- Original Contribution
- Open Access
Trauma recidivism at an emergency department of a Swedish medical center
© The Author(s). 2016
- Received: 8 March 2016
- Accepted: 17 August 2016
- Published: 12 September 2016
To inform targeted prevention, we studied patterns of trauma recidivism and whether a first injury predicts the risk for a recurrent injury.
In a population-based study of 98,502 adult injury events 1999–2012, at the emergency department of Umeå University Hospital, Sweden, we compared non-recidivists with recidivists in terms of patients’ sex, age, type of injury and severity of the injury.
Thirty-six percent of all patients suffered recurrent injuries, which were associated with a higher proportion of inpatient care and more hospital days. Young men and elderly women were at the highest risk for trauma recidivism. At 20 to 24 years, men had a 2.4 (CI 95 % 2.3–2.5) higher risk than women, a 90 years old woman had almost a 10-fold higher risk for another moderate/severe injury than a 20 years old one. A fracture were associated with a hazard ratio of 1.28 (CI 95 % 1.15–1.42) among men younger than 65 years and 1.31 (CI 95 % 1.12–1.54) for men older than 65 years for a subsequent moderate/severe injury. For women younger than 65 years a fracture was associated with a hazard ratio of 1.44 (CI 95 % 1.28–1.62) for a subsequent moderate/severe injury. A sprain carries a higher risk for a new moderate/severe injury for both men and women and in both age groups; the hazard ratio was 1.13 (CI 95 % 1.00–1.26) for men younger than 65 years, 1.42 (CI 95 % 1.01–1.99) for men older than 65 years, 1.19 (CI 95 % 1.05–1.35) for women younger than 65 years and 1.26 (CI 95 % 1.02–1.56) for women older than 65 years. A higher degree of injury severity was associated with a higher risk for a new moderate/severe injury.
Trauma recidivism is common and represents a large proportion of all injured. Age and sex are associated with the risk for new injury. Injury types and severity, also have implications for future injury.
- Descriptive epidemiology
Injury is a major public health problem. In the European Union (EU), 230,000 trauma casualties are recorded annually and everyday more than 112,000 people are admitted to hospital for injuries (Bauer et al. 2014). In the Unites States (US), injury is the dominant killer below the age of 46 (Rhee et al. 2014). But not only severe injuries put a burden of injury on society. Minor and moderate ones are more common and, even if less severe, this injuries account for more than a third of the burden of injuries according to Disability-adjusted life years (DALYs) (Polinder et al. 2012). Improvements in treatment and prevention of cancer and cardiovascular disease during recent decades have left trauma as a proportionally more important cause to the total burden of ill health.
Traditionally injuries have been regarded as occurring by chance and thus not possible to prevent. But injury incidence is not randomly distributed e.g., boys are more injury-prone than girls and elderly women are more affected by injury than elderly men (Saveman and Bjornstig 2011; Hedstrom et al. 2012). We also know that social deprivation, substance abuse and mental illness increase the risk for injury (Gentilello et al. 1995; Wan et al. 2006; Fleury et al. 2010).
Emergency department staff know their recidivist patients. Reiner et al. surveyed ED staff and found that trauma recidivists more often were male, young and under the influence of alcohol or drugs, and introduced the concept of trauma recidivism (Reiner et al. 1990). Several studies have confirmed these results, (Poole et al. 1993; Hedges et al. 1995; Williams et al. 1997; Sayfan and Berlin 1997; Kaufmann et al. 1998; Keough et al. 2001; Caufeild et al. 2004; Worrell et al. 2006; Brooke et al. 2006; Claassen et al. 2007; Toschlog et al. 2007; Kwan et al. 2011; Allan et al. 2013; Rittenhouse et al. 2015). As a response, several trauma centers developed intervention programs. Brief interventions shortly after the injury have been shown to halve the risk of renewed alcohol-related injury (Schermer et al. 2006; Gmel et al. 2007). Violence intervention programs has showed a fourfold to sixfold decrease of trauma recidivism and with good cost-effectiveness (Cooper et al. 2006; Smith et al. 2013; Juillard et al. 2015). Other examples of successful intervention are fall prevention in geriatric care and fracture prevention programs (Lih et al. 2011; Sach et al. 2012; Van der Kallen et al. 2014). Intervention programs seem to working better if implemented shortly after the injury event (Gentilello et al. 1995; Smith et al. 2013).
For targeted prevention knowledge about risk factors is crucial. Most studies on trauma recidivism regard cases of severe injury from urban level 1 trauma US hospitals. There is a lack of knowledge on other aspects on trauma recidivism as only a few studies address trauma recidivism in a rural populations and a wider age perspective. These studies indicate that female sex and older age may be risk factors for trauma recidivism (Williams et al. 1997; Toschlog et al. 2007). Another rural population study showed a recidivism rate of 12 % during just 1-year follow up. The fact that not only severe injuries contribute to the burden of ill health, emphasizes the need for studies, which also including less severe injuries, since these are more common.
Our objective was to estimate the burden of trauma recidivism in a stable population including less severe injuries and to investigate whether certain conditions associated with an injury are risk factors for a subsequent injury.
A retrospective, population-based, longitudinal study on acute adult trauma, 1999–2012.
Since 1993, almost all trauma emergencies handled at Umeå University Hospital are registered in an injury database. The register is part of the European Injury Database (IDB). (http://www.socialstyrelsen.se/register/halsodataregister/patientregistret/idbsverige). The hospital is responsible for all emergency medical services for the Umeå region (n = 145,000), Northern Sweden. The population consists of the residents in the A-64 region registered by the Swedish Tax Agency associated with the municipalities Umeå, Robertfors, Vindeln, Vännäs, Bjurholm and Nordmaling. Only residents registered by the Swedish Tax Agency were included in the study. (The inhabitants of Sweden are obliged by law to inform the Tax Agency about their address of residence within 1 week after moving to a new area) and all visitors from others parts of Sweden or other countries were excluded. In the emergency room, the patient or a proxy fills in a voluntary questionnaire about the circumstances around the injury. If needed, the injury surveillance group at the hospital supplements the information from medical files, ambulance records and police reports. Annually, about 7300 cases per year were recorded. The database was validated every year for missing cases by the hospital’s patient registry for inpatient cases and by billing information from the Emergency Department for outpatient cases. There were 2.3 % missing cases.
We merged injury types into ten main groups; fracture, contusion, wound, sprain, concussion, foreign body, thermo, strain, poisoning and other. The Injury Severity Score (ISS) was used: 1–3 for mild, 4–8 for moderate, and ≥9 for severe injuries (Baker et al. 1974). The ISS is based on the AIS (abbreviated injury scale) (Association for the Advancement of Automotive Medicine 2001), a 6-grade scale, where AIS 1 is a mild injury and AIS 6 denotes a fatal one. An example of AIS 1 is a superficial wound and AIS 3 a femur fracture. The three most severely injured body regions have their AIS number squared and added up to give the ISS score. Recurrent moderate injury was defined as a recurrent injury at least ISS 4 level. This study was approved by the Regional Ethical Review Board in Umeå (2013-61-31 M). A recurrent injury was defined as a new injury after a first injury during the study period.
Incidence of the first and recurrent injury was calculated by dividing the number of injuries by the average midyear population for the A-64 region for each age group during this period (Statistics Sweden). Incidence rate ratios were calculated with 95 % confidence interval. Analyses of mean values (age and mean hospital days) were made with two-sided independent t-test and for comparison of proportions (sex and proportion outpatient/in-hospital) chi-squared test was used. To investigate the association between independent variables (age, injury types and injury severity) and risk of recurrent moderate or severe injury, Cox regression analysis was used. For injury types each category was controlled against all other categories as the reference. An individual with no recurrent injury within 5 years was considered censored. The data of the database were supplemented with death date from the tax agency register. Multiple Cox regression analysis was used to analyze association and to adjust for age as a confounding factor, at a rejection level of 5 %.
Descriptive data for non-recidivists and recidivists, year 1999 to 2012
37,396 (63.5 %)
21,526 (36.5 %)
19,358 (51.8 %)
12,069 (56.1 %)
18,038 (48.2 %)
9,457 (43.4 %)
Age (y), mean ± SD
43.5 year ± 19.3
41.3 year ± 18.7 a
47.9 year ± 21.3
53.2 year ± 22.0 a
Descriptive data for first injury and the recurrent injury, year 1999 to 2012
58,922 (59.8 %)
39,580 (40.2 %)
82.9 %/17.1 %
78.5 %/21.5 %
Hospital days mean ± SD
9.7 ± 17.0
10.6 ± 16.4
Total hospital days
97,768 (52 %)
90,271 (48 %)
14,439 (24.5 %)
9,728 (24.6 %)
12,351 (21.0 %)
9,402 (23.8 %)
13,039 (22.1 %)
8,692 (22.0 %)
10,648 (18.1 %)
5,998 (15.2 %)
1,862 (3.2 %)
1,161 (2.9 %)
1,420 (2.4 %)
1,283 (3.2 %)
730 (1.2 %)
465 (1.2 %)
650 (1.1 %)
448 (1.1 %)
541 (0.9 %)
805 (2.0 %)
3,242 (5.5 %)
1,598 (4.0 %)
Age and sex
Hazard ratio for moderate a second injury in relation to different age groups (age at first injury). The analyses were divided by gender
95 % CI
95 % CI
Age groups, years
90 and older
Hazard ratio for moderate a second injury in relation to first injury type. The analyses were divided by gender and age group and adjusted for age (in years) within each age group. The reference is all other injuries
66 year and older
66 year and older
95 % CI
95 % CI
95 % CI
95 % CI
Hazard ratio for moderate a second injury in relation to first injury level of severity. The analyses were divided by gender and age group and adjusted for age (in years) within each age group
66 year and older
66 year and older
95 % CI
95 % CI
95 % CI
95 % CI
ISS 9 or more
Trauma recidivists account for 36 % of all injured, a higher proportion than found in most other studies. This can be explained by a 14 year follow up which is longer than in most other studies. Moreover, data were collected from the only responsible trauma hospital in the region and from a stable population. Previous studies present a great variety of the contribution of trauma recidivism to the injury burden. The majority of these studies are based on data from major trauma centers in the US (Poole et al. 1993; Hedges et al. 1995; Kaufmann et al. 1998; Brooke et al. 2006; Claassen et al. 2007; Kwan et al. 2011; Allan et al. 2013). Several of these studies show a low proportion of trauma recidivism, which is probably due to the facts that they only include serious injuries or only patients handled by the specific trauma team at the hospital. Moreover, their follow- up was shorter. And also their patients may visit another hospital if injured again. In our study there was just one trauma hospital in the area. Consequently, we claim that many of these studies does not reflect the full burden of injury recidivism. The high proportion of trauma recidivism in our material indicates a significant opportunity for preventive intervention.
Age and sex
We found that young men often suffered proportionately more recurrent injuries than young women. This has also been shown by Kaufmann (Kaufmann et al. 1998). In the elderly population, we found that women were more affected by recurrent injuries, which is consistent with older previously published results, but in contrast to a recently published study from a trauma center in Miami, which found more men among recurrently injured over 65 (Gubler et al. 1996; Allan et al. 2013). In this study, they had 20 years follow up, but found only 1.2 % recurrent injuries, perhaps due to a higher migration rate. This study was done in a major urban trauma center and their recidivists were more susceptible to penetrating trauma, all-terrain vehicle/motorcycle collisions and possibly bicycle accidents, which indicates a completely different case mix compared to our study. Most previous studies have not analyzed the recidivism from an age perspective. Our results show that both age and sex are important factors that affect the risk of recurrent injury.
Type of injury
Fracture and sprain increased the risk for moderate recurrent injury. Clinical examples of sprains are anterior cruciate ligaments injuries, which affect the stability and muscular strength of the knee and anterior dislocation of the shoulder that at younger ages often leads to recurrent dislocation (Cutts et al. 2009; Gardinier et al. 2012). Consequently, the increased risk may be caused by impairment of the functional level of extremities. Wounds and foreign body penetration on the other hand were type of injuries that lowers the risk for future moderate or serious injury. These injuries are likely to heal without compromising the physical function. A Baltimore study with mostly younger trauma patients and a Miami study with elderly both reported that recidivists were more often injured by penetrating trauma the first time (Brooke et al. 2006; Allan et al. 2013). This is in contrast to a study from Nevada, where blunt trauma was more common (Toschlog et al. 2007). Again, we believe that the difference between our results and the results in these studies can be explained by differences in case mix, where our study includes more mild and moderate injuries. That concussion as a first injury increase the risk for subsequent injury is known in sport medicine, in elite football players and in in ice-hockey, football, floorball and hand-ball but has to our knowledge not been shown for a wider population (Nordstrom et al. 2014; Nyberg et al. 2015). To our knowledge this is the first study to show this association in a wider population. Earlier studies have indicated that concussion may compromise cognitive function in the short time interval, which could enhance the risk for subsequent injury (Shores et al. 2008; Peterson et al. 2009). Burman’s study on the other hand indicated that athletes who suffer a concussion might be more injury prone due to a higher willingness to take risks (Burman et al. 2016).
So far, no other study like this one, populations based and longitudinal over 14 years, has previously investigated the type of first time injury as a possible indicator for recurrent injury. Our results show that fractures, sprains and concussion can have an impact on future injury and thus may be a variable to use when selecting individuals for prevention programs.
Level of injury severity
More severe injuries increased the risk for recurrent moderate injury, which is in line with studies on trauma recidivism among the elderly, but in contrast to other studies (Hedges et al. 1995; Gubler et al. 1996). The higher risk after more severe injuries might be explained by a persistent more severe impairment after injury, which affects the risk for subsequent injury (McGwin et al. 2001). However, there might also be behavioral factors contributing to make some individuals more injury prone than others.
Strengths and limitations
The strength of this study is that it provides information for a complete population in a well-defined, stable geographic area with just one hospital. Moreover, during this period no major reorganizations within the health care system were implemented. The database on which this study was based has a high degree of validity and coverage.
A drawback is the selection bias. We define the first injury in the registry (1999-) as the subject’s first injury although they may have suffered one or more injuries before 1999. In addition, a few fatal injuries (immediate deaths at the site of accident) were not included in the study. Both those conditions will have a diluting effect on our results. Some variables. Which have been associated with trauma recidivism like socioeconomic status, chronic disease and use of drugs, were not included in this study.
Trauma recidivism is common and contributes to a considerable part of the trauma burden, which emphasize a need for prevention. Injury type, sex and age can be used as variables in the selection of participants for preventive programs. An acute accident is a pedagogic opportunity. Individual information that the risk for future injury is elevated may motivate the patient to participate in a preventive program.
I thank the Injury Surveillance Group at the Umeå University Hospital and especially Magnus Hellström for providing important information on the database. I also thank Professor Kari Ormstad for linguistic advice.
This research was supported by the County Council of Norrbotten and Umeå University, Sweden.
FR analyzed the data, conducted the statistical analysis, interpreted the results, and wrote the paper. ML provided advice, conducted the statistical analysis, interpreted the results, and wrote the paper. JL, UB and OS contributed to the study design, interpretation of results, and the writing of the paper. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
- Allan BJ, Davis JS, Pandya RK, Jouria J, Habib F, Namias N, Schulman CI. Exploring trauma recidivism in an elderly cohort. J Surg Res. 2013;184(1):582–5.View ArticleGoogle Scholar
- Association for the Advancement of Automotive Medicine. Abbreviated injury scale 1990 revision – update 98. AAAM: Barrington; 2001.Google Scholar
- Baker SP, O’Neill B, Haddon Jr W, Long WB. The injury severity score: A method for describing patients with multiple injuries and evaluating emergency care. J Trauma. 1974;14(3):187–96.View ArticleGoogle Scholar
- Bauer R, Steiner M, Kisser R, Macey SM, Thayer D. Accidents and injuries in the EU : Results of the EuroSafe reports. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz. 2014;57(6):673–80.View ArticleGoogle Scholar
- Brooke BS, Efron DT, Chang DC, Haut ER, Cornwell 3rd EE. Patterns and outcomes among penetrating trauma recidivists: It only gets worse. J Trauma. 2006;61(1):16–9. discussion 20.View ArticleGoogle Scholar
- Burman E, Lysholm J, Shahim P, Malm C. Concussed athletes are more prone to injury both before and after their index concussion: a data base analysis of 699 concussed contact sports athletes. BMJ Open Sport Exerc Med. 2016;2:e000092. doi:10.1136/bmjsem-2015-000092.View ArticleGoogle Scholar
- Caufeild J, Singhal A, Moulton R, Brenneman F, Redelmeier D, Baker AJ. Trauma recidivism in a large urban Canadian population. J Trauma. 2004;57(4):872–6.View ArticleGoogle Scholar
- Claassen CA, Larkin GL, Hodges G, Field C. Criminal correlates of injury-related emergency department recidivism. J Emerg Med. 2007;32(2):141–7.View ArticleGoogle Scholar
- Cooper C, Eslinger DM, Stolley PD. Hospital-based violence intervention programs work. J Trauma. 2006;61(3):534–7. discussion 537–40.View ArticleGoogle Scholar
- Cutts S, Prempeh M, Drew S. Anterior shoulder dislocation. Ann R Coll Surg Engl. 2009;91(1):2–7.View ArticleGoogle Scholar
- Fleury D, Peytavin JF, Alam T, Brenac T. Excess accident risk among residents of deprived areas. Accid Anal Prev. 2010;42(6):1653–60.View ArticleGoogle Scholar
- Gardinier ES, Manal K, Buchanan TS, Snyder-Mackler L. Gait and neuromuscular asymmetries after acute anterior cruciate ligament rupture. Med Sci Sports Exerc. 2012;44(8):1490–6.View ArticleGoogle Scholar
- Gentilello LM, Donovan DM, Dunn CW, Rivara FP. Alcohol interventions in trauma centers. current practice and future directions. JAMA. 1995;274(13):1043–8.View ArticleGoogle Scholar
- Gmel G, Givel JC, Yersin B, Daeppen JB. Injury and repeated injury - what is the link with acute consumption, binge drinking and chronic heavy alcohol use? Swiss Med Wkly. 2007;137(45–46):642–8.Google Scholar
- Gubler KD, Maier RV, Davis R, Koepsell T, Soderberg R, Rivara FP. Trauma recidivism in the elderly. J Trauma. 1996;41(6):952–6.View ArticleGoogle Scholar
- Hedges BE, Dimsdale JE, Hoyt DB, Berry C, Leitz K. Characteristics of repeat trauma patients, san diego county. Am J Public Health. 1995;85(7):1008–10.View ArticleGoogle Scholar
- Hedstrom EM, Bergstrom U, Michno P. Injuries in children and adolescents-analysis of 41,330 injury related visits to an emergency department in northern sweden. Injury. 2012;43(9):1403–8.View ArticleGoogle Scholar
- Juillard C, Smith R, Anaya N, Garcia A, Kahn JG, Dicker RA. Saving lives and saving money: Hospital-based violence intervention is cost-effective. J Trauma Acute Care Surg. 2015;78(2):252–8.View ArticleGoogle Scholar
- Kaufmann CR, Branas CC, Brawley ML. A population-based study of trauma recidivism. J Trauma. 1998;45(2):325–31. discussion 331–2.View ArticleGoogle Scholar
- Keough V, Lanuza D, Jennrich J, Gulanick M, Holm K. Characteristics of the trauma recidivist: An exploratory descriptive study. J Emerg Nurs. 2001;27(4):340–6.View ArticleGoogle Scholar
- Kwan RO, Cureton EL, Dozier KC, Victorino GP. Gender differences among recidivist trauma patients. J Surg Res. 2011;165(1):25–9.View ArticleGoogle Scholar
- Lih A, Nandapalan H, Kim M, Yap C, Lee P, Ganda K, Seibel MJ. Targeted intervention reduces refracture rates in patients with incident non-vertebral osteoporotic fractures: A 4-year prospective controlled study. Osteoporos Int. 2011;22(3):849–58.View ArticleGoogle Scholar
- McGwin Jr G, May AK, Melton SM, Reiff DA, Rue 3rd LW. Recurrent trauma in elderly patients. Arch Surg. 2001;136(2):197–203.View ArticleGoogle Scholar
- Nordstrom A, Nordstrom P, Ekstrand J. Sports-related concussion increases the risk of subsequent injury by about 50 % in elite male football players. Br J Sports Med. 2014;48(19):1447–50.View ArticleGoogle Scholar
- Nyberg G, Mossberg KH, Lysholm J, Tegner Y. Subsequent traumatic injuries after a concussion in elite ice hockey: A study over 28 years. Curr Res Concussion. 2015;2(3):109–12.Google Scholar
- Peterson SE, Stull MJ, Collins MW, Wang HE. Neurocognitive function of emergency department patients with mild traumatic brain injury. Ann Emerg Med. 2009;53(6):796–803. e1.View ArticleGoogle Scholar
- Polinder S, Haagsma JA, Lyons RA, Gabbe BJ, Ameratunga S, Cryer C, Derrett S, Harrison JE, Segui-Gomez M, van Beeck EF. Measuring the population burden of fatal and nonfatal injury. Epidemiol Rev. 2012;34(1):17–31.View ArticleGoogle Scholar
- Poole GV, Griswold JA, Thaggard VK, Rhodes RS. Trauma is a recurrent disease. Surgery. 1993;113(6):608–11.Google Scholar
- Reiner DS, Pastena JA, Swan KG, Lindenthal JJ, Tischler CD. Trauma recidivism. Am Surg. 1990;56(9):556–60.Google Scholar
- Rhee P, Joseph B, Pandit V, Aziz H, Vercruysse G, Kulvatunyou N, Friese RS. Increasing trauma deaths in the united states. Ann Surg. 2014;260(1):13–21.View ArticleGoogle Scholar
- Rittenhouse K, Harnish C, Gross B, Rogers A, Miller JA, Chandler R, Rogers FB. An analysis of geriatric recidivism in the era of accountable care organizations. J Trauma Acute Care Surg. 2015;78(2):409–14.View ArticleGoogle Scholar
- Sach TH, Logan PA, Coupland CA, Gladman JR, Sahota O, Stoner-Hobbs V, Robertson K, Tomlinson V, Ward M, Avery AJ. Community falls prevention for people who call an emergency ambulance after a fall: An economic evaluation alongside a randomised controlled trial. Age Ageing. 2012;41:635–41.View ArticleGoogle Scholar
- Saveman BI, Bjornstig U. Unintentional injuries among older adults in northern sweden--a one-year population-based study. Scand J Caring Sci. 2011;25(1):185–93.View ArticleGoogle Scholar
- Sayfan J, Berlin Y. Previous trauma as a risk factor for recurrent trauma in rural northern israel. J Trauma. 1997;43(1):123–5.View ArticleGoogle Scholar
- Schermer CR, Moyers TB, Miller WR, Bloomfield LA. Trauma center brief interventions for alcohol disorders decrease subsequent driving under the influence arrests. J Trauma. 2006;60(1):29–34.View ArticleGoogle Scholar
- Shores EA, Lammel A, Hullick C, Sheedy J, Flynn M, Levick W, Batchelor J. The diagnostic accuracy of the revised westmead PTA scale as an adjunct to the glasgow coma scale in the early identification of cognitive impairment in patients with mild traumatic brain injury. J Neurol Neurosurg Psychiatry. 2008;79(10):1100–6.View ArticleGoogle Scholar
- Smith R, Dobbins S, Evans A, Balhotra K, Dicker RA. Hospital-based violence intervention: Risk reduction resources that are essential for success. J Trauma Acute Care Surg. 2013;74(4):976–80. discussion 980–2.View ArticleGoogle Scholar
- Toschlog EA, Sagraves SG, Bard MR, Schenarts PJ, Goettler CC, Newell MA, Rotondo MF. Rural trauma recidivism: A different disease. Arch Surg. 2007;142(1):77–81.View ArticleGoogle Scholar
- Van der Kallen J, Giles M, Cooper K, Gill K, Parker V, Tembo A, Major G, Ross L, Carter J. A fracture prevention service reduces further fractures two years after incident minimal trauma fracture. Int J Rheum Dis. 2014;17(2):195–203.View ArticleGoogle Scholar
- Wan JJ, Morabito DJ, Khaw L, Knudson MM, Dicker RA. Mental illness as an independent risk factor for unintentional injury and injury recidivism. J Trauma. 2006;61(6):1299–304.View ArticleGoogle Scholar
- Williams JM, Furbee PM, Hungerford DW, Prescott JE. Injury recidivism in a rural ED. Ann Emerg Med. 1997;30(2):176–80.View ArticleGoogle Scholar
- Worrell SS, Koepsell TD, Sabath DR, Gentilello LM, Mock CN, Nathens AB. The risk of reinjury in relation to time since first injury: A retrospective population-based study. J Trauma. 2006;60(2):379–84.View ArticleGoogle Scholar