|Mar & Sep
1Achutha Menon Centre for Health Science Studies, Sree Chitra Tirunal Institute for Medical Sciences & Technology, Trivandrum, Kerala, India
2Hospital Infection Control (HIC) & Microbiology, Kauvery Group of Hospitals, Chennai, Tamil Nadu, India
Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License (http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-Commercial use, reproduction and distribution of the work without further permission provided the original work is attributed.
Background: Road traffic injuries (RTIs) are a growing public health concern in India. Despite good health infrastructure and better-connected roads, Kerala has the fifth highest incidence of RTI in India. The spatiotemporal analysis of RTI in Trivandrum is undertaken in this context.
Methods: RTI data for 2016 was obtained from the State Crime Records Bureau, Thiruvananthapuram and the population data were taken from the Census 2011. Spatial statistical analysis was used to look for geographical distribution and spatiotemporal clustering of RTIs. Health-care facilities providing emergency trauma care services in the city were identified from secondary sources.
Results: A total of 2,319 people, including pedestrians, got injured in 1,926 accidents in Thiruvananthapuram city in 2016. A Moran’s I value of 0.23 signifies spatial autocorrelation (clustering) of RTIs in Trivandrum. The rates of RTIs were significantly high among senior citizens, male sex, two-wheeler users, and during daytime. Most deaths (48.2%) occur among the active age group of 30 to 60 years. A severity map was generated and using LISA, local hotspots were identified. Eight significant cluster locations were identified among the 100 wards. The mean distance from the accident hotspots to the nearest health-care facility was 1.2 km for the public and private facilities.
Conclusion: There is a significant spatial-temporal clustering of RTIs in Thiruvananthapuram city. Proportionally higher deaths occur among pedestrians, 2-wheeler users, and women, which needs further exploration. The study results could be used to plan and build RTI prevention strategies in Thiruvananthapuram city.
Geographic information system, health care, road traffic injuries
The World Health Organization defines road traffic injury (RTI) as a fatal or nonfatal injury resulting from a collision on a public road involving at least one moving vehicle.1 Children, pedestrians, cyclists, and the elderly are among the most vulnerable road users.1,2
RTIs were the ninth leading cause of death globally and are forecasted to be the fifth leading cause of death by 2030.3,4 Around 94% of road traffic-related deaths are from low- and middle-income countries. Ironically, these countries have half of the world’s registered vehicles due to rapid motorization without adequate infrastructure development or road safety measure.5-7
About 85% of annual deaths and 90% of the disability-adjusted life years are lost due to RTIs.8 This often drives many families to poverty due to the loss of breadwinners. As with most RTIs and deaths, it is often impossible to attach a value to each case of human sacrifice or suffering. The transport sector often faces challenges in designing roads and enforcing traffic regulations to achieve sustainable road transport safety.
Kerala has good health infrastructures and services, a comparatively higher density of population distribution, and connectivity of roads, including those in villages.9 In spite of this, Kerala ranks among the top five states in India with increased RTIs.10
Thiruvananthapuram City Corporation is a fast-growing city in South India, with constant expansions in infrastructures and changes in land use. Most expansions have inadequate planning and lead to narrow roads with traffic congestion. With the development of the information technology hub, numerous start-ups, and franchises, there have been an increased number of vehicles with a disproportionate increase in road infrastructure. The roads in the area are always under excessive pressure, the likelihood of accidents is higher, and the possibility of spatiotemporal prediction of accidents is significant to the traffic police department, transportation planners, and engineers.
The golden hour is regarded as the first hour after an event of RTI and is the most crucial and critical hour. A trauma victim’s chance of survival drastically decreases significantly after the first hour by almost 60%. The National Highway Authority of India (NHAI) guidelines recommend having trauma care centers every hundred kilometers on highways.10 A recent initiative by the government has been not to question the person who has taken the responsibility of bringing the accident victim from the spot of the trauma to the nearest health care facility.
Agnihotri11 mentioned that prevention is the key to dealing with RTIs. In order to minimize morbidity and mortality, a national or regional multidisciplinary trauma system must be developed with all facilities to care for all victims of RTIs. Effective triaging is the key to rapid identification of critically injured victims to match better the victim’s need and the resources available in the particular health-care facility.11
Thiruvananthapuram City Corporation is the capital and headquarter of Kerala. It extends from N 80 21’ 44.485” and E 760 51’ 20.8” to N 80 36’ 25.542” and E 770 1’ 27.119 (Figure 1).
Figure 1. Map of Thiruvananthapuram City With Ward Demarcation and Road Networks.
Source: Geo-spatial resources at AMCHSS.
A road map of Thiruvananthapuram city with all the 100 wards demarcated was created. The road network layers were procured from open sources. All RTIs in 2016 (around 3,000 in number) were obtained from the State Crime Records Bureau (SCRB). A structured data extraction template was used to extract the relevant details for the study from this secondary source. Geocoordinates of the accident locations were captured, and the parameters were mapped over the base map of Thiruvananthapuram city using QGIS software.12
The spatial distribution of the point pattern of accidents and autocorrelation was assessed using Moran’s I statistics (own data using Geoda software13). Spatial-temporal clustering of RTIs was explored by estimating Local Indicators of Spatial Association (LISA) and Clustering Large Applications (CLARA) (own data using R software).
CLARA is an extension of the Partitioning Around Medoids method. CLARA is used for large data sets. It relies on sampling and the clustering process. It tries to minimize sampling bias.
Secondary and tertiary health-care facilities offering emergency trauma care services in Thiruvananthapuram city were identified. The investigator visited all the centers and collected information using a preplanned checklist, including the geocoordinates of the center, which were recorded using a handheld GPS unit.
The protocol for the study was approved by the Institutional Ethics Committee of Sree Chitra Tirunal Institute for Medical Sciences and Technology, Trivandrum. The Ethics Committee is following the Helsinki Declaration. Also, the necessary permission was obtained from the SCRB, Thiruvananthapuram.
After a thorough quality check for missing values, duplicate values, and confining to data within Thiruvananthapuram City Corporation boundaries, the total RTIs reported were 1,926 events during 2016.
Moran’s-I is a statistical test to find spatial autocorrelation. The given set of features and an associated attribute/s evaluates whether the pattern is clustered, dispersed, or random. This is based on feature locations and feature values simultaneously. Moran’s-I was estimated after assigning weights for the events. The calculated value was found to be 0.23. This indicates a positive spatial autocorrelation (Figure 2).
Figure 2. Spatial Autocorrelation (Moran’s-I).
Source: Own data using QGIS software.
Local Indicators of Spatial Association (LISA)
LISA statistics serve 2 purposes. They may be indicators of local pockets of nonstationary or hot spots, similar to the Gi and G*i statistics of Getis and Ord. They are also used to assess the influence of individual locations to identify outliers, as in the Anselin-Moran scatter plot. LISA statistics show hot spots and cold spots. This was calculated after creating a weight for the calculation. LISA was estimated using GeoDa (Figure 3).
Figure 3. Moran’s Plot (Left), Local Indicators of Association (LISA) Plot (Right).
Source: Own data using Geoda software.
Clustering Large Applications (CLARA)
The optimal number of clusters for the data set was found to be 8. A cluster plot (Figure 4) was generated using R.
Figure 4. Cluster Plot.
Source: Own data using R software.
The accident hotspots were identified after augmenting the results from LISA and the cluster plot onto the base map and Google street map layer. The identified hot spots with the maximum clustering were seen in Karamana Bridge, Karamana junction, Sreekaryam, Papanamcode junction, and Kovalam road (Table 1).
Table 1. Distribution of Health-Care Facilities in Thiruvananthapuram City Corporation.
Source: Own data using R software.
A total of 83 health-care facilities were identified. Of this, 20 were government facilities and the rest 63 were private facilities. The health-care facilities were more densely distributed closer to the city than in the peripheries. The mean road distance from the accident hotspots to the nearest health-care facility was calculated and found to be 1.24 km.
There is distinctness in road traffic fatalities and mortality rates between high-income countries vs low- and middle-income countries.4 It is noticed that there is a decrease in rates of fatalities in high-income countries. This could be attributed to implementing a wide range of road safety measures, inclu- ding seat-belt use, vehicle crash protection, traffic-calming interventions, and traffic law enforcement. While the same cannot be said about the middle- and low-income countries with a rise in fatality rates since the 1960s.3
Dr R Adams Cowley first described the golden hour in emergency medicine as “the time period lasting for one h or less following traumatic injury being sustained by a casualty or medical emergency during which there is the highest likelihood that prompt medical treatment will prevent death.”
A few of the other contributing factors to RTIs in India identified by Gopalakrishnan were reckless and high-speed driving, no proper legislation, the attitudes toward the “right of the mighty” bigger and larger vehicles toward smaller vehicles, overburdened public and transport vehicles, poor maintenance of vehicles, drunk and driving, driver fatigue, and encroachment by unauthorized persons and properties.14 These factors were not identified in the present study because they were beyond the scope of the study. The NHAI proposed to setup designated trauma centers every 100 km.15 Though predefined for the highways, this has not been defined for city limits.
Using statistical methods, the geographic information system (GIS) helps identify contributory factors that are usually not identified. GIS inculcates all the factors involved in the event and provides an integrated language to describe the data.14 Sambrani explored advancements in GIS, especially executive research techniques and GIS. Finding the shortest distance requires knowledge of Spatial Decision Support Systems.14 This integration is needed for analysis of the best and shortest route. This can also be used for choosing and analyzing alternative routes. A similar algorithm was used to find the nearest health-care facility to the hot spot in QGIS software. GIS is the tool to manage traffic accident data and augments decision-making in RTIs.15 It was noticed that most of the hot spots identified in the current study were in intersections. This is congruent with Haji Housainlou et al,16 they had identified the importance of intersection crashes in Tehran.
Accidents occur in Thiruvananthapuram city in a clustered pattern over certain locations (Moran’s I = 0.23). We have used spatial analysis techniques to delineate spatial clusters and their locations over time (spatiotemporal clusters). A similar study was done in 2008 by Prasannakumar et al,9 but the factors for temporality taken were the season and proximity of religious and educational institutions to the accident hot spots. The current study takes the distance to the nearest health facility that provides emergency trauma care services.
A survey of the leading accident hot spots with a prestructured checklist concluded that the hot spots lacked police personnel, hazard warnings, and speed breakers within 1 km of the accident hot spot. They also lacked pedestrian crossings. Most of the hot spots had a traffic signal, and the roads were wide, with medians separating the traffic. A few of the hot spots had recent expansions, and it would be interesting to identify the trend of RTIs in these particular hot spots in the coming years or take 2 different points in time. Another interesting observation was all the hot spots had a high overall vehicle density. It would have been interesting if a real-time vehicle density was available for the study. Most of the hot spots are now equipped with street lights. A few of the spots had recently got streetlights in accordance with the road expansion process; hence, it is difficult to document since the study period and the physical surveys were conducted at different times.
The study helped in the value addition of the routinely captured RTI data using the capabilities of open source GIS software. Information about 1,926 accidents in Thiruvananthapuram city in 2016 was obtained from SCRB and geocoded. The investigator visited and collected information (including the geolocations) of all health-care facilities that offer emergency trauma care services within the city and within a 5 km buffer zone around the city limits. On analysis, it became evident that there is geospatial and spatiotemporal clustering of RTI within Thiruvananthapuram city. A preliminary analysis of the hot spots around the city limits revealed many correctable factors like lack of traffic wardens or other traffic calming measures during peak traffic hours that increase RTI proneness. The fatalities, especially in the 18 to 30 and above 60 years, are alarming and require better and probably more stringent license procedures.
The potentials of GIS are endless. GIS can be used to find the ideal locations to emergency centers or hospitals; find black spots; assess the effectiveness of services according to the location.17 They can also be used to find the local dispersion pattern for hospital services. GIS can also help make evidence-based decisions.18 Though the current study was not along these lines, the capabilities of GIS are phenomenal.19,20 We were able to identify the mean distance from the accident hot spot to the nearest health-care facility as 1.27 km to the nearest health-care facility from the hot spots. But we were not able to capture the victim’s choice of health-care facility. This is quite challenging since following an event of RTI, despite a particular health-care facility being in proximity to the event’s location, the victim might choose to approach a distant health-care facility. This often delays and brings challenges to the concept of the golden hour.
The research was done as a part of the Master of Public Health program during 2016-2018, at the Sree Chitra Tirunal Institute for Science and Technology, Trivandrum, by the first author under the guideship of the second author.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The authors received no financial support for the research, authorship, and/or publication of this article.