What do Trauma Surgery Textbooks Teach About Racial Disparities and the Roots of Violence?

This study presents a formal text analysis of trauma surgery textbooks. We examine passages that describe disparities or mechanisms of injury, and we report types of underlying causes and preventative interventions discussed. Trauma textbooks were drawn from an industry-standard list used by medical libraries. Chi-square testing was used to determine whether different types of underlying causes or preventative interventions were discussed by disparity type (those affecting racial minorities vs rural populations) and injury mechanism (accidental injuries vs intentional interpersonal injury). 146 passages were extracted from 7 textbooks, totaling 5576 pages of text. Passages discussing rural disadvantages or unintentional injury were substantially more likely to describe structural risk factors or governmental interventions than those discussing racial disadvantages or intentional injury, respectively. Textbook authors should consider enriching discussion of violence prevention or racial disparities to emphasize structural causes and interventions.

Trauma surgeons, including the American College of Surgeons' Committee on Trauma, have become increasingly involved in efforts to improve public health via violence prevention and disparity reduction. 1 However, it is unclear whether effective preventative strategies for violence and disparities are being formally taught to future generations of surgeons.
Textbooks may have a role in this transmission of knowledge. Teaching about violence prevention and racial disparities deserve consideration both as fundamental public health priorities and as necessary aspects of delivering just medical care. If textbooks are to be responsive to these values, they should discuss violence prevention and racial disparities as thoroughly as they do prevention of unintentional injuries and rural disadvantages, respectively.
This study seeks to describe what the most popular trauma surgery textbooks teach about these matters. We hope that the findings we present will encourage textbook authors to emphasize interventions that help the disadvantaged groups disproportionately affected by both racial disparities and violence.

Methods
Doody's List, an industry-standard ledger of textbooks relied on by medical libraries, was used to identify popular trauma textbooks. All trauma textbooks were included unless they were technical guides or based on non-trauma specialties. A within-text search strategy was developed by a multidisciplinary team to identify discussions of racial/ethnic groups, disparities, concepts of race, injury mechanisms, or injury prevention (Appendix 1 in Supplemental Material). Passages were extracted if they discussed a racial disparity, an urban/rural disparity, injury prevention, or injury mechanisms.
Passages were classified by the types of disparity, injury mechanism, underlying causes/contributors, and corrective interventions discussed. Disparities were grouped into Black vs White, other race/ethnicity vs White, rural disadvantages, and urban disadvantages. Injury mechanisms were classified as penetrating/intentional; unintentional/blunt; intimate partner, sexual or abuse; and suicide. Categories for underlying contributors included governmental policies, other structural factors, socioeconomic factors, individual medical factors, or individual decisional/cultural factors. Passages were also classified based on the type of intervention discussed, if any: structural/binding governmental interventions; police enforcement specifically; other institutional or non-binding governmental interventions; individual medical provider interventions; or individual patient choices.
Three authors performed classifications of underlying causes and interventions: the senior author and two medical students. Assessment of inter-rater agreement included calculations of 3-way inter-rater agreement (proportion of classifications where all 3 classifiers agree), mean 2-way agreement, as well as Fleiss's 3-way Kappa. Chi-square testing was performed to evaluate differences in classifications between rural disadvantages and all other disparities (since urban disadvantages and racial disparities overlap closely), and between intentional injury (exclusive of suicide and intimate partner violence) and unintentional injury.
Doody's list contains 28 trauma texts, of which 7 met inclusion criteria (Appendix 2 in Supplemental Material), totaling 5576 pages of text. From these, 146 passages were extracted, including 58 discussions of disparities and 88 discussions of injury mechanisms or prevention. Passage examples are given in Appendix 3 in Supplemental Material. Texts discussed racial disadvantages (31 passages) more frequently than they discussed urban or rural disadvantages (3 and 24 passages, respectively). Violent and accidental injuries were discussed with roughly equal frequency (37 vs 36 passages), whereas discussions of suicide or intimate partner violence/abuse were less frequent (7 and 8 passages).
For the classifications, the overall 3-way inter-rater agreement was 86.1%, and the mean 2-way agreement was 90.7%. The mean Fleiss Kappa was .61, corresponding to the border between "moderate" and "substantial" agreement according to the original description of the coefficient. Kappa for underlying causes discussed in passages on injury mechanism, underlying causes of disparities, interventions by mechanism, and interventions for disparities were .52, .56, .62, and .72, respectively.
Passages discussing rural disparities were much more likely to describe structural causes of the disparity, such as distance to the nearest hospital, than passages discussing racial or urban disadvantages (56.9% vs 10.8%, P < .001) Figure 1. Underlying causes (A) and corrective/preventative interventions (B) discussed in trauma surgery textbooks by disparity type, with Chi-square P-values, mean rates, and rounded counts. Classification was performed by 3 authors, and rates were averaged across classifiers. "Non-governmental institutional actions" include non-binding governmental actions, such as Centers for Disease Control publication of educational materials.
No statistically significant differences were identified in the type of underlying cause discussed by injury mechanism, except that individual medical factors (such as losing balance in the elderly) were more often mentioned in accidental injury (Figure 2). However, differences were observed in discussion of interventions depending on injury type; governmental interventions were less frequently discussed for interpersonal injuries (8.1% vs 47.2% of passages discussing unintentional injury, eg, P < .001).
Our findings indicate that discussions of racial disparities and prevention of violent injuries could be improved in widely distributed trauma textbooks. Teaching could be substantially enriched if these topics were discussed in as much detail as rural disadvantages and unintentional injuries.
Authors may be making a reasoned decision to avoid these topics based on their own or their publishers' priorities. Textbook authors in other fields frequently avoid politically controversial topics such as evolution and climate change. However, authors should not mistake political controversies for scientific ones. For example, according to the best available evidence, governmental efforts that reduce the availability of firearms do reduce firearm injuries. 2 Only one trauma textbook explicitly discussed this fact. 3 There are multiple approaches textbook authors could take to improving discussion of prevention of disparities Figure 2. Underlying causes (A) and corrective/preventative interventions (B) discussed in trauma surgery textbooks by injury mechanism, with Chi-square P-values, mean rates, and rounded counts. Classification was performed by 3 authors, and rates were averaged across classifiers, but counts were rounded. "Non-governmental institutional actions" include non-binding governmental actions, such as Centers for Disease Control publication of educational materials. and violent injury. Some textbooks have dedicated chapters to rural trauma-a chapter for disparities or violence prevention could close the gap we have described. Alternatively, each discussion of violent injury or disparities could be further developed.
This study has limitations. First, textbooks are becoming less popular among learners, although they remain common sources for surgery residents specifically. 4 Second, there are inherent ambiguities in both the choice of classifications and classifications themselves. We aspire only to prompt reflection on textbook content, not definitively evaluate these passages.
Trauma surgery textbooks are more likely to discuss interventions to ameliorate rural disadvantages or unintentional injuries than racial disadvantages or intentional interpersonal injuries. Textbook authors should consider additional discussion of structural contributors and preventative interventions for racial disparities and violent injury.