Introduction

Anticonvulsive rescue medication is used for the cessation of acute epileptic seizures including occasional seizure disorders such as febrile seizures [2]. These emergency situations frequently occur outside the hospital [1]. For seizures with duration of more than 5 min, administration of rescue medication is recommended [2]. The likelihood of a successful termination of a seizure lies at 96 % in the case of administering rescue medication within the first 15 min after seizure onset; success rates, however, decline to 57 % if the prescribed rescue medication is being administered with undue delay [10]. In case of a seizure, benzodiazepines such as rectal diazepam and buccal midazolam are frequently used for immediate use, especially outside the hospital [7, 9, 13]. As these dosage forms are used only occasionally and most frequently by non-medical professionals such as parents of children suffering from seizures, particular attention should be paid to their correct administration [8]. Inappropriate handling skills in drug administration by the parents might lead to lower drug effectiveness and severe harm to the patient. So far, information on the actual procedure of rescue medication handling by parents is very scarce, even though the stressful conditions may increase the risk of committing handling errors caused by lack of skill or knowledge. Previous studies mainly focused on the need of guidelines regarding choice and dosage of active ingredients and routes and time of administration to treat seizures [11, 13, 16]. In most cases, however, they did not address possible prevention strategies to ensure safe drug handling. In one study, the effect of a training of non-licensed buccal administration of midazolam was evaluated by conducting a telephone interview to the participating parents [6]. Handling errors in routine drug administration of regular long-term medication are accessible by monitoring, whereas the occurrence of a seizure, and thus a real-time setting for the monitoring of an administration of rescue medication, can hardly be predicted. Moreover, for ethical reasons, it is impossible not to intervene in critical situations such as an emergency. However, it is necessary for prescribing physicians to be informed about the parents’ drug administration performance. Only by an appropriate administration, the effect of rescue medication can be ensured [6, 10]. We aimed at investigating parents’ practical performance in administering anticonvulsive rescue medication to their children. For extended information, we used dummy dolls to observe drug handling in addition to parent’s self-report.

Materials and methods

Participants and setting

After obtaining ethical approval for the study protocol from the local ethics committee and written informed consent from the parents concerned, we enrolled parents of patients to participate in this study during their child’s inpatient treatment or outpatient visit in our neuropediatric department. All parents with children suffering from seizures including febrile seizures were invited to take part in this study, if diazepam as rectal tube and/or buccal midazolam was prescribed as rescue medication. Parents who were not able to understand the given task were excluded from the study.

Study protocol

Firstly, an expert panel was set up to predefine and classify typical handling errors possibly occurring in emergency drug administration of frequently prescribed rectal and buccal dosage forms. Secondly, we asked participating parents to fill in a questionnaire to assess their experience of emergency administration. Finally, we assessed their practical performance in rescue medication administration by having them use dummy dolls.

Judgment of potential clinical risk of drug-handling errors by an expert panel

An expert panel consisting of pediatric neurologists and clinical pharmacists with several years of practical experience was set up. The panel predefined a number of potential handling errors regarding rectal tube and oral syringe for buccal administration considering the handling information given in the current version of the respective drug label and further literature as required. To classify the handling errors in respect to their potential clinical risk to harm the patient in an emergency setting, the panel rated the predefined handling errors on a scale of “1” (lowest clinical risk) to “6” (highest clinical risk).

Experience of emergency administration judged by a questionnaire

A questionnaire was developed and forwarded to the parents. It addressed demographic data (age, gender), whether the prescribed rescue medication had been used previously and whether the parents experienced difficulties administering these medications. We assessed whether the parents had been informed by their physician, pharmacist, or nurse, whether they had obtained information from the patient information leaflet, and what kind of procedure would be taken in case the rescue medication failed to terminate the seizure.

Practical performance in emergency administration observed by dummy doll use

We prepared two dummy dolls to facilitate a realistic drug administration of a rectal tube and an oral syringe for buccal administration. By using these types of drug administration, we chose the two dosage forms of anticonvulsive rescue medication that are most frequently prescribed. Parents had to administer the placebo device of the rescue medication that was prescribed to their child to the respective dummy, i.e., to a prepared doll for rectal administration or an artificial mouth reproduction for buccal administration. In case of two prescribed rescue medications with different dosage forms, dummy dolls were used for each dosage form. We performed no actual drug administrations to patients nor did we handle active ingredients. Trained clinical pharmacists acted as monitors to identify and to document the predefined handling errors in administering rescue medications.

Data analysis

Calculations were performed using SPSS (Statistical Package for the Social Sciences, Version 20, IBM, USA). Frequencies are reported as numbers and percentages, continuous data as median with first (25 %) and third (75 %) quartile (Q25/Q75) and minimum/maximum.

For comparison of processes with at least one handling error in rectal or buccal administration, we applied Fisher’s exact test. Additionally, we performed logistic regressions (number of errors per process) for both dosage forms in order to analyze differences between parents who had previously used the rescue medication and those who had no practical experience so far. A p value ≤0.05 was considered to indicate significance.

Results

Characteristics of patients and participating parents

From 95 parents/children fulfilling the inclusion criteria and giving their informed consent, 81 (85 %) participated in the two parts of the study, i.e., filling in the questionnaire and performing administration of rectal and/or buccal placebo devices to the dummy dolls (Table 1). From the children of the enrolled parents, 41 % (33 of 81) were hospitalized and 59 % (48 of 81) were treated as outpatients at the time of the study.

Table 1 Demographic data from 81 participating parents/children
Full size table

Ten percent (8 of 81) of children had exclusively experienced febrile seizures, whereas 90 % (73 of 81) had experienced at least one afebrile seizure. Among them, 84 % (61 of 73) received anticonvulsive long-term treatment.

For 4 % (3 of 81) of children, two different dosage forms (i.e., rectal and buccal administration) had been prescribed at the same time. This way, we assessed 81 questionnaires and 84 observed processes with the dummy dolls, of those 60 of rectal and 24 of buccal administration.

Judgment of potential clinical risk of drug-handling errors by an expert panel

The expert panel predefined seven potential handling errors as being typical for the rectal administration and five for the buccal administration and rated them all as clinically harmful with risk scores reaching from “4” to “6” (as shown in Tables 3 and 4).

Experience of emergency administration judged by a questionnaire

Regarding the rectal medication, 45 % (27 of 60) of the parents reported that they had previously given rectal rescue medication to their child. For 85 % (23 of 27) of those actually administered rectal medications, parents self-reported that they had not yet experienced problems with the administration.

In the assessment of buccal administration, 33 % (8 of 24) of participating parents reported that they had previously used the buccal rescue medication for their child. In 100 % (8 of 8) of these administrations, the corresponding parents reported that they had not yet experienced problems with the emergency administration.

Parents named the physician as the main source of information on emergency drugs, and most frequently reported that they would call an emergency physician as measure in the case of failure of rescue medication (Table 2).

Table 2 Experience of rectal and buccal emergency administration
Full size table

Practical performance in emergency administration observed by dummy doll use

The absolute number and relative frequency of detected handling errors referring to their clinical risk score is shown in Tables 3 and 4. In 97 % (58 of 60), we observed processes with at least one handling error in rectal administrations which were more frequent than in buccal administration (58 %; 14 of 24; p < 0.001).

Table 3 Rectal administration: clinical severity according to expert rating and frequency of predefined handling errors
Full size table
Table 4 Buccal administration: clinical severity according to expert rating and frequency of predefined handling errors
Full size table

A previous use of rescue medication did not have any effect on the practical performance of participating parents: we did not find any differences regarding error-free processes between parents who had previously used rescue medication and those who had not. According to the logistic regressions, the previous use of rescue medication was not a predictor of the number of handling errors per process (logistic regressions for buccal and rectal administration: n. s.).

Discussion

Outside the hospital drug administration to children is regularly not performed by healthcare professionals but by the caregivers, most frequently the parents, of the child [8]. To achieve a safe and efficient drug administration, it is important for healthcare professionals to know possible and individual skill deficits in parents who are responsible for the drug administration in most acute seizures of their children. In case of an emergency, the assessment of handling errors without intervening is not acceptable for ethical reasons. Therefore, we used dummy dolls for frequently prescribed dosage forms, namely a rectal tube and an oral syringe for buccal administration, to identify handling errors in the administration of rescue medication. To assess drug-handling errors in a situation hardly accessible by routine monitoring, our dummy dolls are a close-to-reality approach. This way, we were able to access handling errors in drug administration that might occur in the event of an actual seizure. As far as we know, similar investigations using simulations of drug administration are very limited. Dummy dolls are most frequently used to train first aid measures or as a model in abusive head trauma [18].

Self-reporting in contrast to performance in administration of rectal and buccal placebo devices to dummy dolls

In our observations, irrespective whether rescue medication had already been administered or not, we frequently identified errors in handling rescue medications by parents of children with seizures. Interestingly, those results were in contrast to the parents’ self-reported performance in the questionnaire: 85 % of participants who already conducted rectal administration in an emergency reported not ever having experienced problems. The rate of participants who never experienced problems in buccal administration according to self-estimation was even 100 %. This way, parents’ self-reports have been shown not to be a suitable method to identify drug-related handling errors. This further indicates that parents cannot be expected to seek advice from a healthcare professional because they are not aware of these errors. This is in line with the results of Connolly et al. who have shown that parents who had to administer a buccal rescue medication frequently reported on non-predictable problems that focus on the patient and on physical problems related to epilepsy, such as excessive secretions and vomiting. Those reported problems, however, were not associated with a higher risk to fail in the termination of the seizure [6]. Therefore, we conclude that the errors that parents are aware of may not be the ones impeding a successful administration of the rescue medication. Consequently, drug administration to dummy dolls may help to identify those handling errors that might actually have an impact on the drug safety of rescue medications.

High potential risk according to an expert panel

A high fraction of the identified errors for both dosage forms were rated by the expert panel as the highest risk errors, emphasizing the clinical relevance. For rectal administration, there were more handling errors predefined than for the buccal dosage form going along with an increased risk to actually commit handling errors for rectal administration. Those additional handling errors which can only be committed during rectal administration were rated by the expert panel with a relatively high-risk score of “4.” In fact, our results show that rectal administration is not only more prone to potential but also to actually committed handling errors than buccal administration.

Drug-handling errors particularly frequent in emergency situations

Drug-handling errors deriving from poor administration performance are likely to occur frequently in emergency situations. This might be due to the fact that parents are not as familiar with the drug handling of these special dosage forms as they are in continuous administration of long-term medication. Nevertheless, they need to react immediately. Furthermore, an acute seizure is a particular stressful situation for the parents. Consecutively, the risk of committing an error may further increase.

Hence, there is an urgent need for strategies to assess the parents’ drug-handling performance and to identify skill deficits that might lead to inappropriate drug handling.

Previous administration no predictor of good handling performance

Surprisingly, a previous administration of the rescue medication in the past was not a predictor of less frequent handling errors in both dosage forms. According to our observations with the dummy dolls, a practical experience does not seem to ensure an appropriate administration. Therefore, healthcare professionals should be encouraged to repeatedly inform all parents involved in emergency drug administration about the appropriate use in an emergency situation. In fact, parents would consider a detailed instruction by a trained healthcare professional on dosing and administering rescue medication as valuable [6].

Interventional strategies based on data assessment using dummy dolls

Parents have to take full responsibility for the drug administration to their children after hospital discharge [17]. An education program for parents who administered medication during their child’s hospital stay decreased the rate of handling errors [3]. A training of nurses in appropriate drug handling achieved a significant reduction in drug-handling errors per process [14, 15]. In those previous studies, however, there was no specific focus on rescue medications. The dummy dolls as described in this study facilitate the detection of handling errors in the administration of rescue medication without any risk of harming the patient. They are feasible instruments to support the identification of drug-related handling errors. The identification of those errors is required for successful interventions. Alongside the assessment of parents’ performance in rescue medication handling, it may be used as an instrument to investigate the appropriate administration of rescue medication by other caregivers or even in further interventional studies.

Anticonvulsive rescue medication—errors in the overall medication process

A recent study has shown that in 49 % of patients who were deemed to require rescue medication, no recommendation for such a medication was included in the discharge letters, and 22 % of the respective parents stated that they had no immediate access to rescue medication at all [4]. In addition to these supply shortfalls in rescue medication prescription, our study shows that the administration of rescue medication is also very error-prone. The specific dosage forms in rescue medications require specific skills and knowledge to administer these drugs appropriately. Besides, for total quality management, the entire medication process should be investigated, including both prescription and administration.

So far, administration of rescue medication has rarely been addressed in respective studies. For instance, Callegaro et al. investigated the implementation of guidelines to manage febrile seizures in two pediatric emergency departments, including indications for home administration of rescue medication [5]. In a simulation of a prehospital pediatric emergency, medical service crews had an error rate of 47 % in choosing the right diazepam dosing and 60 % in choosing the right dose for midazolam [12]. Anyhow, practical handling errors were not in the focus of those investigations. Connolly et al. performed a study on parents with children suffering from seizures that focused on a detailed instruction on correct dosing and administration of buccal midazolam, given by trained healthcare professionals to parents. The assessment of errors in the subsequent administration process was performed by a follow-up telephone survey. According to our results, however, self-reports from parents are not an appropriate instrument to identify all errors actually committed in drug administration. Though, the clinical relevance of non-reported errors in drug handling has been shown by our expert risk rating.

Revealing those errors with our dummy dolls made them accessible for targeted interventions. Therefore, our dummy dolls support a successful treatment of acute seizures in an emergency.

Limitations

We monitored simulated drug administrations with dummy dolls and were not able to compare our results to a monitoring of actual administrations of rescue medication in patients. Even if the situation was close-to-reality, a real emergency situation will be more stressful for the patients and their parents, likely resulting in a higher frequency of handling errors. Consequently, problems that are caused by the emergency situation itself, e.g., drug loss because of mishaps in drug handling or because of the patients’ physical problems related to epilepsy, could not have been monitored by using the dummy dolls.

In our study, we did not focus on the different types of seizures that were diagnosed in the children and the associated performances of drug handling of their parents. A differentiated analysis is to be continued in future investigations to determine patient groups with a particular need for close supervision to ensure a successful administration of rescue medications in case of an occurring seizure.

Many seizures might cease spontaneously even though emergency medication was applied in a wrong way, whereas others might result in status epilepticus. It is not predictable which seizures cease spontaneously, if either no medication is administered or medication is administered in the wrong way. Nor it is predictable in which cases an administration of medication is necessary to cease the seizure. A wrong administration of emergency medications, however, poses the risk for prolonged seizures which cannot be ceased. To minimize the risk of non-treatable status epilepticus, emergency medication should be administered timely and in a technically right way.

Our study aimed at investigating the risk of administration errors in emergency medications. This way, we cannot extrapolate to actually occurring adverse drug events as in the questionnaire we focused on knowledge on administration of rescue medication and planned measures in case of an emergency. We did not specifically ask for concrete experiences in the past (e.g., frequency of calling the emergency physician due to failure of the emergency medication or occurrence of adverse drug events such as respiratory depression). The risk of respiratory depression deriving from benzodiazepines can increase with the repeated use of emergency medication. This should be evaluated in a prospective design within a further investigation.

Conclusion

We identified emergency administration of rescue medications by parents to their children with seizures as a highly error-prone process. Rectal administration was accompanied by more errors than buccal administration. Thus, we finally conclude that parents have to be regularly and intensively trained. As parents’ self-reported problems frequently did not include their actually committed handling errors, dummy dolls may be a basic support to identify high-risk handling errors.