This site is intended for health professionals only
M Isabel Vicente-Valor1
J Adrià Bargues-Ruiz1
1Clinical Resident Pharmacist
3Head of the Pharmacy Department
Hospital Universitario de La Ribera
‘High-risk drugs’ are those that when used incorrectly have a high probability of causing serious harm or death to patients.1 Concentrated potassium chloride solutions require specific administration conditions. Potassium chloride can be added to intravenous fluid in a concentration to a maximum of 40mEq/L in order to avoid phlebitis and infused at a rate that does not exceed 10mEq/h to avoid the risk of arrhythmias.2 Non-compliance with specific administration requirements increases the risk of toxicity. In addition, the availability of different concentrations increases confusion.
Concentrated intravenous potassium is one of the most common drugs to be misused or mishandled in some way by healthcare professionals. Several organisations involved in the safe use of medicines include potassium among the list of drugs that most commonly cause safety problems. The United States Pharmacopoeia (USP) keeps a database of medication errors associated with acute hospital care that are reported anonymously and intravenous potassium appears among the top 10 high-alert drugs.3 Errors involving concentrated potassium also appear on the Institute of Safe Medicine Practices (ISMP) List of High-Alert Medications.4
The World Alliance for Patient Safety of the World Health Organization (WHO) has recently considered measures to avoid errors in relation to potassium as a priority and has included it in the ‘Nine Solutions for Patient Safety’.5,6 The nine solutions are based on interventions and action that have reduced problems related to patient safety in some countries. According to the WHO, dosage strategies, units of measure and terminology of concentrated electrolyte solutions used for injection, as well as their preparation and administration, must be standardised.
Spanish guidelines for potassium
Following these worldwide recommendations, Spanish health authorities in cooperation with the Spanish delegation of the ISMP published specific guidelines for the safe use of potassium chloride in Spanish hospitals.7 The recommendations promoted by both institutions are:
To conduct all matters related to the safe use of medicines in the hospital a multidisciplinary Medicines Safety Group (SUMO) was established in June 2009 at the Hospital Universitario de La Ribera in accordance with the hospital’s policy of improving patient’s safety. Members of the SUMO include physicians representing various clinical departments and particularly those where drug use is particularly hazardous or presents specific problems, as well as nursing representatives and the Pharmacy Department. In addition, the Hospital Management Board, Quality Department and Computer Department, key departments that promote the decisions taken by the group, are also represented. Issues related to errors in prevention, detection and reporting, so as to establish guidelines to address the proper use of drugs, are addressed by the team at regular meetings.
Incident reporting system
The first step addressed by the group was the implementation of an incident reporting system to communicate anonymously any real or potential problems occurring during prescribing, dispensing or administering medications to the patient. According to ISMP-Spain’s recommendations, the first decision adopted by SUMO was to withdraw concentrated potassium solutions from ward stock and replace them with ready-to-use infusion products.
To implement this recommendation, four pre-diluted potassium chloride standardised solutions were acquired. Two concentrated presentations with restricted distribution to avoid any confusion in its usage, remained available. The standard solutions of potassium chloride are shown in Table 1.
Standardised intravenous potassium solutions were implemented in the electronic prescription system of our institution so that physicians can prescribe them. In addition, the system includes information about specific drug administration conditions to help in prescribing and administering. Any incident regarding these medications is communicated to the SUMO and evaluated at its regular meetings.
The SUMO considered that some clinical wards presented unique potassium handling characteristics that made it difficult to standardise solutions and concentrated potassium chloride ampoules should therefore be disposed of in these areas. An agreement with the Intensive Care Unit (ICU), Neonatal and Paediatric Unit, Emergency Department (ED), and the Haemodialysis and Centralised Intravenous Additive Service (CIVAS) was made to allow concentrated potassium solutions and specify their conditions of use.
Patients in the ICU and ED are often affected by volume restriction or acute hypokalemia that requires rapid potassium replacement. In the Neonatal and Paediatric Units the variability in the dose of potassium chloride handled makes it difficult to standardise dilutions. Pre-diluted intravenous admixtures administered to the paediatric or neonatal population are lacking but, on the other hand, patients with renal failure and those undergoing hemodialysis procedures usually present electrolyte disbalances. Potassium and other additives are often added to the crystalloid solutions used as replacement fluid in accordance with laboratory pre- and post-dialysis tests. In these wards, clinical staff are trained in handling high-risk drugs. Moreover, specific dilution protocols consistent with their particular characteristics are available to minimise potential risks.
An area in which the clinical pharmacist is involved significantly in error prevention is the CIVAS, where a double-check on drug prescription, drug preparation and administering conditions is applied. The CIVAS keeps concentrated potassium chloride ampoules for the preparation of parenteral nutrition (PN) to meet individual patient needs. In accordance with the institution’s policy, the PN prescription is standardised in order to reduce the minimum variability, optimise handling tasks and to reduce error opportunities and costs. However, standardisation is not always adequate in providing macronutrient requirements and, even more frequently, electrolyte requirements. In these patients, especially those in which PN lasts longer than four days, it is advisable to carry out individual monitoring of nutritional parameters and electrolyte plasma levels including potassium. Periodically monitoring and recording such data allows assessment of the suitability of standard PN to potassium requirements, and to identify situations which affect serum potassium concentrations in patients fed by PN.
Retrospective audit carried out
In this scenario, a retrospective audit in patients who received standard PN as single feeding route during January 2008 was conducted. Demographic data collected included age and sex of patients. The number of days of PN and complications such as abdominal intervention, diarrhoea, vomiting, diabetes or oedema were also registered. Blood clinical tests including serum creatinine level, urea, glucose, electrolytes and pH were implemented to verify biochemistry status and to adjust requirements when needed. In addition, any administration of glucose solution, potassium and any drug that potentially modifies serum potassium levels (see Tables 2 and 3) was recorded.
Hypokalemia was classified as mild (3.0–3.5mEq/L), moderate (2.5–3.0mEq/L) or severe (<2.5mEq/L), and hyperkalemia was defined as mild (5.5–6.0mEq/L) moderate (6.0–7.0mEq/L) or severe (>7.0mEq/L).8,9 The standard potassium content of PN was 32mEq/day in peripheral PN and 60mEq/day in central PN.
During the study period, 50 patients were included (28 men and 22 women) with a mean age of 65 years (range 28–91). A total of 19 patients (38%) developed hypokalemia and just one patient (2%) suffered hyperkalemia (see Figure 1). Of the 16 patients included in the mild hypokalemia group, eight patients presented with it before beginning treatment with PN; six patients developed hypokalemia after receiving peripheral PN for more than four days; and only two patients suffered hypokalemia after receiving central PN. In two cases a moderate hypokalemia before beginning PN treatment was observed (in both cases the reason was losses from the gastrointestinal tract). Just one case was registered as a severe hypokalemia due to severe diarrhoea. Eight patients (16%) who presented with hypokalemia underwent abdominal surgery during their hospitalisation.
In further analysis, furosemide was observed as the drug most frequently involved in hypokalemia development. Furosemide is a well-known drug implicated in potassium losses. In this study 10 patients (20%) were treated with this diuretic, a high prevalence accorded with its widely use. This is probably a subgroup of patients whose potassium levels should be checked before prescribing standard doses, and close monitoring during treatment should be performed.
A mild hyperkalemia (serum potassium level was 5.8mEq/L) was developed by one patient. Hyperkalemia was developed in this patient after receiving a 90mEq/day parenteral dose of potassium chloride for more than four days to correct a previous status of mild hypokalemia. Hypokalemia seems to be more prevalent than hyperkalemia in patients receiving PN. This is to be expected as many patients start PN because of gastrointestinal disease that increases the risk of potassium losses.
Our data indicate that potassium levels should be checked and corrected when required before starting PN. Administration of standard peripheral nutrition for more than four days could induce hypokalemia, as potassium requirements are not satisfied with potassium doses included in PN. In this situation, central PN or the resumption of oral diet should be encouraged as soon as possible as peripheral PN can rarely meet patient’s macro and micronutrients’ requirements. It is advisable to monitor potassium in patients who receive drugs which modify plasma potassium levels or who suffer abdominal surgery.
Potassium dilution and standardisation increases security in potassium handling and avoids fatal errors. However, individual requirements are not always met with standard solutions. Detection of potassium inbalance in patients receiving PN is quite an easy task as patients on nutritional support are routinely monitored. A more ambitious step in patient care is extending potassium monitoring and requirements individually to patients receiving standardised potassium in crystalloids solutions. To address this challenge, preselecting patients receiving drugs that most induce potassium inbalance is a reliable task. Although it seems a time-consuming effort in an environment of scarce human resources, the high level of computerisation at our hospital makes it a manageable task. Drug identification through the electronic prescribing system and the accessibility of potassium levels throughout the laboratory’s electronic programme will allow the extension of the procedure to a higher number of patients.
Hospital pharmacists have a key role in the comprehensive management of patients from the point of view of preventing errors, promoting treatment policies and procedures within their institutions, and individually monitoring patient’s therapy.
1. Cohen MR et al. Medication Errors. 2nd ed Washington DC: American Pharmaceutical Association; 2007, 317–411.
2. Koda-Kimble MA (ed). Applied therapeutics. 8th ed. Baltimore: Lippincott Williams & Wilkins; 2005.
3. Santell JP et al. Drug topics health-system edition. 8 December 2003. Available online at: www.usp.org/pdf/EN/patientSafety/drugSafetyReview2003–12–08.pdf (accessed 6 September 2010).
4. Institute for Safe Medication Practices. Available online at: www.ismp.org/Tools/highalertmedications.pdf (accessed 6 September 2010).
5. World Health Organization. May 2007. Available online at: www.who.int/patientsafety/solutions/patientsafety/Preamble.pdf (accessed 6 September 2010).
6. World Health Organization. Volume 1. Solution 5. May 2007. Available online at: www.who.int/patientsafety/solutions/patientsafety/PS-Solution5.pdf (accessed 6 September 2010).
7. Madrid: Ministerio de Sanidad y Consumo; 2009. Available online at: www.seguridaddelpaciente.es/formacion/tutoriales/MSC-CD4/ (accessed 6 September 2010).
8. Garth D. 2 April2010. Available online at: http://emedicine.medscape.com/article/767448-overview (accessed 6 September 2010).
9. Garth D. 16 April 2010. Available online at: http://emedicine.medscape.com/article/766479-overview (accessed 6 September 2010).