Tropical Medicine and International Health

doi:10.1111/tmi.12562

volume 20 no 11 pp 1578–1587 november 2015

Road traffic injuries in northern Laos: trends and risk factors of an underreported public health problem G€ unther Slesak1,2, Saythong Inthalath3, Annelies Wilder-Smith2,4 and Hubert Barennes5,6,7 1 2 3 4 5 6 7

Tropenklinik Paul-Lechler-Krankenhaus, T€ ubingen, Germany Institute of Public Health, University of Heidelberg, Heidelberg, Germany Luang Namtha Provincial Health Department, Luang Namtha, Laos Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore Agence Nationale de Recherche sur le VIH et les Hepatites, Phnom Penh, Cambodia INSERM, ISPED, Centre INSERM U897-Epidemiologie-Biostatistique, Bordeaux, France Epidemiology Unit, Pasteur Institute, Phnom Penh, Cambodia

Abstract

objectives Road traffic injuries (RTI) have become a leading cause for admissions at Luang Namtha Provincial Hospital (LNPH) in rapidly developing northern Laos. Objectives were to investigate trends, risk factors and better estimates of RTI. methods Repeated annual surveys were conducted with structured questionnaires among all RTI patients at LNPH from 2007 to 2011. Hospital and police data were combined by capture–recapture method. results The majority of 1074 patients were young [median 22 years (1–88)], male (68%), motorcyclists (76%), drove without licence (85%) and without insurance (95%). Most accidents occurred during evenings and Lao New Year. Serious motorbike injuries were associated with young age (1–15 years), male sex (OR 2.2, 95% CI 1.1–4.6) and drivers (OR 2.1, 95% CI 1.1–4.3); more serious head injuries with alcohol consumption (OR 2.5, 95% CI 1.7–3.7), male sex (OR 2.3, 95% CI 1.4–3.7) and no helmet use (OR 2.0, 95% CI 1.2–3.4). No helmet use was associated with young age, time period, pillion passengers (OR 2.7, 95% CI 1.6–4.7), alcohol (OR 1.9, 95% CI 1.2–2.8) and no driver license (OR 2.0; 95% CI 1.1–3.4). Main reasons not to wear helmets were not possessing one, and being pillion passenger. Capture–recapture analysis showed four times higher RTI estimates than officially reported. Mortality rate was 11.6/100.000 population (95% CI 5.1–18.1/ 100.000). conclusions RTI were substantially underestimated. Combining hospital with police data can provide better estimates in resource-limited settings. Preventive programmes and law enforcement have to target male drivers, alcohol, licensing and helmet use, especially among children and pillion passengers. Increased efforts are needed during evening time and special festivals. keywords traffic, road, injury, accident, Laos, hospital survey, capture–recapture

Introduction Road traffic injures (RTI) are the 8th most common cause of death globally [1]. Low- and middle-income countries carry a disproportionate high burden with 91% of the world’s fatalities [2]. In the western Pacific region, RTI have become the leading cause of death among 15to 44-year-olds [3]. In Lao People’s Democratic Republic (Laos), RTI have dramatically increased with the recent economic development from a low- to a lower middle-income country [4, 5]. The majority of its 6.8 million people live in rural areas with underdeveloped infrastructure [4, 5]. An 1578

annual peak of RTI is seen in April during Lao New Year associated with excessive alcohol consumption [6]. With new trans-Asian highway construction, RTI have become the leading cause for hospital admissions in Luang Namtha Province (LNP), northern Laos, in 2006 [7]. Road safety campaigns in 2007 could temporarily reduce severe RTI by 35% and increase helmet use from 11% to 46%, but no further thereafter [7]. Knowledge of the local situation and common risk factors is essential for effective road safety programmes. Most RTI go unrecorded, and official statistics only reflect part of the problem [8, 9]. Some data are already available from the country’s capital Vientiane [8–13] but

© 2015 John Wiley & Sons Ltd

Tropical Medicine and International Health

volume 20 no 11 pp 1578–1587 november 2015

G. Slesak et al. Road traffic injuries in northern Laos

data from rural areas remain scarce. Data validity could be improved if several sources were combined using the capture–recapture method as in other resource-limited settings without reliable surveillance systems [14–18]. This method is commonly used to calculate hidden animal populations; a sample is captured, marked and released. In a later sample, the number of marked (‘recaptured’) individuals should be proportional to the marked individuals in the whole population. The study was conducted to identify characteristics, trends and risk factors associated with RTI in LNP from 2007 to 2010, and to investigate reasons for not wearing helmets. Official police and hospital data were combined to estimate the real RTI incidence.

Methods Study site LNPH is a 50-bed hospital which serves the population (149 000 people in 2007) as the referral centre for five district hospitals and the military hospital of the province (Figure 1). It is the only hospital offering X-ray, bone

Figure 1 Map of the provincial capital Luang Namtha in northern Laos with the new trans-Asian highway R3 constructed from 2004 to 2008. (Reproduced with permission from IHT/NYT Syndicate).

© 2015 John Wiley & Sons Ltd

surgery and intensive care for trauma patients [7]. CT scan is not available. Seriously injured patients, especially with head trauma, need to be transferred by plane to the capital Vientiane. Around 20 000 outpatients and 3300 inpatients are treated per year [19]. Study procedure First, a baseline survey of all RTI was conducted from December 2007 to December 2008. Second, the survey was repeated until 2011 each year from December to January as the two months with previously average RTI numbers. Third, an additional semi-quantitative survey was conducted to identify reasons and characteristics for not wearing helmets. A different time period (March–October 2009) was purposively chosen including months with maximum variation of RTI numbers. All patients were eligible if presenting with injuries from road traffic accidents. Patients were included if informed consent was given. In case of inability to understand or answer the questionnaire, next of kin were interviewed if consenting. In case of death on the road/during transportation, information was collected from witnesses and accompanying persons. The additional helmet survey was performed only among motorcyclists over 12 years of age who were able to answer and consented. A structured questionnaire was adapted from surveys by Handicap International in Vientiane [9–11]. Questions included characteristics of patients (age, sex, district, occupation, alcohol and drug consumption, insurance, driving license), patient transfer to the hospital, type of road user (RU), time and mode of accident, characteristics of accident (vehicle, vehicle’s light, road and weather condition, helmet use, police investigation) and injury characteristics. A simplified pretested anonymous questionnaire was used for the helmet survey. It included accident characteristics, patient’s characteristics (age, occupation, alcohol consumption) and helmet use. Main outcome was investigated by the open question: ‘Why did you not wear a helmet?’ Interviews were performed by trained nurses at the emergency department and ICU in the patient’s room to ensure confidentiality. Most answers were to be ticked, including injury site and severity on a simplified body pictogram. Four categories of severity of injury were defined. (i) superficial injury (minor wound(s)/skin abrasion); (ii) moderate injury (wound that needed suturing/minor surgery, e.g. lacerations); (iii) severe injury (fracture, injury that needed major surgery, e.g. splenectomy, or ICU treatment, e.g. chest drain for hematopneumothorax, intracerebral bleeding); and (iv) death. Vulnerable road 1579

Tropical Medicine and International Health

volume 20 no 11 pp 1578–1587 november 2015

G. Slesak et al. Road traffic injuries in northern Laos

Table 1 Patient characteristics

Time period Number Median age (range*) Male Main occupation Student/pupil Farmer Government employee Worker Vendor Police Soldier Driver Other None/unknown Referred from Accident site Other hospital Home Other Type of RU/vehicle Pedestrian Bicycle Motorbike Tractor Pickup/car (Mini)Van Bus Truck Injury severity No visible injury Superficial Moderate Severe Death

Total (17 months) (%)

December 2007–January 2008 (2 months) (%)

December 2009–January 2010 (2 months) (%)

December 2010–January 2011 (2 months) (%)

1074 22 (1–88) 734 (68.4)

141 22 (2–59) 94 (66.7)

107 23.5 (3–67) 72 (67.3)

183 22 (3–75) 122 (66.7)

380 (35.4) 300 (28.0) 136 (12.7)

53 (37.6) 40 (28.4) 12 (8.5)

33 (30.8) 28 (26.2) 18 (16.8)

73 (39.9) 46 (25.1) 20 (10.9)

123 44 21 18 8 28 15

11 6 4 3 0 4 8

15 3 1 3 0 6 0

22 6 5 3 2 4 2

(11.5) (4.1) (2.0) (1.7) (0.7) (2.6) (1.4)

(7.8) (4.3) (2.8) (2.1) (2.8) (5.7)

(14.0) (2.8) (0.9) (2.8) (5.6)

(12.0) (3.3) (2.7) (1.6) (1.1) (2.2) (1.1)

903 (84.2) 105 (9.8)

117 (83.0) 20 (14.2)

89 (84.0) 10 (9.4)

137 (74.9) 21 (11.5)

63 (5.9) 2 (0.2)

2 (2.1) 1 (0.7)

7 (6.6) 0

25 (13.7) 0

66 43 818 65 23 28 20 10

(6.1) (4.0) (76.2) (6.1) (2.1) (2.6) (1.9) (0.9)

5 6 95 24 1 9 0 1

(3.5) (4.3) (67.4) (17.0) (0.7) (6.4)

(5.6) (0.9) (81.3) (11.2)

(0.7)

6 1 87 12 0 0 0 0

97 560 308 85 17

(9.1) (52.5) (28.9) (8.0) (1.6)

2 74 50 11 2

(1.4) (53.2) (36.0) (7.9) (1.4)

12 47 32 11 4

(11.3) (44.3) (30.2) (10.4) (3.8)

13 9 140 10 5 2 0 4

(7.1) (4.9) (76.5) (5.5) (2.7) (1.1)

11 118 41 9 4

(6.0) (64.5) (22.4) (4.9) (2.2)

(2.2)

*Two-thirds (66.6%) of patients were between 15 and 34 years of age.

sex (OR 2.25, 95% CI 1.38–3.66) and no helmet use (OR 1.99, 95% CI 1.16–3.41). Not wearing a helmet was associated with young age, time period of survey, pillion riding, alcohol consumption and lack of driver license (Table 3).

14–65). Main reasons for not wearing a helmet were not possessing a helmet and being pillion passenger (Figure 3). Factors associated in multivariate analysis with not wearing a helmet were pillion riding (OR 7.74, 95% CI 3.49–17.2) and alcohol consumption (OR 2.08, 95% CI 1.09–3.97).

Reasons for driving without a helmet A total of 201 patients could be included in the additional survey. Ten patients were excluded because of death (3), unconsciousness (1), no consent (2), no answer regarding helmet use (2) and loss of data (2). Most patients were male (71.6%) and young (median age 22 years, range

© 2015 John Wiley & Sons Ltd

Combination of hospital and police data Of 1074 patients, 28% (299) reported police investigations. This proportion was highest in the last survey period (39% vs. previous surveys 27%, OR 1.74, 95% CI 1.24–2.44). RTI with bigger vehicles had the highest 1581

Tropical Medicine and International Health

volume 20 no 11 pp 1578–1587 november 2015

G. Slesak et al. Road traffic injuries in northern Laos

Table 1 Patient characteristics

Time period Number Median age (range*) Male Main occupation Student/pupil Farmer Government employee Worker Vendor Police Soldier Driver Other None/unknown Referred from Accident site Other hospital Home Other Type of RU/vehicle Pedestrian Bicycle Motorbike Tractor Pickup/car (Mini)Van Bus Truck Injury severity No visible injury Superficial Moderate Severe Death

Total (17 months) (%)

December 2007–January 2008 (2 months) (%)

December 2009–January 2010 (2 months) (%)

December 2010–January 2011 (2 months) (%)

1074 22 (1–88) 734 (68.4)

141 22 (2–59) 94 (66.7)

107 23.5 (3–67) 72 (67.3)

183 22 (3–75) 122 (66.7)

380 (35.4) 300 (28.0) 136 (12.7)

53 (37.6) 40 (28.4) 12 (8.5)

33 (30.8) 28 (26.2) 18 (16.8)

73 (39.9) 46 (25.1) 20 (10.9)

123 44 21 18 8 28 15

11 6 4 3 0 4 8

15 3 1 3 0 6 0

22 6 5 3 2 4 2

(11.5) (4.1) (2.0) (1.7) (0.7) (2.6) (1.4)

(7.8) (4.3) (2.8) (2.1) (2.8) (5.7)

(14.0) (2.8) (0.9) (2.8) (5.6)

(12.0) (3.3) (2.7) (1.6) (1.1) (2.2) (1.1)

903 (84.2) 105 (9.8)

117 (83.0) 20 (14.2)

89 (84.0) 10 (9.4)

137 (74.9) 21 (11.5)

63 (5.9) 2 (0.2)

2 (2.1) 1 (0.7)

7 (6.6) 0

25 (13.7) 0

66 43 818 65 23 28 20 10

(6.1) (4.0) (76.2) (6.1) (2.1) (2.6) (1.9) (0.9)

5 6 95 24 1 9 0 1

(3.5) (4.3) (67.4) (17.0) (0.7) (6.4)

(5.6) (0.9) (81.3) (11.2)

(0.7)

6 1 87 12 0 0 0 0

97 560 308 85 17

(9.1) (52.5) (28.9) (8.0) (1.6)

2 74 50 11 2

(1.4) (53.2) (36.0) (7.9) (1.4)

12 47 32 11 4

(11.3) (44.3) (30.2) (10.4) (3.8)

13 9 140 10 5 2 0 4

(7.1) (4.9) (76.5) (5.5) (2.7) (1.1)

11 118 41 9 4

(6.0) (64.5) (22.4) (4.9) (2.2)

(2.2)

*Two-thirds (66.6%) of patients were between 15 and 34 years of age.

sex (OR 2.25, 95% CI 1.38–3.66) and no helmet use (OR 1.99, 95% CI 1.16–3.41). Not wearing a helmet was associated with young age, time period of survey, pillion riding, alcohol consumption and lack of driver license (Table 3).

14–65). Main reasons for not wearing a helmet were not possessing a helmet and being pillion passenger (Figure 3). Factors associated in multivariate analysis with not wearing a helmet were pillion riding (OR 7.74, 95% CI 3.49–17.2) and alcohol consumption (OR 2.08, 95% CI 1.09–3.97).

Reasons for driving without a helmet A total of 201 patients could be included in the additional survey. Ten patients were excluded because of death (3), unconsciousness (1), no consent (2), no answer regarding helmet use (2) and loss of data (2). Most patients were male (71.6%) and young (median age 22 years, range

© 2015 John Wiley & Sons Ltd

Combination of hospital and police data Of 1074 patients, 28% (299) reported police investigations. This proportion was highest in the last survey period (39% vs. previous surveys 27%, OR 1.74, 95% CI 1.24–2.44). RTI with bigger vehicles had the highest 1581

Tropical Medicine and International Health

volume 20 no 11 pp 1578–1587 november 2015

G. Slesak et al. Road traffic injuries in northern Laos

24% 22%

Total

20%

12/2007-1/2008

18%

12/2009-1/2010

16%

12/2010-1/2011

14%

12.8% 12.7%

12% 7.8%

8%

6.4%

6%

4.9%

4.2%

4% 2%

10.0%

9.1%

10%

2.8% 2.6%

2.3% 0.8%

0.8%

3.8% 3.5%

4.2% 4.2%

3.5%

2.1%

0.3% 0.0% 0.2% 0.4% 0.5%

23:00-23:59

22:00-22:59

21:00-21:59

20:00-20:59

19:00-19:59

18:00-18:59

17:00-17:59

16:00-16:59

15:00-15:59

14:00-14:59

13:00-13:59

12:00-12:59

11:00-11:59

10:00-10:59

9:00-9:59

8:00-8:59

7:00-7:59

6:00-6:59

5:00-5:59

4:00-4:59

3:00-3:59

2:00-2:59

1:00-1:59

0:00-0:59

0%

Figure 2 Road traffic injuries stratified by time of the day 2007–2011 (n = 1072). Table 2 Factors associated with serious injuries (= severe injuries and death) in motorbike riders in univariate and multivariate analysis Univariate analysis

Number Median age (range) Age group* 1–15 years 16–17 years 18–19 years 20–21 years 22–23 years 24–27 years 28–33 years 34–42 years >42 years Male* Driver* Driver license*

Serious (%)

Not serious (%)

69 (8.4) 25 (4–60)

749 (91.6) 22 (1–80)

11 3 8 4 5 9 8 9 11 59 58 9

81 80 103 71 84 86 86 84 73 524 525 65

(16.2) (4.4) (11.8) (5.9) (7.4) (13.2) (11.8) (13.2) (16.2) (85.5) (84.0) (15.5)

(10.8) (10.7) (13.8) (9.5) (11.2) (11.5) (11.5) (11.2) (9.8) (70.1) (70.3) (9.4)

OR (95% CI)

Multivariate analysis P

OR (95% CI)

P

1 0.20 0.41 0.26 0.30 0.53 0.46 0.51 0.75 2.23 2.11

0.017 0.079 0.032 0.035 0.202 0.128 0.17 0.54 0.032 0.037

0.11 0.36

2.52 (1.27–5.02) 2.25 (1.16–4.37) 1.76 (0.83–3.75)

0.007 0.014 0.14

(0.05–0.75) (0.15–1.11) (0.08–0.89) (0.10–0.92) (0.20–1.40) (0.17–1.25) (0.19–1.34) (0.29–1.90) (1.07–4.63) (1.05–4.27)

*Variables with P < 0.2 and age group were included in multivariate analysis. (Note: inclusion of all variables resulted in the same model). Variables with P ≥ 0.2 in univariate analysis were no helmet use, alcohol consumption, poor weather, time of the day and time period.

rates of police investigation (trucks 5/10 and buses 19/ 20). Police reported 340 RTI, including 10 deaths in 2008. Hospital’s numbers were twice as high (695). Capture–recapture analysis estimated 1379 RTI (95% CI 1188–1570). Deaths were estimated at 17 in 2008 (95% CI 7.6–27). RTI mortality rate for the whole province was calculated as 11.6 per 100.000 population (95% CI 5.1–18.1 per 100.000). 1582

Discussion The study reveals a much higher burden of RTI than officially reported. Most RTI occurred among younger, male and VRU, especially motorcyclists. RTI frequently happened during the evening and Lao New Year. Determinants for serious injuries in motorcyclists were being young, driver and male. Risk factors for more serious

© 2015 John Wiley & Sons Ltd

Tropical Medicine and International Health

volume 20 no 11 pp 1578–1587 november 2015

G. Slesak et al. Road traffic injuries in northern Laos

Table 3 Factors associated with no helmet use Univariate analysis

Number Median age (range) Age group* 1–15 years 16–17 years 18–19 years 20–21 years 22–23 years 24–27 years 28–33 years 34–42 years >42 years Male* Pillion rider* No driver license* Alcohol* Time period* December 2007–December 2008 December 2009–January 2010 December 2010–January 2011 Occupation* Student/pupil Farmer Government employee Worker Vendor Police Soldier Other Unknown

No helmet (%)

Helmet (%)

645 (81.2) 22 (1–80)

149 (18.8) 28 (5–61)

89 73 100 57 67 74 65 59 60 453 212 557 310

3 9 10 17 18 16 27 29 20 112 20 109 61

(13.8) (11.3) (15.5) (8.9) (10.4) (11.5) (10.1) (9.2) (9.3) (70.2) (32.9) (92.5) (48.4)

472 (73.2) 73 (11.3) 100 (15.5) 263 166 66 89 23 12 12 11 3

(40.8) (25.7) (10.2) (13.8) (3.6) (1.9) (1.9) (1.7) (0.5)

(2.0) (6.0) (6.7) (11.4) (12.1) (10.7) (18.1) (19.5) (13.4) (75.2) (13.4) (79.6) (40.9)

100 (67.1) 10 (6.7) 39 (26.2) 33 37 45 9 8 7 2 6 2

(22.1) (24.8) (30.2) (6.0) (5.4) (4.7) (1.3) (4.0) (1.3)

OR (95% CI)

Multivariate analysis P

OR (95% CI)

P

1 0.35 0.44 0.12 0.16 0.20 0.10 0.09 0.12

(0.09–1.37) (0.11–1.75) (0.03–0.44) (0.04–0.59) (0.05–0.74) (0.03–0.35) (0.03–0.32) (0.03–0.44)

0.13 0.25 0.001 0.006 0.016

Road traffic injuries in northern Laos: trends and risk factors of an underreported public health problem.

Road traffic injuries (RTI) have become a leading cause for admissions at Luang Namtha Provincial Hospital (LNPH) in rapidly developing northern Laos...
233KB Sizes 3 Downloads 11 Views