D Hall, J Riley*, I Swann# A&E Department Royal Alexandra Hospital, Corsebar Road, Paisley, *A&E DepartmentRoyal Infirmary, Glasgow, #Mathematics Department, Glasgow Caledonian University, Cowcadens Road, Glasgow
Correspondence to: Dr David J Hall, A&E Dept, Royal Alexandra Hospital, Corsebar Road, Paisley PA2 9PN
email: davehall12@hotmail.com
SMJ 2005 50(1): 24-26
Abstract:
Objective:To assess whether there is a breath alcohol level (BrAC) below which confusion in the head injured patient should not be attributed solely to the acute effects of alcohol. Method: Based in the Accident and Emergency Ward in Glasgow Royal Infirmary, a prospective observational study was carried out over a five month period. Patients admitted to the ward were recruited for the study if they had a primary diagnosis of head injury. The outcome measures recorded and analysed were sequential 2 hrly BrAC readings (mg/L) and Glasgow Coma Scale findings (Eye opening, motor and verbal responses). The relationship between these was investigated, which revealed additional relevant factors affecting level of consciousness. Results: The breath alcohol analyser was found to be a useful non-invasive, quick and easy to use tool. The results obtained were consistent with the expected pattern of reducing BrAC levels over a 6 hour period. Within this group of patients, a poor correlation was found between each of the three responses of the Glasgow Coma Scale and BrAC readings. For those patients who remained confused, when their BrAC reading was less than 1mg/L, other causes of a lowered level of consciousness were identified. Conclusion: Confusion in the head injured patient with a BrAC of less than 1mg/L, should alert one to the likelihood of causes other than alcohol intoxication.
Keywords: Head injury, breath alcohol analysis, intoxication, confusion
Introduction
The management of the head injured patient who may be intoxicated remains a problem for Accident and Emergency(A&E) staff, since alcohol can compromise the patients’ conscious level, thereby potentially masking the effects of the head trauma and leading to delay in treatment of intracranial complications.1,2,3 Glasgow Royal Infirmary treated 5084 (approximately 8% of attendees) patients with a head injury in 1998, of whom 1221 (24%) were admitted for observation, with 114 requiring transfer to the Institute of Neurological Sciences for further management. It is widely recognised that alcohol plays a major role in the causation of head injury especially in inner city A&E departments such as ours. 1,2,4,6
Breath Alcohol (BrAC) analysis has been advocated in the assessment of these patients, by determining an equivalent blood alcohol level (BAC), and estimating the possible contribution that alcohol has made to the patients’ altered conscious level and behaviour.5,6 In the head injured patient, however, the assumption that alcohol is the cause of a lowering of conscious level is fraught with danger. Various authors have suggested a BAC below which a lowered level of consciousness should not be attributed to the effects of alcohol.1,2 In addition the SIGN guidelines10 on the early management of head injury suggests CT for patients without head injury when GCS is low but fails to improve. Based on these previous findings, the decision was made to study closely patients who remained confused with a BrAC below 1mg/L, which is equivalent to a BAC of 230mg/100ml.
Method
A prospective study of hospitalised head injured patients was carried out to ascertain whether there was a breath alcohol level (BrAC) below which a lowered level of consciousness should not be attributed to the acute effects of alcohol alone. An attempt was made to include all head injured patients who were admitted to the A&E ward for observation during a 20 week period, between May and September 2000. Initial BrAC readings were obtained with first neuroobservation on admission, so that patients who were discharged from A&E, those transferred to the regional neuro-surgical unit, admitted to the ICU or to other wards were automatically excluded. Verbal consent was sought as far as possible. Patients were informed of the nature of the study and information sheets provided. Refusal, rather than inability to provide a reading resulted in exclusion. Confidentiality was maintained by anonymising the data.
A Lions Alcolmeter® S-300, with which the nursing staff on the ward were already familiar, was calibrated and set in the “active mode”, meaning that the patients had to actively blow into the instrument to record a reading. Serial BrAC readings, from the time of admission, were plotted against GCS findings at two-hourly intervals, until either the readings returned to 0mg/L or the patient was discharged/ self-discharged or transferred from the ward. Unsuccessful attempts at using the breath alcohol analyser were noted. A pro-forma was completed for each patient which included details of the mechanism of injury, recent and average alcohol consumption, current prescribed medication, illicit drug use, past medical history, investigations performed during admission and patient outcome.
Results
A total of 145 patients were enrolled in the study, representing 68% of the total number of patients admitted to the A&E observation ward during the study period with a primary diagnosis of head injury. Of these, 115 were male (79%) and 30 were female (21%). Their ages ranged from 16 to 87 with a mean age of 43 years. The mechanisms of injury are summarised in Table I.
Investigation and outcome
Of the 145 patients admitted, four had undergone CT scanning of the brain and one patient was transferred for neuro-surgical intervention. Headache, vomiting, dysphasia and radiological evidence of a skull fracture were the stated indications for scanning. All of these patients had a GCS assessment of 15 when the decision to scan was taken. Three of these patients had essentially normal scans whilst the other required transfer and craniotomy for the evacuation of a sub-acute extradural haematoma. BrAC testing on admission was negative in three of the four patients requiring scans, and was low (0.11mg/L) in the other.
Use of the BrAC analyser
One hundred and eighteen patients were able to successfully provide at least one BrAC reading (81.4%), whilst 27 were unable to provide any readings (18.6%). Overall there were 277 attempts at using the analyser, 239 of which were successful (86.3%) and 38 unsuccessful (13.7%). Only one unsuccessful attempt was the result of instrument failure, whilst the others were due to patient inability to comply with the procedure.
BrAC findings
Sixty seven of the 118 patients who were able to successfully provide a BrAC reading proved positive for alcohol (56.8%), whilst 51 provided only negative readings (43.2%). A further 27 patients were unable to provide a sample. In order to test the maxim that, “confusion cannot be explained by the acute effects of alcohol alone, in the head injured patient with a BrAC reading of less than 1mg/L,” those patients who remained confused, Verbal response of less than five, whilst at the same time registering BrAC readings of less than 1mg/L were studied closely. There were 10 patients in this group, eight of whom were male and two female. Their ages ranged from 16 to 79 and BrAC readings from 0 to 0.98mg/L (Table II). All 10 patients were eventually discharged from the ward and in all cases, excepting one (patient no.10), verbal response returned to five.
Discussion
The danger of attributing confusion to alcohol alone in an A&E setting is well recognised.3 This has resulted in a number of studies involving blood and breath alcohol analysis some of which, have suggested blood alcohol levels below which a lowered level of consciousness should not be attributed to the effects of alcohol. Breath alcohol analysis was chosen for this study because it is rapid, non-invasive and precludes the need for laboratory facilities. There remains international controversy as to the “true” blood:breath alcohol ratio,8 accepted in the U.K as being 2300:1. Galbraith et al’s2 “cut off ” figure of 200mgs/ 100mls BAC translates as an equivalent BrAC of 0.88mg/ L using the British ratio. In the present study, the more easily remembered BrAC of 1mg/L was tested, which is equivalent to 230mgs/100mls BAC, but which would be exactly 200mg/L using the French ratio of 2000:1.
Whilst accepting the manufacturer’s claims as to the accuracy of the analyser, heed was taken of their caveat that it may be less so when used in the “passive mode”, whereby nasal samples are obtained. By insisting on its use in the “active mode”, where the patient has to produce a substantial breath in order to elicit a reading, there is no doubt that the number of samples which could be expected to be obtained was effectively limited. The intention, however, was not to study patients who were comatose or deeply unconscious, since one would hope and expect that they would be receiving close attention in the A&E resuscitation room and many would have a CT scan. The emphasis, rather, was on the more common situation of, the confused patient who may not appear to warrant such attention in the mistaken assumption that he is “just drunk”. These limiting parameters probably explain the low rate of CT scanning amongst the study group, and their relatively benign clinical course.
Technically the breath alcohol analyser was found to be reliable, with only one unobtainable reading being the result of instrument failure. There was, however, significant “user” limitation, whereby 27 (18.6%) of patients were unable to produce a reading. This inability to use the breath analyser was in some patients due to physical disability e.g. respiratory disease, but in others due to a lowered level of effects of alcohol itself. When the three responses comprising the GCS, (eye opening, best motor and best verbal responses) were examined on admission to the ward (time 0), there was no clear relationship between them and BrAC readings. For “Eye opening response” (Fig 1), an association between an inability to use the breath analyser and a lowered response was found. When the “Motor response” at time 0 was examined it was found, understandably, that all patients who were able to use the analyser had a response of six, i.e. were able to obey commands, whilst those unable to use the analyser had a lower response. In effect, the ability to use the analyser was a complex test of motor response. When the “Verbal response” of patients at time 0 was examined (Fig 2), there was no clear relationship between it and BrAC readings, although once more there was a lowered response in those unable to use the analyser.
One might assume that there would be a direct relationship between BrAC readings and level of consciousness, and that at a given BrAC, GCS would return to 15, the BrAC level at which alcohol no longer causes a reduction in level of consciousness. However, this could not be concluded from this study. Brickley and Shepherd1 did demonstrate a correlation between GCS and BAC in a group of patients without head injury, but even then there were exceptions. Nath, Beastal and Teasdale,9 on the other hand, found no correlation between level of consciousness and degree of alcohol intoxication in a group of severely head-injured patients. Likewise, in this present study, no obvious relationship between GCS and BrAC level was found in a group of “mildly” head-injured patients. Why should this be the case? This is explained by the heterogenous nature of the study group, in which the cause of lowered conscious level was multifactorial. As well as the acute effects of alcohol there may have been ongoing concussion from head injury and other accompanying causes of lowered level of consciousness e.g. drug abuse, dementia (Table II). We are convinced that these were at least contributing reasons for ongoing confusion, in patients with BrAC readings below 1mg/L. Glasgow Royal Infirmary has, within its catchment area, one of the most deprived populations in Europe. There is a particularly high incidence of alcohol and drug abuse with many hostels for the homeless in the area, the residents of which are often undernourished, chronic alcoholics who are prone to repeated head injury. It would be prudent, therefore to always consider the many possible causes of confusion in such a population.
Conclusion
In the head injured patient undergoing observation on a short stay ward, who is confused and has a BrAC reading of less than 1mg/L it should not be assumed that the confusion is simply the result of alcohol intoxication. Diagnoses to consider should include postictal state, illicit and prescribed drugs, dementia and of course brain injury.
REFERENCES
1 Brickley M R and Shepherd JP. The relationship between alcohol intoxication, injury severity and Glasgow Coma Scale in assault patients. Injury 1995Vol 20, 311-314.
2 Galbraith S, Murray WR, Patel AR and Knill-Jones R. The relationship between alcohol and head injury and its effect upon the conscious level. Br J Surg Vol. 63 1976,128.
3 Jagger et al. Effect of Alcohol Intoxication on the Diagnosis and Apparent Severity of Brain Injury. Neurosurgery 1984 Sep;15[3]: 303-6.
4 Swann IJ, McMillan R, Strong I. Head injuries at an inner city accident and emergency department. Injury 1981 Jan; 12[4]: 274-8.
5 Cook LS, Levitt AM, Simon B, et al. Identification of ethanol-intoxicated patients with minor head trauma requiring computerised tomography scans. Acc Emerg Med 1994;1:227-34.
6 Walsh ME, Macleod AD. Breath Alcohol Analysis in the Accident and Emergency Department. Injury 1983July;15[1]:62-6.
7 Winek CL, Murphy KL: The rate and kinetic order of ethanol elimination. Forensic Sci Int 1984; 25:159-166.
8 Wright BM, Jones TP and Jones AW. Breath Alcohol Analysis and the Blood:Breath Ratio. Med. Sci. Law 1975 Vol 15 No.3 205-210.
9 Nath FP, Beastal G and Teasdale GM. Injury 1986;17, 150-153.
10 Scottish Intercollegiate Guidelines Network (SIGN), Early Management of Patients with a Head Injury, Edinburgh:SIGN;2000. SIGN Publication no 46.
Ethics: We the authors declare that “Ethics approval was granted by the Glasgow Royal Infirmary Research Ethics Committee on 29/11/99, based on patients giving consent where possible, but recognising that due to the nature of the study written consent was not deemed to be practical in all cases. The committee accepted that the use of the Breath Alcometer was already normal practice in the A&E ward and that the data would all be anonymised. Patient information sheets were made available and patients could withdraw from the study at any time.”