The ICD-9-based illness severity score: a new model that outperforms both DRG and APR-DRG as predictors of survival and resource utilization

Diagnosis-Related Group Weight and Derived Case Mix Index to Assess the Complexity among Twins

OBJECTIVE

Diagnosis-related groups (DRGs) are used to summarize hospital morbidity and mortality. Each DRG has a weight which is important in calculating the case mix index (CMI), a numeric summary of disease complexity in a population of patients. We utilized DRG weight and resultant CMI to compare postnatal outcomes among singletons versus monochorionic and monoamniotic, monochorionic diamniotic, and dichorionic diamniotic twins.

STUDY DESIGN

This single-center and retrospective cohort study evaluated DRGs assigned by the investigators, birth weight, gestational age, length of stay (LOS), NICU admission rate, and mortality in twin births between 2014 and 2016. Twins were analyzed depending on chorionicity and amnionicity. Overall, 3 months of singleton births served as the control. The CMI derived from DRG weights were compared across groups.

RESULTS

Twins (n = 288) had lower gestational ages and birth weights and higher mortality, LOS, NICU admission rates and DRG weights/CMI compared with singletons (n = 327; p < 0.001 for each). The LOS was no different between twin subtypes; monochorionic monoamniotic twins had the highest mortality and DRG weight (p < 0.001).

CONCLUSION

DRG weight and CMI values summarize in-hospital complexity and can be a useful tool to evaluate differences in care complexity among groups of patients.

KEY POINTS

· Using diagnosis-related group and case mix index to assess morbidities.. · Morbidities of twins are monochorionic-monoamniotic versus monochorionic-diamniotic versus dichorionic-diamniotic twins.. · Only seven diagnosis-related group in neonatology make it a valuable tool for clinicians..

Machine Learning in the Prediction of Trauma Outcomes: A Systematic Review

STUDY OBJECTIVE

Machine learning models carry unique potential as decision-making aids and prediction tools for improving patient care. Traumatically injured patients provide a uniquely heterogeneous population with severe injuries that can be difficult to predict. Given the relative infancy of machine learning applications in medicine, this systematic review aimed to better understand the current state of machine learning development and implementation to help create a basis for future research.

METHODS

We conducted a systematic review from inception to May 2021, using Embase, MEDLINE through Ovid, Web of Science, Google Scholar, and relevant gray literature, for uses of machine learning in predicting the outcomes of trauma patients. The screening and data extraction were performed by 2 independent reviewers.

RESULTS

Of the 14,694 identified articles screened, 67 were included for data extraction. Artificial neural networks comprised the most commonly used model, and mortality was the most prevalent outcome of interest. In terms of machine learning model development, there was a lack of studies that employed external validation, feature selection methods, and performed formal calibration testing. Significant heterogeneity in reporting was also observed between the machine learning models employed, patient populations, performance metrics, and features employed.

CONCLUSION

This review highlights the heterogeneity in the development and reporting of machine learning models for the prediction of trauma outcomes. While these models present an area of opportunity as an ancillary to clinical decision-making, we recommend more standardization and rigorous guidelines for the development of future models.

International classification of disease based injury severity score (ICISS): A data linkage study of hospital and death data in Victoria, Australia

BACKGROUND

Surveillance of severe injury incidence and prevalence using ICD-based injury severity scores (ICISS) requires valid, locally applicable diagnosis-specific survival probabilities (DSPs). This study aims to derive and validate ICISS in Victoria, Australia, and compare various ICISS methodologies in terms of accuracy and calculated severe injury prevalence.

METHODS

This study used injury admissions (ICD-10-AM coded) from the Victorian Admitted Episodes Database (VAED) linked with death data (Cause of Death - Unit Record Files: CODURF). Using design data (July 2008 - June 2014; n = 720,759), various ICISS scales were derived, based on (i) in-hospital and (ii) three-month mortality. These scales were applied to testing data (July 2014 - December 2016; n = 334,363). Logistic regression modelling was used to determine model discrimination and calibration.

RESULTS

In the design data, there were 6,337(0.9%) hospital deaths and 17,514(2.4%) three-months deaths; in the testing data, there were 2,700(0.8%) hospital deaths and 8,425(2.5%) three-month deaths. Newly developed ICISS scales had acceptable to outstanding discrimination, with Area Under the Curve ranging from 0.758 to 0.910. Age-specific ICISS scales were superior to general ICISS scales in model discrimination but inferior in model calibration. Calculated severe injury (ICISS ≤0.941) prevalence in the testing data ranged from 2% to 24%, depending on which mortality outcomes were used to calculate DRGs.

CONCLUSIONS

This study provides local, validated ICISS scores that can be used in Victoria. It is recommended that age group stratified ICISS based on the worst-injury method is used. From the comparison of various ICISS scores, reflecting the range of ICISS permutations that are currently in use, care should be taken to compare ICISS methodology before comparing severe injury prevalence per population, injury cause, and time trends.

Pediatric firearm injuries: Racial disparities and predictors of healthcare outcomes

BACKGROUND/PURPOSE

The U.S. has an alarming rate of firearm injuries. Racial disparities among victims and predictors of outcomes are not well established. Our objective was to assess costs, length of stay (LOS), and inpatient mortality among nonfatal and fatal pediatric firearm injuries that required hospitalization.

METHODS

Pediatric (≤18 years of age) hospitalizations with a firearm injury discharge diagnosis were identified from the national Kids' Inpatient Databases (KID) for 2006 through 2012. Firearm injury intent, weapon type, and hospitalization rates by racial groups were examined. Inpatient mortality, costs, and length of stay were examined using regression models.

RESULTS

Of 15,211 hospitalizations, the majority of injuries were due to assault (60%) and the intentions of firearm injury differed by race (p < 0.001). The median cost per hospitalization was $10,159 (interquartile range: $5071 to $20,565), totaling more than a quarter of a billion dollars. On regression analysis, Black (OR: 0.41; CI: 0.30-0.55) and Hispanic (OR: 0.47; CI: 0.34-0.66) patients were less likely to die than White patients.

CONCLUSION

Pediatric firearm injury circumstances and survival vary by race with Whites being more likely to experience unintentional injury and suicide, while Blacks and Hispanics are more likely to experience inflicted injury.

LEVEL OF EVIDENCE

Level II.

TYPE OF STUDY

Clinical Research Study.

Summary perioperative risk metrics within the electronic medical record predict patient-level cost variation in pancreaticoduodenectomy

BACKGROUND

Automated data extraction from the electronic medical record is fast, scalable, and inexpensive compared with manual abstraction. However, concerns regarding data quality and control for underlying patient variation when performing retrospective analyses exist. This study assesses the ability of summary electronic medical record metrics to control for patient-level variation in cost outcomes in pancreaticoduodenectomy.

METHODS

Patients that underwent pancreaticoduodenectomy from 2014 to 2018 at a single institution were identified within the electronic medical record and linked with the National Surgical Quality Improvement Program. Variables in both data sets were compared using interrater reliability. Logistic and linear regression modelling of complications and costs were performed using combinations of comorbidities/summary metrics. Models were compared using the adjusted R² and Akaike information criterion.

RESULTS

A total of 117 patients populated the final data set. A total of 31 (26.5%) patients experienced a complication identified by the National Surgical Quality Improvement Program. The median direct variable cost for the encounter was US$14,314. Agreement between variables present in the electronic medical record and the National Surgical Quality Improvement Program was excellent. Stepwise linear regression models of costs, using only electronic medical record-extractable variables, were non-inferior to those created with manually abstracted individual comorbidities (R² = 0.67 vs 0.30, Akaike information criterion 2,095 vs 2,216). Model performance statistics were minimally impacted by the addition of comorbidities to models containing electronic medical record summary metrics (R² = 0.67 vs 0.70, Akaike information criterion 2,095 vs 2,088).

CONCLUSION

Summary electronic medical record perioperative risk metrics predict patient-level cost variation as effectively as individual comorbidities in the pancreaticoduodenectomy population. Automated electronic medical record data extraction can expand the patient population available for retrospective analysis without the associated increase in human and fiscal resources that manual data abstraction requires.

Impact of Weekend Presentation on Short-Term Outcomes and Choice of Clipping vs Coiling in Subarachnoid Hemorrhage

BACKGROUND

Presentation on a weekend is commonly associated with higher mortality and a decreased likelihood of receiving invasive procedures.

OBJECTIVE

To determine whether weekend presentation influences mortality, discharge destination, or type of treatment received (clip vs coil) in subarachnoid hemorrhage (SAH).

METHODS

We performed a serial cross-sectional retrospective study using the Nationwide Inpatient Sample. All adult discharges with a primary diagnosis of SAH (ICD-9-CM 435) from 2005 to 2010 were included, and records with trauma or arteriovenous malformation were excluded. Unadjusted and adjusted associations between weekend presentation and 3 outcomes (in-hospital mortality, discharge destination, and treatment with clip vs coil) were estimated using chi-square tests and multilevel logistic regression.

RESULTS

A total of 46 093 admissions for nontraumatic SAH were included in the sample; 24.6% presented on a weekend, 68.9% on a weekday, and 6.5% had unknown day of presentation. Weekend admission was not a significant predictor of inpatient mortality (25.4% weekend vs 24.9% weekday; P = .44), or a combined poor outcome measure of mortality or discharge to long-term acute care or hospice (30.3% weekend vs 29.4% weekday; P = .23). Among those treated for aneurysm obliteration, the proportion of clipped vs coiled did not change with weekend vs weekday presentation (21.5% clipped with weekend presentation vs 21.6% weekday, P = .95; 21.5% coiled with weekend presentation vs 22.4% weekday, P = .19).

CONCLUSION

Presentation with nontraumatic SAH on a weekend did not influence mortality, discharge destination, or type of treatment received (clip vs coil) compared with weekday presentation.

Does acute myocardial infarction kill more people on weekends? Analysis of in-hospital mortality rates for weekend admissions in Portugal

Objectives To investigate a possible weekend effect in the in-hospital mortality rate for acute myocardial infarction in Portugal, and whether the delay in invasive intervention contributes to this effect. Methods Data from the National 2011-2015 Diagnostic-Related-Group databases were analysed. The focus was on adult patients admitted via the emergency department and with the primary diagnosis of acute myocardial infarction. Patients were grouped according to ST-elevation myocardial infarction and non-ST-elevation myocardial infarction episodes. We employed multivariable logistic regressions to determine the association between weekend admission and in-hospital mortality, controlling for episode complexity (through a severity index and acute comorbidities), demographic characteristics and hospital identifications. The association between the probability of a prompt surgery (within one day) and the day of admission was investigated to explore the possible delay of care delivery for patients admitted during weekends. Results Our results indicate that in-hospital mortality rates were not significantly higher for weekend admissions than for weekday admissions in both ST-elevation myocardial infarction (STEMI) and non-STEMI episodes. This result is robust to the inclusion of a number of potential confounding mechanisms. Patients admitted on weekends had lower probabilities of undergoing invasive cardiac surgery within the day after admission, but delay in care delivery during the weekend was not associated with worse outcomes in terms of in-hospital mortality. Conclusions There is no evidence for the existence of a weekend effect due to admission for acute myocardial infarction in Portugal, in both STEMI and non-STEMI episodes.

Differences between determinants of in-hospital mortality and hospitalisation costs for patients with acute heart failure: a nationwide observational study from Japan

OBJECTIVES

Although current case-mix classifications in prospective payment systems were developed to estimate patient resource usage, whether these classifications reflect clinical outcomes remains unknown. The efficient management of acute heart failure (AHF) with high mortality is becoming more important in many countries as its prevalence and associated costs are rapidly increasing. Here, we investigate the determinants of in-hospital mortality and hospitalisation costs to clarify the impact of severity factors on these outcomes in patients with AHF, and examine the level of agreement between the predicted values of mortality and costs.

DESIGN

Cross-sectional observational study.

SETTING AND PARTICIPANTS

A total of 19 926 patients with AHF from 261 acute care hospitals in Japan were analysed using administrative claims data.

MAIN OUTCOME MEASURES

Multivariable logistic regression analysis and linear regression analysis were performed to examine the determinants of in-hospital mortality and hospitalisation costs, respectively. The independent variables were grouped into patient condition on admission, postadmission procedures indicating disease severity (eg, intra-aortic balloon pumping) and other high-cost procedures (eg, single-photon emission CT). These groups of independent variables were cumulatively added to the models, and their effects on the models' abilities to predict the respective outcomes were examined. The level of agreement between the quartiles of predicted mortality and predicted costs was analysed using Cohen's κ coefficient.

RESULTS

In-hospital mortality was associated with patient's condition on admission and severity-indicating procedures (C-statistics 0.870), whereas hospitalisation costs were associated with severity-indicating procedures and high-cost procedures (R² 0.32). There were substantial differences in determinants between the outcomes. In addition, there was no consistent relationship observed (κ=0.016, p<0.0001) between the quartiles of in-hospital mortality and hospitalisation costs.

CONCLUSIONS

The determinants of mortality and costs for hospitalised patients with AHF were generally different. Our results indicate that the same case-mix classifications should not be used to predict both these outcomes.

Survival Advantage Associated with Treatment of Injury at Designated Trauma Centers

This article analyzes the effectiveness of designated trauma centers in Florida concerning reduction in the mortality risk of severely injured trauma victims. A bivariate probit model is used to compute the differential impact of two alternative acute care treatment sites. The alternative sites are defined as (1) a nontrauma center (NC) or (2) a designated trauma center (DTC). An instrumental-variables method was used to adjust for prehospital selection bias in addition to the influence of age, gender, race, risk of mortality, and type of injury. Treatment at a DTC was associated with a reduction of 0.13 in the probability of mortality.

High Subarachnoid Hemorrhage Patient Volume Associated With Lower Mortality and Better Outcomes

BACKGROUND

High-volume centers have better outcomes than low-volume centers when managing complex conditions including subarachnoid hemorrhage (SAH).

OBJECTIVE

To quantify SAH volume-outcome association and determine the extent to which this association is influenced by aggressiveness of care.

METHODS

A serial cross-sectional retrospective study using the Nationwide Inpatient Sample for 2002 to 2010 was performed. Included were all adult (older than 18 years of age) discharged patients with a primary diagnosis of SAH admitted from the emergency department or transferred to a discharging hospital; cases of trauma or arteriovenous malformation were excluded. Survey-weighted descriptive statistics estimated temporal trends. Multilevel logistic regression estimated volume-outcome associations for inpatient mortality and discharge home. Models were adjusted for demographic characteristics, year, transfer status, insurance status, all individual Charlson comorbidities, intubation, and all patient-refined, diagnosis-related group mortality. Analyses were repeated, excluding cases in which aggressive care was not pursued.

RESULTS

A total of 32,336 discharges were included; 13,398 patients underwent clipping (59.1%) or coiling (40.9%). The inpatient mortality rate decreased from 32.2% in 2002 to 22.2% in 2010; discharge home increased from 28.5% to 40.8% during the same period. As SAH volume decreased from 100/year, the mortality rate increased from 18.7% to 19.8% at 80/year, 21.7% at 60/year, 24.5% at 40/year, and 28.4% at 20/year. As SAH patient volume decreased, the probability of discharge home decreased from 40.3% at 100/year to 38.7% at 60/year, and 35.3% at 20/year. Better outcomes persisted in patients receiving aggressive care and in those not receiving aggressive care.

CONCLUSION

Short-term SAH outcomes have improved. High-volume hospitals have more favorable outcomes than low-volume hospitals. This effect is substantial, even for hospitals conventionally classified as high volume.

An analysis of the effectiveness of a state trauma system: treatment at designated trauma centers is associated with an increased probability of survival

BACKGROUND

States struggle to continue support for recruitment, funding and development of designated trauma centers (DTCs). The purpose of this study was to evaluate the probability of survival for injured patients treated at DTCs versus nontrauma centers.

METHODS

We reviewed 188,348 patients from the state's hospital discharge database and identified 13,953 severely injured patients admitted to either a DTC or a nontrauma center between 2008 and 2012. DRG International Classification of Diseases-9th Rev. Injury Severity Scores (ICISS), an accepted indicator of injury severity, was assigned to each patient. Severe injury was defined as an ICISS less than 0.85 (indicating ≥15% probability of mortality). Three subgroups of the severely injured patients were defined as most critical, intermediate critical, and least critical. A full information maximum likelihood bivariate probit model was used to determine the differences in the probability of survival for matched cohorts.

RESULTS

After controlling for injury severity, injury type, patient demographics, the presence of comorbidities, as well as insurance type and status, severely injured patients treated at a DTC have a 10% increased probability of survival. The largest improvement was seen in the intermediate subgroup.

CONCLUSION

Treatment of severely injured patients at a DTC is associated with an improved probability of survival. This argues for continued resources in support of DTCs within a defined statewide network.

LEVEL OF EVIDENCE

Epidemiologic study, level III.

Variation in Neonatal Inpatient Charges at the State and Local Level

In an analysis of all Ohio newborn infants discharged home alive between 2007 and 2012, the authors identified that significant variation in hospital charges (among Medicare Severity Diagnostic Related Group categorizations), previously identified nationally, persists at the state and local levels among term and preterm infants (p <.0001). Additionally, the authors identified variation in length of stay among infants with extreme immaturity or respiratory distress syndrome (p <.0001). Charge data remain the best available proxy for closely guarded hospital cost figures; increased pricing transparency would further support comparison of hospital newborn care costs.

Inpatient hospital outcomes following injury in Suriname: lessons for prevention

Traumatic injury is an important and indiscriminant contributor to mortality. Hypothesizing that outcomes from severe injuries do not vary by demographic factors or socioeconomic status, this research analyzed the relationship between race, ethnicity, injury characteristics, and fatality following hospitalization in Suriname. Data were obtained for all hospital episodes in 2008 from the only hospital within the greater Paramaribo area that provides emergency department services. A logistic regression was used to analyze the subset of 544 non-elderly adult trauma victims to assess the contribution of patient demographics and anatomic injury severity to outcome, which was defined as mortality during acute hospitalization. The specific demographics included were patient age, gender, race, and insurance status. Injury severity was measured using the International Classification Injury Severity Score. The results indicate that age, insurance status, injury type, and injury severity were significant predictors for survival. While the uninsured experienced a higher rate of mortality, the model suggests this result is not due to physiologic reasons but behavioral and socioeconomic. The higher mortality is driven by greater injury severity, which increases not only the mortality rate but also the cost of care. Injury severity itself, independent of all other factors, is the most important contributor. The results suggest that a reduction of 10% in injury severity, around the mean, would reduce the probability of mortality by 70%. This suggests that targeting risk-taking behavior, perhaps relating to compliance with safety practices (e.g. seat belt and helmet laws), driver education, and road safety measures can play important roles in reducing mortality and morbidity from injury in Suriname.

Accuracy of the all patient refined diagnosis related groups classification system in congenital heart surgery

BACKGROUND

Administrative data are increasingly used to evaluate clinical outcomes and quality of care in pediatric congenital heart surgery (CHS) programs. Several published analyses of large pediatric administrative data sets have relied on the All Patient Refined Diagnosis Related Groups (APR-DRG, version 24) diagnostic classification system. The accuracy of this classification system for patients undergoing CHS is unclear.

METHODS

We performed a retrospective cohort study of all 14,098 patients 0 to 5 years of age undergoing any of six selected congenital heart operations, ranging in complexity from isolated closure of a ventricular septal defect to single-ventricle palliation, at 40 tertiary-care pediatric centers in the Pediatric Health Information Systems database between 2007 and 2010. Assigned APR-DRGs (cardiac versus noncardiac) were compared using χ2 or Fisher's exact tests between those patients admitted during the first day of life versus later and between those receiving extracorporeal membrane oxygenation support versus those not. Recursive partitioning was used to assess the greatest determinants of APR-DRG type in the model.

RESULTS

Every patient admitted on day 1 of life was assigned to a noncardiac APR-DRG (p<0.001 for each procedure). Similarly, use of extracorporeal membrane oxygenation was highly associated with misclassification of CHS patients into a noncardiac APR-DRG (p<0.001 for each procedure). Cases misclassified into a noncardiac APR-DRG experienced a significantly increased mortality (p<0.001).

CONCLUSIONS

In classifying patients undergoing CHS, APR-DRG coding has systematic misclassifications, which may result in inaccurate reporting of CHS case volumes and mortality.

Trauma center performance indicators for nonfatal outcomes: a scoping review of the literature

BACKGROUND

According to Donabedian's framework, outcomes covering the following six domains should be used to evaluate health care quality: death, adverse events, readmissions to hospital, resource use, quality of life, and ability to function in daily activities. The objective of this study was to identify the nonfatal outcomes that have been used to evaluate the performance of trauma hospitals. Secondary objectives were to describe definitions and methodological quality.

METHODS

We performed a scoping literature review of studies using at least one nonfatal outcome to evaluate the performance of acute care hospitals for the treatment of general trauma populations. We searched MEDLINE, EMBASE, Cochrane central, CINAHL, BIOSIS, TRIP and ProQuest databases. Methodological quality was evaluated using elements of the STROBE statement and the Downs and Black tool.

RESULTS

Of 14,521 citations, 40 were eligible for inclusion. We identified 14 nonfatal outcomes as follows: (i) adverse events including complications (used in 35 evaluations), missed injuries (n = 4), reintubation (n = 2), unplanned intensive care unit admissions (n = 2), and unplanned surgeries (n = 4); (ii) resource use including hospital (n = 19), intensive care unit (n = 15), and ventilator (n = 4) length of stay, inappropriate hospital stay (n = 1), and potentially unnecessary care (n = 1); (iii) hospital readmissions (n = 4); and (iv) ability to function in daily activities including functional capacity (n = 2), and discharge destination (n = 3). No measures of quality of life were identified. There was high heterogeneity in the definitions used. Only 18% of studies had high methodological quality.

CONCLUSION

Among recommended domains of nonfatal outcomes, adverse events and resource use were frequently used to evaluate trauma care, readmissions and function in daily activities were rarely used, and quality of life was never used. In addition, definitions of nonfatal outcomes were variable, and methodological quality was low. There is a need to develop valid and reliable performance indicators based on each domain of Donabedian's framework to evaluate trauma care.

Mental health disorders influence admission rates for pain in children with sickle cell disease

BACKGROUND

Patients with sickle cell disease (SCD) experience a broad range of mental health disorders placing them at risk for more complicated hospitalizations for pain. The current study examined the impact of mental health disorders on admission rates and hospital length of stay (LOS) for vaso-occlusive pain events (VOE) in pediatric patients with SCD.

PROCEDURE

Patients (5-18 years old) with a primary discharge diagnosis of SCD with crisis were acquired through the Pediatric Health Information System and categorized by history of mental health disorders (mood disorder, anxiety disorder, disruptive behavior disorder, and substance use disorder). Using a retrospective cohort design, hospital admission rates for VOE were examined as the primary outcome and LOS as a secondary outcome.

RESULTS

A total of 5,825 patients accounted for 23,561 admissions for SCD with crisis with approximately 8% of the patients having a mental health diagnosis. Longer LOS was found among patients with a history of any mental health diagnosis (P < 0.0001) and within the mood disorder (P < 0.0001), anxiety disorder (P < 0.0001), and substance use disorder (P = 0.01) subtypes. Hospital admissions rates for VOE were higher among patients with a history of any mental health diagnosis (P < 0.0001) and within the mood disorder (P < 0.0001), anxiety disorder (P < 0.0001), disruptive behavior disorder (P = 0.002), and substance use disorder (P < 0.0001) subtypes.

CONCLUSIONS

Pediatric patients with SCD and a history of a mental health diagnosis have longer LOS and higher admission rates for management of VOE. Ultimately, these findings suggest that mental health pose a challenge to the management of sickle cell pain.

Preventable mortality: does the perspective matter when determining preventability?

BACKGROUND

We report a novel approach to mortality review using a 360° survey and a multidisciplinary mortality committee (MMC) to optimize efforts to improve inpatient care.

METHODS

In 2009, a 16-item, 360° compulsory quality improvement survey was implemented for mortality review. Descriptive statistics were performed to compare the responses by provider specialty, profession, and level of training using the Fisher exact and chi-square tests, as appropriate. We compared the agreement between the MMC review and provider-reported classification regarding the preventability of each death using the Cohen kappa coefficient. A qualitative review of 360° information was performed to identify the quality opportunities.

RESULTS

Completed surveys (n = 3095) were submitted for 1683 patients. The possibility of a preventable death was suggested in the 360° survey for 42 patients (1.40%). We identified 502 patients (29.83%) with completed 360° surveys who underwent MMC review. The inter-rater reliability between the provider opinions regarding preventable death and the MMC review was poor (kappa = 0.10, P < 0.001). Of the 42 cases identified by the 360° survey as preventable deaths, 15 underwent MMC review; 3 were classified as preventable and 12 were deemed unavoidable. Qualitative analyses of the 12 discrepancies did reveal quality issues; however, they were not deemed responsible for the patients' death.

CONCLUSIONS

The mortality survey yielded important information regarding inpatient deaths that historically was buried with the patient. Poor agreement between the 360° survey responses and an objective MMC review support the need to have a multipronged approach to evaluating inpatient mortality.

Systematic review of predictive performance of injury severity scoring tools

Many injury severity scoring tools have been developed over the past few decades. These tools include the Injury Severity Score (ISS), New ISS (NISS), Trauma and Injury Severity Score (TRISS) and International Classification of Diseases (ICD)-based Injury Severity Score (ICISS). Although many studies have endeavored to determine the ability of these tools to predict the mortality of injured patients, their results have been inconsistent. We conducted a systematic review to summarize the predictive performances of these tools and explore the heterogeneity among studies. We defined a relevant article as any research article that reported the area under the Receiver Operating Characteristic curve as a measure of predictive performance. We conducted an online search using MEDLINE and Embase. We evaluated the quality of each relevant article using a quality assessment questionnaire consisting of 10 questions. The total number of positive answers was reported as the quality score of the study. Meta-analysis was not performed due to the heterogeneity among studies. We identified 64 relevant articles with 157 AUROCs of the tools. The median number of positive answers to the questionnaire was 5, ranging from 2 to 8. Less than half of the relevant studies reported the version of the Abbreviated Injury Scale (AIS) and/or ICD (37.5%). The heterogeneity among the studies could be observed in a broad distribution of crude mortality rates of study data, ranging from 1% to 38%. The NISS was mostly reported to perform better than the ISS when predicting the mortality of blunt trauma patients. The relative performance of the ICSS against the AIS-based tools was inconclusive because of the scarcity of studies. The performance of the ICISS appeared to be unstable because the performance could be altered by the type of formula and survival risk ratios used. In conclusion, high-quality studies were limited. The NISS might perform better in the mortality prediction of blunt injuries than the ISS. Additional studies are required to standardize the derivation of the ICISS and determine the relative performance of the ICISS against the AIS-based tools.

Age at hip or knee joint replacement surgery predicts likelihood of revision surgery

We compared revision and mortality rates of 4668 patients undergoing primary total hip and knee replacement between 1989 and 2007 at a University Hospital in New Zealand. The mean age at the time of surgery was 69 years (16 to 100). A total of 1175 patients (25%) had died at follow-up at a mean of ten years post-operatively. The mean age of those who died within ten years of surgery was 74.4 years (29 to 97) at time of surgery. No change in comorbidity score or age of the patients receiving joint replacement was noted during the study period. No association of revision or death could be proven with higher comorbidity scoring, grade of surgeon, or patient gender. We found that patients younger than 50 years at the time of surgery have a greater chance of requiring a revision than of dying, those around 58 years of age have a 50:50 chance of needing a revision, and in those older than 62 years the prosthesis will normally outlast the patient. Patients over 77 years old have a greater than 90% chance of dying than requiring a revision whereas those around 47 years are on average twice as likely to require a revision than die. This information can be used to rationalise the need for long-term surveillance and during the informed consent process.

Survival advantage for elderly trauma patients treated in a designated trauma center

BACKGROUND

This article analyzes the effectiveness of designated trauma centers (DTCs) in Florida concerning reduction in the mortality risk of severely injured elderly trauma victims.

METHODS

Inpatient hospital data collected by the Agency for Health Care Administration were used to identify elderly trauma patients. An instrumental variables method was used to adjust for prehospital selection bias in addition to the influence of age, gender, race, risk of mortality, comorbidities, and type of injury. The model was estimated using a bivariate probit full information maximum likelihood model to determine the impact of triage to a trauma center as opposed to a nontrauma hospital.

RESULTS

After adjusting for confounding influences, treatment at a DTC was associated with a statistically significant reduction of 0.072, 0.040, and 0.036 in the probability of mortality for patients in the age groups 65 years to 74 years, 75 years to 84 years, and ≥ 85 years, respectively.

CONCLUSIONS

Treatment of severely injured elderly trauma patients in DTCs is associated with statistically significant gains in the probability of survival.

Are all trauma centers created equally? A statewide analysis

OBJECTIVES

Prior work has shown differences in mortality at different levels of trauma centers (TCs). There are limited data comparing mortality of equivalently verified TCs. This study sought to assess the potential differences in mortality as well as discharge destination (discharge to home vs. to a rehabilitation center or skilled nursing facility) across Level I TCs in the state of Ohio.

METHODS

This was a retrospective, multicenter, statewide analysis of a state trauma registry of American College of Surgeons (ACS)-verified Level I TCs from 2003 to 2006. All adult (>15 years) patients transferred from the scene to one of the 10 Level I TCs throughout the state were included (n = 16,849). Multivariable logistic regression models were developed to assess for differences in mortality, keeping each TC as a fixed-effect term and adjusting for patient demographics, injury severity, mechanism of injury, and emergency medical services and emergency department procedures. Outcomes included in-hospital mortality and discharge destination (home vs. rehabilitation center or skilled nursing facility). Adjusted odds ratios (ORs) for each TC were also calculated.

RESULTS

Considerable variability existed in unadjusted mortality between the centers, from 3.8% (95% confidence interval [CI] = 3.7% to 3.9%) to 24.2% (95% CI = 24.1% to 24.3%), despite similar patient characteristics and injury severity. Adjusted mortality had similar variability as well, ranging from an OR of 0.93 (95% CI = 0.47 to 1.84) to an OR of 6.02 (95% CI= 3.70 to 9.79). Similar results were seen with the secondary outcomes (discharge destination).

CONCLUSIONS

There is considerable variability in the mortality of injured patients at Level I TCs in the state of Ohio. The patient differences or care processes responsible for this variation should be explored.

Revisiting the validity of APACHE II in the trauma ICU: improved risk stratification in critically injured adults

BACKGROUND

Quality and benchmarking initiatives highlight the need for accurate stratified risk adjustment. The stratification of trauma patients has relied on scores specific to trauma populations. While the Acute Physiologic and Chronic Health Evaluation (APACHE) II score has been considered "invalid" in the trauma population, we hypothesized that APAHCE II would more accurately predict outcomes in critically injured patients in whom commonly used trauma scores have inherent limitations.

METHODS

A prospective cohort of critically injured patients was enrolled. Severity scores and their sub-components were collected, and in-hospital mortality was assessed. The area under the receiver operating characteristic (AUROC) curve was used to determine the predictive value of each score. Logistic regression estimated the odds of death associated with incremental changes in severity scores and their subcomponents.

RESULTS

1019 patients were available for analysis. APACHE II was the best predictor of mortality (AUROC 0.77 versus AUROC 0.54 for ISS and 0.64 for TRISS). A unit increase in APACHE II was associated with an OR of death of 1.18 (95% CI 1.14-1.22). The components of APACHE II that contributed the most to its accuracy included temperature, serum creatinine and the Glasgow Coma Scale (GCS).

CONCLUSION

Critically injured patients have physiologic derangements not accurately accounted for by commonly used trauma scores. In this subset a more general ICU scoring system is useful for risk adjustment for research, administrative and quality improvement purposes.

The impact of injury coding schemes on predicting hospital mortality after pediatric injury

OBJECTIVES

Accurate adjustment for injury severity is needed to evaluate the effectiveness of trauma management. While the choice of injury coding scheme used for modeling affects performance, the impact of combining coding schemes on performance has not been evaluated. The purpose of this study was to use Bayesian logistic regression to develop models predicting hospital mortality in injured children and to compare the performance of models developed using different injury coding schemes.

METHODS

Records of children (age < 15 years) admitted after injury were obtained from the National Trauma Data Bank (NTDB) and the National Pediatric Trauma Registry (NPTR) and used to train Bayesian logistic regression models predicting mortality using three injury coding schemes (International Classification of Disease-9th revision [ICD-9] injury codes, the Abbreviated Injury Scale [AIS] severity scores, and the Barell matrix) and their combinations. Model performance was evaluated using independent data from the NTDB and the Kids' Inpatient Database 2003 (KID).

RESULTS

Discrimination was optimal when modeling both ICD-9 and AIS severity codes (area under the receiver operating curve [AUC] = 0.921 [NTDB] and 0.967 [KID], Hosmer-Lemeshow [HL] h-statistic = 115 [NTDB] and 147 [KID]), while calibration was optimal when modeling coding based on the Barell matrix (AUC = 0.882 [NTDB] and 0.936 [KID], HL h-statistic = 19 [NTDB] and 69 [KID]). When compared to models based on ICD-9 codes alone, models that also included AIS severity scores and coding from the Barell matrix showed improved discrimination and calibration.

CONCLUSIONS

Mortality models that incorporate additional injury coding schemes perform better than those based on ICD-9 codes alone in the setting of pediatric trauma. Combining injury coding schemes may be an effective approach for improving the predictive performance of empirically derived estimates of injury mortality.

Surgical quality improvement: a simplified method to apply national standards to pediatric trauma care

BACKGROUND

The emerging "pay for performance" national initiative mandates the development of valid metrics for risk stratification and performance assessment. The International Classification Injury Severity Score (ICISS) predicts survival from injury and is calculated as the product of survival risk ratios (SRRs) for a patient's 3 worst injuries. Survival risk ratios are derived as the proportion of fatalities for every International Classification of Diseases, Ninth Edition, Clinical Modification, diagnosis in a "benchmark" population. We hypothesized that the ICISS prediction model derived from the National Pediatric Trauma Registry (NPTR) would accurately predict mortality in an independent sample from a single pediatric trauma center (PTC) and could be applied to the NSQIP methodology to analyze performance.

METHODS

The ICISS survival probabilities (Ps) were calculated for PTC patients using SRRs computed from 102,608 NPTR records. Records with a single diagnosis and Ps of 1 were excluded from the analysis. Receiver operator characteristics analysis (ROC) was used to evaluate the accuracy of Ps to predict mortality. The Hosmer-Lemeshow statistic was used to determine the degree that the NPTR-derived expected probabilities matched the observed mortality profile at the PTC. Program performance from 2000 to 2004 was then evaluated using Ps adjusted by logit transformation to predict expected mortality (E) for each year cohort. Observed mortality divided by expected mortality (O/E) was calculated for each year group to compare PTC performance to the NSQIP standard of one. The influence of injury severity on these results was determined by evaluating the correlation between O/E and mean Ps of each year cohort.

RESULTS

A total of 1523 records were analyzed. The ROC area under the curve (AUC ) for Ps was .947 (confidence interval, .934-.957). The Hosmer-Lemeshow statistic (chi(2) = 5.102; df = 8; P = .747, not significant) indicated the model fit the data well. Adjusted O/E ratio after logit transformation of Ps for the PTC demonstrated initial performance slightly below standard (1.000778) followed by performance better than expected for the subsequent 4 years (range, .6466-.9784). The ratio of observed (O) to expected (E) demonstrated no correlation to mean Ps (r(2) = .378; P = .208).

CONCLUSION

These data validate the application of injury diagnosis derived survival probabilities as objective metrics for determining performance using the NSQIP methodology. Incorporation of these objective predictors of expected outcome to calculation of the risk adjusted O/E ratio enables trend analysis of program performance over time. The lack of significant correlation between O/E and mean Ps demonstrates that NSQIP does indeed reflect process of care while adjusting for severity of patient pathologic condition.

How well can hospital readmission be predicted in a cohort of hospitalized children? A retrospective, multicenter study

BACKGROUND

Children with complex chronic conditions depend on both their families and systems of pediatric health care, social services, and financing. Investigations into the workings of this ecology of care would be advanced by more accurate methods of population-level predictions of the likelihood for future hospitalization.

METHODS

This was a retrospective cohort study. Hospital administrative data were collected from 38 children's hospitals in the United States for the years 2003-2005. Participants included patients between 2 and 18 years of age discharged from an index hospitalization during 2004. Patient characteristics documented during the index hospitalization or any previous hospitalization during the preceding 365 days were included. The main outcome measure was readmission to the hospital during the 365 days after discharge from the index admission.

RESULTS

Among the cohort composed of 186856 patients discharged from the participating hospitals during 2004, the mean age was 9.2 years, with 54.4% male and 52.9% identified as non-Hispanic white. A total of 17.4% were admitted during the previous 365 days, and among those discharged alive (0.6% died during the admission), 16.7% were readmitted during the ensuing 365 days. The final readmission model exhibited a c statistic of 0.81 across all hospitals, with a range from 0.76 to 0.84 for each hospital. Bootstrap-based assessments demonstrated the stability of the final model.

CONCLUSIONS

Accurate population-level prediction of hospital readmissions is possible, and the resulting predicted probability of hospital readmission may prove useful for health services research and planning.

Development of a health care utilisation data-based index for rheumatoid arthritis severity: a preliminary study

Introduction

Health care utilisation ('claims') databases contain information about millions of patients and are an important source of information for a variety of study types. However, they typically do not contain information about disease severity. The goal of the present study was to develop a health care claims index for rheumatoid arthritis (RA) severity using a previously developed medical records-based index for RA severity (RA medical records-based index of severity [RARBIS]).

Methods

The study population consisted of 120 patients from the Veteran's Administration (VA) Health System. We previously demonstrated the construct validity of the RARBIS and established its convergent validity with the Disease Activity Score (DAS28). Potential claims-based indicators were entered into a linear regression model as independent variables and the RARBIS as the dependent variable. The claims-based index for RA severity (CIRAS) was created using the coefficients from models with the highest coefficient of determination (R²) values selected by automated modelling procedures. To compare our claims-based index with our medical records-based index, we examined the correlation between the CIRAS and the RARBIS using Spearman non-parametric tests.

Results

The forward selection models yielded the highest model R² for both the RARBIS with medications (R² = 0.31) and the RARBIS without medications (R² = 0.26). Components of the CIRAS included tests for inflammatory markers, number of chemistry panels and platelet counts ordered, rheumatoid factor, the number of rehabilitation and rheumatology visits, and Felty's syndrome diagnosis. The CIRAS demonstrated moderate correlations with the RARBIS with medication and the RARBIS without medication sub-scales.

Conclusion

We developed the CIRAS that showed moderate correlations with a previously validated records-based index of severity. The CIRAS may serve as a potentially important tool in adjusting for RA severity in pharmacoepidemiology studies of RA treatment and complications using health care utilisation data.

Do pediatric patients with trauma in Florida have reduced mortality rates when treated in designated trauma centers?

OBJECTIVE

The purposes of the study were to compare the survival associated with treatment of seriously injured patients with pediatric trauma in Florida at designated trauma centers (DTCs) with nontrauma center (NCs) acute care hospitals and to evaluate differences in mortality between designated pediatric and nonpediatric trauma centers.

METHODS

Trauma-related inpatient hospital discharge records from 1995 to 2004 were analyzed for children aged from 0 to 19 years. Age, sex, ethnicity, injury mechanism, discharge diagnoses, and severity as defined by the International Classification Injury Severity Score were analyzed, using mortality during hospitalization as the outcome measure. Children with central nervous system, spine, torso, and vascular injuries and burns were evaluated. Instrumental variable analysis was used to control for triage bias, and mortality was compared by probabilistic regression and bivariate probit modeling. Children treated at a DTC were compared with those treated at a nontrauma center. Within the population treated at a DTC, those treated at a DTC with pediatric capability were compared with those treated at a DTC without additional pediatric capability. Models were analyzed for children aged 0 to 19 years and 0 to 15 years.

RESULTS

For the 27,313 patients between ages 0 and 19 years, treatment in DTCs was associated with a 3.15% reduction in the probability of mortality (P < .0001, bivariate probit). The survival advantage for children aged 0 to 15 years was 1.6%, which is not statistically significant. Treatment of 16,607 children in a designated pediatric DTC, as opposed to a nonpediatric DTC, was associated with an additional 4.84% reduction in mortality in the 0- to 19-year age group and 4.5% in the 0 to 15 years group (P < .001, bivariate probit).

CONCLUSIONS

Optimal care of the seriously injured child requires both the extensive and immediate resources of a DTC as well as pediatric-specific specialty support.

DRG 504: The Effect of 96 Hours of Mechanical Ventilation on Resource Utilization

DRG (diagnosis related group) 504 is utilized for patients with extensive burn injuries with skin grafts along with the recent inclusion of patients with nonextensive full-thickness burns with skin grafts who require >or=4 days of mechanical ventilation. Since patients with extensive burns and/or inhalation injuries often required ventilator support, we elected to compare demographics, length of stay variables, and hospital charges for patients assigned to DRG 504 based upon the length of ventilator support. The American Burn Association's National Burn Repository was queried for all inpatients admitted from January 2000 through December 2001 and who were assigned to DRG 504. Demographic, resource utilization, and financial data were analyzed based upon the need for >or=96 hrs of mechanical ventilation. One hundred seven patients were identified of which 94 (87.9%) required >or=96 hrs of mechanical ventilation. While patients with inhalation injuries required significantly more days of ventilator support, length of stay and hospital charges were nearly identical. Patients who required >or=96 hrs of ventilator support, had a 10-fold greater number of ventilator and intensive care unit days (P < 0.0001) and twice the length of hospitalization (P < 0.005) and hospital charges (P < 0.05) for their care compared to those requiring <96 hrs of ventilator support. Burn patients requiring endotracheal intubation and >or=96 hrs of ventilator support during their acute hospitalization consume significantly greater resources than those who do not require such treatment. The Center for Medicare & Medicaid Services should consider modifying DRG 504 for patients with extensive burns to permit a more appropriate resource-based reimbursement to burn center hospitals.

Canadian benchmarks in trauma

BACKGROUND

Benchmarks are used in trauma care for program evaluation, quality improvement, and research. National outcome benchmarks relevant to the Canadian trauma population need to be defined for evaluation of trauma care in Canada. The purpose of this study was to derive survival probabilities associated with trauma diagnoses using International Classification of Diseases, Ninth Revision (ICD-9) codes.

METHODS

All patients admitted to an acute care hospital with nonpenetrating trauma and submitted to the National Trauma Registry of Canada between 1994 through 2000 inclusively were included in analyses. Both inclusive and exclusive survival risk ratios (SRRs) were calculated for groups of ICD-9 injury codes between 800 to 959.

RESULTS

For the study period, there were 1,003,905 and 803,776 eligible trauma patients used to calculate inclusive SRRs and exclusive SRRs, respectively. Survival probabilities for injuries are given according to ICD-9 codes.

CONCLUSION

This is the first study to define national survival benchmarks for the Canadian trauma population. These results can be used to assess survival of patients using the ICISS [(ICD-9) based Injury Severity Score (ISS)] methodology. With regular updates, these data can further be developed for continual trauma outcome assessment, quality improvement, and research into trauma care in Canada.

Feasibility and validity of International Classification of Diseases based case mix indices

Background

Severity of illness is an omnipresent confounder in health services research. Resource consumption can be applied as a proxy of severity. The most commonly cited hospital resource consumption measure is the case mix index (CMI) and the best-known illustration of the CMI is the Diagnosis Related Group (DRG) CMI used by Medicare in the U.S. For countries that do not have DRG type CMIs, the adjustment for severity has been troublesome for either reimbursement or research purposes. The research objective of this study is to ascertain the construct validity of CMIs derived from International Classification of Diseases (ICD) in comparison with DRG CMI.

Methods

The study population included 551 acute care hospitals in Taiwan and 2,462,006 inpatient reimbursement claims. The 18^th version of GROUPER, the Medicare DRG classification software, was applied to Taiwan's 1998 National Health Insurance (NHI) inpatient claim data to derive the Medicare DRG CMI. The same weighting principles were then applied to determine the ICD principal diagnoses and procedures based costliness and length of stay (LOS) CMIs. Further analyses were conducted based on stratifications according to teaching status, accreditation levels, and ownership categories.

Results

The best ICD-based substitute for the DRG costliness CMI (DRGCMI) is the ICD principal diagnosis costliness CMI (ICDCMI-DC) in general and in most categories with Spearman's correlation coefficients ranging from 0.938-0.462. The highest correlation appeared in the non-profit sector. ICD procedure costliness CMI (ICDCMI-PC) outperformed ICDCMI-DC only at the medical center level, which consists of tertiary care hospitals and is more procedure intensive.

Conclusion

The results of our study indicate that an ICD-based CMI can quite fairly approximate the DRGCMI, especially ICDCMI-DC. Therefore, substituting ICDs for DRGs in computing the CMI ought to be feasible and valid in countries that have not implemented DRGs.

Analysis of trends in the Florida Trauma System (1991-2003): changes in mortality after establishment of new centers

BACKGROUND

This study analyzes trends in hospitalization and outcome for adult, elderly, and pediatric trauma victims in the Florida Trauma System (FTS) from 1991 to 2003, during which time the number of centers nearly doubled from 11 to 20.

METHODS

Administrative data was queried for all admissions with at least one trauma related discharge. Patients were stratified by age as pediatric (age, 0 to 15 years), adult (age, 16 to 64 years), or elderly (age, >64 years). Volume of admissions, severity, and mortality were analyzed over time. A logistic regression model was used to test the existence of an organizational experience curve after the designation of a new trauma center.

RESULTS

Injury-related hospitalizations increased for the elderly, stayed the same for adults, and declined for children. As the system matured, a larger percentage of victims, particularly the most severely injured, were triaged to trauma centers, indicating more effective triage. In contrast to adults and pediatric patients, the majority of elderly trauma victims were managed at non-trauma centers. The trauma mortality rate per 1,000 population among the elderly increased during the study period (P < .01). Multivariate analysis indicated that for adult and pediatric victims it took up to 3 years after the designation of trauma center status before the odds of mortality reached parity with that of established centers.

CONCLUSIONS

The FTS has grown with its population and has matured to treat a larger percentage of trauma victims. Trauma victims transported to established trauma centers (4+ years) have a survival advantage compared to their counterparts transported to newly created centers. The reduction in the odds of mortality does not occur immediately after trauma center designation.

Evaluation of a mature trauma system

INTRODUCTION

An effective trauma system should offer patients triaged to a trauma center (TC) a survival advantage and cost-effective treatment. Three questions were asked: 1) Does treatment at a TC versus a nontrauma center (NTC) improve survival? 2) Is the system cost-effective? 3) Is access to the system equitable?

METHODS

The 2003 Florida discharge database identified patients with ICD9 codes 800 to 959. Survival risk ratios (SRR) were calculated using 1999-2000 data and ICISS were produced for each code. Using 2003 data, mortality rates were calculated for matched patients at TCs and NTCs. Instrumental variables methodology was used to account for differences in mortality risks of patients triaged to TCs versus NTCs. Logistic regression analysis was used to determine differences in mortality. Charge/cost ratios were analyzed to compute the cost care and cost/life saved. Accessibility to a TC within 85 minutes of injury was assessed.

RESULTS

Treatment at a TC was associated with an 18% reduction in mortality. Mean costs of care in TCs and NTCs were $11,910 and $6019, respectively. Dividing the mean cost difference by the reduction in mortality yields a cost of $34,887/life saved. A total of 42% of patients returned to work within 24 months of injury. Using an expected median of 19 years of employment for a 33-year-old individual and proposed state funding figures for the trauma system, a life saved results in an approximate annual cost to the state of between $100 and $500. Currently, 95% of citizens of the state have access to the trauma system within 85 minutes of injury; however, only 38% of trauma patients are triaged to a TC. Addition of 3 TCs would increase these percentages to 99% and 65%.

CONCLUSIONS

Triage to a Florida TC is associated with a decreased risk of death. Moreover, cost/life year saved is favorable when compared with societal expenditures for other health problems. Improved deployment of TCs is necessary to optimize access. This assessment methodology is a useful model for evaluation of mature trauma systems.

A Systematic Review and Meta-Analysis Comparing Outcome of Severely Injured Patients Treated in Trauma Centers Following the Establishment of Trauma Systems

BACKGROUND

The establishment of trauma systems was anticipated to improve overall survival for the severely injured patient. We systematically reviewed the published literature to assess if outcome from severe traumatic injury is improved for patients following the establishment of a trauma system.

METHODS

A systematic literature review of all population-based studies that evaluated trauma system performance was conducted. A qualitative analysis of each study's design and methodology and a meta-analysis was performed to evaluate the evidence to date of trauma system effectiveness.

RESULTS

A search of the literature yielded 14 published articles. Trauma systems demonstrated improved odds of survival in 8 of the 14 reports. The overall quality-weighted odds ratio was 0.85 lower mortality following trauma system implementation.

CONCLUSIONS

The results of the meta-analysis showed a 15% reduction in mortality in favor of the presence of a trauma system. Evaluation of trauma system effectiveness must remain an uncompromising commitment to optimal outcome for the injured patient.

Clinical and economic outcomes in thrombolytic treatment of peripheral arterial occlusive disease and deep venous thrombosis

PURPOSE

Over the past 2 decades the use of thrombolytic therapy in the management of peripheral occlusive diseases, most notably peripheral arterial occlusion (PAO) and deep venous thrombosis (DVT), has become an accepted and potentially preferable alternative to surgery. We examined the period when urokinase was in short supply and subsequently unavailable, to explore potential differences in clinical outcome and economic effect between urokinase and recombinant tissue plasminogen activator (rt-PA).

MATERIAL AND METHODS

Data were obtained from the Premier Perspective Database, a broad clinical database that contains information on inpatient medical practices and resource use. The study population included all patients hospitalized in 1999 and 2000 with a primary or secondary diagnosis of PAO or DVT. Incidence was calculated for common adverse events, including bleeding complications, intracranial hemorrhage, amputation, and death. Cost data were also abstracted from the database, and are expressed as mean +/- SD.

RESULTS

Demographic variables were similar in the urokinase and rt-PA groups. The rate of bleeding complications was similar in the urokinase and rt-PA groups. There were no intracranial hemorrhages in the urokinase group, compared with a rate of 1.5% in the rt-PA PAO group (P = .087) and 1.9% in the rt-PA DVT group (P = .175). The in-hospital mortality rate was lower in the urokinase-treated PAO subgroup (3.6% vs 8.5%; P = .026), but a similar finding in the DVT subgroup did not achieve statistical significance (4% vs 9.8%; P = .069). While pharmacy costs were greater in the urokinase-treated group (US 5472 dollars +/- US 5579 dollars vs US 3644 dollars +/- US 6009 dollars, P < .001; PAO subgroup, US 11,070 dollars +/- US 15,409 dollars vs US 6150 dollars +/- US 12,398 dollars, P = .003), overall hospital costs did not differ significantly between the 2 groups. This finding appears to be explained by a shorter hospital stay and reduced room and board costs in the urokinase-treated group.

CONCLUSION

There were significant differences in outcome in patients with PAO and DVT who received treatment with urokinase and rt-PA. While pharmacy costs were significantly greater when urokinase was used, reduction in length of stay accounted for similar total hospital costs compared with rt-PA. These findings must be considered in the context of the retrospective nature of the analysis and the potential to use dosing regimens that differ from those in this study.

Hospital Factors Associated with Splenectomy for Splenic Injury: A National Perspective

BACKGROUND

The management of patients with splenic injury has shifted from routine splenectomy to attempts at splenic salvage. Using the Healthcare Cost and Utilization Project's National Inpatient Sample (HCUP-NIS), we assessed the patterns of care for splenic trauma. We hypothesized that the processes of care in urban and rural hospitals would differ.

METHODS

Generalized estimating equations were used to identify predictor variables associated with laparotomy and splenectomy from a national, population-based sample of inpatients (HCUP-NIS). Fourteen thousand nine hundred one patients with an International Classification of Diseases, Ninth Revision, Clinical Modification discharge diagnosis code of 865 were selected from the 1998 to 2000 HCUP-NIS data. Exclusion criteria included age greater than 80 years. Analyses were compared using all patients and excluding patients who died during the first 2 hospital days.

RESULTS

Eight thousand five hundred fifty-three patients were treated in urban teaching hospitals. Forty percent underwent a laparotomy and 28% underwent a splenectomy at that time. Another 4,461 patients were cared for in urban nonteaching hospitals. Of these, 46% had a laparotomy and 35% underwent a splenectomy. The remaining 1,887 patients were seen in rural hospitals. Forty-six percent had a laparotomy and 36% had a splenectomy. Patients in urban teaching hospitals had lower risk-adjusted odds of splenectomy in multivariate models controlling for confounders including overall injury severity. Overall splenic salvage increased from 1998 to 2000, primarily because of increased salvage rates among urban teaching hospitals.

CONCLUSION

The management of patients with splenic injury differs among urban teaching, urban nonteaching, and rural hospitals. Surgeons at urban teaching hospitals appear more willing to attempt splenic salvage by means of nonoperative management.

Tracheostomy in Thermally Injured Patients: Does Diagnosis-Related Group 483 Adequately Estimate Resource Use and Hospital Costs?

BACKGROUND

This study compares burn and nonburn patients undergoing tracheostomy, all of whom were assigned to diagnosis-related group 483 to determine hospital reimbursement.

METHODS

We reviewed the records of all inpatients admitted to our hospital from January 2000 through December 2001 who underwent tracheostomy and who were assigned to diagnosis-related group 483. In addition, we compared our burn patient data with that from three other burn centers and the National Burn Repository.

RESULTS

We identified 357 inpatients who had tracheostomies during their hospitalization, only 12 of whom (3.4%) had acute burn injuries. The mean extent of burn in these patients was 43.4% total body surface area. The most frequent primary diagnoses for nonburn patients were injury and poisoning, and circulatory and respiratory disorders. Patients with burn injuries had 39.6 ventilator days, 40.7 intensive care unit days, and 49.2 hospital days compared with 19.8, 17.4, and 29.5 days, respectively, for nonburn patients (p <0.0001). Demographic, resource, and financial data for burn patients treated at the three other burn centers and those reported to the National Burn Repository were not significantly different from burn patients treated at our hospital. Total costs and charges for the care of burn patients were $186,830 and $343,904, respectively, compared with $82,176 and $160,498 for the nonburn patients (p <0.0005).

CONCLUSION

Burn patients requiring tracheostomies during their acute hospitalization consume significantly more resources than patients without burn injuries. More appropriate resource-based reimbursement for the care of these patients appears warranted.

Risk Adjustment for Injured Patients Using Administrative Data

BACKGROUND

Risk adjustment methods are needed for population-based studies of injured patients.

METHODS

Data were obtained from National Hospital Discharge Surveys, 1996 to 2000. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnoses were used to categorize Abbreviated Injury Scale score, Injury Severity Score, ICD-9-CM Injury Severity Score, injury mechanisms, and comorbidities. Regression models for weighted survey data were constructed from combinations of these classifications, plus age and sex, to predict mortality, length of stay (LOS), or discharge to long-term care (LTC).

RESULTS

Increased Abbreviated Injury Scale score, increased Injury Severity Score, or decreased ICD-9-CM Injury Severity Score were similarly associated with mortality, prolonged LOS, or more frequent LTC, as was increased age. Penetrating or burn mechanisms were associated with mortality and longer LOS; penetrating or vehicle mechanisms were associated with less frequent LTC. Different comorbidities affected LOS and LTC. Men had shorter LOS and less frequent LTC than women.

CONCLUSION

Hospital outcomes after injury are predictable from age, sex, and standard diagnosis groupings. Anatomic scales gave similar results when adjusted for other factors.

The national pediatric trauma registry: a legacy of commitment to control of childhood injury

The National Pediatric Trauma Registry represents almost 15 years of effective collaboration among hospitals committed to improving care for the injured child. Its design of providing a "physiologic snapshot" of the injured child on presentation has supported numerous studies that have helped define the epidemiology of childhood injury and refine principles of management. Global analysis of the 103,434 records included in this database suggest that mortality is significantly higher in the very young, that vehicular injury remains a major pediatric public health challenge, and that shock is just as devastating in the child as the adult. Based on this foundation of collaborative commitment, future versions of a pediatric trauma database must harness the emerging internet technology that combines information accrual with human thought, and must extend this effort to include all the children of our world.

Use of a state-wide administrative database in assessing a regional trauma system: the New York city experience1 1No competing interests declared

BACKGROUND

A successful regional trauma care system should concentrate severely injured patients within trauma centers, and should improve severity-adjusted outcomes. We compared injured patients' outcomes in New York City's level 1 trauma centers and nontrauma centers.

STUDY DESIGN

We analyzed 1998-2000 New York Statewide Planning and Cooperative Research System (SPARCS) data for 103,725 adult discharges from 70 New York City hospitals (15 level I trauma centers), using ICD-9CM codes 800-950. Their 227 DRG's were aggregated into 7 clinical injury classes. A severity index was extracted from each refined DRG, and deaths, age, and gender were analyzed. Regression analysis predicted death from age, gender, severity index, and trauma center discharge, with separate analyses of the three largest clinical classes, and estimated excess mortality because of trauma center discharge.

RESULTS

Level 1 trauma centers discharged 48.2% of injured patients, with higher mean annual discharges per hospital (1,046 discharges per TC vs. 437 per NTC, p < 0.001). Trauma centers discharged more than half the central nervous system, general/gastrointestinal, cardiothoracic, and vascular injuries. Trauma center patients were 12.5 years younger than NTC patients (p < 0.0005), and were disproportionately men (64.7% TC vs. 47.2% NTC, p < 0.0005). For the entire patient cohort, and for central nervous system, orthopaedic and general/gastrointestinal injuries, severity, age, and gender adjusted mortality risk was significantly greater at trauma centers than nontrauma centers.

CONCLUSIONS

New York City's trauma system concentrates injured patients in trauma centers on the basis of injury class rather than severity, but does not produce improved adjusted mortality outcomes for injured patients.

Hospital-reported medical errors in premature neonates

OBJECTIVE

To determine the national rate of hospital-reported medical errors in premature neonates and describe the patient and organizational characteristics associated with their occurrence.

DESIGN

Nonconcurrent, cohort study.

SETTING

The Healthcare Cost and Utilization Project (HCUP) contains discharge data collected at community hospitals sited in >20 states.

PATIENTS

All neonatal discharges from the 1997 edition of HCUP were included in these analyses. The definition of prematurity included any hospitalized neonate with a birth weight <2500 g, which corresponds to approximately 37 wks gestation. Medical error was defined as an International Classification of Diseases-9 discharge diagnosis of 996-999 in any of the diagnosis fields associated with the discharge.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

The national rate of hospital-reported medical errors in premature neonates is 1.2 per 100 discharges. There was a significant linear increase in the rate of medical errors based on birth weight (Cochran-Armitage test for trend, p <.001). After we controlled for case mix and organizational characteristics using a logistic regression model, medical errors continued to be associated with birth weight, gender, insurance status, and hospital characteristics.

CONCLUSIONS

The rate of hospital-reported medical errors in premature neonates is lower than that reported in both the adult and pediatric populations. Specific patient and organizational characteristics are associated with an increased risk of medical errors. These characteristics may help to identify opportunities to improve patient safety efforts in this vulnerable population.

Prediction of Mortality in Pediatric Trauma Patients: New Injury Severity Score Outperforms Injury Severity Score in the Severely Injured

BACKGROUND

The Injury Severity Score (ISS) is a widely accepted method of measuring severity of traumatic injury. A modification has been proposed--the New Injury Severity Score (NISS). This has been shown to predict mortality better in adult trauma patients, but it had no predictive benefit in pediatric patients. The aim of this study was to determine whether the NISS outperforms the ISS in a large pediatric trauma population.

METHODS

Admissions in the National Pediatric Trauma Registry between April 1996 and September 1999 were included. The ISS and NISS were calculated for each patient. The study endpoints were mortality at hospital discharge, functional outcome in three domains (expression, locomotion, and feeding), and discharge disposition for the survivors. Predictive ability of each score was assessed by area under the receiver operating characteristic curve.

RESULTS

The NISS and ISS performed equally well at predicting mortality in patients with lower injury severity (ISS < 25), but the NISS was significantly better at predicting mortality in the more severely injured patients. Both scores performed equally well at predicting expression and feeding ability. The NISS was superior to the ISS in predicting locomotion ability at discharge. Thirty-seven percent of patients had an NISS that was higher than their ISS. These patients had a significantly higher mortality and suffered worse functional outcomes.

CONCLUSION

The NISS performs as well as the ISS in pediatric patients with lower injury severity and outperforms the ISS in those with higher injury severity.

Trends in hospitalization after injury: older women are displacing young men

OBJECTIVE

To evaluate trends in hospitalization after injuries in the USA.

DESIGN

National Hospital Discharge Survey data from 1979 to 2000 were evaluated annually by age group, sex, injury severity score (ISS), length of stay, and discharge destination.

SETTING AND SUBJECTS

National probability sample of hospitalized patients.

INTERVENTIONS

None.

MAIN OUTCOME MEASURES

Incidence, duration, outcome, and population based rates of hospital admission after injuries.

RESULTS

The number of young males admitted to hospitals after injuries has decreased dramatically; older females are now the group most frequently admitted. Total days in the hospital have decreased in all age groups, but have declined less in the older population than in the younger population; furthermore, most patients aged 65 and over were formerly discharged home, but now most are discharged to long term care facilities. Overall hospitalization rates after injury have decreased in all age groups, but have declined less in the older population; furthermore, male and female hospitalization rates for serious injury (ISS at least 9, excluding isolated hip fracture) are decreasing in younger age groups while increasing in older age groups.

CONCLUSIONS

Older patients comprise a growing proportion of injuries requiring hospitalization. Trauma systems must address this change, and preventing injuries in older people is increasingly important.