Efficacy of topical capsaicin for cannabinoid hyperemesis syndrome in a pediatric and adult emergency department

Abstract

Introduction:

Cannabinoid hyperemesis syndrome (CHS) is a clinical diagnosis characterized by symptoms of recurrent nausea, vomiting, and severe abdominal pain in the setting of chronic cannabis use. Symptoms of CHS are frequently unresponsive to standard antiemetic therapy. Topical capsaicin applied to the abdomen has been cited as a potential effective agent for CHS however robust evidence is lacking.

Methods:

This was a single-center retrospective cohort study to evaluate the efficacy of topical capsaicin in pediatric and adult patients presenting to the emergency department (ED) with suspected or confirmed CHS. The primary outcome assessed was if utilization of capsaicin for CHS resulted in more patients achieving an “efficacious” result, defined as only requiring ≤1 rescue medication for symptom relief after receiving capsaicin or after administration of the first agent in patients who did not receive capsaicin during their ED course. Secondary outcomes included total ED length of stay, time to discharge after administration of the reference agent (RA), proportion of patients requiring admission, total number of medication doses given for symptom relief, change in pain score and episodes of emesis, and proportion of patients returning to the ED within 24 h for the same complaint. Additional analyses were also performed to explore patient characteristics that may be predictive of capsaicin efficacy.

Results:

201 patients were included in the final analysis of which 25 were <21 years old and seen in the pediatric ED. A greater proportion of patients in the capsaicin group achieved the primary outcome of efficacy as compared to patients who did not receive capsaicin (55% vs 21%, p < 0.001, unadjusted OR 1.44 [95% CI 0.586–0.820]). There were no differences in secondary outcomes except for time to discharge after administration of the RA which was shorter in the capsaicin group (3.72 vs 6.11 h, p = 0.001).

Conclusion:

Significantly more patients in the capsaicin group experienced efficacy compared to patients who did not. Time to discharge after administration of the reference agent was shorter for those who received capsaicin compared to patients who did not. Administration of capsaicin did not influence patients’ total number of medications received or total ED length of stay. Future research is needed to determine capsaicin’s efficacy when utilized earlier in therapy, ideally upon initial diagnosis of CHS, and before additional adjunct medications are administered.

1. Introduction

Cannabinoid hyperemesis syndrome (CHS) is a clinical diagnosis of exclusion characterized by symptoms of significant nausea, cyclical vomiting, and severe abdominal pain with a history of frequent and long-standing cannabis use. There are no formally established diagnosis criteria for CHS although expert consensus guidelines have been created to aid in recognition and management [1]. Major diagnostic characteristics include the above symptoms, history of use, and temporary relief with hot showers/bathing. Because these characteristics rely heavily on patient-reported information, the diagnosis can be missed if an accurate history is not obtained [16].

Once a diagnosis of CHS has been made with a low suspicion for alternative causes, management focuses on symptom relief, in addition to patient education on abstaining from future cannabis use. Antiemetics, antacids, antihistamines, benzodiazepines, antipsychotics, and analgesics (including opioids) have all been cited in the literature with varying degrees of efficacy and symptom relief [2, 16, 19]. Capsaicin, the active component of plants belonging to the Capsicum genus (chili peppers), has also been cited in the literature as a possible treatment to abort CHS symptoms.

The exact mechanism for how capsaicin alleviates CHS symptoms is not known. However, the most widely agreed upon hypothesis is through capsaicin’s activation of the transient receptor potential vanilloid-1 (TRPV-1) receptor and direct depletion of substance P, which produces antiemetic effects (Fig. 1). Capsaicin may also provide local analgesia when applied to the abdominal area.

Fig. 1.

Mechanism of capsaicin for cannabinoid hyperemesis syndrome [11,18] created with BioRender.com.

Capsaicin cream is relatively inexpensive, available over the counter, and associated with few adverse effects, the most notable being a local burning sensation and skin irritation when applied topically. Despite the benefits of capsaicin, it is difficult to justify capsaicin as first-line therapy: most evidence is limited to case reports and abstracts (Table 1), pediatric patients are rarely included, and quantitative outcomes and/or statistical analyses are oftentimes absent [3–12]. The largest study at time of publication is a retrospective cohort by Wagner and colleagues consisting of 43 patients who served as their own controls from prior visits. While no difference was found for the primary outcome of ED length of stay, this study provided further evidence of potential benefit with capsaicin as part of a multi-modal treatment for CHS with minimal adverse effects. Limitations included multiple concentrations of capsaicin and lack of standard dosing and administration protocols. Patients with multiple prior ED visits may also have had less extensive workup and quicker administration of capsaicin in subsequent encounters, potentially creating bias in favor of capsaicin. Therefore, the goal of our research was to examine the effectiveness of capsaicin for management of CHS in a larger cohort population that included both adult and pediatric patients utilizing a single concentration of capsaicin and analysis of outcomes.

Table 1.

Published studies of capsaicin for CHS in the ED.

Study	Design	Population	Main findings/conclusions
Biary et al. (2014) [3] (n = 1)	Case report	Adult	Resolution of N/V and abdominal cramping after 20–30 min
Lapoint J. (2014) [4] (n = 2)	Case series (abstract only)	Adolescent and adult aged 19 and 29 yo	Capsaicin cream associated with decrease in pain score from 8 to 4/10 and 3/10, respectively
Lapoint J. (2014) [4] (n = 5)	Case series (abstract only)	Adult	Resolution of nausea, vomiting, and pain within 22–45 min of administration of capsaicin
Roman et al. (2016) [6] (n = 1)	Case report	Adult	Metoclopramide and granisetron ineffective. Resolution of symptoms after application of capsaicin
Dezieck et al. (2017) [7] (n = 13)	Case series	Adults and adolescent aged 19 (n = 1) to 47 yo	Reported relief of symptoms within 45 min after administration of capsaicin cream
McCloskey et al. (2017)[8] (n = 22)	Retrospective cohort series (abstract only)	Ages unknown	Capsaicin associated longer emergency department length of stay (non-significant)
Graham et al. (2017) [9] (n = 2)	Case series	Adolescents aged 16 and 20 yo	Capsaicin associated with improvement in symptoms of nausea and abdominal pain
Hafez et al. (2017) [10] (n = 4)	Case series (abstract only)	Ages unknown	Patients discharged within hours of receiving capsaicin (also received ondansetron and prochlorperazine)
Moon et al. (2018) [11] (n = 1)	Case report (abstract only)	Adult	Partial resolution of symptoms after first application, complete resolution after 2 additional doses of capsaicin
Sharma U et al. (2018) [12] (n = 1)	Case report	Adult	4-day hospital admission; Droperidol, antacids, PPI’s and capsaicin associated with symptom improvement
Wagner et al. (2020) [13] (n = 43)	Retrospective cohort review	Adult	No significant difference in ED LOS when capsaicin was utilized for CHS 179 vs 201 min (p = 0.33)

2. Methods

2.1. Setting

This single-center retrospective cohort study was conducted at a large academic medical center and included patients from both the pediatric and adult emergency departments. The pediatric ED is comprised of 31 total beds with over 26,000 annual patient visits. The adult ED has 86 total beds and over 100,000 annual patient visits.

2.2. Selection of patients

Electronic health record (EPIC, Crystal Reports®) generated reports were used to identify patients who had a capsaicin prescription order in the ED between October 2014 and June 2019. Additional patients were identified based on a diagnosis code that included the search terms “cannabinoid hyperemesis syndrome”, “cyclical vomiting syndrome”, “cyclic vomiting syndrome, nausea presence unspecified”, “cyclic vomiting syndrome, intractability of vomiting not specified, presence of nausea not specified”, and “synthetic cannabinoid abuse” for the same time period.

Patients were included for analysis if they received capsaicin (active group) for suspected or diagnosed CHS in the ED, or if they were diagnosed with CHS but did not receive capsaicin (control group) in the ED. Suspected CHS was defined as (1) reoccurring vomiting without other explanation and (2) confirmed or suspected cannabis use. Confirmed CHS was defined as having “cannabinoid hyperemesis syndrome” as the diagnosis code in the problem list for the visit. Suspected cannabis use was defined as either the presence of a positive drug screen or a historical chart documentation of cannabis use. Patients were excluded if they had incomplete or insufficient documentation for capsaicin administration and/or symptoms, if a urine drug screen was negative for cannabinoids, if they left against medical advice (AMA), or if capsaicin was used for indication(s) other than CHS (e.g. arthritis or neuropathic pain).

2.3. Capsaicin product

Capsaicin 0.025% cream is the only formulation available at our institution, and was used for this study. Instructions in the electronic orderable state to “wear gloves to apply to [an area specified by the ordering clinician, usually the abdomen], wash hands after applying, and avoid contact with eyes or other sensitive areas of the body.” Typically, nurses are counseled to apply approximately a 1 mm layer of cream to the patient’s abdomen upon dispensing of the product from the ED satellite pharmacy if no other instructions are specified in the order.

2.4. Outcome measures

The primary outcome in this study was to assess if utilization of capsaicin for management of CHS resulted in more patients achieving an efficacious outcome. Efficacy was defined as only requiring ≤1 additional medication for symptom relief after receiving the reference agent (RA). In patients who received capsaicin, the reference agent was capsaicin. In patients who did not receive capsaicin (control), the reference agent was the first agent administered. The rationale for this comparison was that if capsaicin is truly effective for symptoms of CHS, then patients should not require >1 additional rescue medication(s) for symptom relief after capsaicin was applied. This was compared to the first agent given in the control group based on the assumption that the first agent was the one the provider presumed to provide the most benefit for symptom relief and thus chose to give it first. Treatment failure/non-efficacy was assigned if the patients received ≥2 rescue medications after administration of RA.

Secondary outcomes collected included ED length of stay, time to discharge, need for admission due to uncontrolled CHS symptoms, total number of medication doses for symptom relief received defined as administration of a PRN medication designated for nausea, vomiting, or pain, change in pain score before and after administration of the RA, change in episodes of emesis before and after administration of the RA, hospital admission rates, and number of patients returning to the ED within 24 h with the same complaint. Patient encounters which did not have documentation of pain scores or episodes of emesis both before and after administration of RA were considered incomplete records and not included as data points for analysis of that outcome.

Patient demographics, including patient age, sex, cannabis use history, symptom duration, and number and type of CHS symptom(s) present were also collected.

2.5. Post-hoc analyses

Several post-hoc analyses were performed after completion of the study. This included a subgroup analysis of pediatric patients for secondary outcomes, examination of potential relationship between efficacy and return to the ED within 24 h, and examination of patient symptoms and patterns of cannabis use as potential predictors of efficacy in patients who received capsaicin.

2.6. Statistical analysis

Data was recorded using REDCap (Research Electronic Data Capture) hosted at University of Michigan and exported to SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, N.Y., USA) for analysis. Chi-square test was used to evaluate categorical data. Student’s t-test and Mann-Whitney U test were used to evaluate normally and non-normally distributed continuous data, respectively. Time to discharge or admission was analyzed using cox-proportional hazard model with the p-value derived from a log-rank test. Change in pain score and episodes of emesis were evaluated using linear regression.

Post-hoc subgroup analyses were performed using SAS 9.4 (SAS Institute, Cary NC) and the Excel software package. Student’s t-test and Mann-Whitney U test were performed for the pediatric subgroup analysis due to small sample sizes. Relationship between efficacy and proportion of patients returning to the ED was analyzed using chi-square test. Evaluation of patients’ symptoms and patterns of cannabis use as predictors of capsaicin efficacy were performed with both univariate and multivariable binary logistic regression. For tire multivariable logistic regression analysis, the symptom of “relief with hot showers” was used as the main variable of interest as a predictor of efficacy. Symptoms of nausea, vomiting, and abdominal pain were inputted as confounding variables. Relief with hot showers was chosen as the main variable in the multivariable analysis because authors hypothesized that this symptom would be most related with achieving an efficacious outcome in patients who received capsaicin, i.e. the mechanism of capsaicin for CHS works by activating the same TPRV-1 receptor as hot showers. Achievement of statistical significance for relief with hot showers as a predictor of efficacy would strengthen the proposed mechanism of capsaicin for CHS that is stated in the literature. Symptoms of nausea, vomiting, and abdominal pain were chosen as confounding variables as these are often symptoms related to undifferentiated gastrointestinal (GI) etiologies which often confound a CHS diagnosis, and capsaicin would not be effective for other GI ailments not related to cannabis use.

A p value of <0.05 was considered statistically significant. This study was approved by our health system’s internal institutional review board (IRB).

3. Results

3.1. Baseline demographics and cannabis use history

201 patients met the inclusion criteria within the study period (Fig. 2). Of those, 149 patients received capsaicin for suspected or confirmed CHS. The remaining 52 patients included those who did not receive capsaicin for suspected or confirmed CHS during their ED course, and served as the control group.

Fig. 2.

Patient selection.

Baseline characteristics are listed in Table 2. No statistically significant differences were found between the two groups. Forty-six percent of patients in both groups were male (p = 0.34). The median age of patients in the capsaicin (C) and control group (no capsaicin; NC) were 31 and 27 years, respectively (p = 0.19). Patients typically reported more than one type of symptom. A majority of patients reported symptoms of nausea (C: 79.2%, NC: 82.7%, p = 0.59), vomiting (C: 90.6%, NC: 92.3%, p = 0.71), and abdominal pain (C: 72.5%, NC: 75%, p = 0.72) upon presentation with approximately one-third of patients reporting relief of symptoms with hot showers (C: 32.2%, NC: 36.5%, p = 0.57). More than half of the patients in each group reported acute onset of symptoms <1 week (C: 62.4%, NC: 55.8%) and around 20% of patients in each group reported having symptoms for up to 1 month (C: 18.1%, NC: 21.2%) and >1 month (C: 17.4%, NC: 21.2%) prior to their ED visit.

Table 2.

Baseline characteristics

	Received capsaicin (n = 149)	Did not receive capsaicin (n = 52)	p
Sex – Male (%)	68 (45.6)	24 (46.1)	0.34
Age – median, years (range)	31 (16–66)	27 (16–57)	0.19
Symptoms present – No. (%)
Nausea	118 (79.2)	43 (82.7)	0.59
Vomiting	135 (90.6)	48 (92.3)	0.71
Abdominal pain	108 (72.5)	39 (75)	0.72
Relief with hot showers	48 (32.2)	19 (36.5)	0.57
Duration of symptoms – No. (%)			0.69
< 1 week	93 (62.4)	29 (55.8)
≥1 week to < 1 month	27 (18.1)	11 (21.2)
≥1 month	26 (17.4)	11 (21.2)
Unknown/not specified	3 (2)	1 (1.9)
Cannabis length of Use – No. (%)			0.24
< 1 year	4 (2.7)	4 (7.7)
≥1 year	39 (26.2)	11 (21.2)
Unknown	106 (71.1)	37 (71.2)
Frequency of cannabis use – No. (%)			0.91
Less than once daily	39 (26.2)	15 (28.8)
At least once daily	85 (57)	28 (53.8)
Unknown	25 (16.8)	9 (17.3)
Type of cannabis use – No. (%)
Smoked	98 (65.8)	37 (71.2)	0.48
Other (not smoked)	7 (4.7)	3 (5.8)	0.60
Unknown	58 (38.9)	17 (32.7)	0.43
Pediatric emergency department patient – No. (%)	17 (11.4)	8 (15.4)	0.46
Time to administration of capsaicin or first agent – median, hours	3	1	<0.001

Length of cannabis use was available for 58 patients (29% of cohort) and 20% of patients in each group (C: 26.2%, NC: 21.2%) reported using cannabis regularly for at least 1 year. Frequency of cannabis use was available for 167 patients (83% of cohort) and two thirds of patients in each group use cannabis at least once daily (C: 68.5%, NC: 65.1). Type of cannabis used was available for 145 patients (72% of the cohort) and most patients reported smoking cannabis (C: 65.8%, NC: 71.2%, p = 0.48). Other methods of cannabis use included edibles and tinctures (C: 4.7%, NC: 5.8%). Reported use of medical/prescription marijuana was classified as unknown type unless otherwise stated since the specific method of use could not be determined. Some patients reported more than one type of cannabis use. Pediatric patients (<21 years of age) comprised 12.4% of the cohort – 17 (11.4%) pediatric patients received capsaicin and 8 (15%) did not receive capsaicin (p = 0.46).

3.2. Treatment characteristics

An antiemetic was the initial agent administered to a majority of patients in both groups (C:63%, NC:73%, p = 0.08) (Fig. 3). The 14 remaining patients in the control group received either an antipsychotic (9%), “Other” agent (8%), non-opioid analgesic (4%), benzodiazepine (2%), antihistamine (2%), or opioid analgesic (2%) as their first agent for symptom control. The 55 remaining patients in the capsaicin arm received either capsaicin (12%), opioid analgesic, (8%), antipsychotic (6%), antihistamine (5%), benzodiazepine (3%), or a non-opioid analgesic (3%) as the initial agent for symptom control. A common “GI cocktail” in used in our ED consists of lidocaine 2% oral solution with aluminum-magnesium hydroxide-simethicone (Maalox®) 400–400-40 mg, and was classified as “Other”.

Fig. 3.

First agent given.

The median time to administration of the reference agent (capsaicin or first agent) was 3 h in the capsaicin group and 1 h in the control group which was statistically significant (p < 0.001).

3.3. Primary and secondary outcomes

Results of primary and secondary outcomes are listed in Tables 3 and 4. A significantly greater proportion of patients in the capsaicin group met the primary outcome of efficacy, defined as requiring ≤1 additional dose of medication for symptom relief after capsaicin or first agent prior to discharge, compared to patients in the control group (55% vs 21%, p < 0.001, unadjusted OR 1.44 [95% CI 0.586–0.820]).

Table 3.

Primary outcome.

	Received capsaicin (n = 149)	Did not receive capsaicin (n = 52)	p	OR	95% CI	β
Primary outcome
Efficacy (requiring ≤1 additional doses of medication for symptom relief after capsaicin or first agent (control) prior to discharge) – No. (%)	84 (55)	11 (21)	<0.001	1.44	0.586–0.820
	<21 and Received capsaicin (n = 17)	<21 did not receive capsaicin (n = 8)	P	OR	95% CI	β
Primary outcome
Efficacy in patients <21 years of age – No. (%)	9 (53)	3 (60)	0.682		0.081–5.159	−0.434

Table 4.

Secondary outcomes.

	Received capsaicin (n = 149)	Did not receive capsaicin (n = 52)	p	OR	95% CI	β
Secondary outcomes
Total ED length of stay (median) – hours	6.1	5.25	0.372
Total medications for symptom relief received – median, No. (range)	4 (0–60)	4 (0–43)	0.934
Admitted due to uncontrolled CHS symptoms – No. (%)	39 (26.2)	13 (25)	0.868	1.01	0.8190–1.183
Return to ED in 24 h for same symptoms or complaint – No. (%)	8 (5.4)	3 (5.8)	0.913	1.02	0.704–1.479
Time to discharge or admission after administration of capsaicin or first agent (average) – hours	3.72	6.11	0.001		2.799–3.501
Change in pain score after administration of capsaicin or first agent – average (126 complete records)	+0.94	+1.38	0.464		−1.636 - 0.750	−0.443
Change in episodes of emesis after administration of capsaicin or first agent (14 complete records)	+0.03	−0.21	0.423		−3.575–1.624	−0.975

Time to discharge or admission after administration of RA was also significantly shorter in patients who received capsaicin (3.72 vs 6.11 h, p = 0.001, [95% CI 2.799–3.501]). However, no differences were found for remaining secondary outcomes for patients who received capsaicin vs no capsaicin for total ED length of stay (6.1 vs 5.25 h, p = 0.372), admission rates (26.2% vs 25%, p = 0.868), median number of medications received for symptom control (4 vs 4, p = 0.934), or percent of patients who returned to the ED within 24 h (5.4% vs 5.8%, p = 0.913). There were also no differences in average change in pain score (+0.94 vs +1.38, p = 0.464) or change in episodes of emesis (+0.03 vs −0.21, p = 0.423) after administration of the RA, when these data points were available.

3.4. Subgroup and post-hoc analyses

Subgroup analysis was performed in patients <21 years of age for the outcome of efficacy using binary logistic regression (Table 3) but no statistical significance was found for an association between age < 21 years old and achievement of efficacy. Additional analyses were performed for this subgroup for the remaining secondary outcomes (Table 5) as requested during peer review but did not result in any significant findings in capsaicin vs no capsaicin for total ED length of stay (6 vs 4.52 h, p = 0.918), median number of medications received for symptom control (3 vs 2, p = 0.685), admission rates (17.6% vs 12.5%, p = 0.745), time to discharge or admission after administration of RA (3.12 vs 4.92 h, p = 0.333), or percent of patients who returned to the ED within 24 h (0 vs 0, p = 0.724). There were also no difference in average change in pain score after administration of the RA (−0.22 vs −2.4, p = 0.199). Analysis for change in episodes of emesis was not performed for this subgroup because there was no data available for this outcome.

Table 5.

Secondary outcomes in patients <21 years of age.

	<21 and received capsaicin (n = 17)	<21 and did not receive capsaicin (n = 8)	p	OR	95% CI
Secondary outcomes
Total ED length of stay (median) – hours	6	4.53	0.918
Total medications for symptom relief received – median, No. (range)	3	2	0.685
Admitted due to uncontrolled CHS symptoms – No. (%)	3 (17.6)	1 (12.5)	0.745	1.50	(0.131–17.180)
Return to ED in 24 h for same symptoms or complaint – No. (%)	0	0	0.724	0.49	(0.009–26.645)
Time to discharge or admission after administration of capsaicin or first agent (average) – hours	3.12	4.92	0.333
Change in pain score after administration of capsaicin or first agent – average (14 complete records)	−0.22	−2.4	0.199

Additionally, chi-square test was performed to evaluate if there was a relationship between efficacy and proportion of patients returning to our EDs within 24 h for repeat CHS symptom management (Table 6). The rationale for this was that efficacy of treatment should be expected to correlate with patients not requiring a repeat ED visit for symptom management. Of the 95 patients who met the pre-specified definition of an efficacious outcome, regardless of treatment group, a majority did not return to our ED within 24 h of ED discharge for repeat CHS symptom management (88 patients, or 92.6%). Seven patients with an efficacious outcome (7.4%) did return to our ED for repeat CHS symptom management within 24 h of ED discharge. Of the 106 patients who did not meet the pre-specified definition of efficacy, 102 (96.2%) did not return to the ED within 24 h of discharge for repeat CHS symptom management. Four patients who did not have an efficacious outcome (3.8%) did return to the ED for repeat CHS symptom management. No statistical significance was found (p = 0.263) and efficacy did not correlate with percentage of patients who returned to our ED within 24 h for repeat CHS symptom management.

Table 6.

Efficacy and proportion of patients returning to the ED.

		Return to ED within 24 h		p
		Did not return to ED within 24 h – No. (%) (n = 190)	Returned to ED within 24 h – No. (%) (n = 11)
Efficacy	No (n = 106)	102 (96.2)	4 (3.8)	0.263
	Yes (n = 95)	88 (92.6)	7 (7.4)

Univariate binary logistic regressions were performed to identify potential variables predictive of capsaicin efficacy (Table 7) with the following variables: type of symptoms, frequency of cannabis use, length of cannabis use, type of cannabis use, and duration of symptoms. Results revealed that overall duration of symptoms was a significant predictor of efficacy (p = 0.0253). Pair-wise comparisons between duration of symptoms revealed that there was a significant difference between patients who had symptoms for ≥1 month and patients with symptoms <1 week in predicting efficacy. In particular, the odds of an efficacious outcome with capsaicin in patients who had a longer duration of symptoms ≥1 month are 0.2881 (95% CI: 0.1166–0.7120) times the odds of achieving efficacy in patients who had a shorter duration of symptoms <1 week (p = 0.0070). In other words, patients presenting to the ED having had longer duration of symptoms were less likely to experience efficacy with capsaicin. No significant differences were found when comparing the above durations to patients reporting duration of symptoms ≥1 week to <1 month.

Table 7.

Effect of patient characteristics and patterns of cannabis use on likelihood of capsaicin efficacy.

Variable	Effect estimate (Odds Ratio)	95% CI	p
Nausea	2.626	1.087–6.717	0.036
Vomiting	2.2127	0.608–8.499	0.2489
Abdominal pain	0.619	0.271–1.373	0.2434
Relief with hot showers	0.622	0.31–1.401	0.2814
Frequency of cannabis use at least once daily	2.054	0.017–3.013	0.3821
Length of cannabis use ≥1 year	0.521	0.254–1.043	0.0694
Smoked cannabis	0.521	0.254–1.043	0.0694
Duration of symptoms			0.0253
≥1 month vs <1 week	0.288	0.112–0.697	0.007
≥1 week to <1 month vs <1 week	0.672	0.278–1.637	0.376
≥1 week to <1 month vs ≥1 month	2.333	0.779–7.305	0.1351

Presence of nausea was also identified as a significant predictor of efficacy (p = 0.0285). The odds of an efficacious outcome in patients receiving capsaicin reporting symptoms of nausea were 2.4783 (95% CI: 1.1004–5.5814) times the odds of efficacy in patients who did not report symptoms of nausea. Symptoms of vomiting (p = 0.2893), abdominal pain (p = 0.0732), and relief with hot showers (p = 0.2803) were not significant predictors of efficacy.

The remaining variables of smoked cannabis (p = 0.0694), frequency of use (p = 0.4599), and duration of cannabis use (p = 0.3821) were not significant predictors of efficacy in patients who received capsaicin.

The multivariable logistic regression (Table 8) using symptom relief with hot showers as the main predictor of efficacy and symptoms of nausea, vomiting, and abdominal pain, as confounding variables in the model was not found to be a significant predictor of efficacy in patients who received capsaicin (p = 0.2433).

Table 8.

Relief with hot showers and capsaicin efficacy.

Variable	Effect estimate (Odds Ratio)	95% CI	p
Relief with hot showers	0.653	0.318–1.335	0.2433
Nausea	2.317	0.991–5.585	0.0549
Vomiting	1.243	0.369–4.273	0.7232
Abdominal pain	0.572	0.256–1.232	0.1601

4. Discussion

As increasing legalization and subsequent rates of cannabis use are expected to correlate with recognition and diagnoses of CHS, the role of capsaicin for symptom management is a topic of continued interest [14, 15]. The mechanism by which topical capsaicin works for CHS is not fully understood but likely involves the desensitization and downregulation of transient receptor potential vanilloid-1 (TRPV-1} receptors and upregulation of substance P, leading to uncontrolled hyperemesis [11]. Thus, capsaicin is thought to provide additional symptom relief by direct depletion of substance P.

A number of publications have reported on the effects of capsaicin for CHS but have been mostly limited to case reports or series lacking statistical analyses of outcomes and with varied reporting of use of adjunctive medications [3–12]. In addition, few studies have examined capsaicin in an adolescent population [7,9]. Overall, existing literature suggests capsaicin may be beneficial in the management of CHS however its exact place in therapy and patient population is unclear. This study aimed to examine the effect of capsaicin for CHS in a large adult and adolescent population examining unique outcomes of interest with comprehensive statistical analyses.

Our investigation found a statistically significant difference in the primary efficacy outcome – patients who received capsaicin required fewer additional medications after capsaicin application, compared to patients in the control group who required more rescue agents after administration of the reference agent. Additionally, the secondary outcome of time to discharge after administration of RA favored the capsaicin group, with patients being discharged sooner after capsaicin application compared to time to discharge after first reference agent in the control group. Although less additional medications were given after capsaicin application, this study is unable to determine what intervention most influenced resolution of symptoms: use of capsaicin, total number of medications given overall, or simply total length of time after start of medication therapy. Both groups received a median number of 4 medications overall, and median time to administration of capsaicin was 3 h compared to median time of 1 h for RA administration. No difference in patients’ overall ED length of stay was observed. Nonetheless, a greater number of patients required fewer adjunct medications after application of capsaicin.

These findings add to the body of evidence supporting the potential benefit of capsaicin at the very least when used with other agents, and considerations should be made to utilize capsaicin early in therapy, soon after a diagnosis of CHS is made. Our findings also correlate with those in Wagner et al. [13] where utilization of capsaicin resulted in fewer additional medications given but no difference in patients’ ED length of stay, although the former was not statistically significant.

Exploration of patient symptoms and patterns of cannabis use as potential predictors of efficacy in the post-hoc analysis yielded some statistically significant and interesting results. Longer duration of symptoms ≥1 month was associated with being significantly less likely to have an efficacious outcome with capsaicin compared to patients reporting a shorter duration of symptoms <1 week. This suggests that patients who present sooner for treatment of CHS symptoms are more likely to experience efficacy with capsaicin than patients who have had a prolonged duration of symptoms. While this result is interesting, when considering this finding in the context of clinical practice, the duration a patient’s symptoms when presenting to the ED as a sole characteristic would unlikely influence whether or not capsaicin is used for treatment of CHS. For this reason, duration of symptoms was not included in the multivariable analysis even though p < 0.05. Nausea was also identified as a significant predictor of efficacy. However, we do not suspect this translates to a great clinical significance either, as absence of nausea when other symptoms are present in treating a patient with CHS would also unlikely influence the decision to use capsaicin. Therefore, nausea was not used as the primary variable of interest in the multivariable analysis even though p < 0.05, but it was included as a confounding variable.

In deciding which confounding variables would be used in the multivariable logistic regression analysis, p values were not used as the sole inclusion criteria as demonstrated above. Authors felt that presence of relief with hot showers was actually of most interest and clinical significance to examine as the main predictor of capsaicin efficacy because presence of this symptom is a hallmark characteristic of CHS and correlates with the potential mechanism of capsaicin for relieving CHS symptoms as described in the literature. Ultimately however, the results from the multivariable logistic regression were not statistically significant. In other words, presence of relief with hot showers was not predictive of capsaicin efficacy. In comparison with the number of patients reporting symptoms of nausea (n = 118), vomiting (n = 135), and abdominal pain (n = 108), the number of patients reporting relief with hot showers was much fewer (n = 48). Small sample size resulting in fewer data points for relief with hot showers may have contributed to the non-significant finding. This result may also suggest that perhaps the mechanism by which capsaicin works for relief of CHS symptoms is not as closely related to relief with hot showers as hypothesized in proposed mechanisms in the literature, or at the very least, not dependent on it.

This study has several limitations. The single-center, retrospective nature of this study limits the ability to draw definitive conclusions and control for confounding variables and missing documentation e.g. pain scores and episodes of emesis (which may have been better surrogate markers of efficacy). Patients who sought care in another facility after being discharged from our ED would also not have been captured in the “Return to ED within 24 hours” outcome. We were also unable to account for medications patients may have received prior to ED presentation or if patients were prescribed, counseled to, or used capsaicin or other agents after discharge. The adolescent subgroup was also likely not large enough to detect any differences in the primary and secondary outcomes. This study also did not specifically examine adverse effects of capsaicin or other medications given although prior research has suggested minimal adverse effects with capsaicin. Chart review for data collection incidentally uncovered notes mentioning patient complaints of local skin irritation or a burning sensation from capsaicin but these were not counted as data points as it was not a uniform method of documentation that could have been efficiently and accurately collected for all patients to reflect on the true rate of such events.

Strengths of this study include the use of a single formulation of capsaicin across all encounters, robust reporting of demographic information, and comprehensive statistical analysis of outcomes. To our knowledge, this study represents the largest study to date assessing capsaicin in the management of CHS which also includes a relatively large number of adolescent patients.

5. Conclusion

This study found a statistically significant difference in patients experiencing efficacy when capsaicin was utilized for management of CHS. Time to discharge after administration of capsaicin was also shorter compared to patients who did not receive capsaicin. Patients presenting to the ED for CHS management with a longer duration of symptoms ≥1 month may be less likely to experience symptom relief with capsaicin than patients presenting with shorter duration of symptoms <1 week. Administration of capsaicin did not appear to influence patients’ total number of medications received or length of ED stay. For these reasons, capsaicin appears to be a reasonable agent to consider for symptoms of CHS, especially early in the ED management course if possible, and at the very least as part of a multi-modal approach to management. Future research is needed to determine capsaicin’s efficacy when utilized earlier in therapy, ideally upon initial diagnosis of CHS and symptom course, and before additional adjunct medications are administered.

Abbreviations:

AMA

Against medical advice

C

Capsaicin group

CHS

Cannabinoid hyperemesis syndrome

ED

Emergency department

NC

No capsaicin (control group)

RA

Reference agent