Trauma, once considered a sporadic event, has been shown to be a chronic condition (Poole, Griswold, Thaggard, & Rhodes, 1993). It has long been recognized that a proportion of hospitalized trauma patients have previous admissions for trauma—termed trauma recidivism (TR; Reiner, Pastena, Swan, Lindenthal, & Tischler, 1990).
Using alcohol, illicit drugs, and psychotropic medications (Cannon et al., 2014; Cherpitel & Ye, 2012; Nunn, Erdogan, & Green, 2016) is a well-known cause of traumatic events, regardless of the trauma mechanism (e.g., traffic crash, fall). The association between alcohol and TR has been consistently documented (Nunn et al., 2016). However, much less research has focused on the associations of other substances in TR (McCoy, Como, Greene, Laskey, & Claridge, 2013), links that currently remain unclear. Moreover, rigorous comparisons of results from previous studies are hindered by important methodological differences regarding the type of injury or trauma mechanism addressed (e.g., traumatic brain injuries [Vaaramo, Puljula, Tetri, Juvela, & Hillbom, 2014], intentional injuries [Sims et al., 1989], or motor vehicle collisions [Fabbri et al., 2005]), the sources of information used to identify TR and/or exposure to substances (questionnaires [Keough, Lanuza, Jennrich, Gulanick, & Holm, 2001; McCoy et al., 2013] or medical records of trauma [Caufeild et al., 2004; Sims et al., 1989]), or the time frame of recidivism (from 5 to 10 years; Fabbri et al., 2005; Keough et al., 2001; McCoy et al., 2013; Sims et al., 1989; Vaaramo et al., 2014). To our knowledge, findings about associations between consumption of alcohol and other drugs and TR for all types of trauma have not been published.
Prospective longitudinal studies are informative designs, but they are expensive and involve inherent difficulties with long-term follow-up. For this reason, approach based on past trauma history has been used as an alternative (Cordovilla Guardia et al., 2013; McCoy et al., 2013).
The purpose of this study was to quantify the strength of associations between the type of substances detected in patients admitted for traumatic injury and TR, defined as a history of past trauma.
Design and Setting
A cross-sectional study was performed to analyze the concurrent relationship between current drug use and TR in patients admitted to the University Hospital of Granada, Spain. The center is located in Andalusia, a region in southern Spain and southwestern Europe; it is the second largest autonomous community in Spain in terms of area and the largest in Spain in terms of population. Since 2011, our research group has implemented a program in our hospital for Screening, Brief Intervention, and Referral to Treatment (SBIRT) for alcohol- and drug- related trauma patients—a comprehensive approach for treating individuals with substance abuse problems (Babor et al., 2007). The SBIRT program includes screening on admission for alcohol and drugs of all patients aged between 16 and 70 years, admitted for traumatic injuries, and an interview with a brief intervention carried out by nurses during hospitalization. The brief intervention was offered to patients who had no clinical condition that may have interfered with efficacy (posttraumatic brain injury, mental disorder, spinal cord injury, or death during hospital stay). The study was approved by the research ethics committee of the University Hospital of Granada, Spain.
Data used for this study were obtained from the SBIRT database. Routine clinical and monitoring data are available. The number of participants eligible for inclusion in this study were all 1,798 patients admitted for traumatic injuries and recruited while the program was in operation during 32 nonconsecutive months from November 2011 to October 2012, June 2013 to November 2013, and June 2014 to March 2015. Of all admitted patients, 28 (1.6%) experienced a second trauma during the study period and were excluded. Of the remaining 1,770 patients, 1,156 (65.3%) could be screened for alcohol and drug use (see Table, Supplemental Digital Content 1, http://links.lww.com/NRES/A255), 592 patients were not screened, and 22 declined screening. Informed consent was required before screening. In patients who were sedated or unable to collaborate, informed consent was required from relatives or was requested from the patient when his or her clinical situation permitted.
Variables and Measurement
The presence of cannabis, cocaine, amphetamines, methamphetamines, benzodiazepines, opiates, methadone, barbiturates, or tricyclic antidepressants was screened in urine with a fluorescence immunoassay (LifeSign Status First). Review of the patient's medical records and direct questioning were used to distinguish between patients who screened positive for benzodiazepines and opioids as a result of emergency treatment of their injuries and patients who had taken these substances before they sought medical attention. Alcohol consumption was screened by analysis of blood samples, and blood alcohol levels higher than 0.3 g/L were considered positive. For patients from whom a blood sample could not be obtained, those who admitted drinking before the injury were considered positive.
Data for age, gender, TR, diagnosis of psychiatric disorder, mechanism of injury, severity of injury, and hospital mortality were collected during the patient's hospital stay by reviewing his or her medical record. The severity of injury was determined with the Injury Severity Score (ISS; Copes et al., 1988).
For all patients, TR and diagnosis of psychiatric disorder were obtained from the digital medical records maintained since 1999 by the Andalusian Health Service Database (Diraya; González Cocina & Pérez Torres, 2007).
The exposure variable (groups of detected drugs) was first dichotomized (negative vs. positive for any substance) and then categorized into six strata: (a) negative, (b) only alcohol, (c) only cannabis, (d) only psychotropic medications/opioids (including benzodiazepines, tricyclic antidepressants, barbiturates, and/or prescribed opioids), (e) only cocaine/amphetamines/methamphetamines, and (f) multidrug detection, including any combination of two or more of the abovementioned groups. Regarding the outcome variable (TR), patients were classified into three levels: single recidivist (SR) for patients with only one previous trauma, multirecidivist (MR) for patients with more than one previous trauma, and nonrecidivist for first-time trauma patients. The severity of trauma according to the ISS was categorized into three levels: 1–8 = mild, 9–15 = moderate, and 16 or greater = severe.
Baseline comparisons between subgroups were done with Kruskal–Wallis and chi-square tests. To quantify the strength of the association between TR and the presence of drugs, multinomial logistic regression was done with TR as the dependent variable (with the category nonrecidivist as the reference). Crude and adjusted odds ratios (aOR) for SR and MR outcomes were obtained in separate models. In the first model, the independent variable was dichotomous: “positive to any substance” (yes/no). In the second model, we included each of the five drug exposure subgroups defined above, with the “negative” category as the reference. To obtain the aOR, we included age, gender, and psychiatric disorders as independent terms in both models. For all estimates, we calculated 95% confidence intervals (95% CIs). All analyses were done with SPSS package (Statistics Base, Exact Tests, and Regression modules), version 22.0 (IBM Corp., Armonk, NY).
Among the 1,156 patients screened, alcohol was the substance detected most frequently (n = 271, 23.4%), followed by benzodiazepines (n = 181, 15.7%), cannabis (n = 153, 13.2%), tricyclic antidepressants (n = 68, 5.9%), cocaine (n = 64, 5.5%), opiates (n = 39, 3.4%), methadone (n = 23, 2%), barbiturates (n = 9, 0.8%), and amphetamines (n = 7, 0.6%). Five hundred twenty-one patients (45.1%) screened positive for any substance (see Table, Supplemental Digital Content 2, http://links.lww.com/NRES/A256): 159 (13.7%) for alcohol only, 62 (5.4%) for cannabis only, 145 (12.5%) for psychotropic medications/opioids only, 14 (1.2%) for cocaine or amphetamines only, and 141 (12.2%) for combinations of substances (Figure 1). Of all patients who screened positive for alcohol and illicit drugs, 310 (26.8%) were able to receive the motivational intervention from the SBIRT program nurses.
In between-group comparisons of demographic and clinical characteristics according to the type of drug consumed (Table 1), age was greatest in the psychotropic medication/opioid group, with a median age of 54.0 years (Q1 = 43.5, Q3 = 61.0, interquartile range = 17.5). This group was the only one in which women predominated (68.3%), and more than half of the patients (57.2%) were hospitalized for injuries related to falls on a level surface. With regard to injury severity, the largest percentage of severe injuries (ISS ≥ 16) was seen in the polydrug (22%) and alcohol (20.8%) groups, with an especially high level of mortality (7.5%) in the latter.
The recidivism rate (SR or MR) for all patients screened was 51.7% (n = 598). Among the recidivists, 53.5% screened positive for any substance, 28.6% screened positive for alcohol, 18.6% screened positive for benzodiazepines, 17.2% screened positive for cannabis, 7.2% screened positive for cocaine, 7% screened positive for tricyclic antidepressants, 4.3% screened positive for opiates, 2.3% screened positive for methadone, 0.7% screened positive for barbiturates, and 0.5% screened positive for amphetamines.
Prediction of TR
The presence of any substance was significantly associated with SR (aOR = 1.52, 95% CI [1.14, 2.04], p = .004) and MR (aOR = 3.17, 95% CI [2.29, 4.39], p < .001) outcomes (see Table, Supplemental Digital Content 3, http://links.lww.com/NRES/A257). When each substance was considered separately (Table 2), the SR outcome was significantly associated only with the polydrug group in both crude and adjusted analyses (aOR = 2.41, 95% CI [1.52, 3.84], p < .001). However, the MR outcome was significantly associated with all subgroups of substances. The strongest associations were found for the cocaine/amphetamine group (aOR = 7.15, 95% CI [2.02, 25.23], p = .002) and the polydrug group (aOR = 5.24, 95% CI [3.19, 8.59], p < .001).
The results of our study show that, in patients hospitalized for traumatic injury, who screened positive for alcohol, illicit drugs, and/or psychotropic medications/opioids, the frequency of previous traumatic injuries was higher than that in patients who screened negative—regardless of the type of substance detected. Our study design allowed us to identify two important findings not addressed in previous reports. First, it appears that, for any drug exposure considered, the strength of association is stronger for MR than for SR. Second, we identified the subgroups of drugs most strongly related with TR, and our results underscore the association between TR and multidrug use. It is also noteworthy that, among single-drug users, the cocaine/amphetamine subgroup had the strongest association with TR, whereas the weakest association was observed for alcohol users.
The recidivism rate in the sample of our patients who were screened (51.7%) is one of the highest reported to date (Nunn et al., 2016), surpassed only by the 89% TR rate in a study of patients recruited at chemical dependence clinics (Farley, Golding, Young, Mulligan, & Minkoff, 2004). Nevertheless, the percentage of recidivist patients exposed to alcohol (28.6%) was lower than that reported in a recent systematic review as an aggregate weighted estimate of 41.0% (Nunn et al., 2016). The lack of a universally accepted definition of TR and the different approaches used to detect alcohol consumption may explain these differences. In contrast, when any substance including, but not limited to, alcohol was considered, the TR percentage rate increased to 53.5%.
Our results show that alcohol was the substance detected most frequently (23.4%), followed by benzodiazepines (15.7%) and cannabis (13.2%). The frequency of substance use found in our sample of trauma patients was higher than in the general population in Spain, for which the estimated frequencies of consumers are 9.8% for alcohol (daily), 6.8% for sedatives-hypnotics, and 1.9% for cannabis (Observatorio Español de la Droga y las Toxicomanías, 2016). The distribution of drug users according to type of substance in our trauma patients showed variations compared with those found in previous studies (Dunham & Chirichella, 2012), mainly in that a higher proportion of our patients screened positive for benzodiazepines (15.7%). The influence of benzodiazepines on traumatic pathology is gaining interest. On one hand, the use of tranquilizers and sedatives, such as benzodiazepines, is increasing (Novak et al., 2016). Moreover, there is growing evidence that exposure to benzodiazepines is related to a higher risk of traffic crashes (Smink, Egberts, Lusthof, Uges, & de Gier, 2010), falls, and fractures—especially in older people (Xing et al., 2014).
The characteristics of our patients exposed to psychotropic medications/opioids differed from the rest of substance users: older women predominated, and falls were the most common mechanism of injury. This may explain why this group had the lowest proportion of severe trauma like traffic collision. In this study, we adjusted the association between TR and psychotropic medication use for preexisting psychiatric disorders—a well-known risk factor for traumatic injury (Cottrol & Frances, 1993; Keough et al., 2001). Therefore, this association cannot be attributed to the greater percentage of patients with psychiatric disorders in this group (46.9%).
Strengths and Limitations
Our study has several strengths. The design was not affected by the information bias present in studies that obtained data on drug exposure through questionnaires (Keough et al., 2001; McCoy et al., 2013) or from registries based on trauma records (Caufeild et al., 2004; Sims et al., 1989). In addition, our study was based on the implementation of an SBIRT program, with systematic screening of all trauma patients between 16 and 70 years old, without any other selection criteria. This comprehensive approach to drug detection in all patients prevented the selection bias that might have occurred if inclusion had depended on the emergency physician's decision to carry out drug testing or not based on suspected previous exposure.
The main limitation is the observational nature of the study; it does not allow confirmation of the causal nature of the relationships between patterns of drug use and TR. We measured the associations in a backward direction: Exposure was determined after the potential outcome was identified. Another limitation, shared with previous studies (Caufeild et al., 2004; Cherpitel & Ye, 2012; McCoy et al., 2013), is the possible presence of false-negative results for alcohol. Because laboratory testing for exposure to alcohol was not feasible for all patients, we were obliged to rely on self-report of drinking (or not drinking) from some patients. However, we believe that this classification bias would tend to bias the corresponding aOR toward the null.
Patients between the ages of 16 and 70 years who were hospitalized for traumatic injury and who screened positive for alcohol, illicit drugs, and/or psychotropic medications/opioids presented a higher frequency of TR compared with patients who screened negative, independently of age, gender, or previous psychiatric disorders. We believe that a deeper knowledge is needed of the effect that different substances (apart from alcohol) have on TR—especially substances in the cocaine/amphetamine and psychotropic medication/opioid groups. Our results support the potential usefulness of strategies such as the SBIRT program, which aims to prevent drug use in patients hospitalized for traumatic injuries.
Babor T. F., McRee B. G., Kassebaum P. A., Grimaldi P. L., Ahmed K., & Bray J. (2007). Screening, Brief Intervention, and Referral to Treatment (SBIRT): Toward a public health approach to the management of substance abuse. Substance Abuse
, 28, 7–30. doi:10.1300/J465v28n03_03
Cannon R., Bozeman M., Miller K. R., Smith J. W., Harbrecht B., Franklin G., & Benns M. (2014). The prevalence and impact of prescription controlled substance use among injured patients at a Level I trauma center. Journal of Trauma and Acute Care Surgery
, 76, 172–175. doi:10.1097/TA.0b013e3182ab10de
Caufeild J., Singhal A., Moulton R., Brenneman F., Redelmeier D., & Baker A. J. (2004). Trauma recidivism
in a large urban Canadian population. Journal of Trauma
, 57, 872–876. doi:10.1097/01.TA.0000135350.06670.60
Cherpitel C. J., & Ye Y. (2012). Trends in alcohol
- and drug-related emergency department and primary care visits: Data from four U.S. national surveys (1995–2010). Journal of Studies on Alcohol and Drugs
, 73, 454–458. doi:10.15288/jsad.2012.73.454
Copes W. S., Champion H. R., Sacco W. J., Lawnick M. M., Keast S. L., & Bain L. W. (1988). The Injury Severity Score revisited. Journal of Trauma and Acute Care Surgery
, 28, 69–77.
Cordovilla Guardia S., Rodríguez-Bolaños S., Guerrero López F., Lara-Rosales R., Pino Sánchez F., Rayo A., & Fernández-Mondéjar E. (2013). Alcohol
and/or drug abuse favors trauma recurrence and reduces the trauma-free period. Medicina Intensiva
, 37, 6–11. doi:10.1016/j.medin.2012.04.010
Cottrol C., & Frances R. (1993). Substance abuse, comorbid psychiatric disorder, and repeated traumatic injuries. Hospital & Community Psychiatry
, 44, 715–716.
Dunham C. M., & Chirichella T. J. (2012). Trauma activation patients: Evidence for routine alcohol
and illicit drug screening. PLOS ONE
, 7, e47999. doi:10.1371/journal.pone.0047999
Fabbri A., Marchesini G., Dente M., Iervese T., Spada M., & Vandelli A. (2005). A positive blood alcohol
concentration is the main predictor of recurrent motor vehicle crash. Annals of Emergency Medicine
, 46, 161–167. doi:10.1016/j.annemergmed.2005.04.002
Farley M., Golding J. M., Young G., Mulligan M., & Minkoff J. R. (2004). Trauma history and relapse probability among patients seeking substance abuse treatment. Journal of Substance Abuse Treatment
, 27, 161–167. doi:10.1016/j.jsat.2004.06.006
González Cocina E., & Pérez Torres F. (2007). La historia clínica electrónica. Revisión y análisis de la actualidad. Diraya: La historia de salud electrónica de Andalucía
[Electronic medical records. A review and analysis of the current situation. Diraya: Electronic medical records in Andalucía]. Revista Española de Cardiología
, 7, 37C–46C.
Keough V., Lanuza D., Jennrich J., Gulanick M., & Holm K. (2001). Characteristics of the trauma recidivist: An exploratory descriptive study. Journal of Emergency Nursing
, 27, 340–346. doi:10.1067/men.2001.116214
McCoy A. M., Como J. J., Greene G., Laskey S. L., & Claridge J. A. (2013). A novel prospective approach to evaluate trauma recidivism
: The concept of the past trauma history. Journal of Trauma and Acute Care Surgery
, 75, 116–121. doi:10.1097/TA.0b013e31829231b7
Novak S. P., Håkansson A., Martinez-Raga J., Reimer J., Krotki K., & Varughese S. (2016). Nonmedical use of prescription drugs in the European Union. BMC Psychiatry
, 16, 274. doi:10.1186/s12888-016-0909-3
Nunn J., Erdogan M., & Green R. S. (2016). The prevalence of alcohol
-related trauma recidivism
: A systematic review. Injury
, 47, 551–558. doi:10.1016/j.injury.2016.01.008
Poole G. V., Griswold J. A., Thaggard V. K., & Rhodes R. S. (1993). Trauma is a recurrent disease. Surgery
, 113, 608–611.
Reiner D. S., Pastena J. A., Swan K. G., Lindenthal J. J., & Tischler C. D. (1990). Trauma recidivism
. American Surgeon
, 56, 556–560.
Sims D. W., Bivins B. A., Obeid F. N., Horst H. M., Sorensen V. J., & Fath J. J. (1989). Urban trauma: A chronic recurrent disease. Journal of Trauma and Acute Care Surgery
, 29, 940–947.
Smink B. E., Egberts A. C., Lusthof K. J., Uges D. R., & de Gier J. J. (2010). The relationship between benzodiazepine use and traffic accidents: A systematic literature review. CNS Drugs
, 24, 639–653. doi:10.2165/11533170-000000000-00000
Vaaramo K., Puljula J., Tetri S., Juvela S., & Hillbom M. (2014). Head trauma sustained under the influence of alcohol
is a predictor for future traumatic brain injury: A long-term follow-up study. European Journal of Neurology
, 21, 293–298. doi:10.1111/ene.12302
Xing D., Ma X. L., Ma J. X., Wang J., Yang Y., & Chen Y. (2014). Association between use of benzodiazepines and risk of fractures: A meta-analysis. Osteoporosis International
, 25, 105–120. doi:10.1007/s00198-013-2446-y