A chi-square automatic interaction detection (CHAID) analysis of factors determining trauma outcomes

Enhancing real-time conflict identification using trajectory data: Exploring the impact of interactions among traffic flow state variables

OBJECTIVE

This study aims to address the limitations of using historical crash data and trajectory data for crash and conflict identification. Specifically, it focuses on enhancing real-time conflict identification by investigating the influence of traffic flow state variables and their interactions on conflicts.

METHODS

Vehicle trajectory data from HighD were processed, allowing extraction of traffic flow state and corresponding conflict during a specific time interval (10 s). Logistic regression models were further used to verify the impact of variables, including interaction terms, on the conflicts for different lane categories (inner, middle, and outer lanes). Additionally, machine learning techniques were employed to compare conflict identification performance including or excluding variable interactions.

RESULTS

The interaction terms of the traffic flow state variables have significant effects on the conflicts for different categories of lanes. It is therefore essential to consider both the individual effects of traffic variables and their interaction effects to analyze conflict risk. Considering variable interactions leads to improved conflict identification accuracy and reduced identification error rates in comparison to the condition where interaction items are not taken into account.

CONCLUSIONS

The interaction terms of traffic flow state variables significantly affect and enhance conflict identification, improving accuracy and reducing error rates. These findings contribute to advancing the high-precision identification of real-time conflict identification, with implications for improving road safety measures.

How Does the Facilitation Effort of Clinical Educators Interact With Aspects of Organizational Context to Affect Research Use in Long-term Care? Evidence From CHAID Analysis

PURPOSE

Organizational context influences the effect of facilitation efforts on research use in care settings. The interactions of these factors are complex. Therefore, the use of traditional statistical methods to examine their interrelationships is often impractical. Big Data analytics can automatically detect patterns within the data. We applied the chi-squared automatic interaction detection (CHAID) algorithm and classification tree technique to explore the dynamic and interdependent relationships between the implementation science concepts-context, facilitation, and research use.

DESIGN

Observational, cross-sectional study based on survey data collected from a representative sample of nursing homes in western Canada.

METHODS

We assessed three major constructs: (a) Conceptual research utilization (CRU) using the CRU scale; (b) facilitation of research use measured by the frequency of contacts between the frontline staff and a clinical educator, or person who brings new ideas to the care unit; and (c) organizational context at the unit level using the Alberta Context Tool (ACT). CHAID analysis was performed to detect the interactions between facilitation and context variables. Results were illustrated in a classification tree to provide a straightforward visualization.

FINDINGS

Data from 312 care units in three provinces were included in the final analysis. Results indicate significant multiway interactions between facilitation and various aspects of the organizational context, including leadership, culture, evaluation, structural resources, and organizational slack (staffing). Findings suggested the preconditions of the care settings where research use can be maximized.

CONCLUSIONS

CHAID analysis helped transform data into usable knowledge. Our findings provide insight into the dynamic relationships of facilitators' efforts and organizational context, and how these factors' interplay and their interdependence together may influence research use.

CLINICAL RELEVANCE

Knowledge of the combined effects of facilitators' efforts and various aspects of organizational context on research use can contribute to effective strategies to narrow the evidence-practice gap in care settings.

Development and external validation of a prediction risk model for short-term mortality among hospitalized U.S. COVID-19 patients: A proposal for the COVID-AID risk tool

BACKGROUND

The 2019 novel coronavirus disease (COVID-19) has created unprecedented medical challenges. There remains a need for validated risk prediction models to assess short-term mortality risk among hospitalized patients with COVID-19. The objective of this study was to develop and validate a 7-day and 14-day mortality risk prediction model for patients hospitalized with COVID-19.

METHODS

We performed a multicenter retrospective cohort study with a separate multicenter cohort for external validation using two hospitals in New York, NY, and 9 hospitals in Massachusetts, respectively. A total of 664 patients in NY and 265 patients with COVID-19 in Massachusetts, hospitalized from March to April 2020.

RESULTS

We developed a risk model consisting of patient age, hypoxia severity, mean arterial pressure and presence of kidney dysfunction at hospital presentation. Multivariable regression model was based on risk factors selected from univariable and Chi-squared automatic interaction detection analyses. Validation was by receiver operating characteristic curve (discrimination) and Hosmer-Lemeshow goodness of fit (GOF) test (calibration). In internal cross-validation, prediction of 7-day mortality had an AUC of 0.86 (95%CI 0.74-0.98; GOF p = 0.744); while 14-day had an AUC of 0.83 (95%CI 0.69-0.97; GOF p = 0.588). External validation was achieved using 265 patients from an outside cohort and confirmed 7- and 14-day mortality prediction performance with an AUC of 0.85 (95%CI 0.78-0.92; GOF p = 0.340) and 0.83 (95%CI 0.76-0.89; GOF p = 0.471) respectively, along with excellent calibration. Retrospective data collection, short follow-up time, and development in COVID-19 epicenter may limit model generalizability.

CONCLUSIONS

The COVID-AID risk tool is a well-calibrated model that demonstrates accuracy in the prediction of both 7-day and 14-day mortality risk among patients hospitalized with COVID-19. This prediction score could assist with resource utilization, patient and caregiver education, and provide a risk stratification instrument for future research trials.

Paediatric terminology in the Australian health and health-education context: a systematic review

AIM

This study aimed to identify paediatric terminology used in the Australian health and health-education context, propose a standardized framework for Australian use, and compare it with a US-based framework.

METHOD

Australian health and health-education websites were systematically searched using a novel hierarchical domain-specific search strategy to identify grey literature containing paediatric terminology. Webpages published from 2009 to February 2014, with a '.gov.au' or '.edu.au' domain and no advertising, were included. Paediatric terms were analysed with power-law distributions. Age definitions were grouped using a chi-squared test automatic interaction detection analysis (p<0.05).

RESULTS

In total, 34 paediatric terms and 197 unique age definitions were identified in 613 webpages. Terms displayed a language distribution, although definitions had semantic and lexical ambiguity. Age definitions were divided into four statistically different groups (F=245.3, p<0.001). Four paediatric terms with distinct age definitions were proposed based on Australian data: 'infant: 0 to <1 year', 'early childhood: 1 year to <5 years', 'child: 5 years to <13 years', and 'young person: 13 years to <22 years'. These recommendations were broader than the US-based comparison.

INTERPRETATION

This is a starting point for standardizing Australian paediatric terminology, and a method for exploring paediatric terminology in other countries.

Scoring systems of severity in patients with multiple trauma

Trauma is a major cause of morbidity and mortality; hence severity scales are important adjuncts to trauma care in order to characterize the nature and extent of injury. Trauma scoring models can assist with triage and help in evaluation and prediction of prognosis in order to organise and improve trauma systems. Given the wide variety of scoring instruments available to assess the injured patient, it is imperative that the choice of the severity score accurately match the application. Even though trauma scores are not the key elements of trauma treatment, they are however, an essential part of improvement in triage decisions and in identifying patients with unexpected outcomes. This article provides the reader with a compendium of trauma severity scales along with their predicted death rate calculation, which can be adopted in order to improve decision making, trauma care, research and in comparative analyses in quality assessment.

Classification tree analysis of the factors influencing injury-related disability caused by the Wenchuan earthquake

OBJECTIVE

To identify the factors that influenced the risk of injury-related disability caused by the Wenchuan earthquake.

METHODS

A chi-squared automatic interaction detection (CHAID) classification tree analysis was used to retrospectively analyse clinical data from patients who underwent surgical treatment for earthquake-related injuries in the first 5 days after the earthquake. The CHAID classification tree explored the relationships between the development of disability and potential influencing factors including sex, age, time interval between injury and treatment, wound type, preoperative and postoperative haemoglobin levels, and operation time.

RESULTS

A total of 334 patients underwent surgery; of these, 113 (33.8%) were discharged with varying degrees of permanent disability. The CHAID classification tree showed that children (≤ 17 years old), a long time interval between injury and treatment, an open wound and a low preoperative haemoglobin level were significant risk factors for disability.

CONCLUSION

The results of this study can help to stratify patients according to their medical needs and to help allocate the available resources efficiently to ensure the best outcomes for injured patients during future earthquakes.

Early predictors of mortality after spine trauma: a level 1 Australian trauma center study

STUDY DESIGN

Retrospective cohort study.

OBJECTIVE

To identify early independent mortality predictors after spine trauma.

SUMMARY OF BACKGROUND DATA

Spine trauma consists of spinal cord and spine column injury. The ability to identify early (within 24 hours) risk factors predictive of mortality in spine trauma has the potential to reduce mortality and improve spine trauma management.

METHODS

Analysis was performed on 215 spine column and/or spinal cord injured patients from July 2008 to August 2011. Univariate and multivariate logistic regression models were applied to investigate the effects of the Injury Severity Score, age, mechanism of injury, blood glucose level, vital signs, brain trauma severity, morbidity before trauma, coagulation profile, neurological status, and spine injuries on the risk of in-hospital death.

RESULTS

Applying a multivariate logistic regression model, there were 7 independent early predictive factors for mortality after spine injury. They were (1) Injury Severity Score more than 15 (odds ratio [OR] = 3.67; P = 0.009), (2) abnormal coagulation profile (OR = 6; P < 0.0001), (3) patients 65 years or older (OR = 3.49; P = 0.007), (4) hypotension (OR = 2.9; P = 0.033), (5) tachycardia (OR = 4.04; P = 0.005), (6) hypoxia (OR = 2.9; P = 0.033), and (7) multiple comorbidities (OR = 3.49; P = 0.007). Severe traumatic brain injury was also associated with mortality but was excluded from multivariate analysis because there were no patients with this variable in the comparison group.

CONCLUSION

Mortality predictors for spine trauma patients are similar to those for general trauma patients. Spine injury variables were shown not to be independent predictors of spine trauma mortality.

Outcome analysis and outcome predictors of traumatic head injury in childhood: Analysis of 454 observations

Aim:

To determine factors associated with poor outcome in children suffering traumatic head injury (HI).

Materials and Methods:

A retrospective study over an 8-year period including 454 children with traumatic HI admitted in the Intensive Care Unit of a university hospital (Sfax-Tunisia). Basic demographic, clinical, biological and radiological data were recorded on admission and during the ICU stay. Prognosis was defined according Glasgow outcome scale (GOS) performed after hospital discharge by ICU and pediatric physicians.

Results:

There were 313 male (68.9%) and 141 female patients. Mean age (±SD) was 7.2±3.8 years, the main cause of trauma was traffic accidents (69.4%). Mean Glasgow coma scale (GCS) score was 8±3, mean injury severity score (ISS) was 26.4±8.6, mean pediatric trauma score (PTS) was 4±2 and mean pediatric risk of mortality (PRISM) was 11.1±8. The GOS performed within a mean delay of 7 months after hospital discharge was as follow: 82 deaths (18.3%), 5 vegetative states (1.1%), 15 severe disabilities (3.3%), 71 moderate disabilities (15.6%) and 281 good recoveries (61.9%). Multivariate analysis showed that factors associated with poor outcome (death, vegetative state or severe disability) were: PRISM ≥24 (P=0.03; OR: 5.75); GCS ≤8 (P=0.04; OR:2.42); Cerebral edema (P=0.03; OR:2.23); lesion type VI according to Traumatic Coma Data Bank Classification (P=0.002; OR:55.95); Hypoxemia (P=0.02; OR:2.97) and sodium level >145 mmol/l (P=0.04; OR: 4.41).

Conclusions:

A significant proportion of children admitted with HI were found to have moderate disability at follow-up. We think that improving prehospital care, establishing trauma centers and making efforts to prevent motor vehicle crashes should improve the prognosis of HI in children.

Isolated traumatic head injury in children: Analysis of 276 observations

BACKGROUND

To determine predictive factors of mortality among children after isolated traumatic brain injury.

MATERIALS AND METHODS

In this retrospective study, we included all consecutive children with isolated traumatic brain injury admitted to the 22-bed intensive care unit (ICU) of Habib Bourguiba University Hospital (Sfax, Tunisia). Basic demographic, clinical, biochemical, and radiological data were recorded on admission and during ICU stay.

RESULTS

There were 276 patients with 196 boys (71%) and 80 girls, with a mean age of 6.7 ± 3.8 years. The main cause of trauma was road traffic accident (58.3%). Mean Glasgow Coma Scale score was 8 ± 2, Mean Injury Severity Score (ISS) was 23.3 ± 5.9, Mean Pediatric Trauma Score (PTS) was 4.8 ± 2.3, and Mean Pediatric Risk of Mortality (PRISM) was 10.8 ± 8. A total of 259 children required mechanical ventilation. Forty-eight children (17.4%) died. Multivariate analysis showed that factors associated with a poor prognosis were PRISM > 24 (OR: 10.98), neurovegetative disorder (OR: 7.1), meningeal hemorrhage (OR: 2.74), and lesion type VI according to Marshall tomographic grading (OR: 13.26).

CONCLUSION

In Tunisia, head injury is a frequent cause of hospital admission and is most often due to road traffic injuries. Short-term prognosis is influenced by demographic, clinical, radiological, and biochemical factors. The need to put preventive measures in place is underscored.

Predictors of change in quality of life after distributed constraint-induced therapy in patients with chronic stroke

BACKGROUND

There are no reports of predictive models or predictors for quality of life (QoL) after constraint-induced therapy (CIT).

OBJECTIVE

This investigation identified predictors of change in stroke-related QoL after distributed CIT using the Chi-squared Automatic Interaction Detector (CHAID) method.

METHODS

A total of 58 patients with chronic stroke were treated with CIT for 2 hours daily for 3 weeks. The 7 potential predictors were age, gender, side of lesion, time since stroke, cognitive status, motor impairment of upper extremity, and activities of daily living (ADL). QoL was measured by the Stroke Impact Scale (SIS). CHAID analysis was used to examine for associations between the 7 predictors and each SIS domain. The validity of each model generated by the analysis was evaluated.

RESULTS

Daily functional performance as measured by the Functional Independence Measure (FIM) was found to determine SIS outcomes, including overall score (P = .006) and the ADL/instrumental ADL (IADL) domain (P = .004). None of the potential predictors emerged as significant predictors of the strength, memory, emotion, communication, mobility, hand function, and participation domains of SIS. The misclassification risk estimates were small, indicating good validity for the CHAID models.

CONCLUSIONS

The functional independence score of the FIM can predict the overall SIS score as well as the ADL/IADL domain of the SIS in chronic stroke patients who receive CIT, but larger databases are needed to confirm this. CHAID analysis was a useful approach for an exploratory study.

The trauma risk adjustment model: a new model for evaluating trauma care

BACKGROUND DATA

The trauma injury severity score (TRISS) has been used for over 20 years for retrospective risk assessment in trauma populations. The TRISS has serious limitations, which may compromise the validity of trauma care evaluations.

OBJECTIVE

To derive and validate a new mortality prediction model, the trauma risk adjustment model (TRAM), and to compare the performance of the TRAM to that of the TRISS in terms of predictive validity and risk adjustment.

METHODS

The Quebec Trauma Registry (1998-2005), based on the mandatory participation of 59 designated provincial trauma centers, was used to derive the model. The American National Trauma Data Bank (2000-2005), based on the voluntary participation of any US hospitals treating trauma, was used for the validation phase. Adult patients with blunt trauma respecting at least one of the following criteria were included: hospital stay >2 days, intensive care unit admission, death, or hospital transfer. Hospital mortality was modeled with logistic generalized additive models using cubic smoothing splines to accommodate nonlinear relations to mortality. Predictive validity was assessed with model discrimination and calibration. Risk adjustment was assessed using comparisons of risk-adjusted mortality between hospitals.

RESULTS

The TRAM generated an area under the receiving operator curve of 0.944 and a Hosmer-Lemeshow statistic of 42 in the derivation phase. In the validation phase, the TRAM demonstrated better model discrimination and calibration than the TRISS (area under the receiving operator curve = 0.942 and 0.928, P < 0.001; Hosmer-Lemeshow statistics = 127 and 256, respectively). Replacing the TRISS with the TRAM led to a mean change of 28% in hospital risk-adjusted odds ratios of mortality.

CONCLUSIONS

Our results suggest that adopting the TRAM could improve the validity of trauma care evaluations and trauma outcome research.

Severe head injury among children: prognostic factors and outcome

AIM

To determine predictive factors of mortality among children after traumatic brain injury.

METHODS

A retrospective study over 8 years of 222 children with severe head injury (Glasgow Coma Scale score < or = 8) admitted to a university hospital (Sfax, Tunisia). Basic demographic, clinical, biological and radiological data were recorded on admission and during intensive care unit stay.

RESULTS

The study included 163 boys (73.4%) and 59 girls, with mean age 7.54+/-3.8 years. The main cause of trauma was road traffic accident (75.7%). Mean Glasgow Coma Scale score was 6+/-1.5, mean Injury Severity Score (ISS) was 28.2+/-6.9, mean Paediatric Trauma Score (PTS) was 3.7+/-2.1 and mean Paediatric Risk of Mortality (PRISM) was 14.3+/-8.5; 54 children (24.3%) died. Univariate analysis showed that low PTS on admission, high ISS or PRISM, presence of shock or meningeal haemorrhage or bilateral mydriasis, and serum glucose > 10 mmol l(-1) were associated with mortality rate. Multivariate analysis showed that factors associated with a poor prognosis were PRISM > 20 and bilateral mydriasis on admission.

CONCLUSIONS

In Tunisia, head injury is a frequent cause of hospital admission and is most often due to road traffic accidents. Short-term prognosis is poor, with a high mortality rate (24.3%), and is influenced by demographic, clinical, radiological and biological factors.

The value of trauma registries

Trauma registries are databases that document acute care delivered to patients hospitalised with injuries. They are designed to provide information that can be used to improve the efficiency and quality of trauma care. Indeed, the combination of trauma registry data at regional or national levels can produce very large databases that allow unprecedented opportunities for the evaluation of patient outcomes and inter-hospital comparisons. However, the creation and upkeep of trauma registries requires a substantial investment of money, time and effort, data quality is an important challenge and aggregated trauma data sets rarely represent a population-based sample of trauma. In addition, trauma hospitalisations are already routinely documented in administrative hospital discharge databases. The present review aims to provide evidence that trauma registry data can be used to improve the care dispensed to victims of injury in ways that could not be achieved with information from administrative databases alone. In addition, we will define the structure and purpose of contemporary trauma registries, acknowledge their limitations, and discuss possible ways to make them more useful.

Geographical differences in mortality of severely injured patients in Italy

BACKGROUND

In Italy there are no accepted standards for trauma care nor dedicated programs for quality assessment on a national scale, like trauma registries. At the same time there seems to be a north-south gradient in the quality of health care. We hypothesized that geographical inequalities of health-care quality may affect trauma mortality.

METHODS

Retrospective comparison of hospital mortality by Cox regression in three main areas of Italy adjusted for age, Glasgow Coma Scale and source of admission. A leading national database on patients admitted to intensive care units (ICU) in the years 2002-2005 was used. 9162 adult trauma cases admitted to the ICU from the emergency department were included.

RESULTS

There is a significant north-south gradient of risk. Compared to the north, the risk of death is about 60% higher in the south and about 30% higher in the central region. These figures are similar in both referral centres and other hospitals and both in the head-injured only and total injured cases.

CONCLUSION

Despite the limitations of this study, mainly related to sampling issues, risk-adjustment and incomplete follow-up, the large geographic differences in mortality that we found highlight likely deficiencies in the quality of trauma care that deserve further accurate assessment.

Diagnosing osteoporosis by using dental panoramic radiographs: the OSTEODENT project

OBJECTIVES

Measurement of cortical thickness and subjective assessment of cortical porosity on panoramic radiographs are methods previously reported for diagnosing osteoporosis. The aims of this study were to determine the relative efficacy of the mandibular cortical index and cortical width in detecting osteoporosis, both alone and in combination, and to determine the optimal cortical width threshold for referral for additional osteoporosis investigation.

STUDY DESIGN

Six hundred seventy-one postmenopausal women 45 to 70 years of age were recruited for this study. They received dual energy x-ray absorptiometry (DXA) scans of the left hip and lumbar spine (L1 to L4), and dental panoramic radiographic examinations of the teeth and jaws. Three observers separately assessed the mandibular cortical width and porosity in the mental foramen region of the mandible. Cortical width was corrected for magnification errors. Chi-squared automatic interaction detection analysis (CHAID) software was used (SPSS AnswerTree, version 3.1, SPSS Inc., Chicago, IL).

RESULTS

Chi-squared automatic interaction detection analysis showed that the cortical porosity was a poorer predictor of osteoporosis than mandibular cortical width. For the 3 observers, a mandibular cortical width of <3 mm provided diagnostic odds ratios of 6.51, 6.09, and 8.04. The test is therefore only recommended in triage screening of individuals by using radiographs made for purposes other than osteoporosis.

CONCLUSION

When evaluating panoramic radiographs, only those patients with the thinnest mandibular cortices (i.e., <3 mm) should be referred for further osteoporosis investigation.

Developing empirical models to enhance stroke rehabilitation

PURPOSE

Models identifying functional indicators most strongly associated with favourable and unfavourable outcomes may bolster evidence to improve stroke rehabilitation assessment and intervention. This study examined the feasibility of decision analysis methods for developing data-driven models that examined associations between specific functional indicators and global disability.

METHOD

Data were derived from functional assessment of 67 participants 3 months following stroke. Decision analysis methods were used to examine specific activity and body function indicators associated with global disability, and the degree of limitation or impairment that contributed to favourable and unfavourable outcomes, in 2 models. The feasibility of decision analysis methods was evaluated.

RESULTS

Of the 26 activity indicators, dressing was most strongly associated with global disability, followed by bill mailing, shopping and sweeping. Of 15 body function indicators, facial weakness and mental functions were most strongly associated with global disability. The misclassification risk estimates were fair for the two models.

CONCLUSIONS

Findings suggest that decision analysis methods show promise for developing models examining associations between specific functional indicators and disability. Further study with these methods may identify specific priorities for functional assessment and intervention in stroke rehabilitation.

Endemic human fasciolosis in the Bolivian Altiplano

Fasciolosis, caused by trematodes of the genus Fasciola, is an emerging disease of humans. One of the highest levels of human fasciolosis hepatica is found amongst the indigenous Aymaran people of the Northern Bolivian Altiplano. A meta-analysis of epidemiological surveys from 38 communities in the region demonstrates that fasciolosis has been endemic in the region since at least 1984 and is a zoonosis of rural communities. Human and bovine fasciolosis is associated with the communities lying in the plain from Lake Titicaca to La Paz, predominantly in the Los Andes province. In Los Andes incidences of up to 67% of population cohorts were found, and prevalence is age-related with the highest infection rate in children aged 8-11 years.

Prognosis of Traumatic Head Injury in South Tunisia: A Multivariate Analysis of 437 Cases

BACKGROUND

This study aimed to determine predictive factors of mortality after posttraumatic brain injury.

METHODS

A retrospective study conducted over a 3-year period (1997-1999) involved 437 adult patients with head injury admitted to the intensive care unit of a university hospital in Sfax, Tunisia. Basic demographic, clinical, biologic, and radiologic data were recorded at admission and during the intensive care unit stay.

RESULTS

This study included 393 men (90%) and 44 women with a mean age of 36 +/- 17 years. Traffic accidents were the main cause of trauma (85.6%). In 58% of the cases, the injury was serious (Glasgow Coma Score, <8). The mean simplified acute physiology score was 39 +/- 15, and the mean Injury Severity Score was 34.5 +/- 17. Of the 437 patients, 127 (29.1%) died. According to multivariate analysis, the factors that correlated with a poor prognosis were age older than 40 years (p < 0.01), simplified acute physiology score exceeding 40 (p < 0.001), Glasgow Coma Score lower than 7 (p = 0.03), intracranial mass lesion (p = 0.02), a cerebral herniation (p < 0.001), diabetes insipidus (p < 0.001), and blood sugar level higher than 10 mmol/L (p < 0. 001).

CONCLUSIONS

In Tunisia, head injury is a frequent cause of hospitalization, comprising 14.4% of all adult admissions. It is observed most often among young patients involved in traffic accidents. The short-term prognosis is poor, with a high (29%) mortality rate, and determined by demographic, clinical, radiologic, and biologic factors. Prevention is highly advised.

Improving the TRISS Methodology by Restructuring Age Categories and Adding Comorbidities

BACKGROUND

The Trauma and Injury Severity Score (TRISS) methodology was developed to predict the probability of survival after trauma. Despite many criticisms, this methodology remains in common use. The purpose of this study was to show that improving the stratification for age and adding an adjustment for comorbidity significantly increases the predictive accuracy of the TRISS model.

METHODS

The trauma registry and the hospital administrative database of a regional trauma center were used to identify all blunt trauma patients older than 14 years of age admitted with International Classification of Diseases, Ninth Revision codes 800 to 959 from April 1993 to March 2001. Each individual medical record was then reviewed to ascertain the Revised Trauma Score, the Injury Severity Score, the age of the patients, and the presence of eight comorbidities. The outcome variable was the status at discharge: alive or dead. The study population was divided into two subsamples of equal size using a random sampling method. Logistic regression was used to develop models on the first subsample; a second subsample was used for cross-validation of the models. The original TRISS and three TRISS-derived models were created using different categorizations of Revised Trauma Score, Injury Severity Score, and age. A new model labeled TRISSCOM was created that included an additional term for the presence of comorbidity.

RESULTS

There were 5,672 blunt trauma patients, 2,836 in each group. For original TRISS, the Hosmer-Lemeshow statistic (HL) was 179.1 and the area under the receiver operating characteristic (AUROC) curve was 0.873. Sensitivity and specificity were 99.0% and 27.8%, respectively. For the best modified TRISS model, the HL statistic was 20.35, the AUROC curve was 0.902, the sensitivity was 99.0%, and the specificity was 27.8%. For TRISSCOM, the HL statistic was 14.95 and the AUROC curve was 0.918. Sensitivity and specificity were 99.0% and 29.7%, respectively. The difference between the two models almost reached statistical significance (p = 0.086). When TRISSCOM was applied to the cross-validation group, the HL statistic was 10.48 and the AUROC curve was 0.914. The sensitivity was 98.6% and the specificity was 34.9%.

CONCLUSION

TRISSCOM can predict survival more accurately than models that do not include comorbidity. A better categorization of age and the inclusion of comorbid conditions in the logistic model significantly improves the predictive performance of TRISS.

New Zealanders' attitudes to milk: implications for public health

OBJECTIVE

To identify consumer attitudes and beliefs about (liquid) milk that may be barriers to consumption.

DESIGN

Two random-quota telephone surveys conducted in Auckland one year apart. Respondents were questioned about their usual milk intake and their attitudes to milk. The questionnaire included attitude items that reflected the main themes of consumer interest in milk.

SETTING

New Zealand.

SUBJECTS

Seven hundred and thirteen respondents in the baseline survey and a separate sample of 719 respondents in the follow-up survey.

RESULTS

At least one-third of the respondents consumed less than a glass (250 ml) of milk a day. Non-consumption was highest in young women (15%). People's concerns about milk related to what was important in their lives; what threatens them physically and emotionally. Women held more positive attitudes but they were concerned about the fat content of milk. Men were less aware of milk's nutritional benefits and as a result were less appreciative of its value.

CONCLUSIONS

There is an opportunity to develop public health initiatives to address the barriers to drinking milk. Industry-health alliances may be an effective means to provide positive nutrition messages about milk and to engage the support of health professionals.

Survey of trauma patients requiring maxillofacial intervention, ages 56 to 91 years, with length of stay analysis

PURPOSE

The purpose of this study was to analyze trauma patients, ages 55 and older, sustaining multiple injuries including maxillofacial trauma. Factors influencing length of intensive care unit stay (ICUS) and length of total hospital stay were delineated and examined to determine if specific causes of increased length of stay could be elucidated, and, once known, if these causes could translate into recommendations tailored to the oral and maxillofacial surgery trauma practice.

PATIENTS AND METHODS

One hundred ninety-six patients, 55 years of age or older, who received either consultation alone, or consultation with surgical treatment, by oral and maxillofacial surgeons, from January 1991 to August 1998 were included in this study. Variables of interest included location of traumatic event, mechanism of injury, patient age and gender, comorbidities on presentation, Injury Severity Score (ISS), specific injuries incurred, ICUS, length of hospital stay (LOS), surgical interventions, and disposition.

RESULTS

Complications were the statistically significant factor determining length of ICU stay. ICUS, complications incurred, and ISS were the important predictors of total LOS. The significant complications affecting LOS were infectious, respiratory, and hematologic complications.

CONCLUSION

The number of complications the patient incurs after an injury can predict length of ICUS. Length of ICUS, ISS, and number of complications incurred were the strongest predictors for total length of hospital stay. Other variables, including age, gender, living or dead, blunt versus penetrating injury, ISS, fracture site (skull, midface, or lower face), and comorbidities on presentation were not statistically significant in this patient population. Infectious, respiratory, and hematologic complications were the complications most closely correlated with increasing length of ICUS and total hospital stay.

Reporting data following major trauma and analysing factors associated with outcome using the new Utstein style recommendations

STUDY OBJECTIVE

To collect and present retrospectively the recommended core data from the Utstein style, analyse factors associated with outcome in major trauma, and discuss the value of the Utstein style definition of major trauma.

DESIGN

A retrospective trauma cohort study.

SETTING

A Norwegian trauma system with a 1200 bed combined local and referral trauma hospital without a formal trauma registry, covering a population of approximately 2.0 million.

PARTICIPANTS

3391 injured patients admitted 12 months from January 15, 1996.

MAIN OUTCOME MEASURES

Recommended core data from the Utstein style, and factors associated with outcome defined as in-hospital death within 30 days.

RESULTS

225 patients had an injury severity score (ISS)>15. In each of the 225 patients, we were able to obtain at least 47% of the recommended core data. Age >70 years, fall as a mechanism of injury, and a Trauma Score (TS)< or =14 were significantly associated with poor outcome. Of 22 with no major trauma (ISS<16), two died in hospital and 20 had an intensive care unit stay of more than 2 days.

CONCLUSION

We found it difficult to collect retrospectively the recommended core data of the Utstein style. Age and physiological alterations (TS) were significantly related to outcome. The recommended definition of major trauma (ISS>15) did not cover all life-threatening injuries. The implementation of trauma registries based on the Utstein style recommendations could facilitate system evaluation and comparison, but definitions and categorizations should be further developed. Efforts should be made to reduce the number of core data.

The effect of polytrauma in persons with traumatic spine injury. A prospective database of spine fractures

STUDY DESIGN

A mixed cross-sectional survey and cohort study using a prospectively gathered database of persons with traumatic spine injury.

OBJECTIVES

To identify demographic and injury mechanism factors that predict greater injury severity, and to determine the effect of injury severity on outcomes in traumatic spine fracture.

SUMMARY OF BACKGROUND DATA

Traumatic spine fracture outcome studies have focused on defining type and level of vertebral fracture without considering the severity of associated injuries. In the trauma population, greater injury severity has been shown to be related to worse outcome. No studies have been reported on the effect of injury severity on outcome in the traumatic spine fracture population.

METHODS

Prospectively collected data on 830 persons with traumatic spine injury who were admitted to a trauma hospital were reviewed. Patient demographics; injury mechanism; hospital events; and disability, employment, and pain status at discharge, 1 year, and 2 years after injury were recorded. Associations between these factors and trauma severity (Injury Severity Score) were explored using Pearson's correlation and analysis of variance.

RESULTS

Trauma was more severe in patients who had been married previously, who were involved in a motor vehicle accident, were ejected from the vehicle, had loss of consciousness, had higher-level and multiple complicated vertebral fractures, or had neurologic deficit. Those more severely injured had longer lengths of stay, more surgery, more complications, higher mortality, more disability, and less return to work.

CONCLUSIONS

Persons with traumatic spine injury and polytrauma have poorer short- and long-term outcomes. This high-risk group may require aggressive interventions, more hospital resources, and close follow-up observation after discharge from hospital to optimize outcome.

Eye injuries in patients with major trauma

BACKGROUND

A study was performed to determine the type and frequency of ocular injuries in patients with major trauma.

METHODS

All patients with ocular and adnexal injuries (n = 178) among 1,119 patients admitted with major trauma (Injury Severity Score >15) to the Royal Prince Alfred Hospital from July 1990 to December 1997 were analyzed.

RESULTS

Sixteen percent of the major trauma cohort had ocular or orbital trauma. Fifty-five percent of patients with injuries involving the face had ocular or orbital injuries. A range of ocular injuries was seen. Analysis of the major trauma cohort showed that motor vehicle drivers, orbital and base of skull fractures, eyelid lacerations, and superficial eye injuries were strongly associated with vision-threatening injury.

CONCLUSION

Patients with major trauma and facial injuries have a high risk of vision-threatening injury. Patients with orbital fractures, base of skull fracture, eyelid lacerations, and superficial eye injuries should be assessed by an ophthalmologist as part of the early management of their trauma to determine whether an ocular injury is present.

The conundrum of the Glasgow Coma Scale in intubated patients: a linear regression prediction of the Glasgow verbal score from the Glasgow eye and motor scores

BACKGROUND

The Glasgow Coma Scale (GCS), which is the foundation of the Trauma Score, Trauma and Injury Severity Score, and the Acute Physiology and Chronic Health Evaluation scoring systems, requires a verbal response. In some series, up to 50% of injured patients must be excluded from analysis because of lack of a verbal component for the GCS. The present study extends previous work evaluating derivation of the verbal score from the eye and motor components of the GCS.

METHODS

Data were obtained from a state trauma registry for 24,565 unintubated patients. The eye and motor scores were used in a previously published regression model to predict the verbal score: Derived Verbal Score = -0.3756 + Motor Score * (0.5713) + Eye Score * (0.4233). The correlation of the actual and derived verbal and GCS scales were assessed. In addition the ability of the actual and derived GCS to predict patient survival in a logistic regression model were analyzed using the PC SAS system for statistical analysis. The predictive power of the actual and the predicted GCS were compared using the area under the receiver operator characteristic curve and Hosmer-Lemeshow goodness-of-fit testing.

RESULTS

A total of 24,085 patients were available for analysis. The mean actual verbal score was 4.4 +/- 1.3 versus a predicted verbal score of 4.3 +/- 1.2 (r = 0.90, p = 0.0001). The actual GCS was 13.6 + 3.5 versus a predicted GCS of 13.7 +/- 3.4 (r = 0.97, p = 0.0001). The results of the comparison of the prediction of survival in patients based on the actual GCS and the derived GCS show that the mean actual GCS was 13.5 + 3.5 versus 13.7 + 3.4 in the regression predicted model. The area under the receiver operator characteristic curve for predicting survival of the two values was similar at 0.868 for the actual GCS compared with 0.850 for the predicted GCS.

CONCLUSIONS

The previously derived method of calculating the verbal score from the eye and motor scores is an excellent predictor of the actual verbal score. Furthermore, the derived GCS performed better than the actual GCS by several measures. The present study confirms previous work that a very accurate GCS can be derived in the absence of the verbal component.

Resource utilization by injured automobile occupants and pedestrians

BACKGROUND

This study was designed to test the hypothesis that the hospital resources utilized in treating pedestrian trauma would be significantly greater than that for automobile occupants. This was based on previous studies that showed that the demographic features and patterns of injury sustained by the pedestrian population were significantly different from that of automobile occupants.

METHODS

A hospital-based study was designed utilizing retrospective analysis of a prospective trauma database. All primary retrievals of pedestrians (n=547) and automobile occupants (n=597) involved in accidents in Central Sydney from mid-1990 to mid-1995 were included. The length of hospital stay, use of the intensive care unit (ICU) and visits to the operating theatre (Standard Resource Cost) were compared.

RESULTS

The age and injury severity scores were significantly higher for the pedestrian group. The length of stay (days) for the pedestrians (mean, 12 SD 14; median, 7 interquartile range (IQR) 13), was significantly higher (P < 0.0001 ) than that for the automobile occupants (mean, 7 SD 11; median, 2 IQR 6). The ICU utilization (days) for the pedestrians (mean, 1.3 SD 4.0; median, 0) was significantly higher (P < 0.0001) than that of the automobile occupants (mean, 0.6 SD 2.9; median, 0). The average operating theatre utilization per pedestrian (0.65 visits) exceeded that of automobile occupants (0.43) by 50% (P < 0.0001).

CONCLUSIONS

The study confirms that the acute care of pedestrian injury utilizes more hospital resources than that of automobile occupants. Resources should be allocated to meet this need both in terms of hospital reimbursement and overall directives in public health policy.