Measuring transport injuries in a developing country: an application of the capture-recapture method

The purposes of this study are to provide an estimation of the incidence of transport injuries in a defined local community in Nicaragua by using the capture-recapture method, and to compare results using this method when data at different levels of severity are utilized.Two sources of injury data were used to monitor injuries: hospital data (inpatient and outpatient) and traffic police records. Characteristics available for matching included name, age, sex, and date of occurrence. The methodology of capture-recapture was used to estimate the ascertainment degree of both sources of data and the estimate mortality and incidence rates. Estimates were calculated both when all hospital data were taken into account (inpatient and outpatient combined) and when only inpatient records were matched against police records.First, including police records and all hospital data, the mortality and morbidity estimates were 35.5/100000 and 43.7/1000 per year, respectively. Second, when outpatients were excluded from the analysis, the corresponding estimates were 28.6/100000 and 7.5/1000, respectively. In non-fatal cases, the ascertainment-corrected coverage through police records was 2.6% and through hospital surveillance 19.0% when both inpatients and outpatients were included. In fatal cases, the corresponding rates were 56.1 and 22.8%, respectively. The combined data set pointed out pedestrians and cyclists as the main risk groups. Most fatalities were due to head injuries.Our results show that neither police records nor hospital records nor the aggregate database provided acceptable coverage of transport-related injuries. Combining police and hospital data by means of capture-recapture analysis produces more valid estimates, but caution must be given to the issue of severity heterogeneity between the two sources.     

Road accident statistics—A comparison of police and hospital information

Studies in Sweden have raised doubts as to the accuracy of road accident statistics in relation to serious and slight injuries. To explore the reliability of British statistics an analysis has been made of a sample of 1200 patients injured in road accidents and attending hospital. These same cases have been traced in the police records on which official statistics are based. All fatal cases were correctly notified but two types of discrepancy occurred among injuries. In a small number of cases re-classification of “serious” and “slight” seemed to be required. A more important discrepancy was that about one-sixth of serious injuries and one third of slight injuries known to the hospital did not appear in the police notifications. This is not surprising in view of the limited scope of compulsory notification. The police mostly know of accidents by the calling of an ambulance or as a result of allegations of traffic infringements. Thus many cases where an ambulance is not called or in which a driver only is injured and no other vehicle is involved, escape notification. Injuries to pedal cyclists are particularly poorly notified. Less than one quarter of those known to the hospital appeared in the official statistics. It is concluded that similar comparisons of hospital and police information should be made elsewhere to confirm whether this sample is representative of the national rate of notification. In the meantime it is suggested that figures for injuries to pedal cyclists and for slight injuries in general should be used with caution.     

Using multiple datasets to understand trends in serious road traffic casualties

Accurate information on the incidence of serious road traffic casualties is needed to plan and evaluate prevention strategies. Traditionally police reported collisions are the only data used. This study investigate the extent to which understanding of trends in serious road traffic injuries is aided by the use of multiple datasets. Health and police datasets covering all or part of Great Britain from 1996-2003 were analysed. There was a significantly decreasing trend in police reported serious casualties but not in the other datasets. Multiple data sources provide a more complete picture of road traffic casualty trends than any single dataset. Increasing availability of electronic health data with developments in anonymised data linkage should provide a better platform for monitoring trends in serious road traffic casualties.     

Complementing police road-crash records with trauma registry data--an initial evaluation

This paper examines the consistency of hospital and police reporting of outcomes of road traffic crashes using a database of linked police crash reports and trauma registry records. Criteria for inclusion into the trauma registry include trauma-related causes with subsequent stay of more than 24 h or death due to injuries. During the 1997 calendar year there were 497 cases of road-related injuries within the combined trauma registry of Sir Charles Gairdner and Fremantle Hospitals, of which only 82% had matching police records. Linkage rates were associated with gender, injury severity and the number of vehicles involved. Within the road user category, pedestrians were least likely to link. Of the linked records, police classification of injury severity was correct in 78% of cases. Male casualties were more likely to be correctly classified than females, after adjustment for related variables including injury severity. Correct classification of injury by police was also closely related to severity of injury. Identification and targeting of these groups of casualties is vital in refining the road-crash reporting system. Increased crash reporting and availability of data from these two sources will provide road authorities with more reliable measures of injury outcome.     

Total costs of bicycle injuries in Norway: correcting injury figures and indicating data needs

Bicycle injuries and fatalities are reported by the police to Statistics Norway. Fatality records from the police are then corrected with Vital Statistics records. However, there is no complete hospital recording that could provide more correct data for bicycle injuries. Bicycle injuries are underreported in official data. There is a nearly complete omission of single bicycle accidents. This disguises societal accident costs and curtails the identification of black spots and effective infrastructure improvements.

This paper provides an estimate of total bicycle injuries in Norway and the total costs of these injuries. Application of case study hospital data from Norwegian towns enabled an estimation of the relationship between these data and the official data, including the distribution of injuries by severity. Costs were then assessed by applying official monetary values for given levels of injury severity.

Total annual bicycle injury costs are huge, but these costs must be balanced against the benefits of bicycling, related to health and environment. Accident reporting and data should be enhanced to enable a reduction of bicycle injuries.     

Limitations of data compiled from police reports on pediatric pedestrian and bicycle motor vehicle events

Police reports were compared to the information provided by a hospital monitoring system for children under 15 years old injured as pedestrians and bicyclists by moving motor vehicles in Orange County, California. The analysis was limited to identifying caveats in the police report database. Underreporting by police was conservatively estimated at 20% for pedestrians and 10% for bicyclists. Comparison of the pedestrian databases suggested underreporting by police of incidents involving 0-4-year-olds, nontraffic incidents, incidents in which the vehicle was backing up, and cases not involving a child crossing a street. Comparison of the bicyclist databases indicated an underreporting by police of nontraffic cases. These caveats, in part, are related to police agency reporting requirements. The police injury severity scale was found to correlate poorly with a scale based on medical diagnoses, and substantial underreporting by police of serious injuries was demonstrated. We suggest that utilization of police injury severity scales be limited to categories of fatal, injured, and not injured (when available).     

Statistical adjustment for misclassification of seat belt and alcohol use in the analysis of motor vehicle accident data

The use of seat belts and alcohol is often mis-reported in police motor vehicle accident reports for a number of reasons. To avoid penalties, occupants often over report seat belt use and under report alcohol use. Police officers sometimes fail to account for evidence such as presence of belt burn, condition of belts, odor of alcohol, crash patterns, etc. Biased conclusions result when using misclassified accident data to estimate the effectiveness of seat belts in preventing injuries and reducing medical costs. We investigated the effects of misclassification of seat belt and alcohol use on the odds ratio of injury as well as medical costs. A statistical method and a SAS program were developed to adjust odds ratios of injury and medical cost estimates to account for misclassification of seat belts and alcohol use. The method allowed for incorporation of variables that could affect misclassification of seat belt and alcohol use. We conducted a Monte Carlo simulation and found that when there were large differences between the misclassification rates for major and minor injury, the unadjusted odds ratio could have up to a 90% bias while our adjusted odds ratio was effectively unbiased. To illustrate the method, we estimated the misclassification rates of seat belt and alcohol use by comparing merged police and hospital reports from Nebraska motor vehicle accident data sets (1996-1997) and then evaluated the bias of the odds ratio of injury and medical costs estimates due to misclassification. Our results showed that the bias of the odds ratio of injury and medical costs due to misclassification of seat belts and alcohol use depended both on the amount of misclassification and the reported frequencies. Misclassification about seat belt and alcohol use only slightly biased the unadjusted odds ratio estimates and mean hospital charge, while misclassification resulted in approximately a 69% underestimate of the total medical costs savings due to seatbelts. However, due to the small size of the merged Nebraska police and hospital data set used to estimate misclassification rates, these results are likely somewhat imprecise.     

Discrepancies in vehicular crash injury reporting: Northeastern Ohio Trauma Study. IV

People injured in motor vehicle traffic crashes were identified from a population-representative incidence sample of hospital emergency department visits. Matched police reports of crashes were sought in official state records of motor vehicle traffic crashes. Of the emergency department cases, 55% had matched police reports. The frequency of matched reports was highest for drivers (74%), people transported to the hospital by emergency vehicle (69%), and those requiring hospital admission (74%). The frequency was lowest for people younger than 16 years (28%), people injured as occupants of vehicles other than passenger cars (24%), medicaid recipients (33%), and nonresidents of the study region (40%). Motor vehicle traffic injuries are undercounted in police-reported statistics. For many groups, police reporting is less than 50% of the cases identified through emergency departments. The likelihood that a case of motor vehicle traffic injury will have a matched police report depends on demographic, social and crash factors as well as on injury severity.     

In-vehicle distraction and fatal accidents in England and Wales.

The paper describes the coding and analysis of a database of police fatal accident reports to investigate the extent to which in-vehicle distraction is a contributory factor in vehicle crashes. A particular focus has been the involvement of mobile telephones and entertainment systems. Analysis of accidents occurring over the period 1985-1995 shows that in-vehicle distraction is reported as a contributory factor in about 2% of fatal accidents (although this figure may be a conservative estimate). Specific examples of distraction attributed to entertainment systems and telephones have been identified. Electronic driver information systems are also of particular interest, but have not featured in the available data. Work is progressing, on an annual cycle, to obtain, code and analyse further data and this is expected to provide an invaluable source of information for accident researchers.     

Road accident statistics: discrepancies between police and hospital data in a French island

In most developed countries, information on road crashes are routinely collected by the police. However, comparison of police records and hospital data underlines a deficit of the number of road accidents in the routine statistics. In La Réunion, a French overseas dependency, an epidemiological study of injuries leading to hospitalisation or deaths has been performed from June 1993 to June 1994. The comparison between hospital data and police records showed that only 37.3% of non-fatally traffic-injured in-patients were recorded by the police. Length of stay in hospital, physician in charge of the first aid, urban place of the crash, type of vehicle involved, day and time of the crash and blood alcohol concentration were significantly associated with the presence in the police file. Police overestimated the severity of the injuries. Police notified 100 deaths on the 115 counted by the study. In France, non-fatally traffic-injured should be followed 30 days to improve quality of police death records. A capture-recapture method was used to estimate the total number of injured people. The capture-recapture method consists in merging information from several sources of notification to determine the real number of cases in the population and the exhaustivity of each source. We estimated that 346 subjects were injured in one month whereas police data recorded only 87 and hospital data 137. This method seems interesting to use in routine after validation when unique personal identifiers are available.     

Validity of using linked hospital and police traffic crash records to analyse motorcycle injury crash characteristics

Linking hospital discharge and police traffic crash records has been used to provide information on causes and outcomes for hospitalised traffic crash cases. Motorcyclists are particularly vulnerable to injury in a traffic crash, but no published linkage studies have reported in detail on this road user group. The present study examined motorcycle traffic crash injury cases in New Zealand in 2000–2004 by probabilistically linking national hospital discharge records with police traffic crash reports. Injury cases had to have spent at least one night in hospital before being discharged and were defined as serious or moderate based on their International Classification of Disease Injury Severity Scores (ICISS). Despite a robust linkage process, only 46% of cases could be linked to a police record; 60% of the serious injuries and 41% of the moderate. The low linkage was most likely due to under-reporting of crashes to or by the police. While moderate injury cases were expected to be under-reported, the level of under-reporting of cases with serious threat-to-life injuries is concerning. To assess whether the linked dataset could provide valid information on the crash circumstances and injury outcomes of hospitalised motorcycle crash cases, the characteristics of the linked and unlinked hospital discharge cases were compared using chi-square tests and multivariate logistic regression. Serious injury cases were less likely to be linked if only one vehicle was involved, or the injured riders and passengers were younger than 20 years or spent less than one week in hospital. For moderate injury cases, there were also differences in linkage by injured body region and crash month. While these discrepancies need to taken into consideration when interpreting results, the linked hospital-police dataset has the potential to provide insights into motorcycle crash circumstances and outcomes not otherwise obtainable.     

The real number of road traffic accident casualties in The Netherlands: a year-long survey

Between August 1986 and July 1987 more than 24,000 households, containing nearly 67,000 persons, were surveyed by telephone about traffic injuries during the past three months. Expressed on an annual basis, approximately 430,000 people, or about 1 in 34 of the Dutch population, had suffered some sort of injury in a road accident. The road traffic morbidity was, therefore, 2,942 per 100,000 inhabitants. Of these, about 135,000 had to be treated in hospital (20,000 as inpatients). More than 100,000 did not need treatment. Cyclists formed by far the largest category of road user, but mopedists had the highest injury rate per kilometer travelled. 210,000 of these casualties fell within the definition for recording by the police. The police recorded only 49,748 traffic casualties, or about 25%, during the same period. The police data were not representative; the completeness declined according to severity of the injuries: inpatients, about 70%; outpatients 26%; extramural about 11%. Cyclists (11%), children (9%), and single vehicle accidents (5%) were very much underrepresented. The largest category of road user is cyclists, not car occupants as indicated by the police data. A number of recommendations are made for supplementing the police data and the existing hospital inpatient data. These include extending the Home Accident Recording System of outpatients and the General Practitioner Panel to include road accident victims. Together a representative sample of 95% of all those receiving medical treatment would thus be obtained.     

A comparison of self-reported motor vehicle collision injuries compared with official collision data: an analysis of age and sex trends using the Canadian National Population Health Survey and Transport Canada data

OBJECTIVE: The purpose of this study was to compare the age and sex trends of motor vehicle collision injuries between a nationally representative self-report survey and official police motor vehicle collision report data. To do this, population-based estimates of motor vehicle collision injuries were established using data from the National Population Health Survey (NPHS), a prospective, population-based, longitudinal survey that was compared to Transport Canada's official motor vehicle collision report statistics. METHODS: Aggregated mean data (1994-1996) were compared for seven age groups (15-19, 20-24, 25-34, 35-44, 45-54, 55-64, and 65+) from both data sets. RESULTS: No significant differences were found between males' and females' MVC injuries for any age category in the NPHS. A comparison of the NPHS and Transport Canada data found two small (significant) within-sex differences between the data sets, but overall, the analysis largely revealed similar trends for self-reported injury for all age categories and sex. CONCLUSIONS: The results indicate that the incidence of injuries based on self-report data in a nationally representative sample is similar to official sources of reporting and are thus a valid indicator or motor vehicle collision injury incidence. The results also confirm that injury trends differ from fatality trends.     

Estimating crashes involving heavy vehicles in Western Australia, 1999-2000: a capture-recapture method

A two-sample exploratory study of police and hospital records was undertaken to estimate the number of fatalities and serious injuries for heavy vehicle drivers involved in a crash in Western Australia. The capture-recapture method was used to assess differences and similarities in characteristics of heavy vehicle drivers from both sources. Each heavy vehicle driver involved in a crash from the police report was matched against the heavy vehicle driver's hospitalisation record from the Hospital Morbidity Data System, with surname, initials, date of birth, gender, date of crash, road user type and vehicle type as matching fields. The estimated number of fatalities and serious injuries to heavy vehicle drivers from 1st July 1999 to 31st December 2000 was 5 and 59, respectively, which was 25 and 31% higher based on the capture-recapture methodology than the aggregated (non-overlapping) total officially reported to the police and hospitals. No significant age difference (p>0.05) was found for drivers involved in a heavy vehicle crash between the two sources (37 years versus 40 year of age). However, female heavy vehicle drivers were over-represented in the hospital records (11%) compared to the police records (1%). The capture-recapture approach is useful for evaluating the completeness of data sources and identifying biases within datasets. The underestimation of heavy vehicle drivers seriously injured and killed has important implications for heavy vehicle safety management and resource allocation in Western Australia.     

Assessment of hospital and police ascertainment of automobile versus childhood pedestrian and bicyclist collisions.

Information about automobile versus pedestrian or bicyclist collisions differ according to the source of collision ascertainment. Hospital records and police reports focus on different characteristics of a collision, which reflects differences in case finding as well as information regarding risk factors. Pedestrians and bicyclists 0-14 years of age who were involved in a motor vehicle collision in the city of Long Beach, CA, between January 1, 1992 and June 30, 1995, were included in the study. Police reports and medical records were used to identify cases. Police reports were found for 1,015 cases, and medical records identified 474 cases; 379 cases were common to both sources. A capture-recapture model was used to evaluate the degree of overlap between the two sources and to derive "ascertainment corrected" injury rates. The injury rate from combined sources was 333.5/100,000 children per year, the pedestrian injury rate was 210.6/100,000 children per year and the bicyclist injury rate was 122.9/100,000 children per year. The "ascertainment corrected" injury rate overall was 381.3/100,000 children per year, 233.0/100,000 for pedestrians and 153.9/100,000 children per year for bicyclists. Eighty percent of hospital-reported cases were captured in the police database, whereas only 37% of police-reported cases were captured by the hospital database. We found that hospital sources identified younger children, fewer bicyclists, more Asian and Hispanic children, and fewer African-American children when compared to police sources. For more comprehensive surveillance resulting in more accurate incidence rates and more complete information, it is better to use multiple sources of data.     

Understanding the human factors contribution to railway accidents and incidents in Australia

Forty rail safety investigation reports were reviewed and a theoretical framework (the Human Factors Analysis and Classification System; HFACS) adopted as a means of identifying errors associated with rail accidents/incidents in Australia. Overall, HFACS proved useful in categorising errors from existing investigation reports and in capturing the full range of relevant rail human factors data. It was revealed that nearly half the incidents resulted from an equipment failure, most of these the product of inadequate maintenance or monitoring programs. In the remaining cases, slips of attention (i.e. skilled-based errors), associated with decreased alertness and physical fatigue, were the most common unsafe acts leading to accidents and incidents. Inadequate equipment design (e.g. driver safety systems) was frequently identified as an organisational influence and possibly contributed to the relatively large number of incidents/accidents resulting from attention failures. Nearly all incidents were associated with at least one organisational influence, suggesting that improvements to resource management, organisational climate and organisational processes are critical for Australian accident and incident reduction. Future work will aim to modify HFACS to generate a rail-specific framework for future error identification, accident analysis and accident investigation.     

Home injuries mortality: sensitivity and specificity analysis of different data sources and operative definitions

The aim of this longitudinal study was to test different operational definitions of home accident mortality. METHODS: The sources of data were the Emergency Information System, hospital discharge reports and the mortality registry of the Lazio Region, 2000-2001. We selected all emergency room visits for unintentional traumas that occurred at home. A 9-month follow-up was performed to calculate mortality rates. A sensitivity analysis of in-hospital mortality, deaths within 30 days and deaths from home accident E-codes was performed. A gold standard definition of home accident-related deaths was proposed. RESULTS: We observed 598 home accident-related fatalities (29.1% of all the deaths found in the follow-up study). In-hospital mortality, deaths within 30 days and deaths for home accident E-codes had sensitivity values of 63.4%, 63.4% and 59.4%, respectively; positive predictive values were 78.1%, 67.1% and 100%, respectively. CONCLUSIONS: The best operational definition of home injury was based on in-hospital mortality, while mortality based on E-codes reported on death certificates was lacking. In order to measure the real burden of home injuries, hospital and mortality data must be integrated.     

A time comparison study of the New York state safety belt use law utilizing hospital admission and police accident report information

New York state enacted the first safety belt use law in the United States in 1984. We evaluated the effects of the law by reviewing all hospital admissions from motor vehicle crashes in Monroe County, New York. We compared admissions for the 18 months prior to the effective date of the law with those for the 18 months after the law became effective. Police accident reports and hospital records were coupled and intensively reviewed. Motorcyclists, pedestrians, and bicyclists (bicycle collisions with motor vehicles) became controls for the study. Seat belt usage was determined from police and hospital record information. Analysis of the data revealed that safety belt use among patients hospitalized because of motor vehicle accidents increased from 11.2% before the law became effective to 53% after the law became effective. Hospital admissions decreased 11.9% among motor vehicle occupants and increased 2.6% among controls. The ISS decreased from 16.01 to 14.55 for motor vehicle occupants and increased' from 14.77 to 15.11 among controls. Among subjects all injuries decreased except injuries of the spine and abdomen, which increased in the postlaw period.     

Accident prototypical scenarios, a tool for road safety research and diagnostic studies

The concept of prototypical accident scenario has been used since the late 1980s in French road safety research. A prototypical scenario can be defined as a prototype of the accident process corresponding to a series of accidents which are similar in terms of the chain of facts and causal relationships found throughout the various accident stages. This concept provides a means of combining and generalising the knowledge obtained from accident case studies, based on in-depth investigation methods or on detailed analyses of police reports. Applications of this concept are developed in both the field of traffic accident research and safety studies (diagnoses) in preparation for engineering measures or local safety policies. This paper presents the prototypical scenario concept, its theoretical background, and the way it is used for safety research and studies.     

The western australian road injury database (1987-1996): ten years of linked police, hospital and death records of road crashes and injuries

Accurate information about injuries and their causes is essential to road safety research, policy development and evaluation. Such information is most powerful when it is available for all road crashes within a jurisdiction. The Western Australian Road Injury Database achieves this through the on-going linkage of crash details from reports to police with the details of injuries to casualties contained in hospital and death records. Over the 10-year period 1987-1996, 386,132 road crashes involving 142,308 casualties were reported to the police in Western Australia. There were also 47,757 hospital discharge records and 2,906 death records related to road crashes during this period. Of the 142,308 police casualties, 17 848 had a matching hospital discharge record and 2,454 had a matching death registration. Linkage within the hospital records revealed that the 47,757 discharge records involved 43,179 individuals, of whom 39,073 were admitted to hospital once, 3,653 were admitted twice, 374 were admitted three times and 78 were admitted more than three times. Of the 43,179 hospitalised casualties, 817 had a matching death record. Linked police, hospital and death records of road crash casualties provide accurate outcome information for casualties in crashes reported to the police. In addition, estimates of under reporting of crashes for different road user groups can be made by comparing hospital records with and without a matching police record. This article demonstrates the power of a linked system to answer complex research questions related to outcome and under-reporting.