Laparoscopic cholecystectomy in the elderly.

OBJECTIVE: To evaluate the outcomes of laparoscopic cholecystectomy in elderly patients at a single institution. METHODS: A retrospective chart review was conducted of all patients > or = 65 years of age who underwent laparoscopic cholecystectomy over a 5-year period (January 1995 to December 1999). Four-trocar site laparoscopic cholecystectomy using the open Hasson technique were performed in all patients. The demographic data (age, sex), associated comorbidities, American Society of Anesthesiologist's (ASA) score, postoperative morbidity, mortality, and length of stay were recorded for each patient. Statistical analysis was done using Fisher's exact test and chi-square analysis. Statistical significance was defined as P < or = 0.05. RESULTS: The patient cohort included 46 patients with a median age of 71 years (range, 65 to 87). Seventeen (37%) patients were < or = 70 years of age, and twenty-nine (63%) patients were > or = 70 years of age. Twenty-two (48%) patients had ASA scores of > or = 3. Patients > or = 70 had significantly higher ASA scores. Eighteen patients > or = 70 years had ASA > or = 3 compared with 4 patients < or = 70 with ASA > or = 3 (P<0.05). Twenty-two patients > or = 70 and 8 patients < or = 70 required urgent surgery P<0.05). Fifteen (33%) patients presented with acute cholecystitis, and 31 (67%) patients presented with a greater number of chronic symptoms. Four (9%) patients had pancreatitis on presentation, and 6 patients underwent preoperative endoscopic retrograde cholangiopancreatography (ERCP). Two of these 6 patients also underwent sphincterotomy. Urgent surgery was performed in 30 (65%) patients. The mean operative time was 103 +/- 37 (SD) minutes. One (2%) conversion to open cholecystectomy was required. The mean postoperative stay was 7 days (range, 1 to 46). Fourteen (30%) patients had only a 1-night postoperative stay. Patients > or = 70 had significantly longer postoperative stays. Nine patients > or = 70 and only 1 patient < or = 70 stayed in the hospital for more than 7 days. Postoperative complications were noted in 6 (13%) patients, most of which were chest infections. Five patients > or = 70 and only 1 patient < or = 70 developed postoperative complications. No mortalities occurred. CONCLUSION: Laparoscopic cholecystectomy is safe and feasible in elderly patients. Patients > or = 70 years seem to have a longer postoperative stay and slightly more postoperative complications. Age alone should not be a contraindication to laparoscopic cholecystectomy in the elderly patient.

Emergency Medical Services Outcomes Project (EMSOP) II: developing the foundation and conceptual models for out-of-hospital outcomes research.

Development of methodologically acceptable outcomes models for emergency medical services (EMS) is long overdue. In this article, the Emergency Medical Services Outcomes Project proposes a conceptual framework that will provide a foundation for future EMS outcomes research. The "Episode of Care Model" and the "Out-of-Hospital Unit of Service Model" are presented. The Episode of Care Model is useful in conditions in which interventions and outcomes, especially survival and major physiologic dysfunction, are linked in a time-dependent manner. Conditions such as severe trauma, anaphylaxis, airway obstruction, respiratory arrest, and nontraumatic cardiac arrest are amenable to this methodology. The Out-of-Hospital Unit of Service Model is essentially a subunit of the Episode of Care Model. It is valuable for evaluating conditions that have minimal-to-moderate therapeutic time dependency. This model should be used when studying outcomes limited to the out-of-hospital interval. An example of this is pain management for injuries sustained in motor vehicle crashes. These models can be applied to a wide spectrum of conditions and interventions. With the scrutiny of health care expenditures ever increasing, the identification of clinical interventions that objectively improve patient outcome takes on growing importance. Therefore, the development, dissemination, and use of meaningful methodologies for EMS outcomes research is key to the future of EMS system development and maintenance.

An examination of the volume-mortality relationship for New York State trauma centers.

OBJECTIVES: New York State Trauma Registry data were analyzed to determine whether there is a significant relationship between the volume of trauma patients treated by a trauma center and its risk-adjusted inpatient mortality rate. METHODS: Stepwise logistic regression was used to identify significant independent predictors of mortality, their weights, and the probability of in-hospital mortality for each patient. These data were then used to calculate risk-adjusted mortality rates for various ranges of hospital volume. Ranges were identified on the basis of homogeneity of mortality rates, the number of hospitals in each range, and the number of patients in each range. Three volume measures were used: (1) total annual volume of trauma cases > or = 1200 and total annual volume > or = 240 for patients with Injury Severity Score (ISS) > or = 15 (equivalent to American College of Surgeons [ACS] criteria), (2) total annual volume of patients with ISS > or = 15, and (3) total annual volume of cases in the Registry (approximately, inpatients with ISS > or = 9). RESULTS: Results show that the 35 New York State trauma centers not meeting the ACS criteria had lower, but not significantly lower, observed and risk-adjusted mortality rates (7.62% and 8.25%, respectively) than the corresponding rates for the 8 New York State trauma centers that met the ACS criteria (9.36% and 8.83%, respectively). Regarding the other two criteria, hospital ranges representing lower annual volumes tended to have somewhat lower, although not significantly lower, observed and risk-adjusted mortality rates. For example, using a total annual volume for patients with ISS > or = 15, the risk-adjusted mortality rates for the volume ranges 1-150, 151-250, and 251+ were 7.78%, 9.23%, and 8.70%, respectively. CONCLUSIONS: We were unable to document an inverse relationship between hospital volume and inpatient mortality rate for trauma centers in New York State. Volume criteria should not be considered indicators of the quality of trauma care.

Accounting for intubation status in predicting mortality for victims of motor vehicle crashes.

BACKGROUND: Two of the important predictors of mortality for trauma patients are the Glasgow Coma Scale and the respiratory rate. However, for intubated patients, the verbal response component of the Glasgow Coma Scale and the respiratory rate cannot be accurately obtained. This study extends previous work that attempts to predict mortality accurately for intubated patients without using verbal response and respiratory rate. METHODS: The New York State Trauma Registry was used to identify 1994 and 1995 victims of motor vehicle crashes (MVCs). For the subset of patients who were not intubated, we developed two statistical models to predict mortality: one did not contain verbal response or respiratory rate, and the other contained a predicted verbal response. These were compared with a model that did include verbal response and respiratory rate. We also compared the predictive abilities of the first two models for all MVC patients (intubated and nonintubated) and determined the extent to which intubated patients were at increased risk of dying in the hospital after having adjusted for other predictors of mortality. RESULTS: For nonintubated patients, the statistical model without verbal response and the model with predicted verbal response had slightly better discrimination and worse calibration than the model that included verbal response and respiratory rate. Predicted verbal response did not improve the strength of the model without verbal response. For all MVC patients (intubated and nonintubated), predicted verbal response was not a significant predictor of mortality when used in combination with the other predictors. Intubation status was a significant predictor, with intubated patients having a higher probability of dying in the hospital than patients with otherwise identical risk factors. CONCLUSION: Inpatient mortality for intubated MVC patients can be accurately predicted without respiratory rate or verbal response. There appears to be no need for predicted verbal response to be part of the prediction formula, but intubation status is an important independent predictor of mortality and should be used in statistical models that predict mortality for MVC patients.

Predictors of mortality in adult patients with blunt injuries in New York State: a comparison of the Trauma and Injury Severity Score (TRISS) and the International Classification of Disease, Ninth Revision-based Injury Severity Score (ICISS).

BACKGROUND: The purpose of this study was to determine the statistical model that best predicted mortality from blunt trauma using a contemporary population-based database. METHODS: 1994-1995 New York State Trauma Registry data for patients with blunt injuries were used to predict mortality using three statistical models: (1) the original Trauma and Injury Severity Score (TRISS) model based on Major Trauma Outcome Study data, (2) a new TRISS model whose coefficients were derived using New York data, and (3) the International Classification of Disease, Ninth Revision-based Injury Severity Score (ICISS) with predicted survival values obtained from the Agency for Health Care Policy and Research's Health Care Utilization Project. The models were compared with respect to discrimination (using the C statistic) and calibration (using the Hosmer-Lemeshow [H-L] statistic). In addition, the models were tested to see how well they predicted outcomes for each of the three mechanisms of blunt injury. RESULTS: The ICISS model had a significantly higher C statistic (0.878) and a better H-L statistic (29.38) for predicting mortality for all adult patients with blunt injuries. The original TRISS model had very poor calibration (H-L = 687.38). None of the three models predicted mortality accurately for victims of motor vehicle crashes or victims of low falls. When separate models were developed for all motor vehicle crashes, low falls, and other blunt injuries, the ICISS and New York TRISS models both fit well, although the calibration was marginal in most cases. The ICISS model had a statistically significantly higher C statistic for other blunt injuries and for motor vehicle crashes. The New York TRISS model had better calibration for low falls. CONCLUSIONS: The ICISS has promise as an alternative to TRISS, but many more comparative studies need to be undertaken using updated TRISS coefficients. Models should also be developed for mechanisms of injury, not just for blunt and penetrating injuries.

Fatality Analysis Reporting System demonstrates association between trauma system initiatives and decreasing death rates.

BACKGROUND: Trauma registries frequently do not include the deaths of patients who do not get to trauma centers (TCs). Thus, complementary methods of monitoring the impact of trauma system initiatives should be considered. The objective of this study is to use National Highway Safety Traffic Administration's Fatality Analysis Reporting System (FARS) and New York State Department of Motor Vehicles data and to study the impact of state and regional initiatives over a 10-year period in the seven-county Hudson Valley New York (HV) region with one regional TC in Westchester County (WC) and to assess its face validity. METHODS: FARS data for the United States (US), New York State (NY), the HV region, and WC were analyzed from 1987 to 1996. Trauma system initiatives included the following. Statewide: (1) TC standards (1989), (2) TC designation and funding (1990), (3) State Trauma Advisory Committee (1991), (4) BLS triage protocol and trauma registry (1993), and (5) quality improvement site surveys (1994). Regional: (1) one regional and two area TCs (1990), (2) helicopter services (1992 and 1994), (3) two additional area TCs, and (4) E 911 in all three counties (1995). The results were presented to the New York State Trauma Advisory Committee. RESULTS: Although nationally motor vehicle crash deaths/100,000 persons have plateaued since 1991, trauma system initiatives have been temporally associated with death rates continuing to diminish in New York, the HV, and WC. From 1987 to 1996, the HV death rate dropped from 17.00 to 9.45, a 44% drop; and the WC rate dropped from 12.51 to 7.05, a 44% drop compared with United States death rate drop of 16% (p < 0.005). The percentage of seriously injured trauma patients going to the trauma centers increased from 53% in 1990 to 72% in 1995 (p < 0.001). The STAC felt that the data reflected in part effects of New York State trauma system initiatives. CONCLUSION: The drops in motor vehicle crash death rates may reflect injury prevention as well as trauma system initiatives. Thus, although FARS and New York State Department of Motor Vehicles data cannot establish cause and effect relationships, it can monitor the aggregated impact of multiple initiatives. Taken together with increasing percentages of seriously injured trauma patients going to trauma centers and comparisons with national FARS data, the association of decreasing deaths with the implementation of a trauma system seems to have face validity.

Emergency medical services outcomes project I (EMSOP I): prioritizing conditions for outcomes research.

Over the past several years, out-of-hospital EMS have come under increased scrutiny regarding the value of the range of EMS as currently provided. We used frequency data and expert opinion to rank-order EMS conditions for children and adults based on their potential value for the study of effectiveness of EMS care. Relief of discomfort was the outcome parameter EMS professionals identified as having the most potential impact for the majority of children and adults in the top quartile conditions. Future work from this project will identify appropriate severity and outcome measures that can be used to study these priority conditions. The results from the first year of this project will assist those interested in EMS outcomes research to focus their efforts. Furthermore, the results suggest that nonmortality out-come measures, such as relief of discomfort, may be important parameters in determining EMS effectiveness.

Predicting and preventing pressure ulcers in adults with paralysis.

A questionnaire mailed to all 2,295 members of the Eastern Paralyzed Veterans Association measured 45 potential risk factors for pressure ulcers. Logistic-regression analysis and Cox proportional-hazards analyses were used to identify the variables that were independently associated with pressure ulcers. The survey response rate was 42.2%. Among 15 risk factors from a previously published scale by the authors, 7 were independent predictors of pressure ulcer development: level of activity, level of mobility, complete spinal cord injury, urine incontinence or moisture, autonomic dysreflexia, pulmonary disease, and renal disease. In addition, 2 new variables added significant predictive value: being prone to infections that cause breathing problems and paralysis caused by trauma (as opposed to disease). Using these 9 risk factors, a new pressure ulcer risk assessment scale was designed specifically for persons with paralysis who are living in a community setting. It appears to be a more accurate method of predicting pressure ulcers than currently used risk assessment scales.

Predicting survival of victims of motor vehicle crashes in New York state.

This study assesses the relative ability of three different models to predict in-hospital mortality for victims of motor vehicle crashes. The first two models, the trauma and injury severity score (TRISS) and a severity characterization of trauma (ASCOT), are models that have been used in many earlier studies and have been quoted extensively in the literature. The third model, which is developed in this study, uses essentially the same risk factors as the other two studies, but employs them in a different manner. In order to provide a fair comparison, new (logistic regression) model coefficients are fit to the first two models using the study data. The models are compared with respect to typical criteria for assessing the fit of logistic regression models as well as their ability to predict mortality for various subsets of seriously injured patients. The study concludes that the new model provides a substantially more accurate prediction of mortality, and that it may be wise for regions attempting to assess relative outcomes in their subregions to develop statistical models that are tailored to their own patients.

A new pressure ulcer risk assessment scale for individuals with spinal cord injury.

Each year, one-fourth of the 200,000 individuals with spinal cord injury in the United States develop pressure ulcers. No method currently exists, however, to accurately identify which of these individuals are at increased risk for development of pressure ulcers. We studied 219 spinal cord-injured patients, seen at a Veterans Affairs Medical Center, during a 6-yr period. Our goal was to develop a pressure ulcer risk assessment scale, specifically for persons with SCI. Each risk factor had to meet four criteria: (1) statistical association with pressure ulcer development; (2) biologically plausible mechanism; (3) literature support; (4) improved prediction. Among the 219 spinal cord-injured patients evaluated, 176 (80.4 percent) had a history of one or more pressure ulcers. Fifteen risk factors met the four criteria for inclusion into the risk assessment scale. They were as follows: restricted activity level, degree of immobility, complete spinal cord injury, urinary disease, impaired cognitive function, diabetes, cigarette smoking, residence in a nursing home or hospital, hypoalbuminemia, and anemia. Compared with the more general scales available, for quantifying the risk of pressure ulcer development, preliminary results suggest that this new scale is a significant improvement for the spinal cord-disabled.

Early predictors of sepsis in the motor-vehicle crash trauma victim.

INTRODUCTION: Sepsis is a major cause of late morbidity and mortality in the victim of trauma. Currently, there is no method that is clinically practical and accurate for predicting the occurrence of sepsis in trauma victims. METHODS: Data were collected on 3,759 motor-vehicle crash victims from 16 hospitals during a 4 1/2 year period. Retrospective analysis was done to examine the relationship of patient and injury factors known within the first 24 hours of admission on the development of sepsis. RESULTS: Sepsis developed in 154 patients (4.1%) who had a mortality rate of 17.5%. Significant early predictors of sepsis included: 1) certain pre-existing conditions; 2) blood transfusion required; 3) seven or more injuries; 4) Glasgow Coma Scale score <10 and hypotension [corrected]; 5) major blood vessel injury; 6) head trauma; 7) internal injury of the chest or abdomen; 8) spinal-cord injury; and 9) certain fracture types. CONCLUSIONS: These predictors might help target high-risk patients and, thus, promote earlier and more effective treatment for those patients.

Multivariate models for predicting survival of patients with trauma from low falls: the impact of gender and pre-existing conditions.

OBJECTIVE: To determine if pre-existing conditions significantly improve the ability of current (TRISS and ASCOT) methods for predicting survival of patients with trauma from low falls. DESIGN: Retrospective analysis using logistic regression models to identify significant independent predictors of survival. SETTING: Eight hospitals affiliated with New York Medical College. PATIENTS: A total of 1906 patients with trauma from low falls who were admitted to the eight hospitals between July 1987 and June 1989. MAIN RESULTS: Gender and several pre-existing conditions significantly improved the ability of age and the physiologic and anatomic variables contained in the TRISS and ASCOT methodologies to predict survival for trauma patients suffering from low falls, with males experiencing a lower probability of survival. Odds of survival for patients with these pre-existing conditions ranged from 0.18 to 0.59 times the odds of survival for similar patients without the pre-existing conditions when the TRISS variables were used, and from 0.23 to 0.56 times the odds for similar patients when ASCOT variables were used. Furthermore, some substantial differences were found when hospital performance was assessed with and without the benefit of pre-existing conditions. CONCLUSIONS: Pre-existing conditions and male gender are significantly related to survival of patients with trauma from low falls, and should be included along with age and the various physiologic and anatomic measures currently being used to predict survival for those patients.

Incidence of acute care complications in vertebral column fracture patients with and without spinal cord injury.

STUDY DESIGN: This study retrospectively analyzed vertebral column fractures in trauma patients during a 2-year period. Data from a multicenter trauma registry were used. OBJECTIVES: The purpose of this study was to ascertain and describe the initial in-hospital morbidity and mortality rates for patients with vertebral column fractures with and without spinal cord injury. SUMMARY OF BACKGROUND DATA: Patients with vertebral fractures and associated spinal cord injuries experience more medical complications than those without spinal cord injuries. However, the precise incidence and relative risk of complications during acute care hospitalization for these two groups are not well documented. METHODS: Vertebral column fractures in 419 adolescent and adult trauma patients hospitalized during a 2-year period were retrospectively analyzed using data from a multicenter trauma registry. RESULTS: Of the 419 patients, 104 (24.8%) had an associated spinal cord injury. More than half of the spinal cord injury patients (52.9%) and 20.6% of those without spinal cord injury had one or more complications during their hospitalization. Complications resulted in an average of 33.1 extra hospital days, which extrapolates nationally into 1.5 million additional days annually. The four complications differing most significantly in incidence between the spinal cord injury group and the non-spinal cord injury group were: urinary tract infections (24.0% vs. 8.6%), respiratory (23.1% vs. 8.6%), cardiac (11.5% vs. 3.2%), and decubitus ulcer (7.7% vs. 1.0%). Pneumonia, although not statistically different, was high in both groups (13.5% vs. 7.3%). CONCLUSIONS: The incidence of the 25 types of medical complications reported here provides specific and relevant information to assist health professionals in treating patients during their acute care. We estimate that complications during initial hospitalization add $1.5 billion annually to the cost of caring for patients with vertebral fractures in the United States.

Validation of TRISS and ASCOT using a non-MTOS trauma registry.

OBJECTIVES: To validate the Trauma and Injury Severity Score (TRISS) and A Severity Characterization of Trauma (ASCOT) models for patients with blunt injuries using an independent trauma registry, and to develop new TRISS and ASCOT models for types of patients with blunt injuries and examine their fit. DESIGN: Retrospective analysis of clinical data from the Institute for Trauma and Emergency Care (ITEC). MATERIALS AND METHODS: Statistical models were developed using TRISS and ASCOT variables applied to ITEC data for patients with blunt injuries. These models were compared to Major Trauma Outcome Study (MTOS) models with regard to the resulting coefficients and hospital quality assessments. Also, separate models were developed for different groups of blunt injuries, and these models were compared with one another and tested for adequacy of fit. MEASUREMENTS AND MAIN RESULTS: ASCOT performed acceptably well when new coefficients were derived using ITEC data, but TRISS did not. Although the models developed from MTOS and from ITEC coefficients generally yielded similar hospital quality assessments, there were some notable exceptions. Some TRISS and ASCOT variables were not significantly related to survival for some subgroups of blunt injuries, and neither the TRISS nor the ASCOT model was an adequate predictor of survival for patients suffering from low falls. CONCLUSIONS: New TRISS and ASCOT coefficients should be derived if survival for patients with blunt injuries is to be predicted accurately in independent trauma registries. Also, it may be wise to consider developing separate models for subgroups of patients, particularly if hospitals in the registry have different mixes of patient types.

Dual response runs in prehospital trauma care.

The effects of dual responses [Basic Life Support (BLS) and Advanced Life Support (ALS)] on the outcomes of trauma patients were evaluated. Outcomes included changes in physiologic measurements between the scene and the emergency department (ED), and survival to hospital discharge. Data for 2394 patients with penetrating, motor vehicle crash (MVC), or other blunt injuries were included. Changes in physiologic measurements (Revised Trauma Scores) between the prehospital and ED settings were positively associated with documented ALS or dual response care. Survival to hospital discharge among penetrating injury patients was negatively related to dual responses, whereas that among MVC patients was positively associated with dual responses. Parallel results were found for a subset of more severely injured patients. Future research should confirm and refine these results so that protocols for the appropriate use of dual response runs can be developed.

Basic life support versus advanced life support for injured patients with an injury severity score of 10 or more.

To study the value of advanced life support (ALS) compared with basic life support (BLS) for penetrating and motor vehicle crash (MVC) patients, data were collected from eight hospitals over 24 months on 781 consecutive patients with Injury Severity Scores > or = 10 as well as on a subset of 219 hypotensive patients. Initial prehospital Revised Trauma Scores (RTSs) were compared with initial emergency department RTSs. Scene times, total prehospital times, and the use of a pneumatic antishock garment (PASG), intravenous fluids, and endotracheal intubation were also documented. A modified TRISS method was used to compare mortality rates. The MVC ALS patients showed improvement in mean RTSs between prehospital and the emergency department while MVC BLS patients did not. Mean changes in blood pressure (BP) and the percentage of patients with improved BP were significantly higher among patients who received ALS; ALS was associated with increased use of PASGs and IV fluids. There were no differences between groups with respect to observed versus predicted mortality. Similar results were found in the hypotensive subset of patients. No benefit from the use of ALS for trauma patients with total prehospital times of less than 35 minutes was documented.

Is 24-hour operating room staff absolutely necessary for level II trauma center designation?

Recent papers from established trauma centers reported average elapsed times from emergency department (ED) admission to the operating room (OR) of greater than 100 minutes for patients judged to be in immediate need of surgery. This study was undertaken to determine whether patients treated at an institution desiring level II trauma center designation in a geographic area with a low incidence of penetrating trauma suffered any adverse effects because of lack of a 24-hour in-house OR staff. Trauma registry data at The Stamford Hospital, a suburban community teaching hospital without OR nursing staff in-house at night, were reviewed and compared with data from three affiliated level I trauma centers and with established national standards using TRISS methodology. Of 659 major trauma patients, 86 (44 blunt, 42 penetrating) underwent surgery within 12 hours of admission. Patients' injuries were similar in severity to those seen at the affiliated trauma centers and to the Major Trauma Outcome Study population. Mortality rates were also similar. No statistically significant differences were seen in elapsed times from ED arrival to OR arrival even in the subgroup of patients with systolic blood pressure values of < or = 90 mm Hg. No unexpected adverse outcomes could be ascribed to the lack of 24-hour OR staffing in this setting. The estimated cost of providing additional OR staffing is $145,000 per year. Since times to the OR and outcomes were similar to those at level I centers, this expense may not be warranted.

A study of pneumatic antishock garments in severely hypotensive trauma patients.

The effectiveness of a pneumatic antishock garment (PASG) on severely hypotensive trauma patients (BP < or = 50 mm Hg) was studied using two data sets. The first included data from eight hospitals collected over 4 1/2 years; the second included 2 years of data from an additional eight hospitals. Data were collected by trained nurse abstractors whose interrater reliability was extremely high for AIS and ISS scoring. One hundred forty-two patients had blood pressures < 50 mm Hg. The PASG patients had a higher survival rate than non-PASG patients (Pr = 0.055). The PASG appeared to have the most effect on patients with abdominal injuries since no patient with such an injury survived unless a PASG was applied. Controlling for severity using the TRISS method, z scores indicated that the survival rate in the PASG group was significantly higher than expected whereas that in the non-PASG group was similar to that predicted; the same pattern was found when blunt injury and penetrating injury patients were analyzed separately. Improvement in survival among PASG patients occurred despite an average scene time that was 4.7 minutes longer than that for non-PASG patients. No improvement in survival among PASG versus non-PASG patients with blood pressures of 50-70 mm Hg or in those with blood pressures of 90 mm Hg or less was found. We conclude that the use of PASG in severely hypotensive patients (BP < or = 50) should be considered medically acceptable pending randomized controlled studies.

Blood transfusion increases the risk of infection after trauma.

To determine whether blood transfusion influences infection after trauma, we analyzed data on 5366 consecutive patients hospitalized for more than 2 days at eight hospitals over a 2-year period. The incidence of infection was significantly related to the mechanism of injury: penetrating injuries, 8.9%; blunt injuries, 12.9%; and low falls, 21.4%. Stepwise logistic regression analyses of infection using the variables age, sex, respiration rate in the emergency department, Glasgow Coma Scale in the emergency department, Injury Severity Score, shock (systolic blood pressure < 90 mm Hg on admission to the emergency department), and log of total amount of blood transfused during hospitalization showed that amount of blood received and Injury Severity Score were the only two variables that were significant predictors of infection across groups. Even when patients were stratified by Injury Severity Score, the infection rate increased significantly with increases in numbers of units of blood. Blood transfusion in the injured patients is an important independent statistical predictor of infection. Its contribution cannot be attributed to age, sex, or the underlying mechanism of severity of injury.

Association of skull base and facial fractures.

A retrospective analysis of 268 trauma patients with facial fractures who received computed tomography of the head was undertaken to assess an association with skull base fractures. The incidence of skull base fracture was compared to facial fractures of various anatomic locations. Skull base fractures were significantly increased in orbital wall/rim fractures (36.0%, P = .0823). In contrast, skull base fractures related to orbital floor (27.3%, P = .6191) and maxillary/zygomatic (29.4%, P = .1148) fractures were not significantly greater and were infrequently seen with mandible (4.0%, P = .0454) and nasal (7.7%, P = .0345) fractures. The incidence of skull base fracture was directly associated with the number of facial fractures per patient; one facial fracture (21.0%), two facial fractures (30.4%), and three or more facial fractures (33.3%) (P < .05). The incidence of skull base fractures was related to the location of facial fractures and the number of facial fractures per patient. The results provide additional clinical information to facilitate the prompt detection and diagnoses of skull base fracture.