Article Text

Did attending P.A.R.T.Y. change youth perceptions? Results from 148 Queensland schools participating in the Prevent Alcohol and Risk-Related Trauma in Youth Program, 2018–2019
Free
  1. Cate M Cameron1,2,
  2. Rob Eley3,4,
  3. Chantelle Judge3,
  4. Roisin O'Neill5,
  5. Michael Handy1,5
  1. 1 Jamieson Trauma Institute, Metro North Health, Herston, Queensland, Australia
  2. 2 Centre for Healthcare Transformation, Australian Centre for Health Services Innovation (AusHSI), Queensland University of Technology - QUT, Brisbane, Queensland, Australia
  3. 3 Emergency Department, Princess Alexandra Hospital, Woolloongabba, Queensland, Australia
  4. 4 Faculty of Medicine, The University of Queensland, Saint Lucia, Queensland, Australia
  5. 5 Trauma Service, Royal Brisbane and Women's Hospital, Herston, Queensland, Australia
  1. Correspondence to Associate Professor Cate M Cameron, Jamieson Trauma Institute, Metro North Hospital and Health Service, Herston, QLD 4029, Australia; cate.cameron{at}health.qld.gov.au

Abstract

Background Prevent Alcohol and Risk-Related Trauma in Youth (P.A.R.T.Y.) is an immersive 1 day in-hospital injury awareness and prevention programme designed to educate high-school students on the consequences of a variety of risk-taking behaviours. This multisite contemporary analysis examined differences in programme effect and temporal changes on participant knowledge and attitudes.

Methods Metropolitan and rural schools were invited to attend the programme at one of the 11 hospital sites throughout Queensland, Australia. Pre–post study design with participant questionnaires provided at three time periods: immediately preprogramme and postprogramme, and 4 months later. The questionnaire used scenarios to determine a participant’s opinion on the safety of drugs/alcohol, driving and risk-taking activities, using Likert scales.

Results A total of 5999 students participated in the programme between 1 January 2018 and 31 December 2019. Responses to all questions related to safety, harm or risk followed a similar pattern. The immediate postcourse responses demonstrated significant increased awareness of risk or change in action, followed by a decay at 4 months to within 10% of preprogramme levels. Public school students, males and students from Central and North Queensland demonstrated lower risk-aversion (p<0.05).

Conclusion This study demonstrated across more than 100 school sites, the positive change in knowledge and student participant attitudes towards risk-taking behaviours after attending the P.A.R.T.Y. programme. The need to address the significant decay at the 4-month follow-up was identified. Findings offered potential for tailoring of messaging to target key demographic groups/topics where the decay was greatest.

  • education
  • adolescent
  • school
  • alcohol
  • attitudes
  • program evaluation

Data availability statement

Data may be obtained from a third party and are not publicly available. Data obtained from consented participants are held by RBWH, Metro North HHS, Qld, Australia.

Statistics from Altmetric.com

Introduction

Young adults are predisposed to risk-taking and sensation seeking behaviours due to incomplete frontal cortex maturation, resulting in increased motivation to pursue rewards without thought of consequences.1 Throughout the world, young adults are over-represented in preventable deaths and permanent disability from injuries primarily due to transport crashes, violence, drug and alcohol misuse,2 and are disproportionately involved in rural and remote road trauma.3 In 2017–2018 in Australia, injury accounted for 809 deaths and 73 219 hospitalisations for persons aged 15–24 years.4

Prevent Alcohol and Risk-Related Trauma in Youth (P.A.R.T.Y.) is an immersive 1 day in-hospital injury awareness and prevention programme designed to educate high-school students on the consequences of a variety of risk-taking behaviours and to enable them to recognise and mitigate risk. The programme is run in the clinical areas of a hospital by the emergency services and health professionals directly involved in the management of trauma patients. Participants follow the typical journey of a trauma patient, and depending on the hospital site, may visit the Emergency Department, Intensive Care Unit, Trauma Ward, Morgue and Allied Health gyms. In addition to lectures and demonstrations, participants meet with injury survivors who have traumatically sustained a permanent disability.

The P.A.R.T.Y. programme has been running since 2010 in South East Queensland, coordinated from the Royal Brisbane and Women’s Hospital (RBWH) in Brisbane. From 2015, several hospital sites were added, and modifications made to the questionnaire, including additional risk-taking related attitudinal questions. The current programme runs across 11 Queensland hospital sites with over 140 participating schools. In 2019, a school-based programme was trialled in seven schools. No previous P.A.R.T.Y. programmes, in Australia or internationally, have included multisite data comparisons or the new data collection questionnaire. This contemporary analysis enabled more detailed examination of determinants and differences in programme effect and temporal changes on participant attitudes towards risk-taking behaviours after attending the P.A.R.T.Y. programme.

Methods

Study design

Prestudy and poststudy design with participant questionnaires provided at three time periods: preprogramme (T1), postprogramme (T2) and 4 months postprogramme (T3).

Program sites

P.A.R.T.Y. Programmes conducted between 1 January 2018 and 31 December 2019 were included in this analysis. The 11 participating hospital sites were RBWH, Princess Alexandra (PAH), Gold Coast, Logan, Rockhampton, Sunshine Coast, Gympie, Toowoomba, Townsville, Cairns and Mackay Hospitals. Those sites ran 139 events, for 5370 students, from 141 different schools. In addition, in 2019, 7 events for 629 students were run in seven schools in the Brisbane region. A total of 148 schools participated.

Participating schools and students

All high schools within the region surrounding each hospital site were invited to participate in the programme. Individual participants were selected by the school. They were predominately students aged 15–19 attending years 10–12. Schools did however have the discretion to include both year 9-year-old and 14-year-old children. The schools were advised to exclude any student who had been involved in significant personal or family trauma within the previous 2 years. All participants were asked to participate in the research component of the programme. However, students could still attend the programme even if they did not consent to participate in the research component and could withdraw from that component at any time. The school principal, parent/guardian and participants all provided informed consent to participate.

Survey data collection

Data were collected from a single questionnaire administered at the three time points, modified from the original Sunnybrook Hospital P.A.R.T.Y. programme (http://partyprogram.com/) to include other risk-taking related attitudinal questions. Questionnaire changes were based on Stakeholder consultation to establish face validity. Additional demographics consisted of age, school year level, gender, and driver licence status. Information collected on the participating schools identified their geographic location (North, Central and Southeast Queensland), rurality (urban and rural) and school type (public, private or independent). The questionnaire used scenarios to determine a participant’s opinion on the safety of drugs and alcohol use, driving and risk-taking activities, using 5-point Likert scales. Additional questions asked what action the participant would take within the scenario. There were 10 questions, with subsections, for a total 16 items (online supplemental table 1). Participants were asked to complete the preprogramme and postprogramme questionnaires onsite, and the follow-up questionnaire at their respective school. Unique numeric identifiers were added to questionnaires to enable matching all three questionnaires. Only researchers had access to deidentified data for analysis.

Data analysis

Data cleaning and analyses were undertaken using SPSS Statistics, V.24. Pearson χ2 and Fisher exact tests were used to describe school and participant characteristics. Differences in change in knowledge and risk-taking behaviours preprogramme and postprogramme, and at follow-up were assessed using Friedman Test for related samples, followed by Wilcoxon Signed Rank Test for post-hoc analysis among the three time periods; Mann-Whitney U Test to compare gender (male, female) school type (private, public) and rurality (rural, metropolitan); Kruskal-Wallis Test, followed by the post-hoc Mann-Whitney for unrelated samples of region (North, Central and Southeast), and school year (10, 11, 12), at each time period. Alpha level for post-hoc tests were adjusted from 0.05 to reflect number of tests and effect sizes followed Cohen’s d where 0.1 is small, 0.3 medium and >0.5 large. Complete case analysis was conducted. Independent schools were excluded from analysis of school type due to small sample size of only three schools, and year 9 from the school year comparisons as only 12 students were represented.

Results

Baseline characteristics

A total of 5999 students participated between 1 January 2018 and 31 December 2019. The majority of schools (65.5%) and students (58.7%) were from Southeast Queensland (table 1). Public schools represented 56.0% of the schools and 48.0% of the participants. In all geographical regions, more females participated than males. Median age was 16 years with almost half of the students (48.8%) in year 11. Rural schools contributed 32% of the participants. A total 3397 students (56.6%) completed all surveys at the three time points. There were no significant differences in the participant or school characteristics of complete responders compared with those participants who did not complete the 4-month follow-up (p>0.05) (results not shown).

Table 1

Programme locations, participant numbers and characteristics

Response patterns

Significant differences were found in responses at the different time periods (p<0.001). Post hoc analysis between precourse and postcourse, pre-4-months, post-4-months were also significantly different; however, with few exceptions effect sizes were small (<0.3). Responses to all questions related to safety, harm or risk followed a similar pattern. The immediate postcourse responses demonstrated increased awareness of risk or change in action, followed by a decay at 4 months. As exemplified in figure 1, baseline awareness of risk or harm differed according to topic. Only 10% of students considered alcohol consumption (figure 1, Q1) to be extremely risky. Although this increased to 30% immediately postcourse, by 4 months postcourse, responses to this question almost attained precourse levels. In contrast, jumping into a pool from height (Q3) started with a much greater awareness of risk at 67% (figure 1, Q3). Even with this recognised high-risk activity which was also seen in use of drugs methylenedioxymethamphetamine or ecstasy (MDMA) (Q4) and driving under the influence of alcohol (Q6), increased awareness and decay occurred at T2 and T3. Intermediate to these low and high-risk behaviours was smoking marijuana (figure 1, Q2) which was considered to be harmful by 32% of students. The questions related to action taken by the participant, for example, ask a speeding driver to slow down or if willing to drive when potentially under the influence of alcohol, also demonstrated similar profiles at T1, T2 and T3.

Figure 1

Responses to questions 1–3 at each time point.

Percentage change in responses

Table 2 presents the percentage change in responses at the three times for the two points on the Likert scale which reflects the most risk-averse views or most positive action. For example, at T1 75.64% of participants considered alcohol either ‘extremely risky’ or ‘slightly risky’. At T2 this increased to 85.19%; an increase of 12.63%. At T3 the percentage had fallen to 74.04% representing a decay of 15.06% over T2. Relative to T1 the T3 value was a 2.12% decrease.

Table 2

Changes to responses at three time points for the two highest risk-averse options or actions on scales

The smallest change of 0.66% (Q8a) was for the perception of risk if a driver of a car was reading a text message, while driving. Almost all students at all time points considered this to be extremely high risk (>97%), whereas the largest increase immediately postprogramme, 53.90% (Q9b), was not to ride in an overcrowded car. The knowledge about how unsafe overcrowding was (Q9a) 90% at T1 yet at the same time, only 47% said it was unlikely they would get in the car (Q9b).

Decay occurred for every question by 4 months. The extent of decay ranged from 1.34% for Q8a to 24.01% for Q10d which asked about the likelihood of driving to work when potentially still under the influence of alcohol. For most questions, the decay reinstated preprogramme levels to within 10%. The exceptions being Q9b (not riding in an over-crowded car) and Q10a (whether the P.A.R.T.Y. programme would influence choices in the future).

Risk-aversion

Significant differences with small effect sizes were determined for gender, school type, school year and geographic region. Females, private school pupils and year 12 pupils were more risk-averse than males, public school and year 10 and 11 pupils, respectively. The gender difference is demonstrated in figure 2 for Q1. Analysis revealed that males were less risk-averse than females at all three-time periods (Mann-Whitney (Z) p<0.001; effect sizes small 0.14, 0.18, 0.11). Similarly, pupils from private schools were more risk-averse at all three-time periods (eg, Q1 Mann-Whitney (Z) p=0.003, p<0.001, p=0.003; effect sizes small 0.04, 0.05, 0.05).

Figure 2

Response to question 1 at each time point, by gender.

Participants from the Central region, and on some items those from the North, were less risk-averse than Southeast areas of Queensland for questions related to transport, Q7–9. For example, in response to Q8 those pupils were less likely think it was risky pre and post (p<0.001) for a driver to read a phone text message while driving.

There was no difference in response to the question related to the perceived risk of alcohol consumption (Q1) between students without a driving licence and those with either a learner or provisional licence (p>0.17 at all three time periods). However, responses related to alcohol and driving did elicit small differences in response. At T3 students with a licence were slightly more likely (96.5% vs 94.3%) to consider it unsafe (p=0.005) for a friend to drive after consuming alcohol (Q6). As to whether they would drive to work after a night drinking, those with driving licences were more likely to state it was extremely unlikely or unlikely they would drive than those without licences (62.1% vs 57.4%, p=0.025 at T1; 83.8% vs 79.3%, p=0.004 at T2). At T3 both groups had decayed to 65.9% and 66.1%, respectively, with no difference between them (p=0.91).

Males identified consuming more alcohol postcourse (p=0.042) and at 4 months (p=0.030) than did females; however, the effect sizes were small (0.027 and 0.037). Similarly, pretest Y10 students were far more likely to state they never drank alcohol (65%) compared with Y11 (46%) and Y12 (35%). Significantly more public-school students indicated drinking preprogramme and postprogramme (p<0.001); however, again the effect sizes were small (0.093 and 0.018).

In response to the question whether the P.A.R.T.Y. programme would make participants think differently about future choices (Q10) preprogramme, only 19% said it was extremely likely. This increased to 60% immediately postprogramme and decayed to 47% by 4 months. This question elicited a different response from participants attending the seven school-based programmes. Although there was a rise from preprogramme to postprogramme, fewer school-based participants indicated they were extremely likely to be influenced by the programme (effect sizes for Q10A was 0.401 (pre) and 0.361 (post)). For all other questions, the results did not change when school-based participants were excluded from the analysis.

Discussion

This is the first study to collectively analyse P.A.R.T.Y. data from 11 different hospital sites providing the programme to 148 different schools, reaching almost 6000 students across Queensland. This contemporary analysis enabled more detailed examination of school and participant differences in programme effect and temporal changes on participant attitudes towards risk-taking behaviours before and after attending the programme.

P.A.R.T.Y. is one prevention programme that has demonstrated reductions in injury rates among participants.5 6 A Canadian, 10-year longitudinal study demonstrated participants remained injury free for longer compared with the control group (898 days vs 436 days; p<0.0001).5 A Western Australian study demonstrated juvenile justice offenders who participated in P.A.R.T.Y. had significantly fewer traumatic injuries and lower numbers of subsequent traffic or violence related offences compared with a control group (3.6% study group vs 26.8% control; p<0.001).6 A single-site study in Victoria, showed significant improvements among knowledge and attitudes towards safety perceptions of driving after alcohol use, seat belt use and risk-taking activities postprogramme; however, the results demonstrated decay in knowledge 3–5 months later.7 The risk of decay in youth learning from behaviour intervention programmes has been demonstrated in multiple settings8–11 and the benefit of the inclusion of learning booster programmes in behaviour interventions has been suggested for sustaining intervention effects.9 12

The immediate effectiveness of the P.A.R.T.Y. programme in changing knowledge and intent was demonstrated in the current multisite analysis, consistent with other reports from Australia and internationally.6 7 13 Participants showed more awareness for every situation immediately postcourse, the extent of increase depending on the baseline value. It is possible some of the positive responses at all time points may have been the result of social desirability bias where the students answered questions based on what answers they thought were desired by the P.A.R.T.Y. programme.14 15 Despite this immediate positive effect, again consistent with other studies,6 7 16 on every item, there was temporal decay17 in risk-perception and knowledge by the 4-month follow-up.

Addressing this temporal decay is the goal for future work. One proposed method is the provision of a follow-up booster component. If added to the P.A.R.T.Y. programme this may help to reinforce learnings and minimise knowledge decay,9 12 with potential for tailoring of messaging to target key demographic groups/topics where the decay is greatest. Options include targeted messaging and the delivery in an online or app format using smartphone technology. There is potential to include gamification strategies, which have been recently shown to increase the effect of injury prevention messaging and reduction in decay of messaging temporally.18 While there was no participant involvement in this analysis, stakeholder consultation is planned for any programme development or changes.

The analysis of different student and school characteristics, with a mix of public, private schools; a range of geographic regions; as well as urban and rural locations; provided insight to areas in the programme where differences in risk perception and knowledge decay warrant targeted reinforcing of key messages. The findings demonstrated the questions with the greatest decay, related to scenarios for taking action to mitigate risk or act safely, despite retaining a sustained knowledge of that risk being extremely high. It is these areas that warrant further targeted strategies to reinforce learnings to support young people to make safer choices.

While effect sizes were small, some school and student characteristics demonstrated significant differences in risk-aversion. Students from private schools consistently demonstrated higher risk-aversion, at all-time points. This may reflect increased resources, health and welfare service provision in the private school sector which has been shown to be associated with fewer risk-taking behaviours among students.19 There was lower male participation, in both urban and rural areas. Males that did participate consistently demonstrated higher risk-taking responses than females. As students self-selected participation, it is possible that the males that participated were lower risk takers. Had more males high risk-taking students participated, the gender difference may have been more pronounced. Given the disproportionately higher representation of young males across most injury types and injury mechanisms,20 there is a need to develop better strategies to engage young males in prevention programmes and consider new methods for targeted learning and reinforcement.

Students located in Central and North Queensland and rural areas also showed lower risk-aversion on questionnaire items, particularly on driving related risks. These differences potentially reflect the fact that some rural students learn to drive at younger ages on private rural properties and may feel more confident driving than their metropolitan counterparts. Rural student views may also reflect a perceived necessity where travel distances are greater, or there is a lack of available alternative public transport options.

While the number of included schools receiving the modified school-based delivery of the programme was limited, the question on whether the P.A.R.T.Y. Programme ‘will make you think differently about your choices in the future’ showed less effect for school-based delivery compared with hospital programmes. Although the school-based programmes have the capacity to reach greater numbers of students, and in recent and ongoing COVID-19 pandemic times, hospital-based programme delivery may be affected or ceased, strategies to increase the effectiveness of school-based delivery are warranted.

Limitations

The study has several limitations. As with any self-reported responses to questionnaires, there is a potential for bias, including participants may give the more socially acceptable answer rather than being completely truthful. However, as this was a longitudinal study with the same participants at three time points, any bias is likely to remain consistent.

Participants to the programme self-selected through their schools with resultant limitations inherent in that process. It is possible that students with a particular interest, or whose parents instructed them to attend, did so. They thus may not be totally representative of the broader student population. Almost two-thirds of the participants were females. However, sex differences within the response scales were extremely small, thus supporting the contention that the different proportions of males and females did not influence the conclusions. While the non-response rate at 4 months was considerable, there were no significant differences in the participant or school characteristics of complete responders and non-responders, and therefore less likely to have introduced bias.

Those questions which required a response indicating what action a participant would take, used the stated intent as a proxy for risk-aversion and safe actions. An alternative could be including a validated risk propensity tool that measures social and health risks in adolescents.

Conclusion

This study demonstrated across more than 100 school sites, the positive change in knowledge and participant attitudes towards risk-taking behaviours after attending the P.A.R.T.Y. programme, but also clearly showed the significant decay at the 4-month follow-up. The results also identified key demographic groups where the decay was greatest which provides opportunity to tailor messaging in future programmes.

What is already known on the subject

  • Young adults are over-represented in preventable deaths and permanent disability from injuries.

  • Prevent Alcohol and Risk-Related Trauma in Youth (P.A.R.T.Y.) is one prevention program that has demonstrated a reduction in injury rates among participants.

What this study adds

  • Findings demonstrated that the P.A.R.T.Y. program run in multiple venues state-wide, resulted in positive change in knowledge and participant attitudes towards risk-taking behaviours by school students.

  • The significant decay at follow-up highlighted the need to address temporal decay, and findings offered potential for tailoring of messaging to target key demographic groups/topics where the decay was greatest.

Data availability statement

Data may be obtained from a third party and are not publicly available. Data obtained from consented participants are held by RBWH, Metro North HHS, Qld, Australia.

Ethics statements

Patient consent for publication

Ethics approval

The P.A.R.T.Y. programme was first approved by RBWH Ethics and Research (10/QRBW/146). Subsequent ethics approval was obtained for the present analysis (LNR/2018/QRBW/48712).

Acknowledgments

Ms Maura Desmond, previous Statewide P.A.R.T.Y. Program Manager and Ms Rhiannon Ward, Undergraduate Health Sciences student, University of Queensland for research support. All participating hospital sites, program contributors and participating schools and students.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Contributors CMC jointly conceived and assisted with conducting this study. She led the study development and approvals; funding applications; provided interpretation of the results and led the manuscript development. RE jointly conceived and assisted with conducting this study. He provided input into the study development; funding applications; led the statistical analysis, provided interpretation of the results, and contributed to the manuscript development and review. CJ conducted the statistical analysis and contributed to the manuscript revision and review. RO’N provided invaluable insights as the Statewide P.A.R.T.Y. Program Manager, contributed to the study development, interpretation of the results and contributed to the manuscript revision and review. MH jointly conceived the study, contributed to the study development, interpretation of the results and contributed to the manuscript revision and review.

  • Funding The Royal Brisbane and Women’s Hospital (RBWH) hold the P.A.R.T.Y. multi-site licence and provide administration and coordination of all programs in Queensland. The P.A.R.T.Y. Program is funded in partnership with the Queensland Department of Transport and Main Roads and past funding from Australian Associated Motor Insurers Limited (AAMI). This study was supported by a Brisbane Diamantina Health Partners (BDHP) seed grant (no grant number).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.