Abstract
Purpose
To estimate, using a novel interactive voice response (IVR) survey, the incidence of self-reported violence-related experiences of adolescent assault-injured patients in the weeks following Emergency Department (ED) discharge.
Methods
In an urban ED, a prospective cohort study with eight week follow-up IVR survey either weekly, bi-weekly or monthly after discharge was conducted with patients aged 12-19 years presenting with assault-related injuries. Survival analysis methods were used to estimate cumulative risks of self-reported violence experience within four and eight weeks.
Results
Ninety-five patients were enrolled; 42 (44.2%) reported to the IVR survey. As a result of the ED index event, an estimated 18.2% (CI=9.1-34.6%) reported being assaulted (no weapon), 2.9% (CI=0.4-19.1%) had been shot or stabbed, 20.7% (CI=10.9-37.3%) had assaulted someone else (no weapon), and 2.9% (CI=0.4-19.1%) shot or stabbed someone else. Additionally, 54.6% (CI=39.6-70.9%) had avoided going certain places, 47.0% (CI=32.5-64.1%) considered retaliating, 38.1% (CI=24.3-56.3%) had been threatened, and 27.0% (CI=15.4-44.6%) had carried a weapon. Most outcome occurrences happened within four weeks. There was evidence that intent to retaliate when asked at baseline was associated with an elevated risk of several outcomes.
Conclusions
The risk for subsequent violence among assault-injured adolescent ED patients appears high within weeks of discharge.
Keywords: violence, trauma, injury, adolescent, recidivism
Introduction
Assault-related trauma is a chronic, recurrent disease among America's adolescents. The potential for subsequent violence, heightened fear, and perceived need for weapon carrying may be particularly high among assaulted Emergency Department (ED) patients.[1] EDs remain an underutilized resource for identifying adolescents at greatest risk of subsequent injury and providing violence prevention resources.[2] To improve our understanding of risks adolescent patients face and factors that may identify those most needing intervention, we pilot tested an Interactive Voice Response (IVR) survey as a way to accomplish follow-up research with adolescent assault-injured patients.[3] This paper presents results of patients' self-reported violence experiences following ED discharge.
Methods
Recruitment and Baseline Data Collection
A prospective cohort study was conducted by interviewing 12-19 year-old patients treated after interpersonal assault (non-partner) in the ED of an urban, university-affiliated, tertiary care pediatric hospital during 2007-2008. Clinically-supervised research assistants conducted enrollment and explained to patients that our motivation was to ultimately help prevent violence and, more immediately, to learn more about violence-related experiences of adolescents in urban environments. Consenting patients were administered a 10-minute questionnaire using validated items.[4] Intent to retaliate was assessed by asking: “Not all fights are over after someone gets hurt. Do you think you will hurt someone because of this fight?”
Follow-up
Follow-up was accomplished with an IVR survey and randomly assigning patients to report either weekly, bi-weekly, or monthly, over eight weeks following discharge. This involved calling a toll-free telephone number, keying in a self-chosen access password, and keying Yes/No responses to recorded questions about violence-related experiences since discharge or since last reporting (published previously[5]).
Each patient received a gift card to a local convenience store chain. The card initially contained no value. During follow-up, the IVR system notified a study investigator each time a patient completed a report. Within 48 hours, the investigator electronically wired $5 to the card for each report that was made with a $10 bonus if all reports were completed.
Analysis
IVR data were analyzed using the product limit method and 95% confidence intervals (CI) to estimate the cumulative risk (ie, 1-survival) of each outcome within four and eight weeks of ED discharge.[6] Reported outcomes were assumed to have occurred on the date halfway between the report date and the date the patient was discharged from the ED or last reported. Stratified analyses evaluated whether a stated intent to retaliate when asked at baseline was associated with a differential risk of outcomes during follow-up. Stratified survival curves were evaluated for equality using the Peto-Peto-Prentice test[7, 8] with differences presented as relative risks by dividing the intent group risk by the no-intent group risk. No power analysis was conducted given this was a feasibility study. The study was approved by the Institutional Review Board of The Children's Hospital of Philadelphia and a Certificate of Confidential was obtained from the National Institutes of Health.
Results
773 patients were screened for eligibility, 131 were eligible, and 95 (72.5%) participated (Table 1). 42 (44.2%) patients completed at least one IVR report during follow-up and 13.7% made all requested reports. Details of follow-up have been reported.[3]
Table 1. Baseline Characteristics of Subjects.
| Characteristic | Subjects, % (n=95) |
|---|---|
| Sex | |
| Male | 63.2 |
| Female | 36.8 |
| Age, mean (SD) | 15 (1.6) |
| Race | |
| African American | 88.4 |
| Caucasian | 4.2 |
| Other | 7.2 |
| School Performance | |
| Grades in past year | |
| Mostly A's and B's | 24.2 |
| Mostly B's and C's | 41.1 |
| Mostly C's and D's | 17.9 |
| Mostly D's and F's | 8.4 |
| Other | 7.4 |
| Ever suspended | 77.9 |
| Ever skipped or cut school | 37.2 |
| Prior fights | |
| ≥1 other fight in past year | 51.6 |
| ≥1 other fight requiring medical attention in past year | 8.4 |
| Friends' behaviors in past year | |
| Carried knife or gun | 34.7 |
| Got into physical fight | 86.0 |
| Heard gunshot in past year | 85.3 |
| Witnessed someone being beat up in past year | 84.2 |
| Substance use in past 30 days | |
| Tobacco | 11.6 |
| Alcohol | 11.6 |
| Marijuana | 13.7 |
| Weapon used during index assault | |
| Weapon used by patient only | 0.0 |
| Weapon used by assailant only | 17.6 |
| Weapon used by both patient and assailant | 1.1 |
| No weapon used | 81.3 |
| Indend to retaliate | 32.6 |
SD: standard deviation.
Table 2 (top) reports the estimated four-week cumulative risk and eight-week cumulative risk of each of the 11 outcomes during follow-up. The four-week cumulative risk of being threatened, for example, was 30.9% (i.e., an estimated 30.9% had been threatened within four weeks) and an estimated 38.1% had been threatened within eight weeks of discharge. An estimated 18.1% of patients had threatened someone within four weeks and 21.5% of patients had threatened someone else, related to the initial event, within eight weeks. These risk estimates indicate that being threatened and threatening someone were quite common outcomes, and most instances regarding threats occurred within the first four weeks after receiving treatment for an assault-related injury. Several other outcomes were common and also occurred most often within 4 weeks.
Table 2. Self-reported Violence-related Outcomes During Eight Weeks Following Emergency Department Discharge.
| Cumulative Four-week and Eight-week Risks of Violence-related Outcomes | ||||
|---|---|---|---|---|
| Cumulative Risk | ||||
| Outcome | Four-week risk (95% CI) | Eight-week risk (95% CI) | ||
| Threats | ||||
| Been threatened | 30.9% (18.8%-48.0%) | 38.1% (24.3%-56.3%) | ||
| Threatened someone | 18.1% (9.0%-34.5%) | 21.5% (11.3%-38.8%) | ||
| Assaults | ||||
| Been beaten up | 18.2% (9.1%-34.6%) | 18.2% (9.1%-34.6%) | ||
| Beat up somone | 20.7% (10.9%-37.3%) | 20.7% (10.9%-37.3%) | ||
| Weapon-Related Assaults | ||||
| Been shot or stabbed | 2.9% (0.4%-19.1%) | 2.9% (0.4%-19.1%) | ||
| Shot or stabbed someone | 2.9% (0.4%-19.1%) | 2.9% (0.4%-19.1%) | ||
| Other | ||||
| Carried a gun or knife | 27.0% (15.4%-44.6%) | 27.0% (15.4%-44.6%) | ||
| Avoiding places due to index assault | 56.2% (40.8%-72.8%) | 56.2% (40.8%-72.8%) | ||
| Thought of hurting someone due to index assault | 47.0% (32.5%-64.1%) | 47.0% (32.5%-64.1%) | ||
| Sustained injury in a fight that required medical | 12.9% (5.6%-28.4%) | 12.9% (5.6%-28.4%) | ||
| Know someone who had been hurt in a fight | 51.3% (36.7%-67.8%) | 54.6% (39.6%-70.9%) | ||
| Cumulative Eight-week Risks and Relative Risks of Violence-related Outcomes Based on Stated Intent to Retaliate | ||||
|---|---|---|---|---|
| Intent to Retaliate | P-value | Relative risk | ||
| Outcome | Yes | No | ||
| Threatened someone | 53.3% (21.0%-91.6%) | 11.0% (3.7%-30.6%) | 0.016 | 4.9 |
| Carried a gun or knife | 48.6% (20.4%-85.5%) | 21.8% (10.3%-42.4%) | 0.074 | 2.2 |
| Beat up someone | 31.4% (11.3%-69.5%) | 14.5% (5.7%-34.4%) | 0.157 | 2.2 |
| Been beaten up | 31.3% (11.3%-69.5%) | 14.4% (5.7%-34.2%) | 0.182 | 2.2 |
P-values based on the Peto test of equivalence of survival curves.
Relative risks calculated by dividing the risk in the intent group by the risk in the no-intent group.
Table 2 (bottom) shows that among patients who had expressed intent to retaliate during the initial ED interview, the estimated eight-week cumulative risks of threatening someone, carrying a gun or knife, beating up someone, and being beaten up were 53.3%, 48.6%, 31.4%, and 31.3% respectively. These patients were an estimated 4.9 times more likely to threaten someone during the eight weeks following discharge than were patients not having an intent to retaliate. Large relative risk estimates (ie, >2) suggested evidence that an intent to retaliate was also associated with being more likely to carry a gun or knife, beat up someone, or be beaten up during the weeks following discharge, however the null hypothesis of survival curve equality for these outcomes was not rejected.
Discussion
This study found evidence that violence-related experiences among adolescents treated for assault-related injuries are common during the weeks after being discharged from the ED. The results also suggest that simply asking adolescents if they have a plan to retaliate may help identifying those at particular risk to carry a weapon, threaten someone, assault someone, or be assaulted during coming weeks.
Despite limitations that include risks for reporting bias at baseline and from attrition and non-response during follow-up, one study site, a small sample, and using a retaliation intent question having unknown validity and reliability, the research approach used here is compelling and can have advantages over other methods that rely on hospital records to study recidivism, eg [9]. The research approach and remuneration protocol were used based on their success in an earlier study of adult patient victims of intimate partner violence.[10] Additional results regarding feasibility of the present study have been reported.[3] Based on those findings and the evidence here that adolescents' risks of violence-related experience following ED discharge are high, large-scale studies that use IVR to better understand and potentially test effects of interventions that might be administered to this patient population appear warranted. It will be important for those studies to assess the reliability and validity of the IVR method and of questions to assess ED patients' intent to retaliate.
Acknowledgments
The study was completed with Crognale Funds from the Division of Emergency Medicine at The Children's Hospital of Philadelphia. A version of this manuscript was presented at the Annual Meeting of the Pediatric Academic Societies in Vancouver, British Columbia on May 3, 2010.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Fein JA, Kassam-Adams N, Gavin M, Huang R, Blanchard D, Datner EM. Persistence of posttraumatic stress in violently injured youth seen in the emergency department. Arch Pediatr Adolesc Med. 2002;156:836–40. doi: 10.1001/archpedi.156.8.836. [DOI] [PubMed] [Google Scholar]
- 2.Cunningham R, Knox L, Fein J, et al. Before and after the trauma bay: the prevention of violent injury among youth. Ann Emerg Med. 2009;53:490–500. doi: 10.1016/j.annemergmed.2008.11.014. [DOI] [PubMed] [Google Scholar]
- 3.Blackstone MM, Wiebe DJ, Mollen CJ, Kalra A, Fein JA. Feasibility of an interactive voice response tool for adolescent assault victims. Acad Emerg Med. 2009;16:956–62. doi: 10.1111/j.1553-2712.2009.00519.x. [DOI] [PubMed] [Google Scholar]
- 4.Dahlberg LL, Swahn M, Behrens CB. Measuring Violence-Related Attitudes, Behaviors, and Influences Among Youths: A Compendium of Assessment Tools. 2nd. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2005. [Google Scholar]
- 5.Blackstone MM, Wiebe DJ, Mollen CJ, Kalra A, Fein JA. Feasibility of an interactive voice response tool for adolescent assault victims. Acad Emerg Med. 2009;16 doi: 10.1111/j.1553-2712.2009.00519.x. Data Supplement S1. IVR follow-up questions. [DOI] [PubMed] [Google Scholar]
- 6.Rothman KJ, Greenland S. Modern epidemiology. 2nd. Philadelphia, PA: Lippincott Williams & Wilkins; 1998. p. 39.p. 289. [Google Scholar]
- 7.Peto R, Peto J. Asymptotically efficient rank invariant test procedures (with discussion) J Royal Stat Soc A. 1972;135:185–207. [Google Scholar]
- 8.Prentice RL. Linear rank tests with right censored data. Biometrika. 1978;65:167–79. [Google Scholar]
- 9.Sims DW, Bivins BA, Obeid FN, Horst HM, Sorensen VJ, Fath JJ. Urban trauma: a chronic recurrent disease. J Trauma. 1989;29:940–6. discussion 946-7. [PubMed] [Google Scholar]
- 10.Wiebe DJ, Carr BG, Datner EM, Elliott MR, Richmond TS. Feasibility of an automated telephone survey to enable prospective monitoring of subjects whose confidentiality is paramount: a four-week cohort study of partner violence recurrence after Emergency Department discharge. Epidemiol Perspect Innov. 2008;5:1. doi: 10.1186/1742-5573-5-1. [DOI] [PMC free article] [PubMed] [Google Scholar]