Benzodiazepine overtreatment in status epilepticus is related to higher need of intubation and longer hospitalization

Authors


Summary

Benzodiazepine (BDZ), a widely recognized first-line status epilepticus (SE) treatment, may lead to respiratory depression. This cohort study investigates the effect of BDZ doses in SE patients in terms of morbidity and mortality. It considers incident SE episodes from a prospective registry (2009–2012), comparing patients receiving standard BDZ dose to those receiving exceeding doses (>30% above recommended dose), in terms of likelihood to receive intubation, morbidity, and mortality. Duration of hospitalization was assessed for subjects needing intubation for airways protection (not for refractory SE treatment) versus matched subjects not admitted to the intensive care unit (ICU). We identified 29 subjects receiving “excessive” and 173 “standard” BDZ dose; 45% of the overtreated patients were intubated for airways protection, but only 8% in the standard-dose group (p < 0.001). However, both groups presented similar clinical outcomes: 50% returned to baseline, 40% acquired a new handicap, and 10% died. Orotracheal intubation due to airways protection was associated with significantly longer hospitalization (mean 2 weeks vs. 1 week, p = 0.008). In conclusion, although administration of excessive BDZ doses in SE treatment does not seem to influence outcome, it is related to higher respiratory depression risk and longer hospitalization, potentially exposing patients to additional complications and costs.

Status epilepticus (SE) necessitates a prompt and efficacious treatment to reduce mortality and morbidity (Lowenstein et al., 1999; Shneker & Fountain, 2003). However, morbidity related to SE treatment has received relatively little attention. In SE patients, respiratory depression can result both from SE itself and from its treatment, particularly following benzodiazepine (BDZ) use (which represents the evidence-based first-line SE medication), sometimes leading to orotracheal intubation.

We hypothesized that higher than recommended BDZ doses may determine a greater incidence of respiratory failure and thus influence final outcome.

Methods

Cohort definition

We studied incident adult SE patients (excluding hypoxic-ischemic encephalopathy) with all seizure types (i.e., focal with and without consciousness impairment, generalized, and nonconvulsive SE in coma), identified between January 2009 and February 2012 (38 months) in our prospective registry; details have been published elsewhere (Novy et al., 2010). SE was defined as prolonged (>5 min), or repetitive seizures (without full recovery between episodes) (Lowenstein et al., 1999) and diagnosed clinically by neurology consultants (electroencephalography [EEG] confirmation required for nonconvulsive SE forms).

We analyzed the BDZ use as first-line treatment, regardless of the administration modality (intranasal, intravenous, intramuscular or rectal). “Excessive BDZ dose” was defined as a loading medication amount >30% (weight-adjusted) above the recommended initial dose (Table 1), according to the current SE Swiss guidelines (Leppert et al., 2005). We identified SE patients admitted to the intensive care unit (ICU) after receiving orotracheal intubation; the indication (i.e., upper airways protection vs. refractory SE treatment by general anesthetics) was prospectively recorded. The study received full approval from our ethics commission.

Table 1. Recommended doses for the most commonly used BDZ as first-line SE treatment
BDZRecommended dosesa (mg/kg)Excessive dose (mg/kg)
  1. a

    Adapted from (Leppert et al., 2005).

  2. Excessive doses are defined as >30% above the recommended dose.

Lorazepam0.100>0.130
Diazepam0.250>0.325
Clonazepam0.015>0.020
Midazolam0.200>0.260

Variables

Clinical and demographic characteristics of each patient, and outcome at discharge (return to baseline, new handicap, or death) were recorded prospectively. For the estimation of SE severity, a validated clinical score (Status Epilepticus Severity Score [STESS]) was prospectively calculated (Rossetti et al., 2008); etiologies were categorized as potentially fatal as previously detailed (Rossetti et al., 2006).

Finally, we conducted a nested case–control assessment, comparing two subgroups of subjects (matched for STESS, age, gender, and potentially fatal etiology) defined by endotracheal intubation to prevent airways obstruction; hospitalization length (from admission to discharge or transfer to a rehabilitation unit) was retrieved for both groups.

Statistical analysis

The relationship between initial BDZ dose and outcome variables was assessed using t-tests and a two-tailed Fisher's exact tests, as needed. No correction for multiple comparisons was applied, as this was an exploratory study, but significance was conservatively assumed at p < 0.01.

Results

Two hundred eighty-three SE episodes were identified: 3 (<1%) remitted spontaneously, 15 (5%) received no BDZ, and 63 (22%) presented recurrent episodes. Among the 202 remaining incident cases, we identified 29 subjects who received an excessive BDZ dose and 173 who received a standard (or lower than recommended) BDZ dose. There was no difference between the two groups concerning gender, age, history of previous seizures, or etiology (Table 2). In contrast, in the excessive-BDZ-dose group we found significantly more patients who presented with a generalized convulsive SE (73% vs. 44%) and higher STESS score (>3 points in 83% vs. 56%).

Table 2. Characteristics of the cohort patients, stratified after the benzodiazepine doses
 Exposed to excessive BDZ (29)Not exposed to excessive BDZ (173)Testp-value
  1. Bold indicate statistically significant p-values (p < 0.01).

Gender (male)20 (69%)97 (56%)Fisher0.22
Age (years, mean ± SD)56.9 ± 3562.7 ± 36 t 0.11
Previous seizures9 (31%)83 (48%)Fisher0.11
Potentially fatal etiology16 (55%)79 (46%)Fisher0.42
Seizure type    
Absence06 (3%)Fisher <0.001
Focal, consciousness preserved1 (3%)30 (17%)
Focal, consciousness impaired5 (17%)51 (29%)
Generalized convulsive21 (73%)76 (44%)
Non convulsive SE2 (7%)10 (7%)
Consciousness    
Stuporous/comatose26 (90%)100 (58%)Fisher <0.0001
Refractory SE13 (45%)44 (25%)Fisher0.04
STESS ≥324 (83%)96 (56%)Fisher 0.007
Intubation need    
For refractory SE treatment3 (10%)18 (10%)Fisher <0.00001
For airways protection13 (45%)13 (8%)
Outcome at discharge    
Return to baseline15 (52%)89 (51%)Fisher1.0
New handicap11 (38%)64 (37%)
Death3 (10%)20 (12%)

The group of patients exposed to an excessive BDZ dose needed significantly more orotracheal intubation (55% vs. 18% in the standard dose group), particularly for airways protection (45% vs. 8%), whereas intubation for refractory SE treatment through pharmacologic coma induction was equally performed (10%). The clinical outcome did not differ between the two groups: approximately one-half of patients returned to baseline, two of five acquired a new handicap, and one of 10 died.

Concerning hospitalization length (Table 3), the 26 patients admitted to the ICU after needing orotracheal intubation for airways protection (receiving a standard or an excessive BDZ dose) spent a significantly longer time in hospital (mean value of 2 weeks), in contrast to the 26 matched subjects who were not intubated (1 week; p = 0.008).

Table 3. Characteristics of 26 patients who received orotracheal intubation for airways protection after BDZ administration (intubated) and 26 subjects matched for STESS, age, gender, and etiology deadly and who received no intubation (not intubated)
 Intubated (26)Not intubated (26)Testp-value
  1. Bold indicates statistically significant p values (p < 0.01).

Gender (male)19 (73%)14 (54%)Fisher0.25
Age (years, mean ± SD)57.9 ± 3059.1 ± 33 t 0.78
Previous seizures7 (27%)10 (38%)Fisher0.55
Potentially fatal etiology11 (42%)11(42%)Fisher1.0
Seizure type    
Focal, consciousness preserved00Fisher1.0
Focal, consciousness impaired3 (11%)3 (11%)
Generalized convulsive22 (85%)23 (89%)
Nonconvulsive SE1 (4%)0
Consciousness    
Stuporous/comatous24 (92%)25 (96%)Fisher1.0
Refractory SE9 (35%)5 (19%)Fisher0.35
STESS ≥319 (73%)20 (77%)Fisher1.0
Hospitalization length (days, mean ± SD)15.5 ± 257.73 ± 14t 0.008

Discussion

The principal finding of this study is that administration of an excessive BDZ dose for initial SE treatment was not associated with worse outcome at hospital discharge, despite higher prevalence of orotracheal intubation performed for airways protection, and a longer duration of hospitalization.

BDZ represents the only clearly evidence-based SE treatment (Treiman et al., 1998; Alldredge et al., 2001; Silbergleit et al., 2012). The safety profile of BDZs in SE management has been previously addressed in few studies, reporting an estimated BDZ-related respiratory depression incidence of 3–15% in pediatric patients (McIntyre et al., 2005) and a similar incidence of 10–14% in adults (Alldredge et al., 2001; Silbergleit et al., 2012). This seems in line with our findings (17%), where excessive BDZ doses were linked to a higher risk for orotracheal intubation, but somewhat surprisingly without any notable effect on outcome (including mortality). This suggests that prognosis probably depends mostly on intrinsic factors linked to the SE episode and the patient, in particular etiology, age, and mental status impairment (i.e., the STESS score), as suggested by previous observations (Rossetti et al., 2006), but not the global SE-specific medication strategy (Rossetti et al., 2013).

Another potential consequence of BDZ overuse is related to hospitalization length: intubation was more frequently performed after BDZ overtreatment, with the excess proportion being almost exclusively related to the need of airway protection, and correlating with a significantly longer hospital stay. A recent study concerning early convulsive SE management (Silbergleit et al., 2012) reported a hospitalization length of 5–7 days, but no comparison was performed between patients needing ICU care and the others. This is of practical concern, since prolonged hospitalization may expose patients to complications, both general (Sutter et al., 2012) and ICU specific, and may represent a source of health care extra costs.

Factors inducing physicians and paramedics to overcome recommended doses may be related to the perceived SE severity; indeed, SE forms with more profound consciousness impairment were more prone to receive BDZ overdoses, as is illustrated by the higher prevalence (p = 0.04) of refractory SE in the excessive-dose group. In fact, overtreated patients not only were more likely to be intubated, but also to present generalized convulsions, which represent per se a risk factor for developing respiratory depression. A previous study (Alldredge et al., 2001) reported a lower incidence of respiratory side effects following BDZ administration versus placebo, but in contrast to our study low BDZ doses were used in that protocol, and patients had exclusively generalized convulsive SE. In the light of these observations, it seems that a “U-shaped” correlation between respiratory depression and BDZ doses may exist.

Our study has some limitations. First, the definition of BDZ overdose is somewhat arbitrary; however, it is based on previous work on treatment adherence to guidelines (Rossetti et al., 2013). It is possible that higher cutoff doses would have reduced the number of exposed patients, but probably also strengthened the association with the prolonged hospitalization length. Second, it is possible that BDZ overdose “prevented” worse outcome in severe SE forms, balancing off diverging effects and resulting in similar final outcomes. Nevertheless, this hypothesis seems rather unlikely, since our previous study suggested that medication dose has a marginal role in SE prognosis, even after adjustment for major outcome predictors (Rossetti et al., 2013). Third, although we prospectively recorded indications for intubation, this may reflect SE severity and refractoriness, rather than being related to BDZ administration. However, no additional intubations were performed for the specific refractory SE treatment in the overtreated group (which, in contrast, received significantly more intubations for airways protection), thus suggesting a BDZ-overdose specific role. Fourth, BDZ subgroups and administrations ways (e.g., intravenous, intrabuccal, intranasal, oral) were not specifically assessed in view of the low numbers and of the almost exclusive use of intravenous clonazepam. Finally, although most variables were assessed prospectively, data regarding the ICU stay were retrospectively retrieved. On the other hand, we believe that the present study offers an important strength: to our knowledge, this is the first analysis assessing nonselected SE adult patients, that is, without age restriction and with both convulsive and nonconvulsive forms, and is therefore more representative of the general adult SE population.

In conclusion, even if the clinical prognosis does not appear to be affected by BDZ overuse, clinicians should be aware that this could expose patients to potentially avoidable complications. Further studies are needed to expand these findings.

Disclosure

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines

Ancillary