Faculty Bibliography

BACKGROUND: Although intimate partner violence (IPV) is the leading cause of serious injury and the second leading cause of death among reproductive age women in America, effective screening is difficult. Our institution currently screens for IPV during the floor intake assessment by having a registered nurse (RN) ask three unscripted questions about physical, verbal, and sexual abuse during a battery of 81 questions. The patients are frequently in pain, medicated, distraught, or intoxicated, and the RN is juggling multiple responsibilities. We also use a protocol-driven alcohol abuse screen on every trauma admission known as "Screening, Brief Intervention, and Referral for Treatment" (SBIRT). It is conducted by trained counselors when any effects of alcohol are gone in a distraction-free setting after patients have had time to ruminate on their admission. We hypothesized that linking the validated partner violence screening (PVS) to SBIRT would result in higher rates of positive IPV screens than after RN screens. METHODS: This prospective trial was conducted at an urban Level I center. English- and Spanish-speaking female trauma patients underwent the three-question, nonvalidated RN-screen on floor arrival per the local standard of care. Before discharge, they then underwent SBIRT screening per trauma service protocol, after which SBIRT administered the PVS as our investigative intervention. All screens were native language. SBIRT screeners were blinded to the results of the earlier RN screen. If an SBIRT or RN screen was not performed for any reason, it was categorized as a negative screen. Admissions to the surgical intensive care unit had both screens delayed until floor transfer. McNemar's exact test was used for paired categorical data and Fisher's exact test otherwise. Significance was set at an alpha of 0.05. RESULTS: One hundred twenty-five consecutive female inpatients (mean age, 40.9 years+/-17.7 years; Injury Severity Score, 9.8+/-7.5) were enrolled, with 14 (11.2%) screening positive for one or both methods. The SBIRT-linked screen was significantly better at detecting IPV than the RN screen (p=0.01). No association was found between the likelihood of giving a discordant response to the two IPV screens and acute alcohol intoxication or polysubstance abuse at the time of admission, being a Spanish-only speaker, or if the initial admission was directly to the surgical intensive care unit. Despite being mandatory on intake, 23 of 125 patients (18.4%) had no RN screen performed, with 2 of these patients screening positive for IPV by SBIRT personnel. CONCLUSION: Linking an IPV screen to an established alcohol abuse screen results in higher rates of detection of IPV than screening by RNs at intake assessment. At our institution, adoption of this practice should result in detecting and referring approximately 85 additional female trauma inpatients per year for IPV services.

OBJECT: In this paper, the authors' goal was to examine the relationship between transfusion and long-term functional outcomes in moderately anemic patients (lowest hematocrit [HCT] level 21-30%) with traumatic brain injury (TBI). While evidence suggests that transfusions are associated with poor hospital outcomes, no study has examined transfusions and long-term functional outcomes in this population. The preferred transfusion threshold remains controversial. METHODS: The authors performed a retrospective review of patients who were admitted with TBI between September 2005 and November 2007, extracting data such as HCT level, status of red blood cell transfusion, admission Glasgow Coma Scale (GCS) score, serum glucose, and length of hospital stay. Outcome measures assessed at 6 months were Glasgow Outcome Scale-Extended score, Functional Status Examination score, and patient death. A multivariate generalized linear model controlling for confounding variables was used to assess the association between transfusion and outcome. RESULTS: During the study period, 292 patients were identified, and 139 (47.6%) met the criteria for moderate anemia. Roughly half (54.7%) underwent transfusions. Univariate analyses showed significant correlations between outcome score and patient age, admission GCS score, head Abbreviated Injury Scale score, number of days with an HCT level < 30%, highest glucose level, number of days with a glucose level > 200 mg/dl, length of hospital stay, number of patients receiving a transfusion, and transfusion volume. In multivariate analysis, admission GCS score, receiving a transfusion, and transfusion volume were the only variables associated with outcome (F = 2.458, p = 0.007; F = 11.694, p = 0.001; and F = 1.991, p = 0.020, respectively). There was no association between transfusion and death. CONCLUSIONS: Transfusions may contribute to poor long-term functional outcomes in anemic patients with TBI. Transfusion strategies should be aimed at patients with symptomatic anemia or physiological compromise, and transfusion volume should be minimized.