2.7 Interurban road safety research gaps in LMICsTo our knowledge, no review on the epi<strong>de</strong>miology of interurban RTC in LMICs wasavailable. As part of a systematic review of original studies published in Medline® withinterurban road settings (Appendix 1), we only found 31 articles from these countries. Most ofthem used police or health data to assess crash bur<strong>de</strong>n. We did not find any study comparingdifferences in injury reporting between these types of data. Only one study reported injurybur<strong>de</strong>n per 100 million vehicle-km. Situational factors other than the light or weatherconditions were almost never documented in these studies. None of them assessed theinteractions between driver- and road-related crash factors. Only few studies assessed highriskcrash site i<strong>de</strong>ntification methods on inter-urban road sections in LMICs. Finally, few roadinterventions were assessed in these countries, mostly by non comparative methods.3. ObjectivesTo contribute to filling the research gaps on RTCs and RTIs in LMICs, the objectives of thisthesis were:1. To assess the road crash and injury bur<strong>de</strong>n on selected interurban roads in LMICs–Studies I & II;2. To <strong>de</strong>scribe road user groups and situational factors involved in interurban roadcrashes–Studies I & II;3. To assess the association of situational factors with injury crashes on selected roads ofLMICs–Studies III & IV;4. To assess the road hazard perception of the high-risk crash sites in voluntary drivers–Study V.The thesis presents two <strong>de</strong>scriptive studies (I & II) and three analytical studies (III-V) in linewith these objectives. In study I, the crash bur<strong>de</strong>n, number of persons who died or wereinjured per vehicle-km was assessed on Yaoundé-Douala road section in Cameroon. Further,associated crash factors and types were <strong>de</strong>scribed using police reports. In study II, wecompared RTIs per vehicle-km reported to police, ambulance, and hospital on Karachi-Halaroad section during a one-year period. In study III, we assessed situational factors associatedwith injury crash sites using case-control methods on the Yaoundé-Douala road section. Instudy IV, we compared the inci<strong>de</strong>nce <strong>de</strong>nsity rates estimated from events (crash, fatality, andsevere injury) and vehicle-km between highway work zones and other traffic zones onKarachi-Hala road section. In study V, we assessed the hazard perception of high-risk crashsites and those not involved in crashes from the above two road sections by showing theirvi<strong>de</strong>os to voluntary Pakistani drivers. Furthermore, situational factors associated with drivers’hazard perception were assessed using multivariate mo<strong>de</strong>ls. Study I and III have beenpublished whereas manuscript of study IV is un<strong>de</strong>r review and those for study II and V are inpreparation.20
4. Descriptive studiesThe paucity of traffic injury data in African LMICs, as discussed in the literature reviewpublished in 2007 by the head of the research team in which this thesis was conducted [50],ma<strong>de</strong> interesting for us to start our work by <strong>de</strong>scribing this public health problem for somespecific setting. Since few studies were published on interurban traffic safety at that time, westarted by assessing traffic crash bur<strong>de</strong>n on Yaoundé-Douala road section, Cameroon (StudyI). Results from this study permitted us to better appreciate the process of traffic injuryreporting in this country. We repeated a similar study in Pakistan, but this time we were ableto collect RTI data from multiple sources including police, ambulance, and hospitals (StudyII). Comparison of injury reporting was useful to assess discrepancies in police data, often theonly source of crash reporting as evi<strong>de</strong>nced in Cameroon.4.1 Study I: Traffic crash and injury bur<strong>de</strong>n on Yaoundé-Douala roadsection, CameroonLocated in Central and West Africa on the Gulf of Guinea, Cameroon covers an area of475 440 square kilometres and has a population of nearly 19 million inhabitants. Cameroonhas one of the ten highest per capita GDPs—about 2 300 US$ (2008)—in sub-Saharan Africa.The life expectancy at birth is 54 years [69]. The total road network is 50 000 km long, withonly 10% paved [69]. Its interurban road network is highly strategic in the region as it isexpected that its economy will grow substantially in coming years, with increasing oil importsfrom Chad and industrialization [70].Cameroon is in pre-motorization stage with 16.8 registered vehicles per 1 000 persons [2]. Anincreasing trend in crash fatalities has been observed in Cameroon since 1970s [3]. Trafficfatalities increased from un<strong>de</strong>r 400 in 1972 to over 1 150 in 2005 [2]. Over half of those whodied are passengers and drivers of four-wheeled vehicles, whereas pe<strong>de</strong>strians, cyclists, andmotorcyclists account for the rest of <strong>de</strong>aths [2]. A recent estimate suggested that actual trafficfatalities could be around five thousand per year in Cameroon [2]. Speed limit, drunk-driving,seat-belt, and helmet laws are poorly enforced [71].The formal pre-hospital care system is stillin <strong>de</strong>velopment and RTCs are the principal reason for such interventions in urban settings[71].Traffic safety on interurban road network is increasingly becoming the matter of publicconcern in Cameroon. Crashes that drew extensive public attention were mostly thoseinvolving public transportation vehicles with several fatalities [72]. When our study wasstarted, no comparative fatality and injury estimates were available for these road sections.Further, no <strong>de</strong>scription was available in the literature of crash types, causes, and road usersinvolved. Several reasons could explain unavailability of this information: firstly, not allconcerned police reports were forwar<strong>de</strong>d to the <strong>de</strong>partment in charge of traffic injurysurveillance and, secondly, no <strong>de</strong>tailed data were neither collected nor analyzed at the nationallevel, that could provi<strong>de</strong> a better picture of traffic safety on interurban roads [72].21
- Page 1: Université Victor Segalen Bordeaux
- Page 4 and 5: Publications (peer-reviewed).......
- Page 6 and 7: Index of figuresFigure 1. Traffic f
- Page 8 and 9: AbbreviationsAKUAVCIBMIEASESSDALYDW
- Page 10 and 11: AbstractBackground: Interurban traf
- Page 12 and 13: L'objectif de cette thèse était d
- Page 14 and 15: 2. Background2.1 Road injury burden
- Page 16 and 17: 2.4 Multiple factors involved in tr
- Page 18 and 19: Figure 4. Percentage difference of
- Page 22 and 23: ObjectivesThe objectives of this fi
- Page 24 and 25: ResultsCrash burdenA total of 935 R
- Page 26 and 27: Figure 7. Monthly trend of traffic
- Page 28 and 29: Injury outcome patternsMost of inju
- Page 30 and 31: MethodsThe study setting was 196-km
- Page 32 and 33: patients. In the ED, those with NIS
- Page 34 and 35: Table 6. Traffic injuries reported
- Page 36 and 37: 5. Analytical StudiesPrevious liter
- Page 38 and 39: under supervision of Dr. Sobngwi-Ta
- Page 40 and 41: Table 9. Situational variables at c
- Page 42 and 43: MethodsStudy design and settingStud
- Page 44 and 45: to Dec 08 were retrieved and photoc
- Page 46 and 47: normal zones. However, this associa
- Page 48 and 49: MethodsStudy design and settingsThe
- Page 50 and 51: Figure 14. Picture extracted of a h
- Page 52 and 53: located in built-up area in Pakista
- Page 54 and 55: Table 15. Differences in hazard per
- Page 56 and 57: 6. Discussion6.1 Originality of stu
- Page 58 and 59: Although adjustments are possible,
- Page 60 and 61: observational studies on how the de
- Page 62 and 63: to understand the deficiencies in t
- Page 64 and 65: [24] Damsere-Derry J, Afukaar FK, D
- Page 66 and 67: [69] Central Intelligence Agency. T
- Page 68 and 69: [111] Geurts K, Wets G, Brijs T, Va
- Page 70 and 71:
[154] Rosenbloom T, Shahar A, Elhar
- Page 72 and 73:
4. Farooq U, Bhatti JA, Siddiq M, M
- Page 74 and 75:
Appendix 1: Literature review on in
- Page 76 and 77:
they identified a cluster of long b
- Page 78 and 79:
more cost-effective than redesignin
- Page 80 and 81:
Table 18. Traffic injury interventi
- Page 86 and 87:
Appendix 3: Study I supplementary r
- Page 88 and 89:
Appendix 4: Manuscript in preparati
- Page 90 and 91:
BACKGROUNDPakistan, located at the
- Page 92 and 93:
patients were recorded during their
- Page 94 and 95:
This might motivate police officers
- Page 96 and 97:
12. Peden M, Scurfiled R, Sleet D,
- Page 98 and 99:
Table 1. Traffic injuries reported
- Page 100 and 101:
Table 3. Ascertainment of police, a
- Page 102 and 103:
Appendix 5: Article published - Stu
- Page 104 and 105:
104
- Page 106 and 107:
106
- Page 108 and 109:
108
- Page 110 and 111:
Appendix 6: Article under review -
- Page 112 and 113:
1. INTRODUCTIONWith the aging of hi
- Page 114 and 115:
A total of 180 crashes were identif
- Page 116 and 117:
conspicuity at HWZs in Pakistan. 2
- Page 118 and 119:
21. Sobngwi-Tambekou J, Bhatti J, K
- Page 120 and 121:
Table 2. Highway work zone crash fa
- Page 122 and 123:
122
- Page 124 and 125:
ABSTRACTObjectives: Interurban road
- Page 126 and 127:
oad). A matched strategy was used t
- Page 128 and 129:
SitesOut of 131 crash sites identif
- Page 130 and 131:
Similarly, it was shown previously
- Page 132 and 133:
Majdzadeh, R., Khalagi, K., Naraghi
- Page 134 and 135:
Table 2. Characteristics of Pakista
- Page 136 and 137:
Table 4. Factors associated with ha
- Page 138 and 139:
Table 21. Situational factors at hi
- Page 140 and 141:
Table 23. Situational factors assoc