Implementation of a Standardized Management Algorithm for Cannabis Hyperemesis Syndrome in a Pediatric Emergency Department - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Pediatr Emerg Care . Author manuscript; available in PMC: 2026 Apr 22. Published in final edited form as: Pediatr Emerg Care. 2026 Apr 20;42(7):535–543. doi: 10.1097/PEC.0000000000003588 Search in PMC Search in PubMed View in NLM Catalog Add to search Implementation of a Standardized Management Algorithm for Cannabis Hyperemesis Syndrome in a Pediatric Emergency Department Carlton M Kelly Carlton M Kelly 1 Northwestern University Feinberg School of Medicine, Chicago, IL Find articles by Carlton M Kelly 1 , Megan Attridge Megan Attridge , MD, MS 2 Division of Emergency Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL 3 Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL Find articles by Megan Attridge 2, 3 , Yiannis L Katsogridakis Yiannis L Katsogridakis , MD, MPH 2 Division of Emergency Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL 3 Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL Find articles by Yiannis L Katsogridakis 2, 3 , Lauren Mrozek Lauren Mrozek , MSN, RN, CPN, CPEN 4 Department of Nursing, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL Find articles by Lauren Mrozek 4 , Kathryn Jackson Kathryn Jackson , MS 5 Department of Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL Find articles by Kathryn Jackson 5 , Jennifer A Hoffmann Jennifer A Hoffmann , MD 2 Division of Emergency Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL 3 Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL 5 Department of Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL Find articles by Jennifer A Hoffmann 2, 3, 5 Author information Article notes Copyright and License information 1 Northwestern University Feinberg School of Medicine, Chicago, IL 2 Division of Emergency Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL 3 Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL 4 Department of Nursing, Ann & Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL 5 Department of Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL Contributions: Carlton Kelly substantially contributed to the conception and design of the work; the acquisition, analysis, and interpretation of data; and drafted the work. Dr. Jennifer Hoffmann provided substantial contributions to the conception and design of the work; the acquisition and interpretation of data; and reviewed the work critically for important intellectual content. Dr. Megan Attridge, Dr. Yiannis Katsogridakis, Kathryn Jackson, and Lauren Mrozek substantially contributed to interpretation of data and reviewed the work critically for important intellectual content. All authors gave final approval of the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. ✉ Corresponding Author: Carlton M. Kelly, [email protected] , Street address: 251 East Huron Street, Chicago, IL 60611, Fax: (732)-758-8156 Issue date 2026 Jul 1. PMC Copyright notice PMCID: PMC13098712 NIHMSID: NIHMS2133896 PMID: 42003265 The publisher's version of this article is available at Pediatr Emerg Care Abstract Objectives: Cannabis hyperemesis syndrome (CHS) is an increasingly common cause of pediatric emergency department (ED) visits, yet management pathways remain understudied. We evaluated clinical outcomes following the implementation of a CHS management algorithm in a pediatric ED. Methods: We conducted a retrospective study of encounters by adolescents before and after implementation of a CHS management algorithm in an academic pediatric ED from July 2020-July 2024. We examined medications administered, length of stay, disposition, and return visits before and after implementation using chi-square, Fisher’s exact, or Mann-Whitney U tests as appropriate. We used mixed-effects models to examine the association of time period and admission rates, adjusting for age, sex, and emergency severity index level. A similar model examined the association of time period and ED length of stay that was also adjusted for total daily ED arrivals. Results: Of 533 screened encounters, 128 met inclusion criteria, representing 44 unique patients. Following algorithm implementation, administration rates increased for capsaicin (2.7% vs. 22.2%, p<0.001) and metoclopramide (6.8% vs. 42.6%, p<0.001). Frequency of haloperidol administration did not change significantly (20.3% vs. 9.3%, p=0.138), but mean dose decreased (2.7 mg vs. 1.0 mg, p=0.014). The adjusted odds of hospital admission (adjusted OR: 0.57, 95% CI: 0.17, 1.86) and ED length of stay (adjusted beta: −0.01, 95% CI: −0.21, 0.20) did not significantly differ. Conclusions: Algorithm implementation was associated with increased capsaicin and metoclopramide use but no change in admission rates or length of stay. Prospective studies are needed to assess optimal CHS management in children. Introduction As of 2025, 39 states and the District of Columbia (D.C.) have enacted medical cannabis laws, allowing providers to recommend cannabis for specified health conditions, while 24 states and D.C. have legalized recreational cannabis for adult use.( 1 ) Cannabis-related hospitalizations have increased in children’s hospitals, especially in states where recreational cannabis laws have been enacted.( 2 ) A growing proportion of these cannabis-related encounters appear to be driven by cannabis hyperemesis syndrome (CHS), a condition increasingly recognized as a distinct cause of recurrent vomiting and abdominal pain in young people with chronic cannabis use. In the United States, CHS-related visits among individuals aged 15–24 years rose from approximately 7 per million in 2006 to over 170 per million in 2020.( 3 ) Among adolescents aged 13–21 years, a separate analysis reported more than a tenfold increase in CHS-related emergency visits between 2016 and 2023, with a mean yearly rise of nearly 50%.( 4 ) CHS was first described in 2004.( 5 ) The ROME IV criteria define the major diagnostic features of CHS in adults as: regular cannabis use for at least three months, recurrent episodes of nausea/vomiting resembling chronic vomiting syndrome, and relief of vomiting episodes by sustained cessation of cannabis use. Supporting criteria include symptomatic relief with hot showers, weight loss, abdominal pain, and changes in bowel habits.( 6 – 8 ) Evidence to guide CHS treatment in younger patients remains limited and is largely drawn from adult studies. In adults, a triple-blind randomized trial found haloperidol to be more effective than ondansetron for relieving nausea and abdominal pain and for shortening ED stays.( 9 ) Similar findings were reported in a prospective cohort study of adult patients, in which droperidol combined with diphenhydramine reduced nausea and vomiting.( 10 ) Although a mixed adult-pediatric retrospective cohort study observed that capsaicin use was associated with fewer rescue medications, no significant association was observed among patients under 21 years old.( 11 ) Even so, pediatric case series have described improvement with topical capsaicin,( 12 ) and one retrospective single-center analysis found that nontraditional antiemetics, including benzodiazepines, haloperidol, and capsaicin, were more effective than traditional regimens for symptom control in children with CHS.( 13 ) As evidence to inform optimal CHS management for children continues to grow, management strategies remain varied. Care pathways and algorithms have successfully standardized care and improved outcomes for other pediatric emergency conditions;( 14 , 15 ) however, existing CHS algorithms show modest variability in medication sequencing and escalation. Many pathways recommend initial laboratory evaluation, urinalysis, pregnancy testing when indicated, electrocardiogram assessment, and intravenous fluid resuscitation. The Johns Hopkins All Children’s Hospital pathway begins with ondansetron and a single oral challenge, followed by haloperidol or droperidol with diphenhydramine for persistent symptoms,( 16 ) and the Children’s Minnesota guideline lists several agents as first-line, including droperidol, aprepitant, olanzapine, diphenhydramine, topical capsaicin, lorazepam, and ondansetron.( 17 ) A similar approach has been proposed in a recent review, beginning with ondansetron and progressing to haloperidol with diphenhydramine when QTc is normal or lorazepam when prolonged.( 18 ) Our study describes implementation of a stepwise algorithm in a pediatric ED, beginning with intravenous fluids and either ondansetron or the combination of metoclopramide and diphenhydramine. If symptoms persisted, patients were given lorazepam, followed by topical capsaicin and, if refractory, intravenous haloperidol. This study aims to compare medication administration patterns and patient outcomes among adolescents presenting to a pediatric ED with symptoms consistent with CHS before and after the implementation of a CHS management algorithm. Materials and Methods Study design and setting We conducted a retrospective, cross-sectional study of ED visits by patients 13–18 years old with symptoms consistent with CHS at an academic children’s hospital from July 20, 2020, to July 19, 2024. The ED has approximately 60,000 annual visits. An algorithm for the ED management of CHS, featuring treatment recommendations was implemented on July 20, 2022. The algorithm was developed through multidisciplinary consensus among pediatric emergency medicine physicians, pharmacists, and toxicology consultants, using available literature, institutional experience, and safety considerations to guide medication selection and sequencing. Figure 1 outlines the algorithm’s approach to managing CHS for patients ≥13 years old weighing ≥45kg, beginning with intravenous fluids and either ondansetron (0.15 mg/kg, maximum 8 mg) or the combination of metoclopramide (0.15 mg/kg, maximum 10 mg) and diphenhydramine (1 mg/kg, maximum 50 mg). If symptoms persisted, the next steps involved reassessment every 30–60 minutes, followed by administration of lorazepam (0.025 mg/kg, maximum 2 mg). If needed, topical capsaicin (0.025% in a thin layer) was applied, and in cases of refractory symptoms, intravenous haloperidol (1 mg) was administered. Figure 1: Open in a new tab ED cannabis hyperemesis algorithm The study was reviewed and deemed exempt by the hospital institutional review board. Selection of encounters Potentially eligible encounters were identified using three methods: 1) presence of both “GI symptoms” and “cannabis use,” which were defined using chief complaints, diagnosis codes, specific keywords in ED provider notes, and laboratory results, 2) presence of specific keywords related to cannabis hyperemesis in ED provider notes, or 3) the CHS ED order set was activated. See Table , Supplemental Digital Content 1 for additional details about the case identification strategy. Encounters were excluded if they involved a positive pregnancy test, a negative tetrahydrocannabinol (THC) urine test, or a complex chronic condition based on specific diagnosis codes.( 19 ) Potentially eligible encounters were manually reviewed to determine if they met inclusion criteria. Encounters were included if there was documentation in the ED provider note of a clinical diagnosis of CHS or cyclical vomiting patterns combined with frequent cannabis use. Frequent use was defined as either self-reported use of at least three days per week or two positive urine THC tests within three months.( 20 ) Encounters were excluded if the treating clinician documented high suspicion for an alternative diagnosis inconsistent with CHS or if the patient was transferred to another facility. Data Abstraction Procedures Following best practices for electronic health record reviews, a detailed manual of operations with standardized definitions was developed to guide data abstraction.( 21 , 22 ) Data were abstracted into a centralized, standardized REDCap database.( 23 ) A second investigator reviewed 10% of medical records to assess interrater reliability. Variable Definitions For each unique patient, we recorded demographics, substance use (cannabis and tobacco), and mental health history. For each encounter, we recorded presenting symptoms, Emergency Severity Index (ESI) acuity level, total patient arrivals on the encounter day and outcomes. The primary outcome was patient disposition. Secondary outcomes included medications administered, adverse events to medications, ED length of stay, and return visits within 24 and 72 hours. Analysis Encounter-level characteristics were summarized using frequencies (percentages) and medians (IQR) for categorical and continuous variables, for the overall sample and by time period (before versus after algorithm implementation). The first ED encounter for each patient during the study period was considered the index encounter. To avoid over-representation of stable patient characteristics, descriptive analyses were limited to each individual’s index encounter. All encounters were included in regression models to account for repeated visits and evaluate outcomes across encounters. Patient and clinical characteristics during the index visit were compared between time period groups. Run charts were constructed to display quarterly trends in key operational measures over the study period. These included total patient encounters, medication administration rates, admission rates, and ED length of stay. Differences in baseline characteristics, clinical presenting features and in primary and secondary outcomes between the pre-implementation and post-implementation periods were analyzed using χ 2 tests (or Fisher’s exact tests) for categorical variables and Mann-Whitney U tests for continuous variables. A mixed-effects logistic regression model was used to evaluate the relationship between time period and disposition, adjusted for age, sex, and ESI. A generalized linear model with a gamma distribution and log link function evaluated the relationship between time period and length of stay, adjusted for age, sex, ESI, and total daily ED arrivals. For both models, the subject identifier was included as a random effect to account for within-subject correlations from repeated measures. Categories of covariates representing <5% of the study population were collapsed in the models. Model fit was evaluated using Akaike Information Criterion. Likelihood ratio tests were used to compare nested models and determine whether the inclusion of additional predictors significantly improved model performance. Results were reported as unadjusted and adjusted odds ratios and beta coefficients with 95% confidence intervals. A sensitivity analysis was conducted among patients in the post-implementation cohort comparing encounters with confirmed CHS order set use versus those without, assessing differences in admission rates, emergency department length of stay, and the frequency of administered medications. A p-value of <0.05 was considered statistically significant. Analyses were conducted using R version 4.4.2 (R Core Team, Vienna, Austria). Results Characteristics of Study Sample Of 533 screened encounters, 128 encounters met inclusion criteria, representing 44 unique patients. The number of CHS-related encounters per patient ranged from 1 to 13 (median 1) during the study period, with the distribution displayed in Supplemental Digital Content 2 . The interrater reliability for data abstraction had a kappa value of 0.946. Among the 44 index encounters, the median age was 17 (range 14–18) and 61.4% were female ( Table 1 ). A history of anxiety was documented in 20.5% of index encounters, while a history of depression was noted in 18.2%. Daily cannabis use was reported in 25.0%, and positive urine THC tests were completed in 36.4%. Index encounters before and after algorithm implementation had similar characteristics, except for tobacco/nicotine use. Figure 2 presents quarterly trends in CHS patient encounters. Supplemental Digital Content 3 presents encounter-level characteristics before and after algorithm implementation. Table 1: Characteristics of Index Encounters with Cannabis Hyperemesis Syndrome Before and After Algorithm Implementation Overall N (%) Visits Before Algorithm Implementation (7/20/2020–7/19/2022) N (%) Visits After Algorithm Implementation (7/20/2022–7/19/2024) N (%) p value a Overall N=44 N=25 N=19 Age group 13–14 years old 2 (4.6) 1 (4.0) 1 (5.3) 0.115 15–16 years old 14 (31.8) 5 (20.0) 9 (47.4) 17–18 years old 28 (63.6) 19 (76.0) 9 (47.4) Sex Male 17 (38.6) 9 (36.0) 8 (42.1) 0.760 Female 27 (61.4) 16 (64.0) 11 (57.9) Race White 12 (27.3) 6 (24.0) 6 (31.6) 0.951 Black 7 (15.9) 4 (16.0) 3 (15.8) Asian 1 (2.3) 1 (4.0) 0 (0.0) Other/ Multiple races 24 (54.6) 14 (56.0) 10 (52.6) Ethnicity Hispanic 23 (52.3) 21 (47.7) 11 (57.9) 0.557 Non-Hispanic 21 (47.73) 13 (52.0) 8 (42.1) Insurance Payer Private 14 (31.8) 7 (28.0) 7 (36.8) 0.745 Public 30 (68.2) 18 (72.0) 12 (63.2) Language Spoken English 35 (79.6) 21 (84.0) 14 (73.7) 0.467 Spanish 9 (20.5) 4 (16.0) 5 (26.3) Daily Cannabis Use Yes 11 (25) 8 (32.0) 3 (15.8) 0.458 No 20 (45.5) 10 (40.0) 10 (52.6) Unknown 13 (29.6) 7 (28.0) 6 (31.6) Positive Urine THC Test Yes 16 (36.4) 11 (44.0) 5 (26.3) 0.373 Not completed 28 (63.6) 14 (56.0) 14 (73.7) Current Tobacco/ Nicotine Use Yes 3 (6.8) 2 (8.0) 1 (5.3) 0.007 No 11 (25) 2 (8.0) 9 (47.4) Unknown 30 (68.2) 21 (84.0) 9 (47.4) History of Anxiety Yes 9 (20.5) 6 (24.0) 3 (15.8) 0.771 No 35 (79.6) 19 (76.0) 16 (84.2) History of Depression Yes 8 (18.2) 5 (20.0) 3 (15.8) 1 No 8 (18.2) 5 (20.0) 3 (15.8) Open in a new tab a Differences in baseline demographics, substance use and mental health history between the pre-implementation and post-implementation periods were analyzed using χ 2 tests (or Fisher’s exact tests). A p-value < 0.05 was considered statistically significant. N= number of index encounters, tetrahydrocannabinol (THC) Figure 2: Open in a new tab Temporal Trends in Cannabis Hyperemesis Syndrome Patient Encounters Before and After Algorithm Implementation The dashed line indicates the date of algorithm implementation on July 20, 2022. Presenting Features and Outcomes Before and After Algorithm Implementation Nausea and abdominal pain were documented in 73.0% and 89.2% of pre-algorithm encounters and 81.5% and 70.4% of post-algorithm encounters ( Table 2 ). Among encounters where hot shower response was documented, 82.4% of pre-algorithm and 87.5% of post-algorithm cases reported symptom relief. Table 2: Clinical Presentation and Outcome Measures for Emergency Department Encounters Visits Before Algorithm Implementation (7/20/2020–7/19/2022) N (%) Visits After Algorithm Implementation (7/20/2022–7/19/2024) N (%) p value a N=74 N=54 Nausea Yes 54 (73.0) 44 (81.5) 0.166 No 0 (0.0) 1 (1.9) Unknown 20 (27.0) 9 (16.7) Abdominal Pain Yes 66 (89.2) 38 (70.4) 0.025 No 6 (8.1) 11 (20.4) Unknown 2 (2.7) 5 (9.3) Relief with Hot Showers Yes 14 (18.9) 7 (13.0) 0.585 No 3 (4.1) 1 (1.9) Unknown 57 (77.0) 46 (85.2) Emergency Severity Index Level 2 10 (13.5) 6 (11.1) 0.485 Level 3 58 (78.4) 40 (74.1) Level 4 6 (8.1) 8 (14.8) Total Day Arrivals – median (IQR) 98.5 (61.8) 164 (53.2) <0.001 Encounters per Unique Patient – mean (SD) 3.0 (3.3) 2.5 (2.6) 0.419 Percentage of Visits with CHS Order Set Activation - 12 (22.2) Medications Administered Ondansetron 47 (63.5) 29 (53.7) 0.280 Capsaicin 2 (2.7) 12 (22.2) <0.001 Metoclopramide 5 (6.8) 23 (42.6) <0.001 Lorazepam 25 (33.8) 15 (27.8) 0.563 Diphenhydramine 39 (52.7) 29 (53.7) 1 Normal Saline 69 (93.2) 51 (94.4) 1 Haloperidol 15 (20.3) 5 (9.3) 0.138 Haloperidol Intravenous Dosage, mg- mean (SD) 2.7 (1.6) 1 (0.0) 0.014 Adverse Events to Medication 0 (0.0) 0 (0.0) Admission to Hospital 39 (52.7) 24 (44.4) 0.376 ED Length of Stay, min – median (IQR) 244.5 (145.0) 299.0 (174.8) 0.247 Return Visit, b % Within 24 hours 4 (11.4) 2 (6.7) 0.678 Within 72 hours 7 (20.0) 4 (13.3) 0.526 Open in a new tab a Differences between time periods were analyzed using χ 2 tests (or Fisher’s exact tests) for categorical variables and Mann-Whitney U tests for continuous variables. b Among encounters with the disposition of discharged, pre-algorithm group N=35, post-algorithm group N=30. N= Number of encounters, IQR= interquartile range, SD= standard deviation, CHS= cannabis hyperemesis syndrome Figure 3 presents quarterly trends in medication administration rates. Use of capsaicin and metoclopramide increased significantly following implementation (capsaicin: 2.7% vs. 22.2%, p < 0.001; metoclopramide: 6.8% vs 42.6%, p<0.001). Rates of haloperidol administration did not differ significantly between periods (20.3% vs. 9.3%, p = 0.138), although the mean intravenous dose decreased after implementation (2.7 vs. 1.0 mg, p = 0.014). There were no adverse events to medications documented in the clinician notes during the study period. Figure 3: Open in a new tab Quarterly Trends in Medication Administration for Cannabis Hyperemesis Syndrome Encounters Proportion of encounters in which each medication was administered, shown by quarter. The dashed line marks algorithm implementation in July 20, 2022. Admission rates did not significantly differ (52.7% vs. 44.4%, p=0.376). Median ED length of stay did not significantly differ (299 vs. 245 minutes, p=0.247). Among encounters resulting in discharge, there were no significant changes in return visits within 24 and 72 hours. Quarterly trends in admission rates and ED length of stay are presented in Supplemental Digital Content 4 and 5 . Association of Time Period and Admission Rate After algorithm implementation, the adjusted odds of admission were not significantly different compared to before implementation (adjusted OR: 0.57, 95% CI 0.17, 1.86) ( Table 3 ). Encounters categorized as ESI Level 2 had significantly higher adjusted odds of admission compared to encounters with ESI Level 4 (adjusted OR: 8.28, 95% CI: 1.00, 68.45). Table 3: Multivariable Logistic Regression Model for Hospital Admission Rates Unadjusted Odds Ratio for Hospital Admission (95% CI) Adjusted Odds Ratio for Hospital Admission (95% CI) Time Period Post-Algorithm 0.50 (0.15, 1.69) 0.57 (0.174, 1.86) Pre-Algorithm Reference Reference Age 13–15 0.10 (0.01, 1.18) 0.12 (0.01, 1.49) 16–18 Reference Reference Sex Male 0.28 (0.08, 0.92) 0.35 (0.10, 1.25) Female Reference Reference Emergency Severity Index Level 2 15.21 (1.97, 117.73) 8.28 (1.00, 68.45) Level 3 3.27 (0.70, 15.17) 2.19 (0.44, 10.99) Level 4 Reference Reference Open in a new tab a Adjusted odds ratio of admission in the post-algorithm group controlled for age, sex, and emergency severity index. CI= confidence interval In the adjusted model, no significant association was observed between time period and ED length of stay, after adjusting for age group, sex, ESI levels, and total daily arrivals ( Table 4 ). Table 4: Generalized Linear Regression for Emergency Department Length of Stay Unadjusted Beta Coefficient (95% CI) Adjusted Beta Coefficient (95% CI) Time Period Post-Algorithm 0.09 (−0.07, 0.25) −0.01 (−0.21, 0.20) Pre-Algorithm Reference Reference Age 13–15 0.24 (−0.10, 0.59) 0.22 (−0.13, 0.56) 16–18 Reference Reference Sex Male 0.04 (−0.14, 0.23) 0.07 (−0.13, 0.26) Female Reference Reference Emergency Severity Index Level 2 0.13 (−0.20, 0.46) 0.16 (−0.19, 0.51) Level 3 0.13 (−0.12, 0.39) 0.162 (−0.11, 0.43) Level 4 Reference Reference Total Day Arrivals 0.002 (0.00, 0.003) 0.002 (−0.001, 0.004) Open in a new tab a Length of stay was measured from rooming time to disposition. The beta coefficient represents the natural logarithm of the proportional change in length of stay for a one-unit increase in the predictor. b Adjusted odds ratios for admission in the post-algorithm group were controlled for age, sex, emergency severity index, and total day arrivals. CI= confidence interval Sensitivity Analysis: Post-Implementation Encounters by Order Set Activation In the post-implementation cohort, 12 encounters (22.2%) included activation of the CHS order set. Admission rates did not significantly differ between encounters with and without order set use (46.2% vs. 42.9%, p = 0.834). Median ED length of stay was 351 minutes (IQR: 266–426) among encounters with order set activation and 272 minutes (IQR: 201–340) among those without (p = 0.127). Use of lorazepam was significantly higher among encounters with order set activation (61.5% vs. 16.7%, p = 0.003). Rates of capsaicin (38.5% vs. 16.7%, p = 0.129), metoclopramide (61.5% vs. 35.7%, p = 0.099), haloperidol (23.1% vs. 7.1%, p = 0.136), diphenhydramine (69.2% vs. 47.6%, p = 0.173), and ondansetron (46.2% vs. 57.1%, p = 0.487) administration were not significantly different between groups. Discussion To our knowledge, this study is the first to evaluate clinical outcomes after implementing a CHS-specific management algorithm for children. Most notably, the implementation of the algorithm was linked to significant shifts in medication administration patterns, with increases in capsaicin and metoclopramide administration. The clinical and demographic characteristics observed in our study sample align closely with prior studies. Our sample had a female predominance of 61.4%, consistent with prior reports ranging from 62% to 68.4%.( 7 , 13 , 20 , 24 ) This gender difference is notable compared to adult studies, which often report a predominance of male patients.( 7 , 18 , 25 ) We found that females and older adolescents had a higher frequency of multiple visits for CHS during the study period. Additionally, 20.5% of adolescents had a history of anxiety and 18.2% had a history of depression, which are similar to general population estimates and consistent with previous research on adolescent CHS populations.( 7 , 26 ) Traditional antiemetics were commonly administered in both groups. In the pre-algorithm group, ondansetron was the most frequently administered, in approximately two-third of encounters. In the post-algorithm group, ondansetron was administered in just over half of encounters. Despite their frequent use, traditional antiemetics have been shown to be less effective for CHS symptom resolution.( 13 , 27 ) In this algorithm, ondansetron was added as a first line treatment not based on efficacy but rather a nurse-driven triage protocol which was standardized for undifferentiated nausea and vomiting. Among the nontraditional antiemetics captured in this study (benzodiazepines, haloperidol, and topical capsaicin), capsaicin administration increased significantly from 2.7% to 22.2% of encounters after algorithm implementation. In case reports, capsaicin has been reported to have good efficacy with few serious side effects, suggesting it may be a favorable option for CHS treatment.( 12 , 27 ) Haloperidol has been associated with improved CHS symptoms compared to traditional antiemetics in a previous pediatric study as well;( 13 ) however, its use carries an increased risk of QT prolongation. This risk is particularly important to consider in patients with hyperemesis, as associated electrolyte imbalances such as hypokalemia and hypomagnesemia may increase the risk of cardiac arrhythmias.( 9 , 13 ) Following algorithm implementation, the frequency of haloperidol administration did not change significantly, but the average haloperidol dose significantly decreased from 2.7 mg to 1.0 mg. The reduction in average dose of haloperidol could minimize the risk of potential side effects, although the optimal dose for children that balances efficacy and safety remain unknown. Following algorithm implementation, the hospital admission rate decreased from 52.4% to 44.4%, but this difference was not statistically significant, including in models that adjusted for other encounter-level characteristics. There are several potential explanations for this finding. First, it is possible that the medications included on the algorithm are not effective for children with CHS. Second, fewer than half of encounters after algorithm implementation involved administration of each nontraditional antiemetic (lorazepam, capsaicin, and haloperidol). Thus, one or more of these medications might be effective in children with CHS but was not administered frequently enough in the study to shift outcomes. Third, the lack of significance might be attributable to insufficient power, if the magnitude of effect of these medications was too small to be detected with the included sample size. Larger prospective studies are needed to test the effectiveness of medications for CHS in children with respect to symptom control and ability to prevent admissions. In our study, implementation of the algorithm had no significant impact on visit length, suggesting that the availability of a multi-step treatment pathway did not prolong time to disposition decisions. The CHS management algorithm was formally introduced during a division-wide faculty meeting, where inclusion and exclusion criteria, as well as therapeutic components of the order set, were reviewed. Ongoing dissemination efforts included reinforcement of the guideline during nursing huddles and resident team discussions to promote consistent application in clinical practice. Despite these efforts, the CHS order set was activated in only 22.2% of post-implementation encounters, suggesting limited uptake during the early implementation phase. In our sensitivity analysis, encounters with order set activation admission rates did not significantly differ for encounters with versus without order set activation. Although the median ED length of stay was numerically longer among order set users (351 vs. 272 minutes), this difference was not statistically significant. The trend may nonetheless suggest differences in treatment sequencing and response. Lorazepam use was substantially higher among order set users (61.5% vs. 16.7%), reflecting the algorithm’s structure, which places haloperidol as a fourth-line option after lorazepam. Emerging evidence, largely from studies in adult populations, suggests that haloperidol and benzodiazepines may be more effective than traditional antiemetics.( 9 , 13 ) Several newly published pediatric CHS pathways support this approach, placing haloperidol or droperidol ahead of benzodiazepines in the treatment sequence.( 16 – 18 ) Finally, the standardized haloperidol dose of 1 mg in our algorithm may have been insufficient for some patients, as weight-based dosing of 0.05 mg/kg is more commonly used in prior pediatric studies.( 6 , 9 , 17 , 18 , 24 ) This pattern reflects consensus-based practice rather than evidence from high quality pediatric trials. Additional studies are needed to establish the optimal pediatric haloperidol dose for CHS. The absence of validated pediatric-specific diagnostic criteria for CHS continues to present challenges for early recognition and treatment, as most existing frameworks, including the ROME IV criteria, are derived from adult populations.( 28 ) While our study was not designed to establish new diagnostic criteria, these limitations underscore the broader need for age-appropriate validation of existing definitions to improve clinical recognition and management in younger patients. Despite its contributions, this study has several limitations. As a retrospective chart review, it is subject to biases, such as the potential underdiagnosis of CHS due to incomplete documentation of cannabis use patterns or prior episodes of emesis. Differentiating CHS from cyclic vomiting syndrome remains a challenge, which may have influenced case ascertainment. Additionally, repeat encounters occurring at other hospitals were not captured in this study, which could have resulted in underestimation of the frequency of repeat visits. We were unable to assess the effectiveness of specific treatments on symptom resolution. No medication-related adverse events were documented, though events may have occurred without corresponding ED clinician documentation or may have occurred after admission, potentially contributing to underreporting. Lastly, the study period overlapped with broader contextual factors, including the COVID-19 pandemic and legalization of recreational cannabis in Illinois (effective January 1, 2020), which may have influenced cannabis use patterns, ED presentation rates, and clinical recognition of CHS independent of the algorithm. These unmeasured confounders may have contributed to changes in encounter frequencies and diagnostic recognition during the study period. This study offers valuable insights into the role of a CHS-specific management algorithm in pediatric emergency care, addressing a growing yet underexplored condition. By highlighting meaningful changes in clinical practice, the findings underscore the potential of standardized care pathways to change management strategies. Moving forward, larger, multicenter studies are necessary to refine diagnostic criteria, test the effectiveness of treatments, and explore the implications of standardized guidelines on care delivery. The development and evaluation of tailored pediatric CHS guidelines will be critical to guide management strategies and ultimately improve outcomes. Supplementary Material Supplemental Material NIHMS2133896-supplement-Supplemental_Material.docx (360.7KB, docx) List of Supplemental Digital Content Supplemental Digital Content 1. (.doc) Supplemental Digital Content 2. (.doc) Supplemental Digital Content 3. (.doc) Supplemental Digital Content 4. (.doc) Supplemental Digital Content 5. (.doc) Acknowledgements: JAH receives support from the National Institute of Mental Health of the National Institutes of Health under Award Number K23MH135206–01. The funders had no role in the design and conduct of the study. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Abbreviations: CHS cannabis hyperemesis syndrome ED emergency department THC delta-9-tetrahydrocannabinol ESI emergency severity index Footnotes Disclosures: The authors declare no conflicts of interest or financial disclosures related to this study. References 1. Marijuana Legality by State: DISA; 2024. [cited 2024 Oct 01]. Available from: https://disa.com/marijuana-legality-by-state . [ Google Scholar ] 2. Masonbrink AR, Richardson T, Hall M, Catley D, Wilson K. Trends in Adolescent Cannabis-Related Hospitalizations by State Legalization Laws, 2008–2019. J Adolesc Health. 2021;69(6):999–1005. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Jack B, Susi A, Reeves P, Nylund CM. Increasing trends of cannabinoid hyperemesis syndrome in youth: The grass is not always greener. J Pediatr Gastroenterol Nutr. 2025;80(4):638–43. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Toce MS, Monuteaux MC, Fishman MD, Hudgins JD. Emergency Department Visits for Cannabis Hyperemesis Syndrome Among Adolescents. JAMA Netw Open. 2025;8(7):e2520492. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Allen JH, de Moore GM, Heddle R, Twartz JC. Cannabinoid hyperemesis: cyclical hyperemesis in association with chronic cannabis abuse. Gut. 2004;53(11):1566–70. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Lonsdale H, Wilsey MJ. Paediatric cannabinoid hyperemesis. Curr Opin Pediatr. 2022;34(5):510–5. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Zhu JW, Gonsalves CL, Issenman RM, Kam AJ. Diagnosis and Acute Management of Adolescent Cannabinoid Hyperemesis Syndrome: A Systematic Review. J Adolesc Health. 2021;68(2):246–54. [ DOI ] [ PubMed ] [ Google Scholar ] 8. Drossman DA, Hasler WL. Rome IV-Functional GI Disorders: Disorders of Gut-Brain Interaction. Gastroenterology. 2016;150(6):1257–61. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Ruberto AJ, Sivilotti MLA, Forrester S, Hall AK, Crawford FM, Day AG. Intravenous Haloperidol Versus Ondansetron for Cannabis Hyperemesis Syndrome (HaVOC): A Randomized, Controlled Trial. Ann Emerg Med. 2021;77(6):613–9. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Chopra Q, Peyko V, Lee JA, Puhalla L, Gemmel DJ, Bolotin T. Droperidol Plus Diphenhydramine for Symptom Improvement in Suspected Cannabinoid Hyperemesis Syndrome: A Prospective Cohort Study. Open Access Emerg Med. 2024;16:267–73. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Kum V, Bell A, Fang W, VanWert E. Efficacy of topical capsaicin for cannabinoid hyperemesis syndrome in a pediatric and adult emergency department. Am J Emerg Med. 2021;49:343–51. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Graham J, Barberio M, Wang GS. Capsaicin Cream for Treatment of Cannabinoid Hyperemesis Syndrome in Adolescents: A Case Series. Pediatrics. 2017;140(6). [ Google Scholar ] 13. Geraci E, Cake C, Mulieri KM, Fenn NE 3rd. Comparison of Antiemetics in the Management of Pediatric Cannabinoid Hyperemesis Syndrome. J Pediatr Pharmacol Ther. 2023;28(3):222–7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Hoffmann JA, Pergjika A, Liu L, Janssen AC, Walkup JT, Johnson JK, et al. Standardizing and Improving Care for Pediatric Agitation Management in the Emergency Department. Pediatrics. 2023;152(1). [ Google Scholar ] 15. Chen KH, Chen CC, Liu HE, Tzeng PC, Glasziou PP. Effectiveness of paediatric asthma clinical pathways: a narrative systematic review. J Asthma. 2014;51(5):480–92. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Wilsey M, Dees P. Cannabinoid Hyperemesis Syndrome Clinical Pathway: Johns Hopkins All Children’s Hospital; 2025. [cited 2025 Nov 8]. Available from: https://www.hopkinsmedicine.org/-/media/files/allchildrens/clinical-pathways/chs_pathway-2025-08-18.pdf . [ Google Scholar ] 17. Maccabee-Ryaboy N, Puthoor P, Johnson J, Montague A, Weinstein C, Wu B, et al. Cannabinoid Hyperemesis Syndrome (CHS): Evaluation and Emergency Department Management (Age 12+): Children’s Minnesota; 2025. [cited 2025 Nov 8]. Available from: https://www.childrensmn.org/references/CDS/cannabinoid-hyperemesis-syndrome-clinical-pathway.pdf . 18. Ibia IE, Toce MS. Cannabis Hyperemesis Syndrome in Children: A Review of Epidemiology, Pathology, Diagnosis, and Treatment. Pediatr Emerg Care. 2025;41(5):397–405. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr. 2014;14:199. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Shah M, Jergel A, George RP, Jenkins E, Bashaw H. Distinguishing Clinical Features of Cannabinoid Hyperemesis Syndrome and Cyclic Vomiting Syndrome: A Retrospective Cohort Study. J Pediatr. 2024;271:114054. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Kaji AH, Schriger D, Green S. Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies. Ann Emerg Med. 2014;64(3):292–8. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med. 1996;27(3):305–8. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–81. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Lonsdale H, Kimsey KM, Brown JM, Dey A, Peck J, Son S, et al. Pediatric Cannabinoid Hyperemesis: A Single Institution 10-Year Case Series. J Adolesc Health. 2021;68(2):255–61. [ DOI ] [ PubMed ] [ Google Scholar ] 25. Sorensen CJ, DeSanto K, Borgelt L, Phillips KT, Monte AA. Cannabinoid Hyperemesis Syndrome: Diagnosis, Pathophysiology, and Treatment-a Systematic Review. J Med Toxicol. 2017;13(1):71–87. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Substance Abuse and Mental Health Services Administration. 2021 NSDUH annual national report. 2023. 27. Reinert JP, Niyamugabo O, Harmon KS, Fenn NE 3rd. Management of Pediatric Cannabinoid Hyperemesis Syndrome: A Review. J Pediatr Pharmacol Ther. 2021;26(4):339–45. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Quiros JA, Saps M. The Coming Storm: Cannabis Hyperemesis Syndrome in Adolescents. J Adolesc Health. 2021;68(2):223–4. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplemental Material NIHMS2133896-supplement-Supplemental_Material.docx (360.7KB, docx) ACTIONS View on publisher site PDF (1.2 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top