a b s t r a c t

Objectives: Nausea and vomiting (N&V) are among the most common complaints in the emergency department (ED). However, low acuity is assigned to most of these patients at the triage, and waiting for long hours without medication decreases patient safety and satisfaction. We aimed to compare the inhalation of isopropyl alcohol (IPA) with placebo (P) to treat nausea at the triage area of an ED. Methods: In this prospective, randomized and placebo-controlled trial, we used a convenience sample of consec- utive adult (ages 18-65) patients presented to the triage area of the ED with the complaint of N&V, and we ran- domized them to inhale IPA or P embedded gauzes. We used an 11-point (0-10) Numeric rating scale (NRS) to evaluate the degree of N&V before the inhalation, at the baseline, and at 2, 4 and 10 min after the inhalation. Results: We randomized 118 patients (IPA, n = 62; P, n = 56, intent-to-treat), three patients left the ED without being seen, and 115 patients completed the study. IPA and P groups were similar according to age, sex, comor- bidities, and vital signs. We found that patients in the IPA group had significantly lower mean NRS starting with the 2nd minute (robust two-way mixed ANOVA between-subjects, p = 0.008). We also observed a signif- icant within-subjects effect in the IPA group. The mean NRS value was decreased at each consecutive time point in the IPA group (all pairwise comparisons, p < 0.001).

Conclusion: In this study, IPA was significantly more effective than P for N&V at the triage. Moreover, patients in the IPA group had less need for rescue treatment.

(C) 2021

1. Introduction

The use of emergency departments (ED) has increased in most coun- tries in recent years [1]. Nausea and vomiting (N&V) are among the most common and distressing symptoms leading to ED visits [2-4]. acute gastroenteritis is the most common cause of those symptoms; however, more deadly diseases such as stroke, myocardial infarction, or endocrine and metabolic problems may present with N&V [4]. Nau- sea also decreases the quality of life, and vomiting may complicate into a life-threatening situation [5]. Most of the patients who present to ED with N&V have stable vital signs and are triaged as ESI level 4-5 (Emergency Severity Index); therefore, they have to wait extended hours for examination. The primary treatment of N&V is symptomatic; this does not reduce the need to make a diagnosis. However, in symp- tomatic undifferentiated cases, treatment with antiemetics such as do- pamine antagonists, serotonin 5-HT₃ antagonists, and metoclopramide

* Corresponding author.

E-mail address: [email protected] (H. Akoglu).

are warranted. Unfortunately, they are usually withheld until a proper examination is performed [5]. Symptomatic treatment of N&V at the waiting room, while waiting to be seen, with a safe and fast-acting agent may decrease patient pain and discomfort, increase patient satis- faction, reduce the time in the ED, and complication rates secondary to vomiting.

Isopropyl alcohol (IPA) is a colorless liquid frequently used in the chemical industry. Nasal inhalation of IPA has been used for more than 100 years in South America for the treatment of N&V [6]. IPA was studied in postoperative N&V and was found as an effective alternative compared to placebo (P), ondansetron, granisetron, or promethazine [6-14]. A Cochrane review in 2012 included nine studies comprised of six randomized (RCT), and three Controlled clinical trials (CCT) found IPA more useful than P for reducing postoperative N&V, but less effec- tive than standard anti-emetic drugs [15]. Beadle et al. conducted the first RCT in 2015 at an urban ED in the USA and found that IPA increased N&V relief compared to P inhalation in 10 min, but they did not evaluate the effect of time [16].

In this study, we aimed to determine the efficacy and harm of the nasal inhalation of IPA compared to P for the symptomatic treatment

https://doi.org/10.1016/j.ajem.2020.12.052

0735-6757/(C) 2021

Fig. 1. Patient flowchart.

of N&V at the waiting room in ESI triage category 4 or 5 patients and to

find the time of onset of the effect.

1. Materials and methods
   1. Study design, location, and date

We conducted this prospective, randomized, double-blinded, and controlled study in the ED of an academic center with an annual ED cen- sus of 350,000 patients between January 2017 and 2018. This study was approved by the institutional ethics review board, and written informed consent was obtained from each participant.

• 1. Study population, sample, inclusion and exclusion criteria

Triage nurses evaluated all consecutive adult patients presenting to the ED with N&V for eligibility at the triage and included to the study if 1) age was between 18 and 65, 2) ESI triage category was 4 or 5, 3) al- lergy or prior reaction to any inhalation agent was not present, 4) not pregnant, 5) they did not have any antiemetics, psychoactive drugs or medications known to reduce N&V in the last 6 h, and 6) if they scored three or above in the 11-point numerical rating scale (NRS) for N&V. We included patients to the study if any of the researchers were on duty to oversee the procedure and collect the data. Patients who could not complete the inhalation procedure were excluded.

• 1. Randomization and blinding

Patients were randomized to one of the study groups by the study investigators (HC and ES) at the triage before they were referred to an examination room, after obtaining written informed consent. Patients were informed that they would inhale one of the two different sub- stances being tested for their antiemetic activity. The hospital pharmacy prepared IPA or normal saline embedded sterile gauzes (P). They were air-sealed in unmarked, identical, brown envelopes. Envelopes were consecutively numbered according to a simple randomization schedule (www.randomizer.org). The subject and envelope identification num- bers were only available to the statistical consultants. Patients, nurses, and investigators were blinded to the treatment received by each pa- tient. Patients were instructed to complete the sniffing at the other

Table 1

Demographics of the study population.

Variablesa	Total	IPA	Placebo	p
	(N = 118)	(N = 62)	(N = 56)
Age (years), median	30.0	31.5	29.5	0.514
(IQR)	(25.0-45.0)	(24.0-47.0)	(25.5-43.5)
Male, n (%)	44 (37.3)	27 (43.5)	17 (30.4)	0.141
DM, n (%)	11 (9.3)	8 (12.9)	3 (5.4)	0.161
HT, n (%)	7 (5.9)	2 (3.2)	5 (8.9)	0.192
SBP (mmHg)	122.0	126.0	118.0	0.226
	(107.0-139.0)	(108.0-140.0)	(106.5-130.0)
DBP (mmHg)	77.5	78.0	77.0	0.660
	(71.0-88.0)	(70.0-88.0)	(71.0-83.0)
MAP (mmHg)	91.5	93.3	90.0	0.358
	(84.7-101.7)	(84.7-104.7)	(84.7-99.0)
HR (bpm)	89	90.0	86.5	0.541
	(81.0-100.0)	(81.0-101.0)	(79.5-98.5)
Temperature (C)	36.2	36.3	36.2	0.671
	(36.0-36.7)	(36.1-36.7)	(36.0-36.8)
SpO2 (%)	98.0 (97.0-99.0)	98.0 (98.0-99.0)	98.5 (97.0-99.0)	0.824
Presenting symptom, n (%)
Gastroenteritis	68 (59.1)	34 (55.7)	34 (63.0)	0.434
Neurological	19 (16.5)	11 (18.0)	8 (14.8)	0.644
symptoms
Infective symptoms	17 (14.8)	9 (14.8)	8 (14.8)	0.993
Other GIS symptoms	5 (4.3)	3 (4.9)	2 (3.7)	1.000
Other symptoms	6 (5.2)	4 (6.6)	2 (1.7)	0.683
Left without being	3 (2.5)	1 (1.6)	2 (3.6)	0.502
seen, n (%)

IPA: Isopropyl Alcohol, DM: Diabetes Mellitus, HT: Hypertension, SBP: Systolic blood pres- sure, DBP: Diastolic blood pressure, MAP: Mean Arterial Pressure, HR: Heart Rate, SpO₂: Peripheral oxygen saturation.

^a All values are median (IQR) except categorical variables (sex, presence of DM, HT, and Left without being seen).

side of the nurses’ table at the triage area to avoid researchers to detect the scent.

• 1. Intervention and measurement

An 11-point (0-10) NRS was used to measure the level of N&V. Zero, and ten were marked as “no nausea” and “worst nausea imaginable.” If the patient met the inclusion criteria, then a non-study nurse handed the consecutively numbered envelope. The triage nurse instructed the patient to open the envelope, hold the gauze at a 3 cm distance to their nose, start inhaling the gauze, and not sniff beyond 60 s. Then, pa- tients were asked to mark their N&V level on the chart at the 2nd, 4th, and 10th minutes of the procedure’s start. If the patient reported com- plete N&V resolution, they could stop inhalation before 60 s, which was recorded on the sheet. ED physicians were permitted to use any rescue antiemetics for any study patient after the 10th minute.

Table 2

Comparison of the change in mean NRS in study and control groups.

NRS	Baseline	2nd minute	4th minute	10th minute	p?
IPA N	62	62	59	59	0.008
Mean +- SS,	6.4 +- 1.9	4.9 +- 2.7	4.2 +- 2.7	3.7 +- 2.8
(95% CI)	(5.9-6.9)	(4.3-5.6)	(3.5-4.9)	(3-4.4)
Control N	56	56	54	53
Mean +- SS,	6.5 +- 2.0	5.8 +- 2.6	5.8 +- 2.7	5.6 +- 2.9
(95% CI)	(6-7.1)	(5.2-6.5)	(5.0-6.5)	(4.8-6.4)
ppost-hoc??	0.679	0.031	0.003	<0.001

Bold values indicate statistically significant p values.

* Robust two-way mixed ANOVA using trimmed means with 500 bootstrap samples, sppba functions of WRS2 package in R.

?? Yuen’s test for trimmed means with 599 bootstrap samples, yuen and yuenbt func-

tions of WRS2 package in R.

7

Control Group

Isopropyl Alcohol (IPA) Group

6

Numeric Rating Scale for Nausea Mean and Standard Error

5

4

3

Baseline 2 min 4 min 10 min Baseline 2 min 4 min 10 min Time Points

Fig. 2. The mean (and standard error of the mean) NRS in the study and control groups in time according to sex.

10.0

Control Group Pre Post

IPA Group Pre Post

7.5

NRS values, before and after

5.0

2.5

0.0

1 54

1 61

Fig. 3. NRS values each patient in each group at the baseline (red dots) and the 10th minute (black dots) after the intervention. Error plots of NRS values at the baseline (pre) and 10th minutes (post) of each group are given next to the waterfall graphs.

• 1. Primary and secondary outcomes

The primary outcome of this study was the comparison of the change in NRS at the 10th minute. Secondary outcomes were the fre- quency of N&V resolution, rescue treatment, and left without being seen in each group.

• 1. Sample size estimation

We estimated that at least 50 patients were required in each group to achieve a statistically significant mean difference of 2 points (out of 10) with a type 1 error of 5% and power of 90%. We planned to recruit 20% more and targeted a sample size of 120 patients. We used the NRS values reported at the 10th minute by Beadle et al. as our reference points [16].

• 1. Statistical analysis

We used means, standard deviations (SD) and 95% confidence inter- vals (CI) or medians and interquartile ranges (IQR) to summarize nu- merical variables and frequencies and counts for categorical variables. We performed a t-test, paired t-test, Mann-Whitney U test, Wilcoxon test, repeated-measures ANOVA, Friedman test, contingency tables, chi-squared, and Fisher’s exact tests to compare groups. We compared the NRS change in consecutive time points between groups by analysis of variance (ANOVA). We found that normality (checked by the Shapiro Wilks test) and homogeneity of variances (Levene’s test) were violated. Therefore, we used robust functions provided in the WRS2 package of the R to conduct robust two-way mixed ANOVA and paired and non- paired Yuen’s test (which is a robust alternative to t-test with bootstrapping). Statistical analyses were performed by the Medcalc Sta- tistical Software version 19 with a 2-sided significance threshold of 5% (Medcalc Software Ltd., Ostend, Belgium; 2020), R statistical software with WRS2 package was used for the analysis of variance, and graphs were prepared using the R programming language (RStudio: Integrated Development for R, 2020, PBC, Boston, MA URL http://www.rstudio. com/).

1. Results

We recruited 118 patients and randomized 62 to the IPA and 56 to the P groups (intent-to-treat population). Three patients left the ED without being seen after randomization and before or after the inter- vention. We used the CONSORT guideline while preparing this report. We present the patient flowchart in Fig. 1 [17].

The study sample’s median age was 30 (IQR: 25-45), and 37.3% were

male. Demographics of the study population, according to groups, are presented in Table 1. The IPA and P groups were similar according to age, sex, comorbidities, and vital signs (Table 1).

We compared the NRS change in the time between groups with ro- bust two-way mixed ANOVA and observed a statistically significant in- teraction between treatment and time (Table 2). The evaluation of simple main effects revealed a significant difference between treatment arms (between subjects, p = 0.008) and consecutive time points in each treatment arm (within-subjects). Post-hoc comparison of the treatment arms at each time point revealed that a significant difference appeared as early as 2nd minute after the intervention (Table 2). Post-hoc com- parison of the within-subjects’ effects of time demonstrated that none of the pairwise comparisons were significantly different in the P group (all p NS, adjusted with Bonferroni correction). Conversely, pairwise comparisons of the within-subjects’ effects of time on NRS were all sig- nificantly at a level of p < 0.001 in the IPA group (adjusted with Bonferroni correction).

A rescue treatment was administered in 67 (58.3%) of the 115 pa- tients. Metoclopramide, Hyoscine-N-butylbromur, or both were used in 40 (34.7), 2 (1.7), and 25 (21.7) patients, respectively (Figs. 2 and 3).

Table 3

Comparison of rescue treatment rate between study and control groups.

Rescue treatment, n (%)	Total (n = 115)	IPA (n = 61)	Control (n = 54)	p
None	48 (41.7)	34 (55.7)	14 (25.9)	0.004
M	40 (34.8)	19 (31.1)	21 (38.9)
HNB	2 (1.7)	0 (0.0)	2 (3.7)
Both	25 (21.7)	8 (13.1)	17 (31.5)

Bold values indicate statistically significant p values. M: Metoclopramide, HNB: Hyoscine-N-butylbromur.

The percentage of patients in whom a rescue treatment was adminis- tered was significantly higher in the P group (40 (74.1) vs. 27 (44.3), p = 0.004, Table 3).

1. Discussion

In this study, IPA significantly decreased the intensity of N&V com- pared to P starting at the 2nd minute after the inhalation. Besides, the frequency of the need for a Rescue medication was lower in the IPA group compared to P (p = 0.004). In our study, the number of men and women was balanced, and there was no significant difference be- tween the two genders in response to the treatment given [12]. This shows that IPA is equally effective in both sexes.

To the best of our knowledge, only two RCTs evaluated IPA’s efficacy for N&V in the ED. Beadle et al. published the first study, and enrolled 80 patients in either IPA or P groups, and reported a statistically significant median difference of 3 points (out of 10) in VNRS at the 10th minute (p < 0.001). In the second RCT, which was published during the recruit- ment phase of our trial, April at el. found that IPA, with or without oral ondansetron, provided better N&V relief than oral ondansetron alone in ED patients in 30 min [18]. They reported a decrease of 30 points in VAS (out of 100) in 30 min with IPA, which was similar to the findings of Beadle et al. We found a decrease of 2.7 points in NRS at the end of 10 min, which was almost the same compared to previous studies.

• 1. Limitations

This study had several limitations. First, there may have been a selec- tion bias due to overcrowding. Second, the IPA’s efficacy and side effects after 10 min are unknown. Third, the duration between the application of IPA and physical examination is unknown. Despite the significantly lower proportion of patients receiving Rescue therapy with IPA, we do not know if they also had shorter Waiting times. Fourth, we did not follow-up patients for N&V recurrence or need for antiemetics after dis- charge. Fifth, our study population’s mean age is around 30, which is ex- pected in our country for ESI 4-5 patients but may be low for other institutions. Therefore, our results should be used in different age groups with caution. Last, we did not check with a second observer if the triage nurse is stationed at a safe distance from the patient not to smell IPA, and their blinding was not impaired.

1. Conclusion

In this study, IPA was significantly more effective than P for the treat- ment of N&V at the triage. Moreover, patients in the IPA group had less need for rescue treatment. We think that IPA can be used in the ED for the symptomatic treatment of N&V as a cheap and comfortable alterna- tive to other drugs without any significant side effects at the triage and waiting room.

Ethical approval

Ethical approval was confirmed by the Marmara University Institu- tional Review Board at 2.12.2016 (Approval No: 09.2016.616).

CRediT authorship contribution statement

Hande Candemir: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Valida- tion, Writing – original draft, Writing – review & editing. Haldun Akoglu: Conceptualization, Data curation, Formal analysis, Investiga- tion, Methodology, Project administration, Software, Supervision, Vali- dation, Writing – original draft, Writing – review & editing. Erkman Sanri: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Validation, Writing

– original draft. Ozge Onur: Conceptualization, Methodology, Project administration, Supervision, Validation, Writing – review & editing. Arzu Denizbasi: Conceptualization, Methodology, Project administra- tion, Supervision, Validation, Writing – review & editing.

Declaration of competing interest

None declared.

Acknowledgement

We want to thank Dr. Yigit Esref Beyazitli for his help while data collecting.

Funding

None declared.

References

1. Myer PA, Mannalithara A, Singh G, Singh G, Pasricha PJ, Ladabaum U. Clinical and Economic burden of emergency department visits due to gastrointestinal diseases in the United States. Am J Gastroenterol. 2013;108(9):1496-507. https://doi.org/ 10.1038/ajg.2013.199.
2. Egerton-Warburton D, Meek R, Mee MJ, Braitberg G. Antiemetic use for nausea and vomiting in adult emergency department patients: randomized controlled trial comparing ondansetron, metoclopramide, and placebo. Ann Emerg Med. 2014;64

(5). https://doi.org/10.1016/j.annemergmed.2014.03.017 526-532.e1.

1. Skinner HG, Blanchard J, A Elixhauser – … (MD): Agency for Healthcare Research and …. Trends in emergency department visits, 2006-2011: Statistical brief# 179 Healthcare Cost and Utilization Project (HCUP) statistical briefs; 2014.
2. Furyk JS, Warburton DE, Meek RA. Drugs for the treatment of nausea and vomiting in adult patients in the emergency department setting. In: Furyk JS, editor. Cochrane database of systematic reviews; 2012. https://doi.org/10.1002/14651858.CD010106 120(9):261.
3. Hasler WL, Chey WD. Nausea and vomiting. Gastroenterology. 2003;125(6):1860-7. https://doi.org/10.1053/j.gastro.2003.09.040.
4. Anderson LA, Gross JB. Aromatherapy with peppermint, isopropyl alcohol, or pla- cebo is equally effective in relieving postoperative nausea. J Perianesth Nurs. 2004; 19(1):29-35. https://doi.org/10.1016/j.jopan.2003.11.001.
5. Merritt BA, Okyere CP, Jasinski DM. Isopropyl alcohol inhalation: alternative treat- ment of postoperative nausea and vomiting. Nurs Res. 2002;51(2):125-8. https:// doi.org/10.1097/00006199-200203000-00009.
6. Winston AW, Rinehart RS, Riley GP, Vacchiano CA, Pellegrini JE. Comparison of in- haled isopropyl alcohol and intravenous ondansetron for treatment of postoperative nausea. AANA J. 2003;71(2):127-32.
7. Teran L, Hawkins JK. The effectiveness of inhalation isopropyl alcohol vs. granisetron for the prevention of postoperative nausea and vomiting. AANA J. 2007;75(6): 417-22.
8. Cotton JW, Rowell LR, Hood RR, Pellegrini JE. A comparative analysis of isopropyl al- cohol and ondansetron in the treatment of postoperative nausea and vomiting from the hospital setting to the home. AANA J. 2007;75(1):21-6.
9. Spencer KW. Isopropyl alcohol inhalation as treatment for nausea and vomiting. Plast Surg Nurs. 2004;24(4):149-54. https://doi.org/10.1097/00006527- 200410000-00005.
10. Pellegrini J, DeLoge J, Bennett J, Kelly J. Comparison of inhalation of isopropyl alcohol vs promethazine in the treatment of postoperative nausea and vomiting (PONV) in patients identified as at high risk for developing PONV. AANA J. 2009;77(4):293-9.
11. Hunt R, Dienemann J, Norton HJ, et al. Aromatherapy as treatment for postoperative nausea: a randomized trial. Anesth Analg. 2013;117(3):597-604. https://doi.org/10. 1213/ANE.0b013e31824a0b1c.
12. Radford KD, Fuller TN, Bushey B, Daniel C, Pellegrini JE. Prophylactic isopropyl alco- hol inhalation and intravenous ondansetron versus ondansetron alone in the pre- vention of postoperative nausea and vomiting in high-risk patients. AANA J. 2011; 79(4 Suppl):S69-74.
13. Hines S, Steels E, Chang A, Gibbons K. Aromatherapy for treatment of postoperative nausea and vomiting. In: Cochrane Anaesthesia, Critical and Emergency Care Group, editor. Cochrane database syst rev; 2012. https://doi.org/10.1002/14651858. CD007598.pub2 19(4):CD007598.
14. Beadle KL, Helbling AR, Love SL, April MD, Hunter CJ. Isopropyl alcohol nasal inhala- tion for nausea in the emergency department: a randomized controlled trial. Ann Emerg Med. 2016;68(1). https://doi.org/10.1016/j.annemergmed.2015.09.031 1-9. e1.
15. Pandis N, Chung B, Scherer RW, Elbourne D, Altman DG. CONSORT 2010 statement: extension checklist for reporting within person randomised trials. BMJ. 2017;357. https://doi.org/10.1136/bmj.j2835 j2835.
16. April MD, Oliver JJ, Davis WT, et al. Aromatherapy versus oral ondansetron for anti- emetic therapy among adult emergency department patients: a randomized con- trolled trial. Ann Emerg Med. 2018;72(2):184-93. https://doi.org/10.1016/j. annemergmed.2018.01.016.