) and having CKD (OR 1.5, 1.1-2.3) (Supp table 2). The likelihood of developing hyperglycaemia was increased by having diabetes (OR 2.3, 1.4 to 3.5) whilst being older was associated with a decreased likelihood of becoming hyperglycaemic (OR 0.6, 0.4-0.9).
Patients who became hypoglycaemic in the 6 hours following IDex were more likely to have died by the end of the study in August 2018 than those who did not (OR 1.4, 1.1 to 1.9) (Figure 2 and Supp table 3).
Intensive care (ICU) admissions amongst patients treated with IDex
There were 420 ICU admissions by a total of 410 patients who were treated with IDex during the same admission. 208 (49.5%) of these occurred after IDex administration with 177 (42.1%) within 12 hours of IDex. Patients were more likely to be admitted to ICU following IDex if they received multiple doses of Insulin (OR 2.2 95% CI 1.6-3.1) or if they had received a higher than guideline recommended dose (10 units) of insulin (OR 5.6 95% CI 3.1-10.3) (Supp Table 9).
Our results show that emergency treatment of hyperkalaemia varied and in patients treated with IDex, a single dose was insufficient in 4 out of 10 patients, was associated with a risk of hypoglycaemia in 20% of patients and resulted in an increased length of hospital stay compared to hyperkalaemic patients not treated with IDex. Amongst all 171,334 adult patients attending the Emergency Department over this time period 7.9% experienced an episode of hyperkalaemia, whilst moderate-severe hyperkalaemia was present in 3% of adult patients. This is similar to the incidence previously reported in the literature of between 1% and 10% of hospitalised patients [
This retrospective observational study is the first of its kind in the UK to use granular EHR data to provide a detailed review of the epidemiology and emergency management of hyperkalaemia in a large tertiary hospital over three and a half years. The detail available using data abstracted from a comprehensive EHR allows in-depth analysis of prescribing, episodes of glucose dysregulation following insulin treatment and the effect of IDex on serum potassium in a real-world hospital setting.
This study confirms that hyperkalaemia is most common in older patients who are male and with multiple co-morbid conditions, of which CKD and diabetes are associated most strongly with an increased risk of developing hyperkalaemia (Table 2). These data also support the known association between medications such as RAAS inhibitors, beta-blockers and potassium-sparing diuretics and hyperkalaemia, which are commonly used in patients with CKD, heart failure, and diabetes [
This study is one of the largest to date reviewing emergency treatment of hyperkalemia with IDex in 1,284 adult patients and highlights risks associated with this widely used management strategy. Although IDex effectively reduces serum potassium within 1 hour in most patients, this is a transient effect, requiring re-treatment with IDex in over one third of patients and an increased risk of hypoglycaemia. Hypoglycaemia is associated with increased mortality and morbidity [
] and our study suggests an association with increased risk of death amongst those patients who developed hypoglycaemia in the 6 hours following IDex treatment. Twenty percent of patients developed hypoglycaemia and of these, 5% developed severe hypoglycaemia within 6 hours of IDex, which is similar to previously reported data [
], and emphasises the importance of frequent and regular glucose monitoring for a minimum of 6 hours to ensure that patients with decreasing blood glucose levels are treated as soon as possible.
Further limitations of IDex raised by this study include prolonged hospital stay, with the associated healthcare costs, from 6 days for patients with moderate-severe hyperkalaemia not treated with IDex to almost 12 days in those treated with IDex, and an increased likelihood of ICU admission compared to those hyperkalaemic patients not treated with IDex (Supp table 7).
The increased risk of ICU admission following IDex was greatest in those patients receiving multiple doses of insulin and those receiving doses of insulin above the guideline recommend dose (Supp table 9). This highlights the need for ongoing clinican awareness and education in the management of hyperkalaemia in the emergency setting and the need for strategies to limit exposure to IDex. Indeed in the United Kingdom NICE-approved guidleines from the UK Renal Association now recommend the use of newer oral potassium binders as an adjunct to IDex in life-threatening hyperkalaemia [
]. It should be noted that there is limited randomised trial evidence in the emergency setting for this approach.
The primary limitation of this study is its single-centre, retrospective, observational design using real-world data. It is not possible to ascribe causation to any of the factors assessed in our study but merely to highlight associations between hyperkalaemia and treatment with IDex with an increased length of stay in hospital and an increased risk of death. Another limitation is lack of concurrent ECGs for all patients to better evaluate the effect of potassium and the treatments used on the ECG. Of note however, 85% of patients treated with IDex were also given IV calcium salts suggesting clinical concern for the effect of potassium on the ECG in these patients.
Cause of death or indeed the precise date of death was lacking for some patients within our dataset and therefore it is not possible to determine whether any deaths were directly attributable to hyperkalaemia.
Despite these limitations, this study, the largest of its kind to date in the UK, using granular EHR data, enhances the evidence base for emergency hyperkalaemia management and highlights areas that require improvement. Although IDex effectively reduces serum potassium within 1 hour in most patients, this is a transient effect, requiring re-treatment with IDex in over one third of patients. Our study also indicates that glucose dysregulation is a common and serious adverse effect associated with IDex with hypoglycaemia affecting 20% of patients within 6 hours of IDex administration. Care must be taken with each prescription of IDex to ensure that the risk to the individual patient has been assessed, in particular with dosing insulin in elderly patients and those with CKD. Our data suggest that in patients with a lower baseline plasma glucose concentration (<7mmol/L) consideration should be given to using an increased dose of glucose to limit the increased risk of hypoglycaemia, although further randomised studies to determine the safest and most effective IDex dosing strategy are warranted.
In conclusion, this study provides additional insight into the emergency management of hyperkalemia, demonstrates the high risk of hypoglycaemia, increased risk of ICU admission and the prolonged length of hospital stay associated with IDex treatment, highlighting the urgent need for an improved, evidence-based approach to the emergency management of hyperkalaemia. This requires further studies to determine the safest and most effective dosing strategies for IDex and robust trials of emerging treatment strategies such as oral potassium binders in the emergency setting with clinically meaningful endpoints.
Exclusive licence statement
The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, a worldwide licence to the Publishers and its licensees in perpetuity, in all forms, formats and media (whether known now or created in the future), to i) publish, reproduce, distribute, display and store the Contribution, ii) translate the Contribution into other languages, create adaptations, reprints, include within collections and create summaries, extracts and/or, abstracts of the Contribution and convert or allow conversion into any format including without limitation audio, iii) create any other derivative work(s) based in whole or part on the on the Contribution, iv) to exploit all subsidiary rights to exploit all subsidiary rights that currently exist or as may exist in the future in the Contribution, v) the inclusion of electronic links from the Contribution to third party material where-ever it may be located; and, vi) licence any third party to do any or all of the above.
TFH had the original idea for the study and supervised each stage of data management and analysis. TFH, IBW and TJLH were involved in the study design. TJLH undertook the data management and primary analysis, and wrote the first draft of the manuscript. All authors contributed to further drafts and approved the final manuscript.
Access to Data
TJLH and TFH had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
The lead author affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.
This study did not require direct ethical approval as it used anonymised patient data, though it was approved by the Cambridge University Hospitals Research Data Governance Committee. The eCID database and data-sharing agreement are REC approved and research was conducted according to the Declaration of Helsinki ethical principles for research.
Funding and the role of the funders
This study was supported by the NIHR Cambridge Biomedical Research Centre (BRC) and support by a research training fellowship award from AstraZeneca to TJLH. The funding sources were not involved in research study conduct or design; data collection, management of the data. GJ, a co-author and AstraZeneca employee, was involved in data analysis, preparation of the manuscript and the decision to submit the manuscript for publication. The manuscript was reviewed by the funding sources prior to publication however ultimate responsibility for opinions, conclusions and data interpretations lies with the authors.
Data for this study were provided by Cambridge Clinical Informatics (Led by Drs. Afzal Chaudhry and Lydia Drumright).
TFH is supported by the National Institute for Health Research (NIHR) Cambridge Biomedical Research Centre (BRC) and by NIHR 14/49/127, 16/167/120, 17/83/06 and holds awards from Kidney Research UK, the Addenbrooke's Charitable Trust and the April Trust. TFH receives research funding from AstraZeneca. TJLH is a PhD student supported by the Cambridge Experimental Medicine Initiative (EMI) programme and receives research funding from both the NIHR Cambridge BRC and AstraZeneca. GJ is an employee and stockholder of AstraZeneca. IBW is supported by the British Heart Foundation, NIHR Cambridge BRC and receives research funding from AstraZeneca.
Hyperkalemia in hospitalized patients: Causes, adequacy of treatment, and results of an attempt to improve physician compliance with published therapy guidelines.