ORIGINAL CONTRIBUTION

Effect of Intubation for Gamma-hydroxybutyric Acid Overdose on Emergency Department Length of Stay and Hospital Admission Paul Dietze, PhD, Msc, Danielle Horyniak, Paul Agius, Venita Munir, FACEM,* de Villiers Smit, FACEM, Jennifer Johnston, PhD, Craig L. Fry, DrPH, and Louisa Degenhardt, PhD

Abstract Objectives: The objective was to examine the effect of endotracheal intubation on emergency department (ED) length of stay (LOS) and admission rates for patients with gamma-hydroxybutyrate (GHB) overdose. Methods: A 3-year retrospective electronic and paper audit of recreational drug presentations was carried out at two major inner-city EDs in Melbourne, Australia. Different GHB overdose management strategies exist at the respective audit sites, namely: 1) all patients with a Glasgow Coma Scale (GCS) score of 8 or less are intubated or 2) uncomplicated patients with GCS scores of 8 or less are managed without intubation (conservative management), unless further complications arise. This difference allows for comparison of the effects of intubation. All suspected GHB-related cases (defined as cases where GHB or its analogs gamma-butyrolactone or 1,4-butanediol were recorded) in which altered consciousness state was noted as a presenting symptom at triage were selected from all recreational drug-related presentations occurring between January 2008 and December 2010. The relationship between intubation and the primary outcome, ED LOS, was examined using robust regression after adjustment for potential confounders. The relationship between intubation and admission status (admission to hospital versus discharge) was also examined using logistic regression. Results: After adjustment for potential confounders such as GCS score, intubation of GHB-related cases in the ED was associated with an increase in mean LOS of 41% (95% confidence interval [CI] = 19% to 65%) and an increase in the odds of admission to hospital of 9.95 (95% CI = 2.36 to 41.88) at one hospital site, compared to conservative airway management. Conclusions: Conservative airway management (no intubation) is associated with shorter ED LOS in cases of uncomplicated GHB-related coma in the ED and may also be associated with lower admission rates for these patients. ACADEMIC EMERGENCY MEDICINE 2014;21:1226–1231 © 2014 by the Society for Academic Emergency Medicine

From the Centre for Population Health, Burnet Institute (PD, DH, PA), Melbourne; the School of Public Health and Preventive Medicine, Monash University (PD, DH), Melbourne; the Emergency and Trauma Centre, Alfred Health (de VS), Melbourne; the Discipline of Addiction Medicine, Sydney Medical School, University of Sydney (JJ), Sydney; the Centre for Cultural Diversity and Wellbeing, College of Arts, Victoria University (CLF), Melbourne; the Centre for Health & Society, School of Population Health, University of Melbourne (CLF), Melbourne; the National Drug and Alcohol Research Centre, University of New South Wales (LD), Sydney; and the Centre for Health Policy, Programs and Economics, School of Population Health, University of Melbourne (LD), Melbourne, Australia. *Currently unaffiliated. Received February 5, 2014; revision received July 27, 2014; accepted August 24, 2014. Presented at the Australasian Professional Society on Alcohol and Drugs, Melbourne, Australia, November 2012; and Harm Reduction International Conference, Vilnius, Lithuania, June 2013. The project was funded by the Australia National Health and Medical Research Council (NHMRC Grant 452803). DH receives support from the Australian Government through an Australian Postgraduate Award and the NHMRC Centre for Research Excellence into Injecting Drug Use. PD is supported by an ARC Future Fellowship. LD is supported by an NHMRC Principal Research Fellowship. CF is supported by an NHMRC Career Development Fellowship. The authors acknowledge the contribution to this work of the Victorian Operational Infrastructure Support Program. The funding bodies played no role in the study design, data analysis, or preparation of the manuscript for publication. The authors have no potential conflicts to disclose. Supervising Editor: Stephen Smith, MD. Address for correspondence and reprints: Paul Dietze, PhD; e-mail: [email protected]

1226 1226

ISSN 1069-6563 1226 PII ISSN 1069-6563583

© 2014 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12516

ACADEMIC EMERGENCY MEDICINE • November 2014, Vol. 21, No. 11 • www.aemj.org

G

amma-hydroxybutyrate and its analogs (gamma butyrolactone [GBL] and 1,4-butanediol [1, 4-BD], hereafter all collectively termed “GHB”) are used as recreational drugs in many countries including the United Kingdom,1 Spain,2 the United States,3 and Australia.4,5 GHB is a central nervous system depressant with a variety of therapeutic uses.6 Its use has also been documented in drink-spiking events and weight control efforts.3,7 Recreational GHB use and related harms were first documented from the mid1990s as part of the suite of drugs used by nightclub patrons.5,8,9 Surveillance research has revealed significant harms, including overdose and death, associated with GHB use.2,10,11 Surveillance data in Australia suggest that GHB-related adverse events are increasing,12 even though the number of people who report using the drug is relatively small.4 There is a small threshold between a GHB dose that delivers the desired effect for recreational use and overdose.10,13 GHB overdose is characterized by respiratory depression and altered consciousness state, typically with low Glasgow Coma Scale (GCS) scores.11 Overdose may also result in agitation, bradycardia, hypothermia, hypotension, and coma, the onset and clearing of which are often abrupt. If untreated, GHB toxicity may be lethal, with or without cointoxicants.10,11,14 Ambulance data in Melbourne, Australia, comparing GHB- and heroin-related cases, found that most GHBrelated cases presented in an unconscious state and were more frequently transported to hospitals than heroin overdoses.15 The emergency department (ED) is a key setting for managing GHB overdose and other GHB-related harms.2,16 Aside from early recommendations on management of GHB overdose in the ED,3 there have been few published studies on management strategies. Reversal agents have been documented to be ineffective,2 and so most recommendations involve conservative airway management, with rapid-sequence endotracheal intubation (hereafter “intubation”) recommended as required for complicated cases.3,16 However, anecdotal evidence suggests that clinical practices vary across hospitals. In Melbourne, Australia, for example, there are distinct differences between the management strategies for GHB overdose in two major teaching hospitals, with intubation much more likely in one compared to another. These differences are unsurprising given the absence of clear guidelines for the management of GHB overdose. In this study we examined whether intubation (an invasive, time-consuming, labor-intensive procedure that, even when implemented by the experienced operator, may be associated with increased morbidity to the patient17) is associated with decreased ED length of stay (LOS), as reducing time spent in EDs is a key policy target.18 Whether patients were further admitted to hospital was considered as a secondary outcome. If intubation is clinically beneficial for GHB-intoxicated patients with similar clinical profiles, then we would expect that it would be associated with reduced ED LOS and fewer admissions to hospital from the ED, after

1227

controlling for key variables such as hospital site, patient presentation characteristics, and ingestion of other substances. METHODS Study Design This was a retrospective audit of recreational drug presentations occurring between January 2008 and December 2010. The study was approved by ethics committees at both hospitals, as well as Monash University. Study Setting and Population The study was conducted at the EDs of two major inner-city teaching hospitals in Melbourne, Australia. The two hospitals, less than 5 kilometers apart, are located to the north and south of Melbourne’s central business district, respectively, with each seeing between 45,000 and 55,000 presentations per annum. The two hospitals currently have different strategies regarding the management of drug-related presentations. At one site the majority of patients with GCS scores of 8 or less are intubated (consistent with practices in relation to most of their ED presentations), while at the other site, uncomplicated drug overdose patients with GCS scores of 8 or less are managed without intubation unless other indications arise. Study Protocol Case Identification and Data Extraction. An electronic search of medical records (triage notes and discharge diagnosis) using a list of key terms (Data Supplement S1, available as supporting information in the online version of this paper) was used to identify all cases in which recreational drugs (including amyl nitrite, cocaine, GHB, ketamine, LSD, MDA, MDMA, meth/ amphetamine, nitrous oxide, or methylphenidate) were mentioned (disclosure by the patient or by another person or bystander connected to the patient—toxicologic screening data were not available). Full medical records of these patients were then obtained electronically where available or in paper form where not. Information was extracted on a range of pre- and in-hospital variables, including clinical details, patient demographic characteristics, and relevant medical history. Information from records was then entered into a database using Questionnaire Development System Version 2.5 (NOVA Research Company). A subset of cases in which GHB, GBL, or 1,4-BD were confirmed by history (usually established prior to presentation through information available from paramedics or bystanders), and in which altered consciousness state was noted as a presenting symptom at triage), was selected from this database and used in the current analysis. Further details of the overall study have been reported elsewhere.19 Measures The primary explanatory variable was intubation in the ED. The primary outcome measure under examination was ED LOS (measured in hours), with admission status (admitted to hospital or discharged directly home from the ED) included as a secondary outcome.

1228

Sociodemographic and clinical exposures were selected as potential confounders of the relationship between the primary explanatory variable and the outcomes. These included a range of case characteristics as well as those clinical variables that influence the decision to intubate or not in the ED. The resultant list of variables included age ( 1) after OLS regression are excluded; and second, remaining cases are included in weighted regression analyses, where cases with larger model residuals are given proportionately less weight in estimating model coefficients and standard errors. It should be noted that in this analysis, no cases were excluded at the first step.

Dietze et al. • MANAGING GHB IN THE ED

All analyses were undertaken using a complete case approach and based on those providing valid ED LOS data and valid data on key covariates used to test for confounding (n = 335). Admission status was modeled using logistic regression. A total of 325 complete cases were available for the admission analyses. Here, bivariate analyses were undertaken between admission status and each key exposure, with associated (p < 0.1) exposures modeled as covariates in the final multivariable model. Again, an interaction term between hospital site and intubation status was modeled to determine whether the effects of intubation on admission were consistent across hospital site. Analyses were conducted using Stata Version 13.1, and all estimates in final models considered statistically significant at p < 0.05 (Wald tests). RESULTS A total of 1,347 cases of recreational drug use were recorded over the 3-year period. Of these, 480 involved mention of GHB, with 89% (429) having altered consciousness state,11 with 335 cases having complete data on all covariates including ED LOS and 325 cases with complete data including admission status. The characteristics of the complete cases across all of the variables considered in the ED LOS analysis are detailed in Table 1, presented by intubation status. Of 335 total, 66% were male, and 66% were aged under 25 years. Most presentations occurred in the early morning hours of weekend days. Thirty-three percent of patients presented with cointoxicant drugs, and 33% presented with coingestion of alcohol (53% with either, 47% with neither). Fifty-nine percent of patients presented to the ED with GCS scores of 8 or less. ED LOS was 3.6 hours on average. Few patients arrived by means other than ambulance or showed systolic hypotension, bradypnea, or bradycardia when first measured in the ED. A minority (7%, n = 24) of cases resulted in admission, and in 16% of presentations the patient was intubated, with 84% extubated in the ED. This intubation percentage varied according to hospital (6 and 33% for sites 1 and 2, respectively, odds ratio [OR] = 8.05, 95% confidence interval [CI] = 4.03 to 16.09). Effects of Intubation Table 2 shows that intubation was associated with an average increase in ED LOS of 41% (Exp b = 1.41, 95% CI = 1.19 to 1.65) after adjustment for hospital site, GCS score, bradycardia, sex, and age. This translates into a difference of 1.16 in geometric mean hours between intubated cases (geometric mean = 4.01) and nonintubated cases (geometric mean = 2.85). There was no significant interaction between hospital site and intubation status on ED LOS. Table 3 shows that there was a difference in the effect of intubation on admission across hospital sites, as evidenced by the significant interaction term in this multivariable model (adjusted OR = 0.10, 95% CI = 0.02 to 0.65). Intubation increased the odds of admission of patients at Site 1 (AOR = 9.95, 95% CI = 2.36 to 41.88); however, there was no significant association at Site 2 (AOR = 0.99, 95% CI = 0.25 to 3.88), after adjusting for GCS score and age.

ACADEMIC EMERGENCY MEDICINE • November 2014, Vol. 21, No. 11 • www.aemj.org

1229

Table 1 Characteristics of 335 Gamma-hydroxybutyrate–related Presentations With Altered Consciousness State, Broken Down by Intubation Status in the ED, January 2008–December 2010* Not Intubated in ED, n = 282 (84%)

Variables Exposures Age, yr

Effect of intubation for gamma-hydroxybutyric acid overdose on emergency department length of stay and hospital admission.

The objective was to examine the effect of endotracheal intubation on emergency department (ED) length of stay (LOS) and admission rates for patients ...
378KB Sizes 2 Downloads 6 Views