|Home | About | Journals | Submit | Contact Us | Français|
Management of paracetamol overdose (POD) is common in the emergency department (ED) and forms part of the clinical effectiveness audit programme of the British Association for Emergency Medicine. N‐acetylcysteine (NAC) infusion regimens for the treatment of POD are complicated and prescribing and administration errors have been well documented. This study assessed the ability of doctors and nurses to calculate correct doses using manual calculation skills and a weight‐based NAC dosing chart when prescribing and preparing NAC infusions. With manual calculations, errors were made by doctors and nurses in 26% of cases collectively. No errors were made using the dosing chart. The dosing chart ensured 100% accuracy in dose calculations, which may translate into improved patient safety.
Treatment of paracetamol overdose (POD) with N‐acetylcysteine (NAC) prevents hepatic failure. Suboptimal treatment may adversely affect patient outcome. NAC infusion regimens require complex calculations to be made by both prescribers and nurses preparing infusions.1 Errors with intravenous medicines requiring multiple step preparation or complex calculations have been highlighted.2,3 The use of dosing charts to prevent medication errors associated with complex calculations has been proposed by the Department of Health as a risk reduction strategy.4
Typically, doses of NAC, diluent fluids and their volumes and durations of infusions are prescribed by doctors on intravenous infusion charts. In the UK, the preparation and administration of parenteral treatment is usually performed by trained nurses who calculate volumes of a 20% NAC solution (Parvolex®) required to prepare infusions based on doctors' prescriptions. When preparing NAC infusions, calculation errors have resulted in doses varying by more than 50% from the intended dose.4
In 2005, there were 112000 attendances to our emergency department (ED). Approximately 80 patients per month presented with a POD and 25% of these were treated with NAC. The infrequency with which NAC dose and volume calculations are performed in our department may increase the risk of medication errors. Most NAC prescribing for POD is initiated in EDs where pharmacists do not routinely check prescriptions or drug administration to ensure safe medicine use. Incorrect prescriptions for NAC infusions initiated in the ED have been identified and corrected by pharmacists when patients have been reviewed following admission to wards.
The aim of this study was to compare the accuracy of doses and volumes calculated manually by ED staff when prescribing and preparing NAC infusion regimens with those derived from a weight‐based NAC dosing chart.
A weight‐based NAC dosing chart (fig 11),), based on the standard treatment regimen for POD in adults (http://www.spib.axl.co.uk;http://emc.medicines.org.uk/emc/industry/default.asp?page=displaydoc.asp&documentid=1127),5 was designed by pharmacists in conjunction with ED practitioners. Doctors and nurses of all grades involved in NAC prescribing and administration in the ED completed two written assessments. In each assessment, individual doctors were required to calculate and prescribe a NAC infusion regimen for a hypothetical patient weighing 58 kg. Each participating nurse was asked to calculate volumes of NAC required to prepare infusions against an accurate prescription for a hypothetical patient weighing 58 kg. Assessment I (manual) was completed according to current practice by the study participants using their mathematical skills and preferred personal method. TOXBASE (http://www.spib.axl.co.uk), British National Formulary, calculators and supplies of NAC ampoules, including product information, were available. Assessment II was completed with the aid of the dosing chart (fig 11)) which describes doses and volumes of NAC for each of the three regimens for adult patients with weights ranging from 40–110 kg (maximum patient weight for calculation of NAC doses) (http://www.spib.axl.co.uk).
Assessments were completed by 27 doctors and 53 nurses (table 11).). In assessment I, doctors calculated 81 doses on 27 prescriptions. Five dose errors (5/81, 6%) were made by three doctors (3/27, 11%) on three prescriptions. Prescribing errors included two 1000‐fold underdose errors in the 15 min and 4 h regimen, respectively, a 14% underdose in the 4 h regimen, and two overdoses of 50% and 53% in the 16 h regimen.
Nurses performed 159 volume calculations on 53 prescriptions. Thirty‐eight volume miscalculations (38/159, 24%) were made by 18 nurses (18/53, 34%) on 18 prescriptions. Volume errors in the 15 min regimen ranged between −80% and +1230% of the prescribed dose. Errors in both the 4 h regimen and the 16 h regimen ranged between −80% and +1059% of the prescribed dose. Twenty‐one errors were underdoses (21/159, 13%).
In assessment II, no errors were made by doctors or nurses in determining either doses or volumes.
The quantification of error reduction introduced by use of the NAC dosing chart was carried out using the intra‐class correlation coefficient (ICC) and coefficients of variation (CV) for doctors and nurses in each part of the NAC regimen and is illustrated by the Bland‐Altman plot (fig 22).
Table 22 shows the ICC for assessment I (manual) and II (dosing chart) for each part of the NAC regimen. This represents the proportion of the total variability between study participants. An ICC of 0% indicates a substantial correction made by the chart, for both doctors and nurses. Likewise, the CVs (table 33)) are mostly above 20%.
Timely treatment of POD with NAC prevents hepatic failure. For treatment to be safe and effective it is essential that NAC doses are prescribed, prepared and administered accurately. Assuming that the errors identified in this study would have occurred in clinical practice, outcomes may have been compromised in 21 patients (26% (21/80) of cases), 16 of whom would have received underdoses (20%). The chart improved the accuracy of calculations as evident by the finding that five out of six CVs were >20%, indicating a substantial error reduction. The chart was also well received by the staff who used it.
The study does have limitations. Calculations were performed under “test” conditions away from the usual working environment. These conditions are more likely to have favoured good performance, possibly underestimating the error rate compared to busy clinical practice. Additionally, it is not known whether any of these calculation errors would have been identified and corrected by another nurse, as in clinical practice preparation of the infusion is double checked before administration to the patient. This study focuses on the calculation of NAC doses and volumes and assumes that, in order for patients to receive effective treatment, all other stages in the preparation and administration of infusions—including selection of the diluent, drawing up the Parvolex® volume and setting the infusion rate—are carried out correctly.
The prescribing and administration of NAC is a two‐step process. In clinical practice, it is possible that if a doctor makes an error in the dose calculation and therefore prescribes an incorrect dose, the error will not be identified by nurses. The prescribing error will then be carried through to the preparation and administration phase unless the prescriber's calculation is rechecked. The dosing chart ensures accuracy of all calculations and facilitates double‐checking of the prescription by pharmacists and nurses involved in the pharmaceutical care of the patient.
The NAC dosing chart eliminated calculation errors in the staff tested and appeared to be superior to current practice in ensuring accurate dose and volume calculations. This may translate into improved patient safety and outcome. Since the study, the chart has been modified (fig 11)) and incorporated into an integrated POD management pathway in our ED. The effectiveness of this chart in clinical practice is being evaluated.
CV - coefficient of variation
ED - emergency department
ICC - intra‐class correlation coefficient
NAC - N‐acetylcysteine
POD - paracetamol overdose
Competing interests: None declared.