https://orcid.org/0000-0002-5808-8776
Abstract
Performing head CT in intoxicated patients presenting to the emergency department (ED) is common, yet low yield. In our previous study, acute findings were found only in patients with at least one of: neurological Deficits, Intubation, Trauma, Convulsions or Headache. We hypothesise that the absence of these DITCH clinical features rules out the need for immediate CT. We collected data for intoxicated patients attending EDs in our district from March 2021 to May 2022 inclusive. We recorded clinical presentation features, agents, disposition and head CT results. There were 1,308 intoxicated patients out of a total of 219,273 ED presentations. Median age was 38 years. Common ingestions were ethanol, stimulants, GABA-ergics and opioids. 407 patients (31%) had head CT, with 31 patients (7.6%) having 36 acute findings: 19 intracranial haemorrhages, 8 fractures, 2 cerebral oedema and 7 haematomas. All patients with acute CT findings had one of the DITCH features, and all had trauma. The sensitivity and NPV of the hypothesis were 100%. In patients that were not scanned, there were no re-presentations.Our internal validation study is further evidence that acute findings on head CT in intoxicated patients presenting to the ED are only found in those who had one of the DITCH features.
Introduction
Poisoned patients constitute 1% of emergency department (ED) attendances in Australia [Citation1]. Non-contrast head computed tomography (CT) in intoxicated patients is a common, resource intensive and low yield investigation [Citation2, Citation3]. Acute pathological findings on head CT are detected in 4 − 4.8% of cases in recent large series [Citation4, Citation5]. Other studies of alcohol-intoxicated patients with head trauma, 8% of patients had a clinically important CT finding [Citation6, Citation7]. Drug and alcohol intoxicated patients prove to be a difficult cohort for early decision-making in the ED. These patients may be over-triaged despite their injury severity and receive unnecessary radiation from CT [Citation8]. Alcohol-intoxicated patients are no more likely to have an abnormal head CT than non-alcohol related trauma patients presenting to the ED [Citation9, Citation10].
Our retrospective study found that head CT revealed acute pathology only if the poisoned patient had at least one of the following five clinical features which we now summarise with the mnemonic DITCH: Deficits or focal neurological signs; need for Intubation; history of Trauma or head injury; presence of Convulsion or seizure activity; and Headache [Citation4]. While these features were consistent in our retrospective cohort, a prospective study was required to validate the criteria with the potential for a decision aid rule. We hypothesize that the absence of any DITCH criteria can be used as a clinical decision instrument (CDI) to safely rule out the need for immediate head CT in intoxicated patients presenting to the ED.
Methods
Study setting
Our toxicology unit in western Sydney covers a region comprising a population of about 1.5 million with four EDs having access to tertiary toxicology and neurosurgical services. The study received ethics approval for human research from the district research governance committee.
Study design and patient selection
A mixed prospective and retrospective observational study was conducted in the Western Sydney Toxicology Service and our health district EDs. All intoxicated patients with a Glasgow Coma Score (GCS) ≤ 14 that presented to our four EDs between March 2021 and May 2022 inclusive were included in the study. To maximise data capture and minimise any missed cases of intoxication, a multi-modal approach was implemented. Firstly, all intoxicated patients with GCS ≤ 14 referred to the Western Sydney Toxicology Service had data collected prospectively. Then, we retrospectively screened two other cohorts for intoxicated patients: 1. all neurosurgical admissions, and 2. all ED presentations that were discharged without admission to hospital, that had GCS ≤ 14, or the patient record included the words “intoxicated”, “overdose”, “delirium”, “sedated” or “drowsy” in the context of drug overdose. Head CT was ordered at the discretion of the treating clinician.
Data collection & analysis
Data collectors were not involved in clinical decision making with respect to scanning and were blinded to the “rule out” hypothesis. The radiologists reporting the head CT were not blinded to the nature of the clinical presentation but were unaware of the study hypothesis being tested. Pre-specified information was collected on a spreadsheet and stored on our secure database. Data from the electronic medical record (Cerner Millennium, Kansas City, MO) were extracted for demographics, clinical presentation, intoxicating agents, disposition and Poison Severity Score (PSS) [Citation11]. In line with our previous study, neurological deficits were defined as focal signs that are unexpected in an overdose and likely to indicate intra-cranial pathology [Citation4]. For retrospective analysis, missing information about a parameter was regarded as negative for the purposes of calculating statistics. We calculated descriptive statistics including sensitivity, specificity, positive predictive value and negative predictive value – true positive (acute finding on head CT and ≥1 DITCH feature); false positive (no acute finding on CT, but ≥1 DITCH feature); true negative (no acute finding on CT and no DITCH feature); and false negative (acute finding on CT, but no DITCH feature). Patients were followed up by searching their electronic medical record for re-presentations and further imaging. To minimise losses to follow up, we also searched all neurosurgical admissions.
Outcomes
We classified CT findings into four prespecified categories. These included 1. normal CT, 2. chronic changes not requiring follow up (such as age-related pathology not related to the poisoning), 3. Incidental findings requiring follow up, but not related to the poisoning (such as meningioma or aneurysm) and 4. acute pathological findings related to the current presentation (such as intracranial bleeding or cerebral oedema). A data dictionary was supplied to data collectors to promote consistency.
Results
During the 15-month study period, 219,273 patients attended EDs in our health district. summarises the results. After applying inclusion and exclusion criteria, there were 1,308 intoxicated patients with GCS ≤ 14 (58% male). 497 cases were referred to our toxicology service; 796 were discharged directly from the ED; and 15 were neurosurgical admissions. Median age was 38 years (IQR 24,48), with a range of 15 to 100 years. The median PSS was 1 (IQR 1,2), n = 1168. Agents involved () were: alcohol in 501 cases, stimulants 173, sedatives (benzodiazepines, GHB) 155, opioids 85, anti-psychotics 74, anti-depressants 67 and other agents 350. Head CT was performed on 407 of 1308 intoxicated patients (31%). Males comprised 64% of those scanned. Among patients who were not scanned, there were no re-presentations to our district due to deterioration. Two of the patients under the neurosurgical service died and a further two underwent operative procedures for their haemorrhages. All the other patients were treated non-operatively. CT reported findings are presented in .
Figure 1. Intoxicated patients screened for head CT.
Figure 2. Agents involved in intoxicated cases.
Other includes patients with ingestants that were either non-psychotropic, or few in number (<1%):
• Non-opioid analgesics (acetaminophen, aspirin and NSAIDs)
• Anti-diabetic medications
• Chemicals and gases
Recreational substances (Caffeine, LSD)
Table 1. Non contrast head CT findings in intoxicated patients.
All 31 patients (74% male) with acute findings on head CT had at least one of the five DITCH clinical features (). Specifically, all cases with acute head CT findings had trauma either as a history of fall or physical signs of head injury. In addition, 8 had headache, 6 were intubated, 3 had a convulsion and 1 focal neurological deficit (arm weakness). 13 patients had 2 features and 3 patients had 3 features. shows the predictive relationship and sensitivities between the DITCH clinical features and acute head CT findings.
Table 2. Frequencies of acute head CT findings and the accuracy of five DITCH clinical features.
shows the proportion of patients with DITCH clinical findings for both those who underwent head CT and those who did not. Each feature is more common in the group who underwent head CT.
Table 3. DITCH clinical features in poisoned patients according to whether they did or did not undergo head CT.
At the end of the study period, we followed up the 901 patients who did not undergo head CT by searching the electronic medical record. We found no evidence of re-presentations for a missed acute intracranial finding. 841 patients (93%) were alive with no record of re-presentation for neurological deterioration. However, 66 patients (7% of 901) did not have further health service contact and we were unable to confirm their clinical status. One patient died a week after initial assessment due to pneumonia-related respiratory failure.
Discussion
We propose a clinical decision rule (CDI) for the selective use of head CT in intoxicated patients – that head CT is not indicated in intoxicated patients if the none of the five DITCH clinical features are present. If we combine the results from our previous retrospective cohort and the current prospective study, we see that there are a total of 70 acute findings out of 1549 head CT scans [Citation4]. Therefore, we have yet to find an exception to the rule that acute findings on head CT only occur in the presence of one of the DITCH clinical features. While this supports the hypothesis that we can “ditch” the head CT in the absence of any DITCH features, further prospective validation of this CDI is required in larger, external, multi-centre cohorts with diverse patient populations prior to safe, widespread adoption. In this study we tested a “rule out” CDI which had a high sensitivity. However, the low specificity of the criteria invites consideration of whether multi-variate analysis could improve the specificity of the criteria.
In this study, we found all acute findings on head CT were associated with trauma, either via a history of fall or physical signs of head injury. This raises the issue of whether we can simplify the CDI down to trauma only and eliminate the other four key clinical features from the tool. This may not be prudent given that, in our retrospective series, 22/39 patients (56%) with acute findings on head CT had no signs of trauma [Citation4]. We suggest keeping all five clinical criteria based on the finding that we would miss some acute pathology if we just used trauma as a sole criterion in the decision tool. For some cases, gaps in the anamnesis may make it challenging to exclude a history of trauma. It is suggested that the CDI only be applied to patients in whom it is unlikely that trauma has occurred.
Although we propose a CDI in which the absence of the five clinical features “rules out” the need for head CT, we are not conversely recommending that all patients with one of the DITCH features undergo head CT, due to the low specificity and positive predictive value (PPV) of the CDI. Of those patients who had a head CT, only 31 of the 240 patients with DITCH features had an acute finding, giving a low PPV of 13% (31/240). If a patient does have a DITCH feature, we recommend applying clinical judgement in deciding if a head CT is warranted.
In addition to lowering patients’ radiation risk and decreasing costs, fewer head CTs would increase efficiency in an often-overcrowded ED. As a prediction of how this CDI may lower the number of head CT performed, we note that in our study 167 head CT were performed on patients who did not have any DITCH features. If our rule-out criteria had been applied to our cohort, 167/407 (41%) head CTs may have been avoided.
Limitations
As an internal validation, this study involved the four EDs covered by a single toxicology service. However, we are confident that all cases of neurosurgical abnormality have been captured within our cohort. Of those that did not have head CT, we were unable to trace 66 patients (7%) through a search of their electronic medical record. While these patients tended to be younger, and not expected to deteriorate, we were unable to confirm their clinical status. Missed pathology would decrease the sensitivity and NPV of the rule if the patients had no DITCH clinical features. It is unlikely any of these patients had a poor outcome because local processes concerning any unexpected death in the community of a recently discharged patient would be expected to trigger a notification to the hospital. However, measuring re-presentation rates of those not scanned may not exclude all haemorrhages. A limitation of retrospective chart reviews is missing data, which in this study were assumed to be negative – an example would be that if there was no mention of headache, it was counted as no headache. An underestimate of the number of DITCH clinical findings would decrease the true negative rate and increase the false positive rate. The effect of this would be to overestimate the number of scans that could be saved with this rule out CDI.
Conclusions
Building on our previous research, this CDI validation study is further evidence that acute findings on head CT in intoxicated patients presenting to the ED are only found in those who had at least one of the five DITCH clinical features: Deficits, Intubation, Trauma, Convulsions or Headache. Our “rule out” CDI supports the hypothesis that in the absence of any DITCH features, a head CT will not be of immediate benefit. Further external validation would assist in changing clinical practice guidelines.
Previous presentation
Presented at the EAPCCT conference in May 2023.
Abstract published previously in Clinical Toxicology, 61:sup1, 1-129, DOI:10.1080/15563650.2023.2192024
Acknowledgments
Drs Lakmali Abeywickrama, Pramod Chandru, Amanda De Silva, Nina Dhaliwal, Mariez Gorgi, Hyungwoo Kim, Paloma Lopez, Pakeer Markandu, Farida Parvin, Samuel Phillips, Istabraq Raashed, James Sutton, Sakunthala Wijeratne and Samoda Wilegoda are thanked for help with data collection
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
Data available upon reasonable request
Additional information
Funding
References
- Emergency Department Care 2016-17: Australian hospital statistics. Canberra: Australian Institute of Health and Welfare; 2017.
- Narayanan V, Keniston A, Albert RK. Utility of emergency cranial computed tomography in patients without trauma. Acad Emerg Med. 2012;19(9):1–7. doi: 10.1111/j.1553-2712.2012.01437.x.
- Godbout BJ, Lee J, Newman DH, et al. Yield of head CT in the alcohol-intoxicated patient in the emergency department. Emerg Radiol. 2011;18(5):381–384. doi: 10.1007/s10140-011-0966-3.
- McNulty R, Bandaranayake L, Wong T, et al. Utility of non-contrast head computed tomography in poisoned patients. Emerg Med Australas. 2021;33(5):888–892. doi: 10.1111/1742-6723.13751.
- Tay EM, Preisz P, Day RO. Role and impact of brain computed tomography in the management of drug overdoses and guideline recommendations. Emerg Med Australas. 2019;31(6):1053–1058. doi: 10.1111/1742-6723.13316.
- Easter JS, Haukoos JS, Claud J, et al. Traumatic intracranial injury in intoxicated patients with minor head trauma. Acad Emerg Med. 2013;20(8):753–760. doi: 10.1111/acem.12184.
- Cook LS, Levitt MA, Simon B, et al. Identification of ethanol-intoxicated patients with minor head trauma requiring computed tomography scans. Acad Emerg Med. 1994;1(3):227–234. doi: 10.1111/j.1553-2712.1994.tb02436.x.
- Weber CD, Schmitz JK, Garving C, et al. The alcohol-intoxicated trauma patient: impact on imaging and radiation exposure. Eur J Trauma Emerg Surg. 2019;45(5):871–876. doi: 10.1007/s00068-018-0945-4.
- Grzelczak AC, Ceccon A, Guetter CR, et al. Evaluation of traumatic brain injury patients with signs of alcohol intoxication. Rev Col Bras Cir. 2019;46(5):e20192272. doi: 10.1590/0100-6991e-20192272.
- Bracken ME, Medzon R, Rathlev NK, et al. Effect of intoxication among blunt trauma patients selected for head computed tomography scanning. Ann Emerg Med. 2007;49(1):45–51. doi: 10.1016/j.annemergmed.2006.04.013.
- Persson HE, Sjöberg GK, Haines JA, et al. Poisoning severity score. Grading of acute poisoning. J Toxicol Clin Toxicol. 1998;36(3):205–213. doi: 10.3109/15563659809028940.