Managing prolonged seizures in the pre-hospital setting is an undertaking that presents significant challenges to clinicians. Patients can present with airway and breathing compromise, transport can be dangerous and there is the potential for refractory (medication resistant) status epilepticus, which has been shown to become more likely as ‘onset to drug’ time increases (Goodkin and Kapur, 2003).
Medications of the benzodiazepine class are psychoactive, all acting to increase the effect of the inhibitory neurotransmitter gamma-aminobutyric acid (GABA). Pertinent to pre-hospital care is the potent anti-convulsant effect of these medications, the efficacy of which varies across different formulations, which are further categorised depending on their duration of action.
National Ambulance Service guidelines stipulate that the first-line paramedic-led treatment for generalised seizures lasting in excess of 5 minutes should be 10 mg of intravenous diazepam, with the per rectum (PR) route considered where intravenous (IV) access is difficult (Joint Royal Colleges Ambulance Liaison Committee, 2016). Diazepam is known to be effective in the termination of seizures in 50% to 80% of cases (Treiman, 1989), although the drug redistributes into adipose tissue quickly, and as such has a short duration of action of approximately 20 minutes (Greenblatt and Divoll, 1983). In addition, it has been shown that seizure recurrence occurs in up to 50% of patients, leading to the administration of repeat diazepam doses (Cock, 2002).
Upon review of the clinical evidence in support of benzodiazepine use for the termination of seizures, there is a general consensus that no single drug of this class is particularly superior in terms of efficacy. However, there may be other reasons to look into the use of benzodiazepine medications other than diazepam as an alternative first-line treatment option for paramedics.
Pathophysiology of seizures
Status epilepticus is the manifestation of a disorder in which the mechanisms of seizure termination fail. It is a complex condition in itself, and is by no means limited to convulsive seizures. To demonstrate this, Shorvon (1994) proposed the following definition: ‘Status epilepticus is a disorder in which epileptic activity persists for 30 minutes or more, causing a wide spectrum of clinical symptoms, and with a highly variable pathophysiological, anatomical, and aetiological basis.’
The mechanism by which seizure activity initiates and terminates is not well understood, but there is general consensus that the most likely cause is an imbalance between excitatory and inhibitory neurotransmission. These imbalances are thought to lead to abnormal chaotic neural impulses, which then start a seizure. Convulsive seizures have long been associated with cellular damage that is significant enough to render a patient with life-limiting or life-threatening brain dysfunction. It is thought that this is due, in part, to the failure of homoeostatic compensation seen after approximately 30 minutes of active convulsions (Fountain and Lothman, 1995), which inevitably lead to cerebral hypoxia.
Pharmacodynamics of benzodiazepines
Benzodiazepines target the allosteric sites of the GABAa receptor. These receptors are ligand gated ion channels (LGICs), that exist in all neuronal tissues, and play a part in almost all aspects of brain function. The receptors' function when activated is to open a pore and allow chloride anions to reach the associated neuron. The result of this ion influx is a stabilisation or hyperpolarisation of the neuronal cell, making depolarisation by excitatory neurotransmitters (and therefore the generation of action potential) less likely.
The endogenous ligand of GABAa receptors is the inhibitory neurotransmitter gamma-aminobutyric acid (GABA). Benzodiazepines act to increase the affinity of the GABA molecule at the receptor's a subunit by modifying the molecular structure of the receptor itself. Drugs such as diazepam therefore indirectly inhibit excitation, and have particular value in controlling the excessive neuronal activity that is commonly associated with seizures.
Review methods
The focus of this review is to compare benefits of diazepam and lorazepam in the pre-hospital care setting. As there is limited evidence, the search criteria were widened to allow data from other medical settings. The search plan and search strategy are shown below (see Table 1, Table 2).
| Research topic |
| To compare the effectiveness of diazepam to that of lorazepam in the management of convulsive status epilepticus in adults. |
| Category definitions |
| Populations and/or problem |
| Intervention |
| Comparison |
| Outcome |
| Research question |
| Is diazepam or lorazepam the most effective benzodiazepine for use in paramedic management of convulsive seizures in adults? |
| Search restrictions |
| Databases were accessed using both the Oxford Bookes University library and South Central Ambulance Service OpenAthens portal to ensure full text versions of each article could be reviewed. Only studies in English were included. |
| Search start: January 2016 at 10:32 |
| Inclusion criteria |
| Studies that compared diazepam and lorazepam. |
| Exclusion criteria |
| Studies published prior to 1963. |
| Search string |
| (diazepam AND lorazepam AND (seizure* OR convuls* OR fit* OR generalised OR status epilepticus)) NOT (child or paediatric or pediatric) |
| Search results |
|
CINAHL
|
A search was undertaken in January 2016 across the databases Cumulative Index of Nursing and Allied Health Literature (CINAHL) and PubMed (inclusive of MEDLINE). These databases were primarily targeted as they contain a sufficient breadth of indexed journals to provide for quality data. This means that there was a reduced likelihood of omitting high quality papers from this review as a result of poor indexing.
CINAHL provides an extensive index of 3000 English language journals and 2.3 million records focused on nursing and allied healthcare from North America and Europe. CINAHL was selected for this review because of its wide selection of indeces, which includes 27 titles identified within the CINAHL ‘coverage list’ (EBSCO, 2016) that contain the search terms ‘Paramed*’, ‘Emergency’, ‘Pre-hospital’, ‘Prehospital’ and ‘Pharmacology’.
The PubMed database contains in excess of 25 million publications including those indexed by MEDLINE, additional life science journals and biomedical literature. Indexing also includes those publications not yet included in the MEDLINE database, which consequently improves coverage of recently submitted research.
Directly comparative studies
Alldredge et al (2001) conducted the only study in this series which presents results from the out-of-hospital environment. Recruitment into the trial was based on a paramedic-led diagnosis of status epilepticus in the presence of generalised tonic-clonic convulsive activity lasting 5 minutes or more.
Alongside its primary outcomes, this study also measured the occurrence of functional cardiorespiratory changes prior to arrival at hospital, finding that lorazepam and diazepam administration related to complications in 10.6% and 10.7% of cases, respectively. In contrast, 22.5% of cases treated with placebo presented with complications meeting the same criteria. (p=0.08). As the team identify, this suggests that cardiorespiratory complications are less likely to occur when patients are treated with benzodiazepines.
No significant differences were found in rates of survival to discharge and quality of neurological outcome between groups; however, patients arriving at hospital still in status epilepticus were more likely to require intensive care admission than those in whom pre-hospital termination of seizures had been successful (73% versus 32% respectively, chi-square <0.001). Odds ratio for the successful outcome measure was 1.9, 95% CI (0.8–4.4) in the lorazepam versus diazepam comparison. Lorazepam versus placebo and diazepam versus placebo ratios were 4.8, 95% CI (1.9–13.0) and 2.3, 95% CI (1.0–5.9) respectively.
In the study conducted by Treiman et al (1998), direct comparison between medications is not possible due to diazepam and phenytoin being administered consecutively as a single study agent. Despite this, the results demonstrate significant differences in success across the treatment regimes studied (p=0.02). In a pairwise comparison, lorazepam was effective on many more occasions than phenytoin alone (p=0.002), whereas the latter versus diazepam with phenytoin demonstrated no significant difference.
Leppik's results suggest that lorazepam demonstrated better efficacy for the termination of seizures of all types when compared to diazepam; however, this difference reportedly does not attribute statistical significance (unfortunately, the basis for statistical significance in relation to this study is not documented). There was an equal incidence of adverse clinical effects (defined as respiratory depression or arrest) of 13% (41) versus 12% (40) across the two study groups.
Cock and Schapira conduct the only UK based study in this series as a retrospective case note audit. The authors identify significant differences between groups receiving diazepam or lorazepam with the latter offering a higher likelihood of treatment success where it is the first benzodiazepine administered in hospital (p=0.042). It should be noted that there was no difference in seizure termination efficacy, but recurrence after 12 hours was less likely in the lorazepam group (p=0.005). Interestingly, the team identify that problems with guideline implementation may affect adherence to new evidence-based protocols relating to seizure management, and suggest that more extensive use of lorazepam by paramedic staff may help to increase familiarity with the agent and foster its widespread adoption. A breakdown of each study is provided in Table 3 and Table 4.
| Author, date and country | Patient group | Cohort size | Trial type |
|---|---|---|---|
| Alldredge et al, 2001. United States. | Adults (>18 years) with an out of hospital diagnosis of status epilepticus | 205 patients | RCT |
| Treiman et al, 1998. United States. | Adults (>18 years) with an in-hospital diagnosis of overt or subtle status epilepticus | 384 patients (verified overt SE) | RCT |
| Leppik, 1983. United States. | Adults (>18 years) with convulsive, absence or partial status epilepticus | 70 episodes | RCT. |
| Cock and Schapira, 2002. United Kingdom. | Patients >16 years of age with diagnostic code; ‘epilepsy’, ‘fits’ or ‘status epilepticus’ | 72 episodes | Retrospective case note audit |
| Author, date and country | Outcomes | Key results | Weaknesses |
|---|---|---|---|
| Alldredge et al, 2001. United States. | Termination of convulsive seizure activity by the time of ED arrival |
|
|
| Treiman et al, 1998. United States. | Treatment successful if all clinical and pre-defined electrical (EEG) evidence of seizure activity stopped within 20 mins of the start of infusion, and if there was no recurrence between 20-60 minutes' post-treatment |
|
|
| Leppik, 1983. United States. | Seizure termination following one or two IV doses of study drug |
|
|
| Cock and Schapira, 2002. United Kingdom. | Cessation of seizures without recurrence within the following 12 hours |
|
|
Additional observations
Efficacy: seizure termination
In several studies, lorazepam has been shown to successfully terminate seizures in a greater number of patients than diazepam (Treiman et al, 1998; Alldredge et al, 2001). A literature review by Appleton, Macleod and Martland is supportive of lorazepam's efficacy in children when compared with other common benzodiazepines, and presents a case for intranasal use (2008).
Cock and Schapira (2002) found that both drugs were similarly effective in the termination of convulsive status epilepticus; however, there were fewer recurrences of seizures (and therefore less repeat dosages required), and fewer occurrences of pharmacological sedation and endotracheal intubation due to refractory status epilepticus where lorazepam was used. This provides clinical evidence as to the real-world benefits of lorazepam's extended half-life when compared to diazepam – a result of the different distribution characteristics of the drug (Greenblatt and Divoll, 1983).
Timing: administration to action delay
Both drugs have a rapid duration of onset when administered as an IV bolus, with negligible differences in terms of the time taken in achieving a meaningful clinical effect (Greenblatt et al, 1989; Treiman et al, 1998; Trinka, 2009). The biggest differences occur when the two medications are delivered by IV infusion.
Administration routes
Both drugs can be administered by various routes with differing absorption and distribution characteristics (Lowenstein and Cloyd, 2007). IV is the most commonly recommended and utilised in pre-hospital care, and as such, forms a part of current clinical guidelines (Joint Royal Colleges Ambulance Liaison Committee, 2016).
Both drugs are rapidly absorbed into the brain, reaching peak plasma concentrations within a time frame of 1–5 minutes when administered intravenously (Arendt et al, 1983). However, diazepam is known to rapidly redistribute into adipose tissue due to its high lipid solubility, which accounts for its reduced duration of action in comparison with lorazepam. Alternative routes suitable for emergency administration such as buccal, nasal, rectal, oral/sublingual and intramuscular are compatible with both drugs. Time to maximum plasma concentrations following IM administration of diazepam and lorazepam is reportedly delayed (Greenblatt et al, 1982; Wermeling et al, 2001), which limits the potential for ‘auto-injector’ type device development.
Diazepam is more rapidly absorbed via rectal administration due to its higher lipid solubility (Malinovsky et al, 1993; Scott et al, 1998; Wermeling et al, 2001), reaching therapeutic blood plasma concentrations within 15 minutes in adults (Lombroso, 1989; Cloyd et al, 1998).
Studies that have considered intranasal administration of anticonvulsants demonstrate advantages in the rapid absorption of diazepam over lorazepam (Malinovsky et al, 1993; Burstein et al, 1997; Wermeling et al, 2001; Knoester et al, 2002; Riss, Kriel and Croyd, 2006); however, in relation to the comparatively large volume of diazepam required to reach effective plasma concentrations, it has been found that resultant leakage decreases the amount of agent absorbed (Burstein et al, 1997; Knoester et al, 2002).
Duration of action
Mitchell and Crawford (1990) demonstrate that lorazepam has a much longer duration of action than diazepam, which leads to an improvement in outcome for patients treated with the drug. It is thought that this is due to the marked reduction in seizure recurrence associated with lorazepam administration alongside the lesser need for repeat doses of the drug in order to maintain a seizure-free state. Secondary to this results a reduced incidence of pharmacological sedation and intubation/mechanical ventilation retrospectively. It should be noted however, that diazepam pharmacokinetics relies on the microsomal oxidation pathway. This implicates a number of highly active metabolites with extended half-lives (primarily desmethyldiazepam), of which clearance is highly dependent on variables such as age and liver function (Breimer, Jochemsen and von Albert, 1980). In contrast, lorazepam is metabolised via the glucuronidation pathway and has no active metabolites. Together these observations make any differences in absolute peak plasma concentrations less significant in terms of therapeutic efficacy. Additionally, it could be suggested that lorazepam presents a lower risk of undesired substrate interaction which improves its safety profile.
Safety
Both drugs are strong benzodiazepines and carry the primary risk of respiratory depression and arrest if used in high doses. There are documented cases of both adverse effects (often resulting in intensive therapy unit [ITU] management). Incident monitoring suggests that lorazepam results in fewer adverse events than diazepam. Appleton et al (2008) document a reduction in lorazepam linked respiratory depression in context, with no comparable differences in unexpected physiological side effects like anaphylaxis.
Storage
There are no regulatory constraints on the pre-hospital storage of diazepam or lorazepam with regards to environmental control; however, studies specifically focused on pre-hospital systems have been conducted, and these demonstrate a much improved stability profile in lorazepam than diazepam across a wide range of operating temperatures. Gottwald et al (1999) demonstrated that lorazepam maintained 90% of its original concentration after 120 days of unrefrigerated on-ambulance storage, compared with diazepam retaining the same concentration at a maximum of 30 days in a study conducted in San Francisco. In addition, Gottwald showed that if lorazepam was refrigerated between 4–10 degrees Celsius, a 0% reduction in concentration could be achieved at 210 days, in comparison to a 17% concentration loss in diazepam.
Cost
The per-ampoule dose of each drug varies. Diazepam as an emulsion (Diazemuls) 5 mg/mL (presented as 10mg in 2 mL) costs £0.91 per ampoule. Lorazepam 4mg/mL costs £0.35 per ampoule. All prices correct as of 01/09/16 (Joint Formulary Committee, 2016).
Limitations
Although these studies have a tendency towards lorazepam as the more effective agent, the dosages used between studies vary. There is no officially documented equivalent dose of lorazepam to diazepam although the latter is commonly administered in greater quantities or in combination with other anticonvulsants such as phenytoin.
There is a lack of literature comparing anticonvulsant therapies out of hospital. Studies suggest that significant challenges exist in evidence translation between in-hospital and out-of-hospital based research, particularly when risk/benefit ratios are considered in the often more complex working environments presented to ambulance professionals (Bigham and Welsford, 2015). A useful measure of efficacy in the pre-hospital environment is cessation of seizure activity prior to arrival at the emergency department, which consequently reduces the risk of cardiorespiratory complications that present challenges to paramedics.
‘Previous reviews suggest that lorazepam is a more effective and safe treatment option, but there is still some ambiguity as to whether lorazepam should replace diazepam in practice.’
The use of midazolam for convulsive status epilepticus management was not considered in this review. However, with its short duration of action and promising absorption profile via intramuscular, buccal and intranasal routes, midazolam could provide an efficacious alternative to intravenous benzodiazepine therapy in status epilepticus.
Summary
Most clinical guidelines throughout the World (including those published by NICE in the UK), recommend lorazepam as the first-line benzodiazepine treatment for status epilepticus. There is much evidence in support of lorazepam when compared with diazepam in the treatment of the condition (and indeed for non-status epilepticus convulsive seizures). Previous reviews have collated evidence on this topic, and there is a tendency to support lorazepam as a more effective and safer treatment option (Prasad et al, 2007), (Rogalski and Rogalski, 2015); however, there still exists some ambiguity as to whether lorazepam should replace diazepam in practice (Brigo et al, 2016).
There is emerging evidence to support the use of pre-hospital midazolam, which is more commonly associated with sedation and anaesthesia. Preliminary results from a trial conducted by Silbergleit et al (2012) in the United States appear promising. They have demonstrated that the use of intramuscular midazolam is safe and potentially as effective as intravenous benzodiazepines in the control of convulsive seizures.
Conclusion
Research in the realm of pre-hospital care is a relatively new venture for the paramedic profession. Randomised controlled trials offering reliable evidence of pre-hospital pharmacological interventions are limited, and as such, clinical guidelines and legislation may benefit from review in light of new developments in pharmacology.
Lorazepam and alternative benzodiazepines already exist as first-line treatments for convulsive status epilepticus in preference to diazepam, and although ambulance guidelines facilitate some variation in practice, legislation prohibits the autonomous use of these alternatives. Facilitation of randomised controlled trials that directly compare anticonvulsants in the pre-hospital environment should be encouraged in the interest of continuous guideline and legislative review towards maintaining high standards of patient care.