bs_bs_banner

RESEARCH REPORT

doi:10.1111/add.12626

Hazardous birthday drinking among young people: population-based impacts on emergency department and in-patient hospital admissions Russell C. Callaghan1,2,3, Marcos Sanches4, Jodi M. Gatley1,2, Lon-Mu Liu5 & James K. Cunningham6,7 Northern Medical Program, University of Northern British Columbia (UNBC), Prince George, British Columbia, Canada,1 Human Brain Laboratory, Centre for Addiction and Mental Health (CAMH), Toronto, Ontario, Canada,2 Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada,3 Biostatistical Consulting Unit, Centre for Addiction and Mental Health (CAMH), Toronto, Ontario, Canada,4 Department of Economics and Public Economics Center, National Taiwan University, Taipei, Taiwan,5 Department of Family and Community Medicine, The University of Arizona, Tucson, AZ, USA6 and Native American Research and Training Center, The University of Arizona, Tucson, AZ, USA7

ABSTRACT Background and Aims There is growing concern about the possible adverse health impacts of binge drinking during birthday celebrations among adolescents and young adults. We estimate the impacts of birthday alcohol use on adolescent and young adult in-patient/emergency department (ED) hospital admissions. Design We employed Autoregressive Integrated Moving Average (ARIMA) intervention analysis to assess whether the rate of ICD-10 alcohol-use-disorder (AUD) events per 1000 in-patient/ED admissions increased significantly during birthday weeks. Setting All in-patient/ED admissions in Ontario, Canada from 1 April 2002 to 31 March 2007. Participants Individuals aged 12–30 years. Measurements AUD events per 1000 in-patient/ED admissions by age in weeks. Findings Multiple increases were found. The largest occurred during the birthday week of 19 years of age, the beginning of the minimum legal drinking age (MLDA) in Ontario: AUD admission rates increased (spiked) by 38.30 per 1000 total admissions [95% confidence interval (CI) = 34.66–41.94] among males (a 114.3% increase over baseline), and by 28.13 (95% CI = 25.56–30.70) among females (a 164.0% increase). Among both genders, the second largest birthday-week spikes occurred during ages 20–22 years, followed by somewhat lower but still pronounced birthday-week spikes during ages 23–26 years and 30 years (all these spikes: P < 0.05). Birthday-week spikes occurred as early as age 16 years for males and 14 years for females (both spikes: P < 0.05). Conclusions There appears to be an increase in alcohol-related adverse events from drinking around the time of one’s birthday among young adults in Canada. Keywords morbidity.

Adolescents, alcohol, binge drinking, birthday celebrations, emergency department, hospitalization,

Correspondence to: Russell C. Callaghan, Northern Medical Program, University of Northern British Columbia, 3333 University Way, Prince George, BC, Canada V2N 4Z9. Email: [email protected] Submitted 26 July 2013; initial review completed 28 October 2013; final version accepted 16 May 2014

INTRODUCTION Usually defined as single-occasion consumption of more than four drinks for females or five drinks for males, binge drinking is common among young adults in North America [1]. In the United States in 2012, for example, prevalence of past-month binge drinking reached 40% in college samples aged 18–22 years and 45% among slightly older (21–25 years) young adults [2]. Approximately 20% of Canadian undergraduate students © 2014 Society for the Study of Addiction

reported typically drinking five or more drinks per drinking occasion in the past year, and 19% engaged in heavy episodic drinking of five or more drinks at least twice a month [3]. In addition, a national survey found that among young Canadians aged 15–24 years in 2011 who were past-year drinkers, 18% had binge drank in the past week [4]. Such patterns of alcohol consumption pose an increased risk for harms in this population, as many studies have demonstrated a link between binge drinking and a range of negative consequences, such as injury, Addiction, 109, 1667–1675

1668

Russell C. Callaghan et al.

unprotected sexual intercourse, alcohol-impaired driving and assault [3,5–11]. Recently, researchers have attempted to identify specific time-periods (e.g. Spring Break) or calendar events (e.g. New Year’s Eve) during which binge drinking frequently occurs so as to develop event-specific strategies to reduce hazardous alcohol consumption. In the United States, studies have found ‘extremely extreme’ binge drinking associated with celebrations of the 21st birthday—the minimum legal drinking age (MLDA) in the United States [12]. In comparison to alcohol consumption on all other holidays or calendar events (e.g. Christmas, Spring Break) [13,14], the heaviest patterns of alcohol consumption appear to occur during 21st-birthday celebrations. American college survey studies have found that not only do more than 80% of young adults turning 21 years old drink alcohol [12,13,15,16], but also a large majority (∼70%) of these young adults binge drink on this occasion [16], with average maximum estimated blood alcohol concentrations (BACs) reaching 0.20% [17,18] or greater [12]—levels which are associated with potentially serious medical outcomes. Some studies have even demonstrated that approximately 12% of females and 22% of males drink more than 21 drinks in quick succession on their 21st birthdays (as part of the youth social tradition known as ‘power hour’ during late-night birthday celebrations)—a practice which has resulted in several highprofile fatalities among college students in the United States [19]. At least in our understanding, only US-based studies have assessed alcohol consumption during birthday celebrations. However, a single non-US study, based on a sample of birthday cards from Ireland, found that approximately 50% of birthday cards for the 18th birthday (Ireland’s MLDA) referenced ‘alcohol’ or contained explicit references to alcohol consumption or intoxication [20]. Understanding the impact of alcohol use during birthday celebrations throughout adolescence and early adulthood will help to guide and justify event-specific prevention efforts in this group—a population incurring pronounced alcohol-related harms [21–23]. To this end, the current study uses Autoregressive Integrated Moving Average (ARIMA) intervention analysis to generate population-level estimates of the impact of birthdayweek alcohol use on in-patient/emergency department (ED) hospital admissions related to alcohol-use disorders (AUDs) in Ontario, Canada. A critical advantage of using medical information from the Ontario health-care system is that it allows for near-census estimates of all target in-patient/ED hospital events, without potential bias contingent on medical-insurance status. We expected a significant increase in AUD events to occur in the birthday week for 19-year-olds—the beginning of the MLDA in Ontario; and upturns, albeit less pronounced, to occur © 2014 Society for the Study of Addiction

in other birthday weeks during adolescence and early adulthood.

METHODS The current study was approved by the Centre for Addiction and Mental Health (CAMH) Ethics Review Board. Data source All Ontario in-patient and ED events from 1 April 2002 to 31 March 2007 of individuals aged 12–30 years were drawn from the Canadian Institute for Health Information (CIHI) Hospital Morbidity Database (HMDB; a national database capturing all hospital in-patient events) and the National Ambulatory Care Reporting System (NACRS; a national database designed to capture information on client visits to ambulatory care—the present study uses ED department visits extracted from the database). During the study period, the HMDB included information from approximately 170 acute-care facilities in Ontario [24], and the NACRS included data from approximately 180 EDs across the province [25]. These data provided near-census (∼100%) coverage for all in-patient/ED events in the Ontario target population during the study period. Re-abstraction studies have demonstrated the validity of HMDB diagnostic information [26,27]. Identification of outcomes AUDs based on ICD-10 classification We defined AUD episodes as those in-patient/ED events resulting in the ICD-10 codes F10.x (mental and behavior disorders due to alcohol use; see Table 1 for description) in any diagnostic position in the medical record. To

Table 1 ICD-10 definitions [28] of alcohol-use disorder (AUD) conditions. ICD-10 diagnostic codes F10.0 F10.1 F10.2 F10.3 F10.4 F10.5 F10.6 F10.7 F10.8 F10.9

Description of condition Acute intoxication Harmful use Dependence syndrome Withdrawal state Withdrawal state with delirium Psychotic disorder (during or immediately after alcohol use) Amnesic syndrome Residual and late-onset psychotic disorder Other mental and behavioral disorders Unspecified mental and behavioral disorders

Addiction, 109, 1667–1675

Impacts of birthday drinking

ensure that we did not double-count AUD episodes occurring for individuals admitted with AUDs to the ED, then transferred to in-patient hospital settings, we only selected one episode across such transfers.

Analytical plan Construction of age variable We used age in weeks—rather than days—as our primary unit of analysis for age. It is likely that a substantial portion of adolescents and young adults do not celebrate their actual birthday on their date of birth. For example, individuals with birthdays occurring early in the week (Monday, Tuesday, Wednesday) may be more likely to celebrate their birthday on the subsequent weekend (Friday, Saturday)—2–5 days after their actual birthday. In addition, even for individuals celebrating their birthday on their actual birth date, heavy and prolonged alcohol consumption during birthday celebrations may extend past midnight into the next day, and any resulting ED visits or in-patient admissions for AUDs would be tagged with a date stamp of the day following their exact birthday. Thus, the age-in-weeks variable allows for more accurate estimation of the impact of birthday celebrations on hospital morbidity associated with alcohol consumption.

Construction of series Each series in this study begins with the first week in the 12th year of age and ends with the last week of the 30th year of age—988 weeks altogether. The values in the series are AUD rates: each rate is the number of AUD events within an age bin per 1000 in-patient/ED admissions for patients of a given week of age (i.e. both the numerator and denominator were constructed using only patients with the given week of age, the number of AUD events was divided by the total number of admissions within the specified age bin and multiplied by 1000). Patterns—upsurges, trends and so on—for weekly raw count series of AUD admissions in Supporting information, Fig. S1, approximate those for rates in Fig. 1. When serial patterns in admission rates and admission counts are approximately similar, it is reasonable to conduct analyses which focus on the rates [28]. Note that a calendar year has 52 weeks plus an extra 1 or 2 days. As units of even length are generally needed for ARIMA intervention analysis, we included each year’s extra 1 or 2 days in the year’s week 20; the rate for week 20 was the number of AUD events per 1000 in-patient/ED episodes during the 8 or 9 days assigned to the week. © 2014 Society for the Study of Addiction

1669

ARIMA intervention analysis This study’s age–series (Fig. 1) are similar to time– series—which is natural, as week of age is determined according to calendar time—and thus can be expected to have many of the characteristics known to make statistical modeling of time–series challenging: drift, trend (including local trend), outliers, seasonality (variation associated with a part or parts of a yearly period) and serially correlated errors. Recognizing this, we use ARIMA intervention analysis, an analytical procedure appropriate to these issues. ARIMA intervention analysis has been used to examine impacts on drug-related time– series [29–31]. To our knowledge, the present study is the first to use ARIMA intervention analysis to examine alcohol-/drug-related age series. The analysis examines two types of impact: (i) whether the rate of AUD admissions rose significantly during a birthday week (a significant rise during 1 week is referred to here as an upsurge or spike) and (ii) whether the level of the series shifted in association with the MLDA (which began in the birthday week of the 19th year). Regarding coding to test for spikes, all weeks preceding a birthday week were coded 0, the birthday week was coded 1 and weeks following the birthday week were coded 0. With this approach, each birthday week (ages 13–30 years) had its own separate variable. Regarding coding for the MLDA level shift, a step function was used: all weeks preceding the 19th year were coded 0, and all subsequent weeks were coded 1. ARIMA intervention analysis was used to estimate the effects of birthday-week drinking (which we considered the ‘intervention’) on two standardized parameters: a change in level (a ‘step’ change) at the intervention— the birthday week, in our case; and a change in trend (a ‘pulse’ change) before and after the intervention. A key advantage of ARIMA intervention analysis is that the procedure takes into account drift, trends (including local trends), outliers, seasonality and serially correlated errors in the series of AUD admissions across age bins [32]. A ‘step’ change is defined as the difference between the observed level at a specified intervention time-point and the corresponding time-point predicted by the preintervention time trend; and a ‘pulse’ trend is defined as the difference between the post- and pre-intervention slopes. Given that our central aim is to estimate the acute and short-term effects of hazardous birthday alcohol consumption on AUD admissions, our paper focuses primarily on estimating ‘pulse’ changes associated with birthday weeks. None the less, we also estimated the ‘step’ change associated with the birthday at the MLDA—19 years of age in Ontario. In detailing our findings in the subsequent Results and Discussion sections, we use the term ‘spike’ to describe the statistically Addiction, 109, 1667–1675

1670

Russell C. Callaghan et al.

Alcoho-related admission rate

60

Total sample

50

19th birthday week Legal drinking age begins

40 30 20 10 0

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

23

24

25

26

27

28

29

30

23

24

25

26

27

28

29

30

Age

Alcohol-related admission rate

60

Females

50 19th birthday week Legal drinking age begins

40 30 20 10 0

12

13

14

15

16

17

18

19

20

21

22 Age

Alcohol-related admission rate

60

Males

19th birthday week Legal drinking age begins

50 40 30 20 10 0

12

13

14

15

16

17

18

19

20

21

22 Age

Figure 1 Rate of alcohol-related admissions (alcohol-related admissions per 1000 emergency department (ED/in-patient hospital admissions) by age in weeks: total sample, females, and males.The horizontal axis is age in weeks; for clarity and ease of interpretation, the birthday weeks are labeled with the corresponding age in years.

significant increases (‘pulse’ changes) in AUD admissions occurring at birthday weeks. All the series were non-stationary, as indicated by the following: they did not exhibit a fixed mean level and had high, positive autocorrelations that decreased slowly. To address this, each series was first-order differenced. We used an iterative outlier detection and adjustment procedure to obtain joint estimates of model parameters and outlier effects [33]. Sample autocorrelations and Box– Ljung Q-tests for the first 24 lags indicated that the models all had residuals consistent with white noise. The ARIMA intervention analyses were performed with SCA software [34]. ARIMA intervention model Using appropriate model identification methods, the general ARIMA model for the series is: (1 − B)Yt = © 2014 Society for the Study of Addiction

(1 − θ1B)at; where B is the backshift operator such that BYt = Yt−1, and at are independent normally distributed random errors. Under this basic model we included the intervention ω1, which is the impact during a birthday week (e.g. ω19 is the impact associated with the birthday week for 19-year-olds), and the intervention γ, which is the MLDA impact on the level of the series.

RESULTS Series overview Including all data aggregated across the 5-year study period, at 12 years of age there were approximately 6000 admissions per week and then, after a gradual increase, there were approximately 10 000 admissions per week from 19–30 years of age (see Supporting information, Fig. S2). Among all the F10.x AUD codes identified in the Addiction, 109, 1667–1675

Impacts of birthday drinking

1671

Table 2 Birthday week and the minimum legal drinking age (MLDA) impacts on alcohol-use disorder admissions per 1000 in-patient/ED admissions: Autoregressive Integrated Moving Average (ARIMA) intervention analysis parameter estimates. Total sample

Females

Males

Birthday week

Parameter

Est.

CI

ta

Est.

CI

t

Est.

CI

t

13th 14th 15th 16th 17th 18th 19th 20th 21st 22nd 23rd 24th 25th 26th 27th 28th 29th 30th MLDA ARIMA

ω13 ω14 ω15 ω16 ω17 ω18 ω19 ω20 ω21 ω22 ω23 ω24 ω25 ω26 ω27 ω28 ω29 ω30 γ θ1

0.69 3.35 1.96 4.03 4.79 7.22 33.14 15.90 14.14 9.83 7.07 7.55 6.42 5.83 1.44 2.09 2.90 6.32 1.78 0.80

−1.35–2.72 1.31–5.39 −0.08–4.00 1.98–6.07 2.75–6.83 5.18–9.26 30.99–35.29 13.86–17.94 12.10–16.17 7.79–11.87 5.03–9.11 5.51–9.59 4.38–8.46 3.79–7.88 −0.60–3.47 0.05–4.12 0.86–4.93 4.28–8.36 0.41–3.16 0.76–0.84

0.66 3.22 1.89 3.87 4.60 6.94 30.20 15.27 13.59 9.45 6.79 7.26 6.17 5.59 1.38 2.01 2.78 6.07 2.55 41.63

0.45 5.46 2.74 3.85 2.21 6.07 28.13 12.80 8.95 7.34 6.60 6.15 4.42 3.82 1.70 1.71 2.37 5.47 1.73 0.79

−1.98–2.88 3.03–7.90 0.30–5.17 1.41–6.29 −0.22–4.64 3.64–8.50 25.56–30.70 10.37–15.23 6.52–11.38 4.91–9.77 4.17–9.04 3.72–8.58 1.98–6.85 1.38–6.26 −0.74–4.13 −0.72–4.14 −0.06–4.80 3.04–7.90 0.06–3.41 0.75–0.83

0.36 4.40 2.21 3.10 1.78 4.89 21.44 10.31 7.21 5.92 5.32 4.96 3.56 3.07 1.37 1.38 1.91 4.41 2.03 40.34

0.84 1.58 1.36 4.72 7.35 8.44 38.30 19.40 20.04 12.50 7.57 9.35 8.80 8.53 1.35 2.22 3.51 7.48 2.22 0.88

−2.69–4.37 −1.95–5.11 −2.17–4.89 1.19–8.25 3.82–10.88 4.91–11.96 34.66–41.94 15.87–22.93 16.51–23.56 8.97–16.03 4.05–11.10 5.82–12.88 5.27–12.32 5.00–12.07 −2.18–4.88 −1.31–5.75 −0.01–7.04 3.96–11.01 0.39–4.06 0.85–0.91

0.47 0.88 0.75 2.62 4.08 4.69 20.60 10.78 11.13 6.94 4.21 5.19 4.89 4.74 0.75 1.23 1.95 4.16 2.38 57.27

CI = 95% confidence interval; Est. = estimate. at ≥ 1.96 is statistically significant (P ≤ 0.05). ω is the impact estimate of the change in the number of alcohol-use disorder admissions (per 1000 admissions) associated with a birthday week (subscripts indicate year of the birthday week); γ is the impact estimate of the change in the series level associated with the minimum legal drinking age (MLDA). θ1 is the moving average estimate.

medical records (see Table 1), approximately 64% were F10.0, 23% were F10.1, 8% were F10.2 and 5% were for other F10.x conditions. Total admissions rates and admissions rates for females and males all had pronounced spikes in the week of the 19th birthday—the first week of the MLDA (Fig. 1). Substantial spikes also occurred in the birthday weeks of the 20th, 21st and 22nd years, although they were smaller than the 19th-year birthday-week spikes. A resurgent, but relatively small, spike occurred during the birthday week of the 30th year. Birthday-week spikes also occurred in years before the 19th year, but to a substantially lesser degree. Gender-specific patterns Among males and females, AUD admission rates on birthday weeks roughly followed that for the total sample, although there were some noticeable interactions by gender. For example, it can be seen in Fig. 1 that during the birthday week of the 19th year, males had a higher AUD admission rate than did females [of 4813 total male admissions that week, 276 were AUD events—a rate of 57.3 admissions per 1000; while of 5248 total female admissions that week, 211 were AUD events—a rate of 40.2 admissions per 1000 (proportions test: Z = 4.01, © 2014 Society for the Study of Addiction

P < 0.05)]. In contrast, during the birthday week of the 14th year, females had a higher rate than did males [males had 20 AUD admissions out of 3556 total admissions that week—a rate of 5.6 admissions per 1000, while females had 41 AUD admissions out of 2915 total admissions—a rate of 14.1 admissions per 1000 (Z = −3.50, P < 0.05)]. Given these differences in patterns across genders, we chose to estimate ARIMA models separately for males and females. ARIMA intervention analysis findings It can be seen in Table 2 (third column, ‘Est.’ for ‘19th birthday week’) that, for the total sample, the AUD admission rate spiked by 33.1 per 1000 admissions during the 19th birthday week (the MLDA birthday week). That is, compared to the series level, the model indicated that the rate at the 19th birthday week had an additional 33.1 AUD admissions for every 1000 admissions overall that occurred. The spike in the rate for males during the 19th birthday week was greater than that for females—38.3 cases per 1000 admissions versus 28.1, respectively. However, females had a greater percentage increase in the AUD admission rate during the 19th birthday week than did males—164.0 versus 114.3%, respectively (these percentages were constructed by comparing the Addiction, 109, 1667–1675

1672

Russell C. Callaghan et al.

model’s estimated increase in rate during the 19th birthday week to a baseline consisting of the average rate during the 3 weeks preceding the birthday week [35]). Finally, the model indicated that the introduction of the MLDA was associated with a small but statistically significant impact on the level of the series (i.e. compared to the series’ pre-MLDA level, the post-MLDA level experienced a small increase): specifically, the level of the AUD admission rate series increased by 1.8, 1.7 and 2.2 admissions for the total sample, females and males, respectively.

DISCUSSION Multiple birthday weeks were found to be associated with spikes in AUD in-patient/ED hospital admissions. The largest occurred in association with the 19th birthday week—the beginning of the Ontario MLDA: AUD in-patient/ED admissions among males and females spiked 114.3 and 164.0%, respectively. For both males and females, the next largest birthday-week spikes occurred during the 20–22nd years. Significant, albeit lesser, birthday-week spikes were also found for both genders during the 23–26th years of age, at age 30 years and as early as age 16 years for males and age 14 years for females. These results constitute the first empirical estimates demonstrating the consequences of birthdayrelated drinking on population-level hospital-based services. Our results support the development and implementation of strategies to reduce hazardous birthday drinking not only associated with the MLDA (aged 19 years in Ontario), but with other ages throughout adolescence and early adulthood, especially those typically associated with the college years (ages 19–22 years). Starting in the 16th year of age, the absolute magnitude of the spikes among females during birthday weeks was generally less than that for males. This lower magnitude was consistent with the lower prevalence of binge or hazardous drinking among female youth in Canada. For example, among Canadian undergraduate students, 13% of females and 21% of males engaged in binge drinking in the last month, and 38% of males were characterized as hazardous/dependent drinkers versus 28% of females [3]. In a population-based sample of Canadian drinkers (students and non-students) aged 15–24 years, low-risk drinking guidelines for chronic drinking were exceeded in the previous week by 11% of females and 18% of males, and for acute/binge drinking by 7% of females and 14% of males [4]. Although there is some evidence of increasing levels of binge drinking among female Canadian undergraduates [36], the larger absolute size in AUD spikes among males might be accounted for by their higher levels of binge drinking. That being said, this study also found birthday-week impacts of concern particular © 2014 Society for the Study of Addiction

to females. Specifically, statistically significant birthdayweek AUD spikes began 2 years earlier for females than for males (as noted above), and the percentage increase associated with the 19th birthday-week spike was greater for females than males. The MLDA in Ontario, Canada was associated with small but statistically significant increases in the levels of the AUD admission rate series; specifically, increases of 1.8, 1.7 and 2.2 admissions for the total sample, females and males, respectively. The impact of the MLDA, consequently, was not restricted solely to the birthday spike during the 19th year of age already discussed. Instead, it had a compound impact: both an initial (birthday) spike and an increase in the AUD admissions level. The findings of the current study must be interpreted in light of a number of limitations. Our study only captured those serious alcohol-related events tagged with an AUD diagnosis in an in-patient/ED setting and, as a result, our outcome does not include the potentially wide range of alcohol-related harms unrecorded in the administrative records of hospital-based services. Also, the interpretation of the male and female patterns in our results may be affected by gender differences in healthcare service utilization. Prior evidence has shown that women tend to have significantly more annual visits to primary-care clinics and request more diagnostic services [37]. However, there is no evidence of gender differences in visits to specialty clinics, hospital or ED settings [37]. In particular, a recent study found no evidence of differences in the rates of ED usage among females and males in Ontario [38]. Also, the impacts of birthday hazardous drinking may be more pronounced in other regions, as Ontario general-population patterns of hazardous alcohol consumption (defined as levels exceeding national low-risk drinking guidelines) are lower than national estimates [4]. This study’s finding of pronounced spiking in AUD in-patient/ED admission rates during birthday weeks underscores the need for event-specific prevention (ESP). In the last decade, a range of studies have assessed various ESP techniques, which aim to reduce alcoholrelated harms during a specific period of time or welldefined event in which risks of alcohol-related adverse events are elevated [39]. The earliest ESP studies to target hazardous 21st-birthday drinking used birthday-card interventions [16,40–44]. The majority (four of six) of the available studies found that the cards had no impact on 21st-birthday alcohol consumption or celebration plans [16,40–44]. However, the remaining two studies [41,42] supported some positive effects, with students who received the card reporting some decreases in alcohol consumption, self-reported drunkenness and lower peak BACs. The mixed results may have been due to the generally low follow-up response rates, differing card Addiction, 109, 1667–1675

Impacts of birthday drinking

content/designs across studies, self-report bias, differing campus demographics/cultures, an absence of card personalization and a lack of assessment of card-content retention. Three birthday-card interventions measured alcohol-related negative consequences attributed to 21stbirthday drinking; however, there were no significant effects from any card interventions [40,43,44]. Recently, another ESP approach has incorporated the Brief Alcohol Screening and Intervention for College Students (BASICS) [45] into in-person or web-based interventions to reduce birthday drinking among college students in the United States [46,47]. This line of research by Neighbors and colleagues [46,47] has demonstrated that well-supported general prevention strategies such as BASICS (which incorporates personalized feedback on drinking intentions, BAC information, normative information and protective behavioral strategies [46]) can be incorporated into in-person or web-based interventions to reduce birthday drinking. In particular, this line of research has demonstrated that web-based and in-person interventions designed specifically to reduce birthday drinking are effective in reducing BACs and self-reported alcohol-related negative consequences (e.g. arrest, hangover, blackout), especially among those students intending to drink at the riskiest levels on their birthdays [46,47]. In addition, one web-based/in-person BASICS study found that most treatment conditions (with the exception of the web-based 21st-birthdayspecific BASICS) reduced overall alcohol-related negative consequences experienced relative to the control group by 23–29% [47]. While a recent review of effective strategies to reduce hazardous alcohol consumption among college students found that in-person interventions and those incorporating motivational interviewing principles and personalized feedback were the most effective [48], the work of Neighbors et al. [46,47] opens up the possibility of using campus e-mail systems to deliver webbased population-level interventions at relatively low cost. Understanding the impact of alcohol use can help to guide effective prevention efforts. Previous research has found that adolescents and young adults incur pronounced alcohol-related harms [21–23]. That said, this study is the first to show that birthday-related drinking is associated with upsurges in population-level in-patient/ED admissions among young people. These new findings demonstrate the need for event-specific prevention—not just at the time of the MLDA birthday, but at the times of multiple birthdays throughout adolescence and early adulthood.

Declaration of interests None. © 2014 Society for the Study of Addiction

1673

Acknowledgements R.C.C. had full access to all the data in the study, and takes responsibility for the integrity of the data and the accuracy of the data analysis. This study was funded through salary support from the Centre for Addiction and Mental Health (CAMH) and the Northern Medical Program (University of Northern British Columbia) to the first author (R.C.C.) and M.S. J.M.G. was supported through a studentship from the Northern Medical Program (University of Northern British Columbia). None of the supporting institutions (CAMH, Northern Medical Program) had any role in any aspect of the study: design, data collection, data analyses/interpretation, manuscript preparation or approval to submit the final manuscript. J.K.C. and L.M.L. report no funding sources. References 1. Courtney K. E., Polich J. Binge drinking in young adults: data, definitions, and determinants. Psychol Bull 2009; 135: 142–56. 2. Substance Abuse and Mental Health Services Administration. Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings. Rockville MD: Substance Abuse and Mental Health Services Administration; 2013. 3. Adlaf E. M., Demers A., Gliksman L. Canadian Campus Survey 2004. Toronto: Centre for Addiction and Mental Health; 2005. 4. Health Canada. Canadian Alcohol and Drug Use Monitoring Survey (CADUMS): Summary of Results for 2011. Ottawa, ON: Health Canada; 2012. 5. Flowers N. T., Naimi T. S., Brewer R. D., Elder R. W., Shults R. A., Jiles R. Patterns of alcohol consumption and alcoholimpaired driving in the United States. Alcohol Clin Exp Res 2008; 32: 639–44. 6. Naimi T. S., Nelson D. E., Brewer R. D. Driving after binge drinking. Am J Prev Med 2009; 37: 314–20. 7. Jennison K. M. The short-term effects and unintended longterm consequences of binge drinking in college: a 10-year follow-up study. Am J Drug Alcohol Abuse 2004; 30: 659– 84. 8. Wechsler H., Lee J. E., Kuo M., Lee H. College binge drinking in the 1990s: a continuing problem results of the Harvard School of Public Health 1999 College Alcohol Study. J Am Coll Health 2000; 48: 199–210. 9. Brewer R. D., Swahn M. H. Binge drinking and violence. JAMA 2005; 294: 616–8. 10. Mohler-Kuo M., Dowdall G. W., Koss M. P., Wechsler H. Correlates of rape while intoxicated in a national sample of college women. J Stud Alcohol 2004; 65: 37–45. 11. Hall J. A., Moore C. B. Drug facilitated sexual assault—a review. J Forensic Leg Med 2008; 15: 291–7. 12. Rutledge P. C., Park A., Sher K. J. 21st birthday drinking: extremely extreme. J Consult Clin Psychol 2008; 76: 511–6. 13. Neighbors C., Atkins D. C., Lewis M. A., Lee C. M., Kaysen D., Mittmann A. et al. Event-specific drinking among college students. Psychol Addict Behav 2011; 25: 702–7. 14. Foster D. W., Rodriguez L. M., Neighbors C., DiBello A., Chen C. The magic number 21: transitions in drinking. Addict Newsl 2011; 18: 17–9. Addiction, 109, 1667–1675

1674

Russell C. Callaghan et al.

15. Neighbors C., Oster-Aaland L., Bergstrom R. L., Lewis M. A. Event- and context-specific normative misperceptions and high-risk drinking: 21st birthday celebrations and football tailgating. J Stud Alcohol 2006; 67: 282–9. 16. Neighbors C., Spieker C. J., Oster-Aaland L., Lewis M. A., Bergstrom R. L. Celebration intoxication: an evaluation of 21st birthday alcohol consumption. J Am Coll Health 2005; 54: 76–80. 17. Wetherill R. R., Fromme K. Subjective responses to alcohol prime event-specific alcohol consumption and predict blackouts and hangover. J Stud Alcohol Drugs 2009; 70: 593– 600. 18. Brister H. A., Wetherill R. R., Fromme K. Anticipated versus actual alcohol consumption during 21st birthday celebrations. J Stud Alcohol Drugs 2010; 71: 180–3. 19. Be Responsible About Drinking (B.R.A.D. 21). B.R.A.D. 21 Be Responsible About Drinking. 22 January 2013. Available at: http://www.brad21.org/ (accessed 24 April 2014; archived at http://www.webcitation.org/6P54rgKfm on 24 April 2014). 20. Loughran H. Eighteen and celebrating: birthday cards and drinking culture. J Youth Stud 2010; 13: 631–45. 21. Gore F. M., Bloem P. J., Patton G. C., Ferguson J., Joseph V., Coffey C. et al. Global burden of disease in young people aged 10–24 years: a systematic analysis. Lancet 2011; 377: 2093–102. 22. Shield K. D., Taylor B., Kehoe T., Patra J., Rehm J. Mortality and potential years of life lost attributable to alcohol consumption in Canada in 2005. BMC Public Health 2012; 12: 91. 23. Shield K. D., Kehoe T., Taylor B., Patra J., Rehm J. Alcoholattributable burden of disease and injury in Canada, 2004. Int J Public Health 2012; 57: 391–401. 24. Canadian Institute for Health Information. Executive Summary: Data Quality Documentation, Hospital Morbidity Database (HMDB) 2004–2005. Ottawa, ON: Canadian Institute for Health Information; 2006. 25. Canadian Institute for Health Information. Database Background Documentation: National Ambulatory Care Reporting System (NACRS) FY 2003–2004. Ottawa, ON: Canadian Institute for Health Information; 2004. 26. Health Results Team for Information Management. Reabstraction Study of the Ontario Case Costing Facilities: For Fiscal Years 2002/2003 and 2003/2004. Toronto, ON: Canadian Health Information Management Association, Government of Ontario; 2005. 27. Canadian Institute for Health Information. Data Quality of the Discharge Abstract Database Following the First-Year Implementation of ICD-10-CA/CCI: Final Report. Ottawa, ON: Canadian Institute for Health Information; 2004. 28. World Health Organization (WHO). The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Vienna: WHO; 2013. 29. Callaghan R. C., Cunningham J., Victor J. C., Liu L. Impact of Canadian federal methamphetamine precursor and essential chemical regulations on methamphetamine-related acute-care hospital admissions. Drug Alcohol Depend 2009; 105: 185–93. 30. Cunningham J. K., Liu L. M. Impacts of federal ephedrine and pseudoephedrine regulations on methamphetaminerelated hospital admissions. Addiction 2003; 98: 1229–37. 31. Cunningham J. K., Liu L. M. Impacts of federal precursor chemical regulations on methamphetamine arrests. Addiction 2005; 100: 479–88. © 2014 Society for the Study of Addiction

32. Box G. E. P., Tiao G. C. Intervention analysis with applications to economic and environmental problems. J Am Stat Assoc 1975; 70: 70–9. 33. Chen C., Liu L. Joint estimation of model parameters and outlier effects in time series. J Am Stat Assoc 1993; 88: 284– 97. 34. Liu L., Hudak G., Box G., Muller M., Tiao G. Forecasting and Time Series Analysis Using the SCA Statistical System, vol. 1. Chicago, IL: Scientific Computing Associates Corporation; 1992. 35. McCleary R., Hay R. A. Applied Time Series Analysis for the Social Sciences. London: Sage Publications; 1980. 36. Balodis I. M., Potenza M. N., Olmstead M. C. Binge drinking in undergraduates: relationships with sex, drinking behaviors, impulsivity, and the perceived effects of alcohol. Behav Pharmacol 2009; 20: 518–26. 37. Bertakis K. D., Azari R., Helms L. J., Callahan E. J., Robbins J. A. Gender differences in the utilization of health care services. J Fam Pract 2000; 49: 147–52. 38. Li G., Lau J. T., McCarthy M. L., Schull M. J., Vermeulen M., Kelen G. D. Emergency department utilization in the United States and Ontario, Canada. Acad Emerg Med 2007; 14: 582–4. 39. Neighbors C., Foster D. W., Fossos N., Lewis M. A. Windows of risk: events and contexts associated with extreme drinking. In: Correia C. J., Murphy J. G., Barnett N. P., editors. College Student Alcohol Abuse: A Guide to Assessment, Intervention, and Prevention. Hoboken, NJ: John Wiley & Sons; 2012, pp. 53–80. 40. Lewis M. A., Neighbors C., Lee C. M., Oster-Aaland L. 21st birthday celebratory drinking: evaluation of a personalized normative feedback card intervention. Psychol Addict Behav 2008; 22: 176–85. 41. Labrie J. W., Migliuri S., Cail J. A night to remember: a harmreduction birthday card intervention reduces high-risk drinking during 21st birthday celebrations. J Am Coll Health 2009; 57: 659–63. 42. Hembroff L., Atkin C., Martell D., McCue C., Greenamyer J. T. Evaluation results of a 21st birthday card program targeting high-risk drinking. J Am Coll Health 2007; 56: 325–32. 43. Smith B. H., Bogle K. E., Talbott L., Gant R., Castillo H. A randomized study of four cards designed to prevent problems during college students’ 21st birthday celebrations. J Stud Alcohol 2006; 67: 607–15. 44. Glassman T., Dodd V., Kenzik K., Maureen M. E., Sheu J. Social norms vs. risk reduction approaches to 21st birthday celebrations. Am J Health Educ 2010; 41: 38–45. 45. Dimeff L. A., Baer J. S. Brief Alcohol Screening and Intervention for College Students (BASICS): A Harm Reduction Approach. New York, NY: Guilford Press; 1999. 46. Neighbors C., Lee C. M., Lewis M. A., Fossos N., Walter T. Internet-based personalized feedback to reduce 21stbirthday drinking: a randomized controlled trial of an event-specific prevention intervention. J Consult Clin Psychol 2009; 77: 51–63. 47. Neighbors C., Lee C. M., Atkins D. C., Lewis M. A., Kaysen D., Mittmann A. et al. A randomized controlled trial of eventspecific prevention strategies for reducing problematic drinking associated with 21st birthday celebrations. J Consult Clin Psychol 2012; 80: 850–62. 48. Carey K. B., Scott-Sheldon L. A., Carey M. P., DeMartini K. S. Individual-level interventions to reduce college student drinking: a meta-analytic review. Addict Behav 2007; 32: 2469–94. Addiction, 109, 1667–1675

Impacts of birthday drinking

Supporting information Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Figure S1 Number of alcohol-use disorder (AUD) in-patient/emergency department (ED) hospital admissions by age in weeks: total sample, females and males. The horizontal axis is age in weeks; for clarity and ease of

© 2014 Society for the Study of Addiction

1675

interpretation, the birthday weeks are labelled with the corresponding age in years Figure S2 Number of all in-patient/emergency department (ED) hospital admissions by age in weeks: total sample, females and males. The horizontal axis is age in weeks; for clarity and ease of interpretation, the birthday weeks are labelled with the corresponding age in years

Addiction, 109, 1667–1675