Objectives Cannabis is a widely used illicit substance that has been associated with acute injuries. This study seeks to provide near real-time injury estimates related to cannabis and other substance use from the electronic Canadian Hospitals Injury Reporting and Prevention Program (eCHIRPP) database. Methods Data from the eCHIRPP database, years 2011 to 2016, were analyzed via data mining, descriptive, logistic regression, and sensitivity analyses. Drug use trends over time for cannabis and/or other substances (alcohol, illicit drugs, and medications) were assessed. Descriptive statistics (intent, external cause, and nature of injury) and proportionate injury ratios (PIR) associated with cannabis use are presented. Results Cannabis use was observed in 184 cases/100,000 eCHIRPP cases, and related injuries were mostly identified as unin- tentional (66.8%). Poisoning (68.5%) and intoxication (69.4%) were the external cause and nature of injury most associated with these events, and hospitalization was recorded for 14.3% of cases. Per 100,000 eCHIRPP cases, cannabis was used alone in 72.4 cases, and in combination with alcohol, illicit drugs, or medications in 74.6 cases, 11.3 cases, and 7.9 cases, respectively. Relative to non-use, the PIR of hospitalization was not significant for cannabis-only users of either sex (males: PIR 1.0, 95% CI 0.6–1.7, females: PIR 0.9, 95% CI: 0.5–1.7). Conclusion Cannabis use injuries are rare, but can occur when cannabis is used with or without other substances. As Canada considers legislative changes, our finding of cases related to unintentional injury, poisoning, and intoxication suggests areas that might benefit from health literacy efforts. Résumé Objectifs Le cannabis, une substance illicite largement consommée, est associé à des blessures aiguës. Notre étude vise à présenter en quasi temps réel les blessures estimatives liées à la consommation de cannabis et d’autres substances d’après la base de données de la plateforme électronique du Système canadien hospitalier d’information et de recherche en prévention des traumatismes (eSCHIRPT). Méthode Nous avons analysé les données de la base eSCHIRPT pour les années 2011 à 2016 au moyen de techniques de forage de données et d’analyses descriptives, de sensibilité et de régression logistique. Nous avons estimé les tendances de la consommation de cannabis avec ou sans autres substances (alcool, drogues et médicaments) au fil du temps. Nous présentons les statistiques descriptives (intention, cause externe et nature de la blessure) et les rapports proportionnels de blessures (RPB) associés à la consommation de cannabis. Résultats La consommation de cannabis a été observée dans 184 pour 100,000 cas dans eSCHIRPT, et les blessures associées étaient principalement non intentionnelles (66,8%). L’empoisonnement (68,5%) et l’intoxication (69,4%) étaient la cause externe * Deepa P. Rao email@example.com Surveillance and Epidemiology Division, Centre for Surveillance and Applied Research, Public Health Agency of Canada, Rm 707B1, 785 Carling Avenue, Ottawa, ON K1A 0K9, Canada Cannabis Legalization and Regulation Secretariat, Health Canada, Ottawa, ON, Canada 156 Can J Public Health (2018) 109:155–163 et la nature de la blessure les plus souvent associées à ces épisodes, et une hospitalization a été enregistrée dans 14,3% des cas. Pour 100,000 cas dans eSCHIRPT, le cannabis avait été consommé seul dans 72,4 cas, et en combinaison avec de l’alcool, de la drogue ou des médicaments dans 74,6, 11,3 et 7,9 cas, respectivement. Par rapport à la non-consommation, les RPB de l’hospitalization n’étaient pas significatifs pour les consommateurs et consommatrices de cannabis seul (hommes: RPB 1,0, IC de 95%: 0,6–1,7; femmes: RPB 0,9, IC de 95%: 0,5–1,7). Conclusion Les blessures dues à la consommation de cannabis sont rares, mais elles peuvent se produire, que le cannabis soit consommé seul ou avec d’autres substances. Étant donné les modifications législatives envisagées au Canada, nos constatations sur les cas de blessures, d’empoisonnements et d’intoxications involontaires indiquent que des efforts pour renforcer l’information en matière de santé sur ces aspects pourraient être bénéfiques. . . . . Keywords Cannabis Substance use Injury Poisoning Emergency department . . . . Mots-clés Cannabis Consommation de substances Lésion Intoxication Service d’urgence Introduction Washington (Azofeifa et al. 2016;Reed 2016;Northwest High Intensity Drug Trafficking Area 2016). In advance of Cannabis (marijuana) is the most commonly used illicit drug any similar proposed legislation to legalize cannabis for worldwide (United Nations Office on Drugs and Crime 2016). non-medical purposes in Canada (Task Force on Cannabis use is prevalent in Canada; in 2015, 12% of the Marijuana Legalization and Regulation 2016), this study population aged 15 and older reported past-year cannabis aims to provide a baseline description of injuries related use (Health Canada 2015), up from 11% in 2013, and 41% to cannabis and other substance use using data from the reported having used it at least once in their lifetime (Health electronic Canadian Hospitals Injury Reporting and Canada 2013; Health Canada 2014). Among children in Prevention Program (eCHIRPP) database. grades 7 to 12, cannabis had the highest prevalence of use after alcohol with nearly 17% reporting use in the year pre- Methods ceding the survey (Health Canada 2014). A majority of these children perceived regular cannabis use as associated with Data source Bgreat risk^ (58%), while fewer (25%) identified irregular use as being of similar risk (Health Canada 2014). Among children younger than 12 years, there have been reports of The eCHIRPP (Crain et al. 2016) is a dynamic web-based unintentional exposures (Wang et al. 2013). injury and poisoning surveillance system currently operating Acute ingestion of cannabis may result in time-limited cog- in 11 pediatric and 6 general emergency departments (ED) nitive, perceptual, and psychomotor perturbations which, across Canada. It has been used to examine a number of health when combined with physical activities such as driving, have issues (Kang et al. 2013; McFaull et al. 2016; Health been linked to mild, serious, or even potentially fatal physical Surveillance and Epidemiology Division, Public Health injuries (e.g., bruises, fractures, concussions, death) (Hall Agency of Canada 2006). Patients’ accounts of pre-injury 2015; Hall and Degenhardt 2014; Volkow et al. 2014). Even circumstances (narratives of Bwhat went wrong^) are collect- though studies have been inconsistent in suggesting a causal ed using the Injury Reporting form, a questionnaire completed relationship between cannabis consumption and injury (Vitale during their visits to the ED. The attending physician, or other and van de Mheen 2006;Elvik 2013;Mura etal. 2003), asso- hospital staff, adds clinical data to the form and data coders ciations have nevertheless been reported with motor vehicle extract other information found in patients’ narratives. crashes (MVC) (Fischer et al. 2016; Ramaekers et al. 2004), Consequently, the eCHIRPP captures a broader assessment interpersonal violence (Copeland et al. 2013), neighbourhood of an injury event, one that includes risk and protective factors crime (de Looze et al. 2015), self-harm (Silins et al. 2014), and and non-admitted cases, than other databases such as hospital the use of other illicit drugs (Kaar et al. 2015). administrative or mortality data alone (which often use less Eight US states (Colorado, Washington, Oregon, specific ICD codes). It captures injuries severe enough to re- Alaska, California, Maine, Massachusetts, and Nevada) quire medical care, including both those that result in hospital and the District of Columbia (DC) have thus far legalized admission as well as those that do not (Mackenzie and Pless cannabis for non-medical purposes. Recent US public 1999). eCHIRPP covers cases presenting to most major pedi- health surveillance data show increasing prevalence of can- atric centres across Canada, but only select general hospitals. nabis use, mainly among adults, with increases in cannabis- However, previous research has shown that it can represent associated fatal MVCs being reported in Colorado and general injury patterns among Canadian youth (Kang et al. Can J Public Health (2018) 109:155–163 157 2013; Pickett et al. 2000). Records between April 1, 2011 and terms), self-harm (11IN and key terms), ERP involvement May 27, 2016 were extracted for the current analysis (N = (19IN and key terms), sexual assault (12IN and key 636,931 records). terms), and maltreatment (13IN and 14IN). Definitions External cause of injury The external cause of injury variable code was used to describe the mechanism (external cause, EC Analyses are provided for all ages, children (ages 17 years and codes) of injury, and key words were used to mine the narra- below), and adults (ages 18 years and above). Key terms used tive text as follows: poisoning (211EC, 210EC, 301EC, and for narrative text mining are available on request. key terms), fall (201EC and key terms), assault (4001EC, 400EC, and key terms), transport (100EC, 101EC, 102EC), Substances and external agent (202EC, 203EC, 205EC, 209EC, 302EC, 305EC, 309EC, and key terms). The substance use variable code was screened for a listing of either ‘yes’ or ‘unknown’. Following this initial screen, other Nature of injury Nature of injury variable codes (NI codes) variable codes and/or narrative text were used to refine cases. were used to identify the type of injury among cases as fol- Iterative data mining techniques were used to optimize queries lows: intoxication (50NI or key terms), external wound (10NI, of the narrative text. Criteria to identify specific substance use 11NI, 20NI, 22NI), internal wound (24NI, 25NI, 26NI, 27NI, cases are detailed below: 52NI, 53NI, 60NI, 77NI), brain injury (41NI, 42NI, 43NI) or fracture, sprain, or strain (12NI, 13NI, 14NI, 15NI, 16NI, Cannabis use: (1) The substance ID and/or (2) narrative 17NI, 75NI, or key terms). text contains terms like cannabis or marijuana. Criterion 1 was deemed sufficient since these would be based on Severe injuries These were defined as those injuries that re- clinical impressions, and cases where criterion 2 was pos- quired admission to hospital (treatment codes 700T, 800T, or itive but criterion 1 was negative were manually screened. 900T). Admission to hospital was used as a proxy for injury Alcohol use: (1) The substance ID and/or (2) narrative severity (dichotomized yes or no). text contains terms like alcohol. Criteria were assessed as described in section for cannabis use. Illicit drug use: (1) The substance ID and/or (2) narrative Statistical analyses text contains terms like illicit drugs. Cases of cannabis use only were screened out. Criteria were assessed as Data mining syntax (PERL regular expressions (Zhang described for cannabis use. 2011)) wasusedtoassess narrative text, andananalyst Medication use: (1) The substance ID and/or (2) narrative optimized the query language through an iterative pro- text contains terms like medication (either over-the- cess of comparing random samples of cases identified counter or prescription-based). Criteria were assessed as through data mining techniques with their corresponding described for cannabis use. narrative text. Manual resolution was conducted to en- sure accuracy and precision of identified events. To as- Combination of substances were analyzed as binary vari- sess the validity of the methodology chosen to identify ables where instances of combined use were analyzed relative cases of cannabis use, i.e., by either the use of substance to individual or non-use (e.g., analyses of cases of cannabis ID and/or narrative coding, a sensitivity analysis was use alongside alcohol use were made relative to cases of can- performed to estimate the sensitivity, specificity, positive nabis only, alcohol only, or no substance use). predictive value (PPV), and negative predictive value (NPV) of each. Descriptive estimates of substance use Injury characteristics relative to all eCHIRPP cases are reported as a propor- tion relative to 100,000 eCHIRPP cases. Injury charac- Intent of injury event Intent was examined to describe the teristics related to cannabis use (intent, external cause, external or environmental circumstances of the injury and nature of injury) are described as a proportion of event. Intents involving police, emergency medical ser- all cannabis use cases. Among pediatric cases, age- vices, or other such professional staff are identified as adjusted proportionate injury ratio (PIR) estimates cases where emergency response personnel (ERP) were (Breslow and Day 1987) and 95% confidence intervals involved. Intents were categorized based on intention of (CI) were calculated for each sex to examine the likeli- injury codes (IN codes) in combination with narrative text hood of severe injury based on substance use relative to as follows: unintentional injury (10IN, 16IN, and key severe injury among all other non-substance use cases terms), physical assault and/or aggression (15IN and key presentingtotheED. 158 Can J Public Health (2018) 109:155–163 Results and second most frequent intent, respectively (Fig. 2). Narrative text associated with cases of self-harm indicated Between 2011 and 2016, 1170 cases of cannabis use cases thoughts or attempts at suicide from a variety of methods were observed, representing an overall frequency of 184 and included description of depression or feelings of sadness cases/100,000 eCHIRPP cases. Among children, a total of among the patient. For physical assault and/or aggression, 911 cases were observed for a frequency of 257 cases/ cases involved description of altercations, for example getting 100,000 eCHIRPP cases, and among adults, there were 258 into a fight with someone or being punched, or letting out cases for a frequency of 170 cases/100,000 eCHIRPP cases. aggression by punching an inanimate object. When examining For all ages, cannabis use was more frequent among males the external cause, poisoning was the main route of injury for (57.1 versus 42.9% among females; children: 51.0% males all ages. Cases of sexual assault involved cases where canna- and 49.0% females; adults: 79.1% males and 20.9% females) bis and/or other substances were in an individual’s system and among 15 to 19 years old (representing 58.3% of cannabis (either volitionally or because it was given to them) and they use cases, compared with 1.5% for those less than 10 years of were sexually touched or raped. Finally, maltreatment cases age, 21.6% for 10 to 14 years, 9.5% for 20 to 29 years, 4.9% involved aggressive behaviour or negligence by a partner or for 30 to 39 years, 2.7% for 40 to 49 years, and 1.6% for 50- to caregiver. 64-year-old patients). Examining other substances, alcohol use Falls accounted for a high proportion of external causes of was observed at a frequency of 1107 cases/100,000 eCHIRPP injury for both age groups; however, assault was more fre- cases, medication use at 701/100,000 eCHIRPP cases, and quent in child cases whereas transport was in adult cases. illicit drug use at 130 cases/100,000 eCHIRPP cases. Time External causes involving external agents included circum- trends for these various substances are shown in Fig. 1. stances that were both intentional, such as choosing to self- A sensitivity analysis was conducted to assess the validity of harm, or unintentional, such as bumping into an object our case identification method. The substance ID variable was (Fig. 3). With regards to the nature of injury, cases of intoxi- found to be sensitive to 94.1% of cases, specific for 100.0% of cation were the highest ranked among children, while for cases, had a PPV of 100.0%, and an NPVof 99.9%. Narrative adults, it was fractures, sprains, or strains. Of note, brain inju- text mining was sensitive to 83.8% of cases, specific for 99.7% ries were recorded for children, which included either minor of cases, had a PPVof 37.3%, and an NPVof 100.0%. All 1101 head injury, concussion, or intracranial injury, and were the cases identified based on the substance ID variable were kept third highest ranking nature of injury (Fig. 4). Finally, per as cannabis use cases, while another 69 cases were identified 100,000 eCHIRPP cases, cannabis use only (i.e., without through mining of narrative codes. Narrative and substance ID any combination of substance) was observed among 72.4 codes were concordant in 77.9% of cases. cases, while cannabis in combination with alcohol was seen in 74.6 cases, with illicit drugs in 11.3 cases, and with medi- Examining cannabis cases, we observed that the intent, or external or environmental circumstances, of injury varied cations in 7.9 cases (not including cases where 3 or more based on age. Unintentional injury was the leading intent of substances were combined). injury for all ages, and cases included a broad range of cir- Severe injury was observed in 14.3% of all cannabis use cumstances such as MVC, poisoning, and hallucinations. cases (10.1% among children and 29.1% among adults) and Self-harm was the second most frequent intent among chil- occurred due to the following external causes: poisoning dren, and third among adults, while assault came in as the third (51.5%), transport (22.8%), external agent (9.6%), assault (a) Children (b) Adults 201 201 201 201 201 2011 2012 2013 2014 2015 1 2 3 4 5 Cannabis 340 290 220 240 360 Cannabis 180 170 170 140 130 Alcohol 4850 4480 3350 3990 4090 Alcohol 570 630 480 420 410 Medicaons 680 340 710 570 750 Medicaons 690 790 960 630 550 Drugs 350 280 130 300 540 Drugs 80 80 90 100 100 Fig. 1 Time trend of substance use cases presenting to emergency departments among (a) children and (b) adults, eCHIRPP, 2011–2015. Records for 2016 were suppressed. eCHIRPP: electronic Canadian Hospitals Injury Reporting and Prevention Program Number of cases/ 100,000 eCHIRPP cases Number of cases/ 100,000 eCHIRPP cases Can J Public Health (2018) 109:155–163 159 #1 #1 #1 #2 #3 20 #2 #3 #2 #3 Physical ERP Unintenonal assault/ Self-harm Sexual assault Maltreatment involvement Aggression All ages 66.8 13.0 12.1 7.2 0.6 0.3 Children 69.1 10.5 11.3 8.2 0.8 0.1 Adults 58.9 21.7 14.7 3.5 0.0 1.2 Fig. 2 Distribution of intents of injury event among cannabis use-related Personnel (police, emergency medical services, paramedic, etc.). cases, eCHIRPP, 2011–2016*. Numbers denote ranking of leading exter- eCHIRPP: electronic Canadian Hospitals Injury Reporting and nal causes listed within the database. Maltreatment refers to cases by a Prevention Program. *Records entered on or before May 27, 2016 parent or caregiver, or by a spouse or partner. ERP: Emergency Response (8.4%), and falls (7.8%) (data not shown). Due to differences Discussion observed on the basis of sex, PIR estimates are presented for each sex and were examined only among pediatric cases The results of our study describe trends in substance use (Table 1). The PIR of having a severe injury among male (cannabis, alcohol, medication, and illicit drugs) related ED cannabis-only users, relative to severe injuries among males presentations in the eCHIRPP database. Examining the four for all other injuries where no substance (cannabis, alcohol, substances independently, the most frequent substance re- drugs, medication) was used, was 1.0, 95% CI 0.6–1.7 and sponsible for injury among children was medications, while similarly among females was 0.9, 95% CI 0.5–1.7. When can- for adults, it was alcohol; cannabis was the third and fourth, nabis was used in combination with another substance, the respectively. A variety of intents, external causes, and natures PIRs became significant, except for cases of cannabis use in of injury associated with cannabis use were observed. combination with alcohol among males. Combined medication Poisoning and intoxication, in particular, stood out as a lead- and cannabis use among males, for instance, resulted in a sig- ing external cause and nature of injury, respectively. The ma- nificantly higher PIR of severe injury at 4.2, 95% CI 2.3–7.8 jority of injuries were unintentional in nature, although other than observed among cannabis-only males, and similarly for intents included physical assault and self-harm. External females was 4.4, 95% CI 2.3–8.1 (Table 1). To provide com- causes of injury were mostly attributed to poisoning; howev- parison, the likelihood of a severe injury was examined inde- er, falls and transport-related injuries were also observed. pendently for each of the other substances analyzed. We ob- While intoxication once again was the main nature of injury, served that for both sexes, the likelihood of severe injury was injuries such as fractures and open wounds were identified in significantly higher in cases of medications only and illicit the database. Finally, relative to all eCHIRPP cases not in- drugs only when compared with cannabis only (Table 1). volving cannabis use, cannabis use was not significantly #1 #1 #1 30 #2 #3 #3 #2 #2 #3 Poisoning Falls Assault Transport External agent All ages 68.5 9.7 9.4 6.5 5.9 Children 79.8 5.9 6.8 2.7 4.7 Adults 28.3 23.3 18.6 19.8 10.1 Fig. 3 External cause of injury among cannabis use-related cases, used for self-harm. eCHIRPP: electronic Canadian Hospitals Injury eCHIRPP records, 2011–2016*. Numbers denote ranking of leading ex- Reporting and Prevention Program. *Records entered on or before May 27, 2016 ternal causes listed within the database. External agent includes agents Proporon (%) Proporon (%) 160 Can J Public Health (2018) 109:155–163 #1 #1 40 #1 #2 30 #3 #2 #3 #2 #3 Fracture, sprain Intoxicaon External wound Internal wound Head injury or strain All Ages 69.4 13.4 3.2 5.3 8.7 Children 82.9 8.9 1.7 3.8 2.7 Adults 21.7 29.5 8.5 10.5 29.8 Fig. 4 Nature of injury among cannabis use-related cases, eCHIRPP, 2011–2016* Numbers denote ranking of leading external causes listed within the database. eCHIRPP: electronic Canadian Hospitals Injury Reporting and Prevention Program. *Records entered on or before May 27, 2016 associated with hospitalization (i.e., severe injury), but its use individuals may have had limited awareness regarding risks in combination with illicit drugs and/or medications was. It is associated with cannabis use. According to a recent survey, worth noting that of all eCHIRPP cases, when examining more than a third (37.9%) of Canadians report that cannabis each substance independently, alcohol, illicit drug, and med- use should be permitted since they perceive it to not be a dan- ication use, but not cannabis use, were significantly associat- gerous drug, with males agreeing to this statement significantly ed with hospitalization. more than females (p < 0.001) (Health Canada 2006). Given Our observed frequency of injuries related to substance use this low perception of risk, it is also likely that many do not is similar with nationally reported substance use patterns for consider indirect harms associated with cannabis use such as adults: alcohol use was reported among 77% of Canadians falls or wounds, which accounted for a large number of injuries ages 15 and older, psychoactive pharmaceutical drugs at among adults. Even though harms from cannabis use are con- 22%, cannabis at 12%, and illicit drugs (excluding cannabis) sidered to be less likely than those associated with other psy- at 2%. It appears that while cannabis use is known to be choactive agents (Nutt et al. 2010), they are still present. prevalent in Canada (Health Canada 2015), many intents, ex- While differences in the proportion of cannabis use cases ternal causes, and injuries observed within eCHIRPP were between the sexes were not observed in younger ages, adult related to overconsumption and/or toxic exposure to cannabis males weremorelikelytobeinvolvedinaninjuryassociated and its known or unknown combined substances. This may be with cannabis use, which is consistent with usage patterns for indicative of health literacy regarding cannabis, i.e., that each sex (Center for Behavioral Health Statistics and Quality 2015). Data from the US National Survey on Drug Use and Health show that males, aged 12 to 24, were more likely than Table 1 Age-adjusted proportionate injury ratios by sex for severe females to list cannabis as their primary substance of abuse injury based on substance use among children, eCHIRPP, 2011–2016 (Center for Behavioral Health Statistics and Quality 2014). We observed a higher proportion of injury cases among indi- Males Females viduals aged 10 to 14 years; a demographic where recent PIR 95% CI PIR 95% CI findings suggest the importance of considering acute cannabis intoxication in cases of an altered level of consciousness Cannabis only 1.0 0.6–1.7 0.9 0.5–1.7 (Murray et al. 2016). with alcohol 1.0 0.6–1.7 1.7 1.2–2.6 Findings of injuries related to physical assault and self- with drugs 1.9 1.0–3.9 2.4 1.1–5.0 harm are consistent with previous literature suggesting as- with medications 4.2 2.3–7.8 4.4 2.3–8.1 sociations of cannabis use with the development of anxiety Alcohol only 1.6 1.3–1.9 1.5 1.3–1.8 disorders, depression, suicide ideation, and interpersonal Illicit drugs only 2.1 1.4–3.1 2.7 2.0–3.7 violence (Copeland et al. 2013). These instances may have Medications only 2.2 1.9–2.5 5.0 4.7–5.4 been unintentional, likely related to a low awareness of harms (Health Canada 2006), but many were actually in- Severe injury defined as those cases where the individual was admitted to tentional and were consistent with behaviours to indicate the hospital distress (Doyle et al. 2017). The associations of cannabis PIR proportionate injury ratio, CI confidence interval, eCHIRPP electron- ic Canadian Hospitals Injury Reporting and Prevention Program use with intentional self-harm and with forms of assault Records entered on or before May 27, 2016 also reveal important risks of harm and aggression, as Proporon (%) Can J Public Health (2018) 109:155–163 161 described previously (Shorey et al. 2014;Husseyetal. Currently available sources of information regarding cannabis 2006). ERP involvement was observed in a number of in- use and public health outcomes in Canada include national stances and includes cases where illegal activities were surveys, which mainly report on prevalence of use (Health occurring, where individuals contacted the police regard- Canada 2015; Health Canada 2013;Boak etal. 2015), and ing someone under the influence, or where police needed the Canadian Surveillance System for Poison Information, to detain someone for criminal behaviour. A recent which provides information regarding toxoid exposures in- Canadian study reported that crime rates in school cluding those that do not present to EDs. The findings report- neighbourhoods were indeed associated with cannabis ed in this study help to shed light on external causes and use among adolescents (de Looze et al. 2015). Cannabis natures of cannabis-related injuries based on ED data. Future use is also a well-established risk factor for MVCs (Fischer assessments of eCHIRPP data against the current findings will et al. 2016), and this latter external cause of injury did assist in keeping track of how legalization may have affected appear among cases of injury among adults. Studies from observed cases. British Columbia and Quebec suggest that between 12% and 14% of drivers involved in MVCs had cannabis in Strengths and limitations their system while driving (Senate Special Committee on Illegal Drugs 2002). A recent poll by the Canadian A main strength of this study is the utility of the eCHIRPP Automobile Association found that 26% of Canadians be- database to capture and describe cases requiring medical care, tween the ages of 18 to 24 years believed that their driving but not necessitating hospital admission. The eCHIRPP is an under the influence of cannabis was either the same or active surveillance system where data collection is systematic, better. These misconceptions may explain why almost uses standardized coding, and has been ongoing for over two thirds of Canadians were concerned that roads would 25 years. It is able to capture cases presenting to most major become more dangerous with the legalization of cannabis pediatric centres across Canada. As such, it is a useful data (Canadian Automobile Association 2016). source for examining trends and detecting signals in the pedi- Among pediatric cases, severe injury was not found to be atric population. significantly associated with cases of cannabis use for either Based on the nature and geography of participating centres, sex. Consistent with previous findings (Sewell et al. 2009), the eCHIRPP platform likely under-represents older teen, the use of cannabis in combination with alcohol resulted in adult, aboriginal, rural, and fatal cases. Therefore, extrapola- greater severity of injury than with cannabis alone among tions for these subgroups are discouraged. Since this platform males, though not significantly so. Recent reports suggest that is also specific to ED presentations, it also does not capture prescription medication abuse is the fastest growing drug- mild or moderate cases that may not have sought care or that related problem in the USA (Sarker et al. 2016), and results may have been dealt with through resources such as poison from our study also showed a significantly higher proportion centres. The data do not distinguish between medical and non- of severe cases among users of medication and cannabis in medical cannabis use. Analyses were restricted to available combination, as well as medication alone, compared to can- variable codes, thereby restricting the level of detail available nabis alone. Similarly, our observation that severe injury to describe cannabis use or injury descriptors. Since eCHIRPP among users of illicit drugs in combination with cannabis as is not population based, and since cannabis is not currently more likely than with cannabis alone might reflect patterns legal in Canada for non-medical use, under-reporting of inju- describing users of these two substances in combination as ries related to cannabis use is possible. Misclassification bias having a reduced perception of the risks associated with is a possibility given the lack of objectively collected sub- them. Definitions used in this analysis for medications and stance use information. Definitions employed for data mining for illicit drugs both contain agents that could be described as created a known bias towards accepting substance ID coding opioids and may therefore lend a perspective to previously as confirmation of cannabis use; nevertheless, there was described cases of the opioid crisis observed within the strong sensitivity and specificity of substance use and narra- eCHIRPP dataset (Government of Canada 2017). tive codes for identifying cases. Since there was no known Information gathered from US States that have legalized bias placed on examination of narrative codes, the specificity cannabis for non-medical purposes (e.g., Colorado, and NPV of using narrative codes point to the advantage of Washington) report an increased number of ED visits and using this approach to correctly identify non-cases. A recent admissions to hospitals associated with possible cannabis ex- examination of the validity of self-reported substance use posure, increased calls to poison control centers mentioning among emergency room populations and possible self- human cannabis exposure, as well as increased numbers of selection bias found that self-reported alcohol and substance fatalities among drivers positive for THC-only or THC-in- use was actually preferable to other objective methods since combination with alcohol or other drugs (Reed 2016; the former provided more accurate information regarding ac- Northwest High Intensity Drug Trafficking Area 2016). tual use (Vitale et al. 2006). 162 Can J Public Health (2018) 109:155–163 Doyle, L., Sheridan, A., & Treacy, M. P. (2017). Motivations for adoles- Conclusions cent self-harm and the implications for mental health nurses. J Psychiatr Ment Health Nurs, 24,134–142. Cannabis use injuries were observed in the eCHIRPP database Elvik, R. (2013). Risk of road accident associated with the use of drugs: A with a variety of intents, external causes, and natures of inju- systematic review and meta-analysis of evidence from epidemiolog- ical studies. Accid Anal Prev, 60,254–267. ries. The current findings serve to describe a subsample of Fischer, B., Imtiaz, S., Rudzinski, K., & Rehm, J. (2016). Crude estimates injuries that were significant enough to need medical attention of cannabis-attributable mortality and morbidity in Canada- but that did not always necessitate hospital admission. As implications for public health focused intervention priorities. J Canada moves towards the legalization of cannabis for non- Public Health (Oxf), 38,183–188. Government of Canada (2017) Joint Statement of Action to Address the medical purposes, our observation of cases related to uninten- Opioid Crisis. Available at: https://www.canada.ca/en/health- tional injury, poisoning, and intoxication suggests areas that canada/services/substance-abuse/opioid-conference/joint- might benefit from health literacy efforts. statement-action-address-opioid-crisis.html. Accessed 14/28, 2017. Hall, W. (2015). What has research over the past two decades revealed Compliance with ethical standards about the adverse health effects of recreational cannabis use? Addiction, 110,19–35. Hall, W., & Degenhardt, L. (2014). The adverse health effects of chronic Conflict of interest The authors declare that they have no conflict of cannabis use. Drug Test Anal, 6,39–45. interest. Health Canada. Canadian Addiction Survey (CAS), A National Survey of Open Access This article is distributed under the terms of the Creative Canadians’ Use of Alcohol and Other Drugs: Public Opinion, Commons Attribution 4.0 International License (http:// Attitudes and Knowledge. Ottawa ON: Canadian Centre on creativecommons.org/licenses/by/4.0/), which permits unrestricted use, Substance Abuse, 2006; HC Pub 4944: ISBN: 0–662–44537-6. distribution, and reproduction in any medium, provided you give appro- Health Canada. Summary of results for 2013: Canadian Tobacco Alcohol priate credit to the original author(s) and the source, provide a link to the and Drugs Survey (CTADS). Available at: http://healthycanadians. Creative Commons license, and indicate if changes were made. gc.ca/science-research-sciences-recherches/data-donnees/ctads- ectad/summary-sommaire-2013-eng.php. Accessed 10, 2016. Health Canada. Summary of Results Canadian Student Tobacco, Alcohol and Drugs Survey 2014–15. Available at: http://healthycanadians. References gc.ca/science-research-sciences-recherches/data-donnees/cstads- ectade/summary-sommaire-2014-15-eng.php. Accessed 10, 2016. Azofeifa, A., Mattson, M. E., Schauer, G., McAfee, T., Grant, A., & Health Canada. Summary of results for 2015: Canadian Tobacco, Alcohol Lyerla, R. (2016). National Estimates of marijuana use and related and Drugs Survey (CTADS). Available at: http://healthycanadians. indicators - National Survey on drug use and health, United States, gc.ca/science-research-sciences-recherches/data-donnees/ctads- 2002-2014. MMWR Surveill Summ, 65,1–28. ectad/summary-sommaire-2015-eng.php. Accessed 11, 2016. Boak A, Hamilton HA, Adlaf EM, Mann RE,. Drug use among Ontario Health Surveillance and Epidemiology Division, Public Health Agency students, 1977–2015: Detailed OSDUHS findings. Toronto ON, of Canada. Injuries associated with Backyard Trampolines: Centre for Addiction and Mental Health, 2015. Research Canadian Hospitals Injury Reporting and Prevention Program Document Series No. 41. (CHIRPP) database. 1999–2003 (cumulative to February 2006). Breslow NE, Day NE. Statistical methods in cancer research. 1987; Lyon, All ages, 2,705 records. Update 2004–2006, 1749 cases. Available France: IARC Scientific Publications No. 82. at: http://www.phac-aspc.gc.ca/injury-bles/chirpp/injrep-rapbles/ pdf/trampolines-eng.pdf. Accessed 29 Aug 2016. Canadian Automobile Association (2016) Canadians Worried Roads Will Hussey, J. M., Chang, J. J., & Kotch, J. B. (2006). Child maltreatment in Be Unsafe When Marijuana Legalized. Available at: https://www. the United States: Prevalence, risk factors, and adolescent health caa.ca/canadians-worried-roads-will-be-unsafe-when-marijuana- consequences. Pediatrics, 118,933–942. legalized/. Accessed 11/14, 2016. Center for Behavioral Health Statistics and Quality (2014) The TEDS Kaar, S. J., Gao, C. X., Lloyd, B., Smith, K., & Lubman, D. I. (2015). Report: Gender Differences in Primary Substance of Abuse across Trends in cannabis-related ambulance presentations from 2000 to Age Groups. Rockville MD, U.S. Department of Health and Human 2013 in Melbourne. Australia Drug Alcohol Depend, 155,24–30. Services. Kang, J., Hagel, B., Emery, C. A., Senger, T., & Meeuwisse, W. (2013). Center for Behavioral Health Statistics and Quality (2015). Behavioral Assessing the representativeness of Canadian hospitals injury Health Trends in the United States: Results from the 2014 National reporting and prevention Programme (CHIRPP) sport and recrea- Survey on Drug Use and Health. Rockville MD, U.S. Department of tional injury data in Calgary, Canada. Int J Inj Control Saf Promot, Health and Human Services. HHS Publication No. SMA 15–4927, 20,19–26. NSDUH Series H-50. Mackenzie, S. G., & Pless, I. B. (1999). CHIRPP: Canada's principal Copeland, J., Rooke, S., & Swift, W. (2013). Changes in cannabis use injury surveillance program. Canadian hospitals injury reporting among young people: Impact on mental health. Curr Opin and prevention program. Inj Prev, 5,208–213. Psychiatry, 26,325–329. McFaull, S., Subaskaran, J., & Thompson, W. (2016). Emergency de- Crain, J., McFaull, S., Thompson, W., Skinner, R., Do, M. T., Frechette, partment surveillance of injuries and head injuries associated with M., & Mukhi, S. (2016). Status report - the Canadian hospitals baseball, football, soccer and ice hockey, children and youth, ages injury reporting and prevention program: A dynamic and innova- 5 to 18 years, 2004 to 2014. Health Promot Chronic Dis Prev Can., tive injury surveillance system. Health Promot Chronic Dis Prev 36,13–14. Can, 36,112–117. Mura, P., Kintz, P., Ludes, B., Gaulier, J. M., Marquet, P., Martin-Dupont, de Looze, M., Janssen, I., Elgar, F. J., Craig, W., & Pickett, W. (2015). S., Vincent, F., Kaddour, A., Goulle, J. P., Nouveau, J., Moulsma, Neighbourhood crime and adolescent cannabis use in Canadian ad- M., Tilhet-Coartet, S., & Pourrat, O. (2003). Comparison of the olescents. Drug Alcohol Depend, 146,68–74. prevalence of alcohol, cannabis and other drugs between 900 injured Can J Public Health (2018) 109:155–163 163 drivers and 900 control subjects: Results of a French collaborative Shorey, R. C., Stuart, G. L., Moore, T. M., & McNulty, J. K. (2014). The temporal relationship between alcohol, marijuana, angry affect, and study. Forensic Sci Int, 133,79–85. dating violence perpetration: A daily diary study with female college Murray, D., Olson, J., & Lopez, A. S. (2016). When the grass isn't students. Psychol Addict Behav, 28,516–523. greener: A case series of young children with accidental marijuana Silins, E., Horwood, L. J., Patton, G. C., Fergusson, D. M., Olsson, C. A., ingestion. CJEM, 18,480–483. Hutchinson, D. M., Spry, E., Toumbourou, J. W., Degenhardt, L., Northwest High Intensity Drug Trafficking Area. Washington State Swift, W., Coffey, C., Tait, R. J., Letcher, P., Copeland, J., & Marijuana Impact Report. Seattle 2016. Available from http:// Mattick, R. P. (2014). Cannabis cohorts research consortium. msani.org/wp-content/uploads/2016/11/NWHIDTA-Marijuana- Young adult sequelae of adolescent cannabis use: An integrative Impact-Report-Volume-1.pdf. Accessed on 8 Sept 2016. analysis. Lancet Psychiatry, 1,286–293. Nutt, D. J., King, L. A., & Phillips, L. D. (2010). Independent scientific Task Force on Marijuana Legalization and Regulation. Toward the legal- committee on drugs. Drug harms in the UK: A multicriteria decision ization, regulation and restriction of access to marijuana: Discussion analysis. Lancet, 376,1558–1565. Paper. Ottawa, Government of Canada; 2016; Available from Pickett, W., Brison, R. J., Mackenzie, S. G., Garner, M., King, M. A., https://www.canada.ca/content/dam/hc-sc/healthy-canadians/ Greenberg, T. L., & Boyce, W. F. (2000). Youth injury data in the migration/health-system-systeme-sante/consultations/legalization- Canadian hospitals injury reporting and prevention program: Do marijuana-legalisation/alt/legalization-marijuana-legalisation-eng. they represent the Canadian experience? Inj Prev., 6,9–15. pdf. Accessedon15Sept2017. Ramaekers, J. G., Berghaus, G., van Laar, M., & Drummer, O. H. (2004). United Nations Office on Drugs and Crime. World Drug Report 2016. Dose related risk of motor vehicle crashes after cannabis use. Drug New York: United Nations publication, Sales No. E.16.XI.7. Alcohol Depend, 73,109–119. Vitale, S., & van de Mheen, D. (2006). Illicit drug use and injuries: A Reed, J.K. Marijuana Legalization in Colorado: Early Findings. Denver review of emergency room studies. Drug Alcohol Depend, 82,1–9. CO, Department of Public Safety; 2016. Available from https:// Vitale, S. G., van de Mheen, H., van de Wiel, A., & Garretsen, H. F. cdpsdocs.state.co.us/ors/docs/reports/2016-SB13-283-Rpt.pdf. (2006). Substance use among emergency room patients: Is self- Accessed on 8 Sept 2016. report preferable to biochemical markers? Addict Behav, 31, Sarker, A., O'Connor, K., Ginn, R., Scotch, M., Smith, K., Malone, D., & 1661–1669. Gonzalez, G. (2016). Social Media Mining for Toxicovigilance: Volkow,N.D.,Baler,R.D.,Compton,W.M.,&Weiss, S.R.(2014). Automatic monitoring of prescription medication abuse from twitter. Adverse health effects of marijuana use. N Engl J Med, 370, Drug Saf, 39,231–240. 2219–2227. Senate Special Committee on Illegal Drugs. Cannabis: Our position for a Wang, G. S., Roosevelt, G., & Heard, K. (2013). Pediatric marijua- Canadian public policy - Chapter 8. Report of the Senate Special na exposures in a medical marijuana state. JAMA Pediatr, 167, Committee on Illegal Drugs. 2002. 630–633. Sewell, R. A., Poling, J., & Sofuoglu, M. (2009). The effect of cannabis Zhang, Y. (2011). Perl regular expression in SAS macro programming. compared with alcohol on driving. Am J Addict, 18,185–193. SAS Global Forum, 159-2011,1–7.
Canadian Journal of Public Health
– Springer Journals
Published: Feb 26, 2018