- Email: [email protected]
Catastrophic household costs due to injury in Vietnam
Injury, Int. J. Care Injured 44 (2013) 684–690
Contents lists available at SciVerse ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Catastrophic household costs due to injury in Vietnam Ha Nguyen a,b, Rebecca Ivers a,*, Stephen Jan a, Alexandra Martiniuk a,c, Cuong Pham b a
The George Institute for Global Health, University of Sydney, Australia The Center for Injury Policy and Prevention Research, Hanoi School of Public Health, Viet Nam c Dalla Lana School of Public Health, University of Toronto, Canada b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 5 May 2012
Objective: Little is known about the costs of injury and their impact on injured persons and their families in Vietnam. This study aimed to examine the cost of injury in hospitalised patients and to identify the most costly injuries and those more likely to result in catastrophic household expenditure. Method: A prospective cohort study was conducted, recruiting individuals admitted to Thai Binh General Hospital due to injury in Vietnam from January to August 2010. During the hospitalisation period, data on expenditure including direct medical, direct non-medical and indirect costs were collected. Demographic and injury characteristics were also obtained. The associations between the risk of catastrophic expenditure and injury cause, severity and principal injured region were examined by modified Poisson regression approach. Payment of more than 40% of the household non-subsistence spending was considered a catastrophic expenditure. Results: Of 918 patients approached, 892 (97%) were recruited. Total costs for all participants during the hospitalisation period were US$ 325,812. Patients admitted for road injury accounted for the largest number of injuries (n = 477, 53%), and the largest percentage of the total costs (US$ 175,044, 57%). This was followed by individuals hospitalised due to falls, representing 29% of the sample (n = 261) and 31% of the total costs (US$ 103,128). In terms of cost per hospital stay, burn injuries were the most costly (US$ 427), followed by falls (US$ 395) and road crashes (US$ 367). Of all sample, 26% experienced catastrophic expenditure due to their injuries. Factors significantly associated with increased risk of catastrophic expenditure were having more severe or higher MAIS injuries (RR = 2.02, 95% CI: 1.14–3.57), principal injured region to lower extremities (RR = 3.34, 95% CI: 1.41–7.91) or head (RR = 3.21, 95% CI: 1.37–7.52), longer hospital stay (RR = 1.09, 95% CI: 1.07–1.10), older age, lower income and not having insurance (RR = 1.63, 95% CI: 1.21–2.21). Conclusion: A high proportion of households experienced catastrophic expenditure following injury, highlighting the important need for programmes to prevent injuries, road traffic and fall-related injuries in particular. Furthermore, expansion of health insurance coverage may help individuals cope with the financial consequences of injury. ß 2012 Elsevier Ltd. All rights reserved.
Keywords: Cost of injuries Economic burden Cost of hospitalisation Catastrophic expenditure Vietnam
Introduction Injury is a serious public health issue in Vietnam, the thirteenth most populous country in the world with a population of 85.8 million.1 Every day in Vietnam, injuries claim almost 100 lives, or about 35,000 lives per year.2,3 It has been estimated that the number of potential life years lost due to injuries in Vietnam is more than double those due to non-communicable diseases, and more than six times higher than those due to communicable diseases.4 In addition to fatalities, injuries are also the cause of
* Corresponding author at: The George Institute for Global Health, PO Box M201, Missenden Road, NSW Australia 2050. E-mail address: [email protected] (R. Ivers). 0020–1383/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.injury.2012.05.006
hundreds of thousands of hospital admissions.4 Because the largest proportion of injuries occur among people under 60 years of age5 who comprise the majority of the labour force, injuries may have a serious economic impact at the country level. The impact not only results from the loss of productivity, but also from significant expenses for medical treatment, rehabilitation and recovery.6 Injuries can potentially lead to catastrophic financial losses to injured persons and their families, leading to a substantial risk of impoverishment. The idea of financial catastrophe comes from an ethical position that no one ought to spend more than a given of fraction of their income on health care.7 In Vietnam, health care costs are paid directly from the income of patients and their families. According to Ministry of Health estimates, health expenditure in Vietnam consists of private out-of-pocket payments (67%), public sources including from central government
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
budgets (6%), provincial government budgets (10%), official development assistance funds (2%), and social insurance and primary health insurance (9%).8 During 2001–2006, despite fluctuations, the average household expenditure on health contributed more than 60% of the total health expenditure in the country.8 In a study comparing 59 countries, Xu et al. found that Vietnam had the highest proportion of households facing catastrophic payments for health. Specifically, more than 10.5% of households had health expenditures exceeding 40% of the household capacity to pay.9 In a study on costs of traumatic brain injury, Hoang et al. found that 84% of the sample, which included 35 households, faced catastrophic expenditure.10 By converting the impact of injuries into monetary terms, cost of injury studies can provide a common language for policy and decision makers.11 Cost of injury studies have been widely conducted, particularly in high or middle income country settings. For instance, Meerding et al. reported total health care costs due to injuries in the Netherlands in 1999 were s 1.2 billion or 3.7% of the total health care budget.12 In Australia, a snap shot of injuries presented by the Australian Bureau of Statistics show that, resources spent on injured people during 2000–2001 were 8% of total allocated health expenditure, almost AU$ 4 billion.13 In China, Zhou et al. estimated the economic cost of injury in 1999 at US$ 12.5 billion, almost four times the total public health services budget of China.6 More recently in South Korea, estimated medical treatment costs for injuries in 2006 accounted for 9.5% of the national health expenditure, and the annual economic burden associated with injury was more than US$ 39.8 billion.14 A recent World Health Organisation Global Status Report on Road Safety also highlighted a need for comprehensive cost estimates of injuries in addition to cost-effectiveness of various interventions to inform preventive action and mobilise support.15 In Vietnam, very little is known about the costs of injury and their impact on injured persons and their families. Existing knowledge on costs of injury in Vietnam comes from work conducted by Thanh et al. in 2000 and Hoang et al. in 2008. However, the first study was limited in identification of injury cause, which was self-reported by respondents,16 and he later study had a small sample size,10 limiting the conclusions that can be drawn from these works. The present study is an effort to fill in knowledge gaps by estimating the costs of injuries to the injured persons and their families during the hospitalisation period. These will be examined in greater detail in terms of injury characteristics, patient demographics, insurance status, external causes of injury, severity, and body region injured. This study also aims to identify the most costly injuries and those most likely to result in a catastrophic expenditure for the household. Methods A prospective cohort study was used to examine the economic burden of injuries. The study was conducted in Thai Binh General Hospital, the largest trauma hospital in Thai Binh province, with 440 beds. The Thai Binh province is in the Red River delta, approximately 100 km south of Hanoi, the capital city of Vietnam. In 2009, the population of this province was 1,900,000.17 According to the National Household Living Standards survey in 2010, the average monthly per capita income in Thai Binh province was VND 1,129,300, equivalent to US$ 57.90.18 There are nine hospitals at the provincial level and twelve at the district level in the province. As the largest trauma hospital at the provincial level, the Thai Binh General Hospital receives the majority of trauma patients in the province either directly or indirectly – those transferring from lower level hospitals or from same level hospitals in the province. Study participants were those individuals who were admitted to Thai Binh General Hospital due to an injury. An injury is defined
685
as physical damage that results when a human body is suddenly or briefly subjected to intolerable levels of energy. An injury may result from acute exposure to energy in amounts that exceed the threshold of physiological tolerance, or an impairment of function resulting from a lack of one or more vital elements (i.e. air, water, warmth), such as in drowning, strangulation or freezing.19 Additional inclusion criteria also included inpatient hospital treatment for at least one day, age of 18 years or older, current residential address within Thai Binh province area and consent to participate in the study. Fatal cases were not included. In Vietnam, there is a traditional belief that a person should die at home20 and therefore those patients not expected to survive are often taken home by their families. Participants were recruited from 01 January 2010 to 31 August 2010 by trained research assistants and doctors in the hospital. After critical treatment and hospital admission, 918 injured persons meeting inclusion criteria were approached; 892 (97%) consented to participate into the study. The International Classification of Diseases 10th revision (ICD-10) was used to code injuries into categories of external causes prior to reporting the results. Diagnoses also included the principal injured body region, which is the most severely injured region, and injury severity measured by the abbreviated injury score (AIS), which is an anatomical scoring system representing the ‘threat to life’ associated with an injury. The numerical ranking of the AIS ranges from 1 to 6: 1 (minor injury), 2 (moderate), 3 (serious), 4 (severe, life threatening), 5 (critical, survival uncertain) to 6 (un-survivable).21 The most severe injury was scored using the AIS during data collection, and thus the measure of severity used is the maximum abbreviated injury score (MAIS).21 Demographic information (e.g. age, gender, occupation) and injury context (e.g. place, time of injury, specific external cause) were collected after hospital admission by face to face interview. Questions on injury context were adapted from the World Health Organisation Guideline for conducting community surveys on injuries and violence.19 All data collection took place in the hospital by hospital nurses trained as research assistants in this study. Costing method The economic impact of injury was measured in terms of outof-pocket costs from the individual and family perspective. All costs incurred by participants and their household members associated with treatment during hospitalisation were reported. They were categorised into direct medical, direct non-medical and indirect costs.22 Cost data were also collected by means of face to face interview with all 892 injured persons and their caretakers the day before their hospital discharge. These include direct and indirect costs. Direct costs included expenditure associated with treatment and care for the injured person. Taking a similar approach to other studies on cost of injuries,12,23–26 cost items associated with treatment included during hospitalisation were emergency service, surgery or treatment, paramedical or diagnostic examination tests (such as X ray, CT scan), medication (prescribed and overthe-counter drugs), equipment (wheelchair, splint) and rehabilitation in the hospital. In addition to items directly associated with treatment for the injured person, information on non-medical costs incurred by the injured person and their relatives including transportation to the hospital, accommodation and meals were also collected. Indirect costs refer to lost productivity because of injury treatment and recovery.10,12,23,25,26 Using the human capital approach from the individual and family perspective, the indirect costs during the hospitalisation period were estimated by the product of the total days off work over this period and the average
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
686
daily income of the injured person and their relatives. For those without a formal income, such as students, unpaid homemakers or the elderly, opportunity costs were calculated using the average per capita income16,27 for Thai Binh province, which was the average monthly per capita income estimated in the National Household Living Standards survey in 2010. Catastrophic expenditure The out-of-pocket costs required for medical treatment are paid for from the resources available to the household. When these are large relative to available resources, they can lead to financial hardship and impair the living standard of the household.28 To date, there is not complete consensus regarding the specific fraction for defining catastrophic expenditure.9,28 In this study, similar to Hoang et al.’s study on the costs of traumatic brain injury due to motorcycle accidents in Vietnam,10 a threshold of 40% was used. This threshold has also been used in studies analysing the catastrophic health expenditure across multiple countries.9,10,28–30 In our study, participants and their households were classified as experiencing a catastrophic expenditure if the treatment costs accounted for more than 40% of the household’s capacity to pay, which is the total income of household subtracting the expenditure on food.10 Catastrophic expenditure if: Treatment costs > 40% o f ½Income o f household expenditure on food Data analysis Descriptive statistics were used to describe study participants including age, gender, occupation, and external cause of injuries. Costs of injury by external cause were computed in terms of mean, median and interquartile range to present the distribution of data. To explore predictors of risk of catastrophic expenditure, we used a modified Poisson regression approach estimating relative risk. Modified Poisson regression is a Poisson regression applied to binomial data using a robust error variance. This approach estimates relative risk consistently and efficiently.31,32 In our multivariable model, all variables hypothesised a priori to potentially affect the risk of catastrophic expenditure were included: external cause, injury severity, principal injured region, length of stay, gender, age group, occupation and insurance status. Collinearity was assessed by examining scatter plots, correlation matrix and variance inflation factors. Plausible interactions examined were those between length of hospital stay (LOS) with external cause, injury severity, principal body region injured, gender, age, insurance status and between age group with insurance status. The backward stepwise approach was then employed to remove non-significant interactions in the full models. For all analyses, a p-value of less than 0.05 was considered as statistically significant. All analyses were performed using STATA statistical package version 11 (Stata Corporation, College Station, TX, USA). The ethics application for the study was reviewed and approved by the Human Ethics Committee in University of Sydney, Australia, the Hanoi School of Public Health Ethics Committee and Thai Binh General Hospital in Vietnam. Results Table 1 summarises the demographic characteristics of the study sample. Younger (18–29 years) and older (60 years plus) participants accounted for the majority of the sample (27.8% and
Table 1 Demographic characteristics of participants.
Total Age 18–29 30–39 40–49 50–59 60+ Gender Male Female Occupation White-collar worker Blue-collar worker Farmer Retiree/unemployed Insurance Yes No
Frequency
%
Average monthly income (US$)
892
100.0
73.9
248 134 162 133 215
27.8 15.0 18.2 14.9 24.1
80.7 81.8 70.4 74.7 63.3
634 258
71.1 28.9
76.5 67.4
74 139 464 215
8.3 15.6 52.0 24.1
118.3 98.1 58.7 75.8
224 668
25.1 74.9
70.8 74.9
24.1% respectively). Males made up a disproportionate share with more than twice the number of females, 634 males (71.1%) in comparison with 258 females (28.9%). In terms of occupation, more than 50% were farmers. This was more than twice the percentage of the second most common group (24.1%), which included those without a formal occupation or those already retired. In the study sample, only 25.1% were covered by some sort of insurance. In terms of economic status, the average income of participants was approximately US$ 74 per month. Farmers had the lowest reported income (US$ 58.7) and white-collar worker had the highest income (US$ 118.3). Injury characteristics are presented in Table 2. The predominant causes were road crashes (477/892 or 53%) and falls (261/892 or 29%). More than 80% of participants were ranked at a MAIS of 2 and 3, or moderate and serious injuries respectively. More severe injury or MAIS of more than 3 was highest among participants hospitalised due to road crashes, falls and blunt object injuries compared to other injury types. In terms of principal injured regions, injuries to the head accounted for the highest proportion of principal injured regions in the sample. This was followed by injury to the hip and lower extremity region. The body region least impacted by injury was the spinal region, where there were 41 cases accounting for 4.6% of the sample. Among road traffic injuries (RTIs), the most common principal injured region was the head (43.4%). Among falls, however, the most common principal injured regions were hip and lower extremity (44.4%). Regarding length of hospital stay, persons sustaining burn had the longest average hospital stay (10 days) and those injured by a sharp object stayed in the hospital for 6 days on average. Those hospitalised due to falls or road crashes were on average hospitalised for 8.5 days and 7.7 days respectively. Table 3 describes total and average costs of injury by external causes. For the 892 hospitalised injury patients recruited, the total costs during the hospitalisation period were US$ 325,812. This is equivalent to approximately 3929 average person-month salaries (US$ 73.9 per person per month, see Table 1). Costs incurred by road crash victims accounted for the largest proportion of the total costs, more than 57% or with an amount of more than US$ 275,000, equivalent to almost 2115 average person-month salaries. This is followed by the costs incurred by participants hospitalised due to falls, which represent 31% of the total costs. In terms of mean costs per case, burns cost the most to the participants and their families, averaging US$ 321 on medical costs, almost US$ 50 on non-medical costs and approximately US$ 57 on indirect costs. Following burns were road crashes and falls,
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
687
Table 2 Injury characteristics.
Total MAIS 1 2 3 4 5 Principal injured region Head Face Spine Thorax/abdomen Upper extremity Hip/lower extremity Length of hospital stay
RTI (V01–V89)
Fall (W00–W19)
Burn (X00–X19)
Sharp object (W25– W27)
Blunt object (W20– W22)
Assault (X85–Y09)
Other injuries
n
n
n
n
n
n
n
%
477
%
261
%
22
%
32
%
33
%
43
All injuries
%
24
n
%
892
33 203 184 38 19
6.9 42.6 38.6 8.0 4.0
15 96 126 19 5
5.7 36.8 48.3 7.3 1.9
4 11 6 1 0
18.2 50.0 27.3 4.5 .0
2 16 13 1 0
6.3 50.0 40.6 3.1 .0
6 9 14 4 0
18.2 27.3 42.4 12.1 .0
9 14 17 1 2
20.9 32.6 39.5 2.3 4.7
6 9 9 0 0
25.0 37.5 37.5 .0 .0
75 358 369 64 26
8.4 40.1 41.4 7.2 2.9
207 50 12 35 58 115
43.4 10.5 2.5 7.3 12.2 24.1
73 5 26 17 24 116
28.0 1.9 10.0 6.5 9.2 44.4
0 6 0 3 6 7
.0 27.3 .0 13.6 27.3 31.8
2 2 0 0 23 5
6.3 6.3 .0 .0 71.9 15.6
9 4 3 1 11 5
27.3 12.1 9.1 3.0 33.3 15.2
10 13 0 2 17 1
23.3 30.2 .0 4.7 39.5 2.3
2 2 0 3 15 2
8.3 8.3 .0 12.5 62.5 8.3
303 82 41 61 154 251
34.0 9.2 4.6 6.8 17.3 28.1
7.7 days
8.5 days
10.0 days
which incurred total cost of US$ 367 and US$ 395 respectively for the injured persons and their families. Assaults or violence were the least costly with average cost per case just under US$ 254. Of the total costs per case during the hospitalisation period, direct medical costs accounted for the largest share, US$ 270.6 or more than 74% on average. The share of direct medical costs was highest among individuals with sharp object related injuries, US$ 229.8 or 78.7%. On the other hand, this share was lowest among assault related injuries, US$ 166.9 or 65.8%. Regarding other components, direct non-medical costs and indirect costs on average represented more than 12.8% of the total costs (US$ 46.6) and 13.2% (US$ 48.1) respectively.
6.0 days
7.6 days
6.3 days
7.5 days
7.9 days
Overall, 27% of participants had catastrophic expenditure due to their injury (Fig. 1). That is, 27% of participants spent more than 40% of their household capacity to pay for treatment of their injuries. Catastrophic household expenditure was most common among participants hospitalised due to falls, burns, sharp object related injuries and road crashes. Those with lowest proportions of catastrophic payment were victims of assaults and other injuries. Results of univariable and multivariable regression analyses examining predictors of the risk of catastrophic expense due to injury care and treatment are shown in Table 4. In terms of injury characteristics, univariable analyses show that the risk of catastrophic expense was statistically significantly associated
Table 3 Costs of injury (US$) per case (mean, median, interquartile range (IQR)) and all cases by external causes. External causes
Per case
All cases
Direct costs
Indirect costs
Total
Direct costs
Indirect costs
Total
Medical
Non-medical
Medical
Non-medical
RTI (V02–V89)
Mean Median IQR
270.2 186.2 262.4
46.0 30.8 46.2
50.7 32.5 46.1
367.0 288.7 331.4
128,890
21,962
24,192
175,044
Fall (W00–W19)
Mean Median IQR
297.7 208.2 284.1
50.7 34.4 48.2
46.7 33.1 33.3
395.1 291.2 330.9
77,710
13,235
12,183
103,128
Burn (X00–X19)
Mean Median IQR
321.2 179.7 381.2
48.8 51.3 52.6
56.7 51.6 69.4
426.7 283.7 372.0
7066
1074
1248
9388
Sharp object (W25–W27)
Mean Median IQR
229.8 204.2 143.2
31.5 24.4 30.8
30.7 22.2 24.0
291.9 252.1 176.5
7353
1008
981
9342
Blunt object (W20–W22)
Mean Median IQR
224.1 187.9 174.7
36.3 20.5 38.5
43.4 35.4 26.7
303.7 269.3 236.0
7394
1197
1432
10,023
Assault (X85–Y09)
Mean Median IQR
166.9 165.1 132.1
43.1 30.8 33.8
43.6 34.8 16.2
253.6 235.9 109.4
7175
1854
1876
10,905
Others RTI (V02–V89)
Mean Median IQR
240.4 166.9 203.8
51.2 30.8 41.0
40.7 33.8 37.4
332.3 212.3 223.0
5770
1229
977
7976
All injuries
Mean Median IQR
270.6 193.3 246.0
46.6 30.8 46.2
48.1 33.3 38.4
365.3 275.2 308.1
241,358
41,559
42,894
325,812
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
688
40% 33.0% 30%
31.8%
31.3%
29.2%
26.4%
27.5%
21.2% 20%
10%
4.7%
0% RTI*
Burn
Fall
Sharp object
Blunt object
Assault
Others
All
Fig. 1. Catastrophic expenditure. *RTI, road traffic injury.
with fall external cause of injury, higher MAIS, principal injured region to the spine/thorax/abdomen, extremities and head and longer hospital stay. For instance, the risk of catastrophic expenditure among individuals injured due to falls was 1.54 (95% CI: 1.09–2.18) times higher than those injured due to causes other than fall or RTI. Each additional day staying in hospital was associated with 9% increased in the risk of catastrophic expenditure. Regarding demographic characteristics, older individuals were at statistically significantly higher risk of catastrophic expenditure while those without insurance were at statistically significantly higher risk. Specifically, in comparison with injuries Table 4 Modified Poisson regression of predictors of the risk of having catastrophic expenditure. Unadjusted
Adjusteda
RR
95% CI
RR
95% CIb
1.00 1.54 1.23
1.09–2.18 0.90–1.73
1.00 1.03 0.95
0.71–1.33 0.68–1.50
1.00 2.05 2.62 2.99 2.88
1.08–3.88 1.39–4.93 1.49–6.00 1.28–6.48
1.00 2.11 2.30 2.27 1.86
1.19–3.73 1.31–4.03 1.25–4.11 0.73–4.75
1.00 2.89 4.05 6.21 4.82
1.12–7.47 1.66–9.89 2.61–14.74 2.02–11.47
1.00 2.33 2.94 3.34 3.21
0.91–5.98 1.24–6.93 1.41–7.91 1.37–7.52
Length of hospital stay (days) Gender Male Female Age 18–39 40–59 60+ Occupation White-collar worker Blue-collar worker Farmer Retiree/unemployed Insurance Yes No
1.09
1.07–1.11
1.09
1.07–1.10
1.00 1.31
1.05–1.63
1.00 1.08
0.88–1.34
1.00 1.46 1.73
1.12–1.90 1.33–2.26
1.00 1.09 1.27
0.84–1.41 0.94–1.71
1.00 0.74 1.60 1.25
0.35–1.53 0.85–3.00 0.65–2.40
1.00 0.47 0.59 0.49
0.23–0.96 0.31–1.14 0.25–0.96
1.00 1.58
1.17–2.12
1.00 1.63
1.21–2.21
Monthly incomec
0.89
0.87–0.92
0.90
0.88–0.93
Predictors
External causes All other injuries combined Fall (W00–W19) RTI (V01–V89) MAIS 1 2 3 4 5 Principal injured region Face Spine/thorax/abdomen Upper extremity Hip/lower extremity Head
a b c
Adjusting for all other variables listed. 95% CI: 95% confidence interval. The unit of monthly income was in US$ 10.
among individuals aged 18–39 years, the risk of catastrophic expenditure was 1.46 times higher (95% CI: 1.12–1.90) and 1.73 times higher (95% CI: 1.33–2.26) among individuals aged 40–59 years and 60 years plus respectively. Risk of catastrophic expenditure for those without insurance was 1.58 times (95% CI: 1.17–2.12) higher than those with insurance. In addition, those with higher income were also statistically significant at lower risk of catastrophic expenditure (RR = 0.89, 95% CI: 0.87–0.92). In the multivariable regression model, no interaction term was found to be statistically significant and increased risk of catastrophic expenditure was statistically associated with higher MAIS, principal injured region to extremities and head, longer hospital stay, older age, not having insurance and lower monthly income (Table 4). For example, when controlling for other variables in the model, an individual with MAIS of 2, i.e. moderate injury, had double the risk (RR = 2.02, 95% CI: 1.14–3.57) of experiencing a catastrophic expenditure than those with MAIS of 1 or minor injury. After controlling for other characteristics, injured persons without insurance were more likely (RR = 1.63, 95% CI: 1.21–2.21) to have catastrophic expenditure compared to those without insurance. Discussion This study investigated the cost of injuries for individuals admitted to a provincial hospital in Vietnam, Thai Binh province. On average, the total cost of an injury was US$ 365, which was equivalent to an economic loss of more than the average income for 6 months. We found that burns, falls and RTIs were the most costly injuries. This is partly due to a longer hospital stay for these injuries. In addition, more severe injuries were also associated with increased costs. In terms of mean costs per case, the most costly injury appeared to be burns. However, this was explained by the longest hospital stays for burns than for any other injuries. When examining the mean costs per case per day, burns were not as costly as falls or for RTIs. In addition, median costs show that costs were highest among falls and RTIs. With the largest number of cases, falls and RTIs accounted for the highest total economic cost of all injuries. Of US$ 325,812 incurred by 892 injuries during the 8-month period, more than 50% was accounted for by costs incurred by RTIs and more than 30% incurred by fall-related injuries. This finding of highest attributable costs due to RTIs and falls confirms previous findings from the US and the Netherlands.12,23 In this study, the largest component of total costs per case was found to be due to direct medical costs, in contrast to prior studies, where indirect costs were found to be the largest component.12,23,33 However, in this study costs were estimated only during the hospitalisation period, whereas prior studies also included costs incurred during the period beyond hospitalisation.
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
Our study provides the most recent estimates on the cost of injury in Vietnam. In 2001, Thanh et al. estimated that the average total costs were just US$ 13.70. These included costs for health care, transportation for care and treatment of the injury, and costs for patients and their relatives’ time.16 The increased costs potentially reflect the expansion of user fee policy, which has helped reduce hospitals’ dependence on limited government budgets, but resulted in increased out-of-pocket money from individuals, and may also have increased a financial incentive to keep patients in hospitals for longer. Consequently, it has led to a disproportionate burden on the individual and the family, especially disadvantaged people who do not have any insurance.8 In our study sample, only 25% were covered by insurance for their treatment costs. Other explanations for the higher estimates in our study are the methodology. By including only in-patient hospitalised injuries, our sample includes a larger proportion of more severe injuries than those included in Thanh’s study, which included both hospitalised and non-hospitalised injuries reported by participants.16 In addition, instead of collecting data at the injured persons’ home, we obtained cost information in the hospital, thus minimising recall bias. The large burden of injury costs and the current health financing mechanism often results in catastrophic expenditure for the household. In our sample, approximately 26% of participants faced this issue. The largest proportion of catastrophic expenses was incurred by participants injured due to falls and road crashes, likely due to the severity of injury sustained. Our estimates for catastrophic payments were understandably smaller than the 84% found in the Hoang et al. study on the costs of traumatic brain injuries related to motorcycle accidents in 200810; their estimates were costs over a one year period and only included traumatic brain injuries, which we found to result in the highest costs. The proportion of 26% of catastrophic health expenditure found in this study is however substantially higher than the proportion facing catastrophic payments due to all diseases combined among a representative sample in Vietnam, estimated at about 10.5%.9 The impact of catastrophic payments can be severe, impacting on a household’s ability to pay for education and/or housing and/or other non-subsistence items, and potentially contributing to a slide into poverty for families.28 According to the 2010 Vietnam Living Standard Survey, typical households spend 10% of the nonfood expenditure on health care.34 An increased health expenditure due to injury treatment (from 10% to 40% or more) would reduce resources available for other needs. These figures highlight the important need for injury prevention, particularly for road traffic and fall-related injuries. Enforcing traffic laws and occupational health and safety regulations may help to reduce these injuries. In addition, the coverage of health insurance as a mechanism for financial risk protection needs to expand, especially for farmers and blue-collar workers. As reported in 2008, the health insurance system covers approximately 50% of the total Vietnamese population including the formally employed, the poor and children under 6 and leaves the other 50% paying for their health care costs by out-of-pocket money.8 In fact, a study on health insurance in Vietnam demonstrated that, varying by levels of facility, the average amount of saving on out-of-pocket spending was 24% among those with compulsory health insurance and 15% among those with voluntary health insurance.35 This study adds substantially to a small number of studies that have described the burden of injury and the economic costs of injury in Vietnam. The first national injury survey with data on morbidity and mortality highlighted the importance of injury as the leading cause of years of potential life lost in Vietnam.4 This national survey was large in terms of scale, but was limited to injury morbidity and mortality and did not provide any estimates on the cost of injury. In 2003, Thanh et al. reported for the first time
689
the economic burden of unintentional injuries in Vietnam.16 By sampling study subjects from a community and therefore relying on respondents’ knowledge, recall bias might limit the accuracy of information collected on cause, severity and costs of injury. In 2008, Hoang et al. studied costs of traumatic brain injury due to motorcycle accidents in Hanoi.10 By recruiting participants from the hospital, they likely increased the accuracy of information on cause and severity of injuries but there were only 35 participants included.10 The current study recruited participants from the provincial General Hospital, and data on cause and severity of injuries were collected with support from doctors’ diagnoses. Measurement errors in case identification and injury characteristics were likely reduced compared to community based studies, which rely on respondents’ knowledge. In addition, by recruiting participants from the largest trauma hospital in the province, a large proportion of hospitalised injuries with a wide range of causes were included in the study. Indeed, 892 participants in our study sample accounted for approximately 80% of the hospitalised injuries among people aged 18 years or older who had required in-patient treatment in the province during the eight month period.36 Despite improvements over prior research, there are a number of limitations in our study. First, we were limited in studying patients hospitalised for injuries, or more severe injuries, and we were unable to describe costs of non-hospitalised injuries. Second, our study was conducted in a single province, and is t therefore not fully representative of Vietnam. Compared to national figures on injury mortality and morbidity (MOH data)3 and figures on income and health expenditure (GSO data),18 Thai Binh province experiences slightly less injury cases compared to the national average. Therefore, injury costs for Vietnam, both per case and for all cases are likely larger than our estimates. Finally, the costs analysed in this study included only those incurred during the hospitalisation period, while costs would still accumulate beyond this period. In conclusion, these study results have shown that injuries lead to a significant economic burden on the family in Vietnam. Injuries due to falls and road crashes were found to have the largest costs, and as the most prevalent injuries they are causing the largest economic burden on individuals and their families. While resources are limited, injury prevention programmes need to take into account these injuries. In addition, programmes that aim to reduce the incidence and severity of falls and road crashes would likely result in significant savings for the family as well as society. Finally, to reduce the likelihood of catastrophic health payments, it is likely that the expansion of health insurance schemes would be an effective measure to protect them from impoverishment due to injuries. Conflict of interest statement H. Nguyen was funded by Atlantic Philanthropies for the study and the George Institute for Global Health, University of Sydney for student scholarship. R. Ivers and S. Jan were funded by the National Health and Medical Research Council of Australia. A. Martiniuk was funded in part by an unrestricted educational Fellowship from Merck Inc. in 2011. C. Pham was funded in part by Atlantic Philanthropies. Acknowledgements The researchers would like to thank participants and their relatives in providing information used in this study and gratefully acknowledge the huge effort made by individual doctors and nurses in Thai Binh General Hospital in collecting data.
690
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
References 1. General Statistics Office of Vietnam. The 2009 Vietnam population and housing census: major findings. 2010. 2. Administration of Preventive Medicine MoH. Injury mortality statistics in 2007. Hanoi: Ministry of Health; 2008. 3. Administration of Preventive Medicine MoH. Injury mortality statistics in 2008. Hanoi: Ministry of Health; 2009. 4. Le VA, Le CL, Pham VC. Vietnam multi-center injury survey. Hanoi: Hanoi School of Public Health; 2003. 5. Peden M, McGee K, Krug E. Injury: a leading cause of the global burden of disease, 2000. Geneva: World Health Organization; 2002. 6. Zhou Y, Baker TD, Rao K, Li G. Productivity losses from injury in China. Injury Prevention Journal of the International Society for Child and Adolescent Injury Prevention 2003;9:124–7. 7. Wagstaff A, van Doorslaer E. Catastrophe and impoverishment in paying for health care: with applications to Vietnam 1993–1998. Health Economics 2003;12:921–34. 8. Ministry of Health and Health Partnership Group. Joint Annual Health Review 2008 – Health financing in Vietnam. 2009. 9. Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, Murray CJ. Household catastrophic health expenditure: a multicountry analysis. Lancet 2003;362: 111–7. 10. Hoang HTM, Pham TL, Vo TTN, Nguyen PK, Doran CM, Hill PS. The costs of traumatic brain injury due to motorcycle accidents in Hanoi, Vietnam. Cost Effectiveness and Resource Allocation 2008;6:17–23. 11. Rice DP. Cost of illness studies: what is good about them? Injury Prevention 2000;6:177–9. 12. Meerding WJ, Mulder S, van Beeck EF. Incidence and costs of injuries in The Netherlands. European Journal of Public Health 2006;16:272–8. 13. Australian Bureau of Statistics. Injury in Australia: a snapshot, 2004–05. Canberra; 2004. 14. Lim SJ, Chung WJ, Cho WH. Economic burden of injuries in South Korea. Injury Prevention 2011;17:291–6. 15. World Health Organization. Global status report on road safety: time for action. Geneva: World Health Organization; 2009. 16. Thanh NX, Hang HM, Chuc NT, Lindholm L. The economic burden of unintentional injuries: a community-based cost analysis in Bavi, Vietnam. Scandinavian Journal of Public Health Supplement 2003;62:45–51. 17. General Statistics Office. Results of the survey on household living standards 2008. Hanoi; 2009. 18. General Statistics Office of Vietnam. Results of the survey on household living standards 2010. 2011.
19. Sethi DHS, McGee K, Peden M, Bennett S, Hyder AA, Klevens J, et al. Guidelines for conducting community surveys on injuries and violence. Geneva, Switzerland: World Health Organization; 2004. 20. Hallenbeck JL. Communication across cultures. Journal of Palliative Medicine 2004;7:477–80. 21. Association for the Advancement of Automotive Medicine. Abbreviated Injury Scale (AIS) 2005. Barrington, IL; 2005. 22. Akobundu E, Ju J, Blatt L, Mullins CD. Cost-of-illness studies: a review of current methods. Pharmacoeconomics 2006;24:869–90. 23. Corso P, Finkelstein E, Miller T, Fiebelkorn I, Zaloshnja E. Incidence and lifetime costs of injuries in the United States. Injury Prevention 2006;12:212–8. 24. Corso PS, Mercy JA, Simon TR, Finkelstein EA, Miller TR. Medical costs and productivity losses due to interpersonal and self-directed violence in the United States. American Journal of Preventive Medicine 2007;32:474–82. 25. Rice DP, MacKenzie EJ, and associates.Cost of injury in the United States: a report to Congress. San Francisco, CA: Institute for Health and Aging, University of California and Injury Prevention Center, The Johns Hopkins University; 1989. 26. Riewpaiboon A, Piyauthakit P, Chaikledkaew U. Economic burden of road traffic injuries: a micro-costing approach. Southeast Asian Journal of Tropical Medicine and Public Health 2008;39:1139–49. 27. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. 3rd ed. Oxford: Oxford University Press; 2005. 28. Xu K, Evans DB, Carrin G, Aguilar-Rivera AM, Musgrove P, Evans T. Protecting households from catastrophic health spending. Health Affairs 2007;26:972–83. 29. Gotsadze G, Zoidze A, Rukhadze N. Household catastrophic health expenditure: evidence from Georgia and its policy implications. BMC Health Services Research 2009;9:69. 30. Thuan NT, Lofgren C, Chuc NT, Janlert U, Lindholm L. Household out-of-pocket payments for illness: evidence from Vietnam. BMC Public Health 2006;6:283. 31. McNutt LA, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. American Journal of Epidemiology 2003;157:940–3. 32. Zou G. A modified poisson regression approach to prospective studies with binary data. American Journal of Epidemiology 2004;159:702–6. 33. Rice DP. Cost-of-illness studies: fact or fiction? Lancet 1994;344:1519–20. 34. General Statistics Office of Vietnam. Results of the survey on household living standards 2008. 2009. 35. Sepehri A, Sarma S, Oguzoglu U. Does the financial protection of health insurance vary across providers? Vietnam’s experience. Social Science and Medicine 2011;73:559–67. 36. Thai Binh Department of Health. Report on statistics of injuries in 12-month 2010. 2011.
Contents lists available at SciVerse ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Catastrophic household costs due to injury in Vietnam Ha Nguyen a,b, Rebecca Ivers a,*, Stephen Jan a, Alexandra Martiniuk a,c, Cuong Pham b a
The George Institute for Global Health, University of Sydney, Australia The Center for Injury Policy and Prevention Research, Hanoi School of Public Health, Viet Nam c Dalla Lana School of Public Health, University of Toronto, Canada b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 5 May 2012
Objective: Little is known about the costs of injury and their impact on injured persons and their families in Vietnam. This study aimed to examine the cost of injury in hospitalised patients and to identify the most costly injuries and those more likely to result in catastrophic household expenditure. Method: A prospective cohort study was conducted, recruiting individuals admitted to Thai Binh General Hospital due to injury in Vietnam from January to August 2010. During the hospitalisation period, data on expenditure including direct medical, direct non-medical and indirect costs were collected. Demographic and injury characteristics were also obtained. The associations between the risk of catastrophic expenditure and injury cause, severity and principal injured region were examined by modified Poisson regression approach. Payment of more than 40% of the household non-subsistence spending was considered a catastrophic expenditure. Results: Of 918 patients approached, 892 (97%) were recruited. Total costs for all participants during the hospitalisation period were US$ 325,812. Patients admitted for road injury accounted for the largest number of injuries (n = 477, 53%), and the largest percentage of the total costs (US$ 175,044, 57%). This was followed by individuals hospitalised due to falls, representing 29% of the sample (n = 261) and 31% of the total costs (US$ 103,128). In terms of cost per hospital stay, burn injuries were the most costly (US$ 427), followed by falls (US$ 395) and road crashes (US$ 367). Of all sample, 26% experienced catastrophic expenditure due to their injuries. Factors significantly associated with increased risk of catastrophic expenditure were having more severe or higher MAIS injuries (RR = 2.02, 95% CI: 1.14–3.57), principal injured region to lower extremities (RR = 3.34, 95% CI: 1.41–7.91) or head (RR = 3.21, 95% CI: 1.37–7.52), longer hospital stay (RR = 1.09, 95% CI: 1.07–1.10), older age, lower income and not having insurance (RR = 1.63, 95% CI: 1.21–2.21). Conclusion: A high proportion of households experienced catastrophic expenditure following injury, highlighting the important need for programmes to prevent injuries, road traffic and fall-related injuries in particular. Furthermore, expansion of health insurance coverage may help individuals cope with the financial consequences of injury. ß 2012 Elsevier Ltd. All rights reserved.
Keywords: Cost of injuries Economic burden Cost of hospitalisation Catastrophic expenditure Vietnam
Introduction Injury is a serious public health issue in Vietnam, the thirteenth most populous country in the world with a population of 85.8 million.1 Every day in Vietnam, injuries claim almost 100 lives, or about 35,000 lives per year.2,3 It has been estimated that the number of potential life years lost due to injuries in Vietnam is more than double those due to non-communicable diseases, and more than six times higher than those due to communicable diseases.4 In addition to fatalities, injuries are also the cause of
* Corresponding author at: The George Institute for Global Health, PO Box M201, Missenden Road, NSW Australia 2050. E-mail address: [email protected] (R. Ivers). 0020–1383/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.injury.2012.05.006
hundreds of thousands of hospital admissions.4 Because the largest proportion of injuries occur among people under 60 years of age5 who comprise the majority of the labour force, injuries may have a serious economic impact at the country level. The impact not only results from the loss of productivity, but also from significant expenses for medical treatment, rehabilitation and recovery.6 Injuries can potentially lead to catastrophic financial losses to injured persons and their families, leading to a substantial risk of impoverishment. The idea of financial catastrophe comes from an ethical position that no one ought to spend more than a given of fraction of their income on health care.7 In Vietnam, health care costs are paid directly from the income of patients and their families. According to Ministry of Health estimates, health expenditure in Vietnam consists of private out-of-pocket payments (67%), public sources including from central government
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
budgets (6%), provincial government budgets (10%), official development assistance funds (2%), and social insurance and primary health insurance (9%).8 During 2001–2006, despite fluctuations, the average household expenditure on health contributed more than 60% of the total health expenditure in the country.8 In a study comparing 59 countries, Xu et al. found that Vietnam had the highest proportion of households facing catastrophic payments for health. Specifically, more than 10.5% of households had health expenditures exceeding 40% of the household capacity to pay.9 In a study on costs of traumatic brain injury, Hoang et al. found that 84% of the sample, which included 35 households, faced catastrophic expenditure.10 By converting the impact of injuries into monetary terms, cost of injury studies can provide a common language for policy and decision makers.11 Cost of injury studies have been widely conducted, particularly in high or middle income country settings. For instance, Meerding et al. reported total health care costs due to injuries in the Netherlands in 1999 were s 1.2 billion or 3.7% of the total health care budget.12 In Australia, a snap shot of injuries presented by the Australian Bureau of Statistics show that, resources spent on injured people during 2000–2001 were 8% of total allocated health expenditure, almost AU$ 4 billion.13 In China, Zhou et al. estimated the economic cost of injury in 1999 at US$ 12.5 billion, almost four times the total public health services budget of China.6 More recently in South Korea, estimated medical treatment costs for injuries in 2006 accounted for 9.5% of the national health expenditure, and the annual economic burden associated with injury was more than US$ 39.8 billion.14 A recent World Health Organisation Global Status Report on Road Safety also highlighted a need for comprehensive cost estimates of injuries in addition to cost-effectiveness of various interventions to inform preventive action and mobilise support.15 In Vietnam, very little is known about the costs of injury and their impact on injured persons and their families. Existing knowledge on costs of injury in Vietnam comes from work conducted by Thanh et al. in 2000 and Hoang et al. in 2008. However, the first study was limited in identification of injury cause, which was self-reported by respondents,16 and he later study had a small sample size,10 limiting the conclusions that can be drawn from these works. The present study is an effort to fill in knowledge gaps by estimating the costs of injuries to the injured persons and their families during the hospitalisation period. These will be examined in greater detail in terms of injury characteristics, patient demographics, insurance status, external causes of injury, severity, and body region injured. This study also aims to identify the most costly injuries and those most likely to result in a catastrophic expenditure for the household. Methods A prospective cohort study was used to examine the economic burden of injuries. The study was conducted in Thai Binh General Hospital, the largest trauma hospital in Thai Binh province, with 440 beds. The Thai Binh province is in the Red River delta, approximately 100 km south of Hanoi, the capital city of Vietnam. In 2009, the population of this province was 1,900,000.17 According to the National Household Living Standards survey in 2010, the average monthly per capita income in Thai Binh province was VND 1,129,300, equivalent to US$ 57.90.18 There are nine hospitals at the provincial level and twelve at the district level in the province. As the largest trauma hospital at the provincial level, the Thai Binh General Hospital receives the majority of trauma patients in the province either directly or indirectly – those transferring from lower level hospitals or from same level hospitals in the province. Study participants were those individuals who were admitted to Thai Binh General Hospital due to an injury. An injury is defined
685
as physical damage that results when a human body is suddenly or briefly subjected to intolerable levels of energy. An injury may result from acute exposure to energy in amounts that exceed the threshold of physiological tolerance, or an impairment of function resulting from a lack of one or more vital elements (i.e. air, water, warmth), such as in drowning, strangulation or freezing.19 Additional inclusion criteria also included inpatient hospital treatment for at least one day, age of 18 years or older, current residential address within Thai Binh province area and consent to participate in the study. Fatal cases were not included. In Vietnam, there is a traditional belief that a person should die at home20 and therefore those patients not expected to survive are often taken home by their families. Participants were recruited from 01 January 2010 to 31 August 2010 by trained research assistants and doctors in the hospital. After critical treatment and hospital admission, 918 injured persons meeting inclusion criteria were approached; 892 (97%) consented to participate into the study. The International Classification of Diseases 10th revision (ICD-10) was used to code injuries into categories of external causes prior to reporting the results. Diagnoses also included the principal injured body region, which is the most severely injured region, and injury severity measured by the abbreviated injury score (AIS), which is an anatomical scoring system representing the ‘threat to life’ associated with an injury. The numerical ranking of the AIS ranges from 1 to 6: 1 (minor injury), 2 (moderate), 3 (serious), 4 (severe, life threatening), 5 (critical, survival uncertain) to 6 (un-survivable).21 The most severe injury was scored using the AIS during data collection, and thus the measure of severity used is the maximum abbreviated injury score (MAIS).21 Demographic information (e.g. age, gender, occupation) and injury context (e.g. place, time of injury, specific external cause) were collected after hospital admission by face to face interview. Questions on injury context were adapted from the World Health Organisation Guideline for conducting community surveys on injuries and violence.19 All data collection took place in the hospital by hospital nurses trained as research assistants in this study. Costing method The economic impact of injury was measured in terms of outof-pocket costs from the individual and family perspective. All costs incurred by participants and their household members associated with treatment during hospitalisation were reported. They were categorised into direct medical, direct non-medical and indirect costs.22 Cost data were also collected by means of face to face interview with all 892 injured persons and their caretakers the day before their hospital discharge. These include direct and indirect costs. Direct costs included expenditure associated with treatment and care for the injured person. Taking a similar approach to other studies on cost of injuries,12,23–26 cost items associated with treatment included during hospitalisation were emergency service, surgery or treatment, paramedical or diagnostic examination tests (such as X ray, CT scan), medication (prescribed and overthe-counter drugs), equipment (wheelchair, splint) and rehabilitation in the hospital. In addition to items directly associated with treatment for the injured person, information on non-medical costs incurred by the injured person and their relatives including transportation to the hospital, accommodation and meals were also collected. Indirect costs refer to lost productivity because of injury treatment and recovery.10,12,23,25,26 Using the human capital approach from the individual and family perspective, the indirect costs during the hospitalisation period were estimated by the product of the total days off work over this period and the average
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
686
daily income of the injured person and their relatives. For those without a formal income, such as students, unpaid homemakers or the elderly, opportunity costs were calculated using the average per capita income16,27 for Thai Binh province, which was the average monthly per capita income estimated in the National Household Living Standards survey in 2010. Catastrophic expenditure The out-of-pocket costs required for medical treatment are paid for from the resources available to the household. When these are large relative to available resources, they can lead to financial hardship and impair the living standard of the household.28 To date, there is not complete consensus regarding the specific fraction for defining catastrophic expenditure.9,28 In this study, similar to Hoang et al.’s study on the costs of traumatic brain injury due to motorcycle accidents in Vietnam,10 a threshold of 40% was used. This threshold has also been used in studies analysing the catastrophic health expenditure across multiple countries.9,10,28–30 In our study, participants and their households were classified as experiencing a catastrophic expenditure if the treatment costs accounted for more than 40% of the household’s capacity to pay, which is the total income of household subtracting the expenditure on food.10 Catastrophic expenditure if: Treatment costs > 40% o f ½Income o f household expenditure on food Data analysis Descriptive statistics were used to describe study participants including age, gender, occupation, and external cause of injuries. Costs of injury by external cause were computed in terms of mean, median and interquartile range to present the distribution of data. To explore predictors of risk of catastrophic expenditure, we used a modified Poisson regression approach estimating relative risk. Modified Poisson regression is a Poisson regression applied to binomial data using a robust error variance. This approach estimates relative risk consistently and efficiently.31,32 In our multivariable model, all variables hypothesised a priori to potentially affect the risk of catastrophic expenditure were included: external cause, injury severity, principal injured region, length of stay, gender, age group, occupation and insurance status. Collinearity was assessed by examining scatter plots, correlation matrix and variance inflation factors. Plausible interactions examined were those between length of hospital stay (LOS) with external cause, injury severity, principal body region injured, gender, age, insurance status and between age group with insurance status. The backward stepwise approach was then employed to remove non-significant interactions in the full models. For all analyses, a p-value of less than 0.05 was considered as statistically significant. All analyses were performed using STATA statistical package version 11 (Stata Corporation, College Station, TX, USA). The ethics application for the study was reviewed and approved by the Human Ethics Committee in University of Sydney, Australia, the Hanoi School of Public Health Ethics Committee and Thai Binh General Hospital in Vietnam. Results Table 1 summarises the demographic characteristics of the study sample. Younger (18–29 years) and older (60 years plus) participants accounted for the majority of the sample (27.8% and
Table 1 Demographic characteristics of participants.
Total Age 18–29 30–39 40–49 50–59 60+ Gender Male Female Occupation White-collar worker Blue-collar worker Farmer Retiree/unemployed Insurance Yes No
Frequency
%
Average monthly income (US$)
892
100.0
73.9
248 134 162 133 215
27.8 15.0 18.2 14.9 24.1
80.7 81.8 70.4 74.7 63.3
634 258
71.1 28.9
76.5 67.4
74 139 464 215
8.3 15.6 52.0 24.1
118.3 98.1 58.7 75.8
224 668
25.1 74.9
70.8 74.9
24.1% respectively). Males made up a disproportionate share with more than twice the number of females, 634 males (71.1%) in comparison with 258 females (28.9%). In terms of occupation, more than 50% were farmers. This was more than twice the percentage of the second most common group (24.1%), which included those without a formal occupation or those already retired. In the study sample, only 25.1% were covered by some sort of insurance. In terms of economic status, the average income of participants was approximately US$ 74 per month. Farmers had the lowest reported income (US$ 58.7) and white-collar worker had the highest income (US$ 118.3). Injury characteristics are presented in Table 2. The predominant causes were road crashes (477/892 or 53%) and falls (261/892 or 29%). More than 80% of participants were ranked at a MAIS of 2 and 3, or moderate and serious injuries respectively. More severe injury or MAIS of more than 3 was highest among participants hospitalised due to road crashes, falls and blunt object injuries compared to other injury types. In terms of principal injured regions, injuries to the head accounted for the highest proportion of principal injured regions in the sample. This was followed by injury to the hip and lower extremity region. The body region least impacted by injury was the spinal region, where there were 41 cases accounting for 4.6% of the sample. Among road traffic injuries (RTIs), the most common principal injured region was the head (43.4%). Among falls, however, the most common principal injured regions were hip and lower extremity (44.4%). Regarding length of hospital stay, persons sustaining burn had the longest average hospital stay (10 days) and those injured by a sharp object stayed in the hospital for 6 days on average. Those hospitalised due to falls or road crashes were on average hospitalised for 8.5 days and 7.7 days respectively. Table 3 describes total and average costs of injury by external causes. For the 892 hospitalised injury patients recruited, the total costs during the hospitalisation period were US$ 325,812. This is equivalent to approximately 3929 average person-month salaries (US$ 73.9 per person per month, see Table 1). Costs incurred by road crash victims accounted for the largest proportion of the total costs, more than 57% or with an amount of more than US$ 275,000, equivalent to almost 2115 average person-month salaries. This is followed by the costs incurred by participants hospitalised due to falls, which represent 31% of the total costs. In terms of mean costs per case, burns cost the most to the participants and their families, averaging US$ 321 on medical costs, almost US$ 50 on non-medical costs and approximately US$ 57 on indirect costs. Following burns were road crashes and falls,
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
687
Table 2 Injury characteristics.
Total MAIS 1 2 3 4 5 Principal injured region Head Face Spine Thorax/abdomen Upper extremity Hip/lower extremity Length of hospital stay
RTI (V01–V89)
Fall (W00–W19)
Burn (X00–X19)
Sharp object (W25– W27)
Blunt object (W20– W22)
Assault (X85–Y09)
Other injuries
n
n
n
n
n
n
n
%
477
%
261
%
22
%
32
%
33
%
43
All injuries
%
24
n
%
892
33 203 184 38 19
6.9 42.6 38.6 8.0 4.0
15 96 126 19 5
5.7 36.8 48.3 7.3 1.9
4 11 6 1 0
18.2 50.0 27.3 4.5 .0
2 16 13 1 0
6.3 50.0 40.6 3.1 .0
6 9 14 4 0
18.2 27.3 42.4 12.1 .0
9 14 17 1 2
20.9 32.6 39.5 2.3 4.7
6 9 9 0 0
25.0 37.5 37.5 .0 .0
75 358 369 64 26
8.4 40.1 41.4 7.2 2.9
207 50 12 35 58 115
43.4 10.5 2.5 7.3 12.2 24.1
73 5 26 17 24 116
28.0 1.9 10.0 6.5 9.2 44.4
0 6 0 3 6 7
.0 27.3 .0 13.6 27.3 31.8
2 2 0 0 23 5
6.3 6.3 .0 .0 71.9 15.6
9 4 3 1 11 5
27.3 12.1 9.1 3.0 33.3 15.2
10 13 0 2 17 1
23.3 30.2 .0 4.7 39.5 2.3
2 2 0 3 15 2
8.3 8.3 .0 12.5 62.5 8.3
303 82 41 61 154 251
34.0 9.2 4.6 6.8 17.3 28.1
7.7 days
8.5 days
10.0 days
which incurred total cost of US$ 367 and US$ 395 respectively for the injured persons and their families. Assaults or violence were the least costly with average cost per case just under US$ 254. Of the total costs per case during the hospitalisation period, direct medical costs accounted for the largest share, US$ 270.6 or more than 74% on average. The share of direct medical costs was highest among individuals with sharp object related injuries, US$ 229.8 or 78.7%. On the other hand, this share was lowest among assault related injuries, US$ 166.9 or 65.8%. Regarding other components, direct non-medical costs and indirect costs on average represented more than 12.8% of the total costs (US$ 46.6) and 13.2% (US$ 48.1) respectively.
6.0 days
7.6 days
6.3 days
7.5 days
7.9 days
Overall, 27% of participants had catastrophic expenditure due to their injury (Fig. 1). That is, 27% of participants spent more than 40% of their household capacity to pay for treatment of their injuries. Catastrophic household expenditure was most common among participants hospitalised due to falls, burns, sharp object related injuries and road crashes. Those with lowest proportions of catastrophic payment were victims of assaults and other injuries. Results of univariable and multivariable regression analyses examining predictors of the risk of catastrophic expense due to injury care and treatment are shown in Table 4. In terms of injury characteristics, univariable analyses show that the risk of catastrophic expense was statistically significantly associated
Table 3 Costs of injury (US$) per case (mean, median, interquartile range (IQR)) and all cases by external causes. External causes
Per case
All cases
Direct costs
Indirect costs
Total
Direct costs
Indirect costs
Total
Medical
Non-medical
Medical
Non-medical
RTI (V02–V89)
Mean Median IQR
270.2 186.2 262.4
46.0 30.8 46.2
50.7 32.5 46.1
367.0 288.7 331.4
128,890
21,962
24,192
175,044
Fall (W00–W19)
Mean Median IQR
297.7 208.2 284.1
50.7 34.4 48.2
46.7 33.1 33.3
395.1 291.2 330.9
77,710
13,235
12,183
103,128
Burn (X00–X19)
Mean Median IQR
321.2 179.7 381.2
48.8 51.3 52.6
56.7 51.6 69.4
426.7 283.7 372.0
7066
1074
1248
9388
Sharp object (W25–W27)
Mean Median IQR
229.8 204.2 143.2
31.5 24.4 30.8
30.7 22.2 24.0
291.9 252.1 176.5
7353
1008
981
9342
Blunt object (W20–W22)
Mean Median IQR
224.1 187.9 174.7
36.3 20.5 38.5
43.4 35.4 26.7
303.7 269.3 236.0
7394
1197
1432
10,023
Assault (X85–Y09)
Mean Median IQR
166.9 165.1 132.1
43.1 30.8 33.8
43.6 34.8 16.2
253.6 235.9 109.4
7175
1854
1876
10,905
Others RTI (V02–V89)
Mean Median IQR
240.4 166.9 203.8
51.2 30.8 41.0
40.7 33.8 37.4
332.3 212.3 223.0
5770
1229
977
7976
All injuries
Mean Median IQR
270.6 193.3 246.0
46.6 30.8 46.2
48.1 33.3 38.4
365.3 275.2 308.1
241,358
41,559
42,894
325,812
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
688
40% 33.0% 30%
31.8%
31.3%
29.2%
26.4%
27.5%
21.2% 20%
10%
4.7%
0% RTI*
Burn
Fall
Sharp object
Blunt object
Assault
Others
All
Fig. 1. Catastrophic expenditure. *RTI, road traffic injury.
with fall external cause of injury, higher MAIS, principal injured region to the spine/thorax/abdomen, extremities and head and longer hospital stay. For instance, the risk of catastrophic expenditure among individuals injured due to falls was 1.54 (95% CI: 1.09–2.18) times higher than those injured due to causes other than fall or RTI. Each additional day staying in hospital was associated with 9% increased in the risk of catastrophic expenditure. Regarding demographic characteristics, older individuals were at statistically significantly higher risk of catastrophic expenditure while those without insurance were at statistically significantly higher risk. Specifically, in comparison with injuries Table 4 Modified Poisson regression of predictors of the risk of having catastrophic expenditure. Unadjusted
Adjusteda
RR
95% CI
RR
95% CIb
1.00 1.54 1.23
1.09–2.18 0.90–1.73
1.00 1.03 0.95
0.71–1.33 0.68–1.50
1.00 2.05 2.62 2.99 2.88
1.08–3.88 1.39–4.93 1.49–6.00 1.28–6.48
1.00 2.11 2.30 2.27 1.86
1.19–3.73 1.31–4.03 1.25–4.11 0.73–4.75
1.00 2.89 4.05 6.21 4.82
1.12–7.47 1.66–9.89 2.61–14.74 2.02–11.47
1.00 2.33 2.94 3.34 3.21
0.91–5.98 1.24–6.93 1.41–7.91 1.37–7.52
Length of hospital stay (days) Gender Male Female Age 18–39 40–59 60+ Occupation White-collar worker Blue-collar worker Farmer Retiree/unemployed Insurance Yes No
1.09
1.07–1.11
1.09
1.07–1.10
1.00 1.31
1.05–1.63
1.00 1.08
0.88–1.34
1.00 1.46 1.73
1.12–1.90 1.33–2.26
1.00 1.09 1.27
0.84–1.41 0.94–1.71
1.00 0.74 1.60 1.25
0.35–1.53 0.85–3.00 0.65–2.40
1.00 0.47 0.59 0.49
0.23–0.96 0.31–1.14 0.25–0.96
1.00 1.58
1.17–2.12
1.00 1.63
1.21–2.21
Monthly incomec
0.89
0.87–0.92
0.90
0.88–0.93
Predictors
External causes All other injuries combined Fall (W00–W19) RTI (V01–V89) MAIS 1 2 3 4 5 Principal injured region Face Spine/thorax/abdomen Upper extremity Hip/lower extremity Head
a b c
Adjusting for all other variables listed. 95% CI: 95% confidence interval. The unit of monthly income was in US$ 10.
among individuals aged 18–39 years, the risk of catastrophic expenditure was 1.46 times higher (95% CI: 1.12–1.90) and 1.73 times higher (95% CI: 1.33–2.26) among individuals aged 40–59 years and 60 years plus respectively. Risk of catastrophic expenditure for those without insurance was 1.58 times (95% CI: 1.17–2.12) higher than those with insurance. In addition, those with higher income were also statistically significant at lower risk of catastrophic expenditure (RR = 0.89, 95% CI: 0.87–0.92). In the multivariable regression model, no interaction term was found to be statistically significant and increased risk of catastrophic expenditure was statistically associated with higher MAIS, principal injured region to extremities and head, longer hospital stay, older age, not having insurance and lower monthly income (Table 4). For example, when controlling for other variables in the model, an individual with MAIS of 2, i.e. moderate injury, had double the risk (RR = 2.02, 95% CI: 1.14–3.57) of experiencing a catastrophic expenditure than those with MAIS of 1 or minor injury. After controlling for other characteristics, injured persons without insurance were more likely (RR = 1.63, 95% CI: 1.21–2.21) to have catastrophic expenditure compared to those without insurance. Discussion This study investigated the cost of injuries for individuals admitted to a provincial hospital in Vietnam, Thai Binh province. On average, the total cost of an injury was US$ 365, which was equivalent to an economic loss of more than the average income for 6 months. We found that burns, falls and RTIs were the most costly injuries. This is partly due to a longer hospital stay for these injuries. In addition, more severe injuries were also associated with increased costs. In terms of mean costs per case, the most costly injury appeared to be burns. However, this was explained by the longest hospital stays for burns than for any other injuries. When examining the mean costs per case per day, burns were not as costly as falls or for RTIs. In addition, median costs show that costs were highest among falls and RTIs. With the largest number of cases, falls and RTIs accounted for the highest total economic cost of all injuries. Of US$ 325,812 incurred by 892 injuries during the 8-month period, more than 50% was accounted for by costs incurred by RTIs and more than 30% incurred by fall-related injuries. This finding of highest attributable costs due to RTIs and falls confirms previous findings from the US and the Netherlands.12,23 In this study, the largest component of total costs per case was found to be due to direct medical costs, in contrast to prior studies, where indirect costs were found to be the largest component.12,23,33 However, in this study costs were estimated only during the hospitalisation period, whereas prior studies also included costs incurred during the period beyond hospitalisation.
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
Our study provides the most recent estimates on the cost of injury in Vietnam. In 2001, Thanh et al. estimated that the average total costs were just US$ 13.70. These included costs for health care, transportation for care and treatment of the injury, and costs for patients and their relatives’ time.16 The increased costs potentially reflect the expansion of user fee policy, which has helped reduce hospitals’ dependence on limited government budgets, but resulted in increased out-of-pocket money from individuals, and may also have increased a financial incentive to keep patients in hospitals for longer. Consequently, it has led to a disproportionate burden on the individual and the family, especially disadvantaged people who do not have any insurance.8 In our study sample, only 25% were covered by insurance for their treatment costs. Other explanations for the higher estimates in our study are the methodology. By including only in-patient hospitalised injuries, our sample includes a larger proportion of more severe injuries than those included in Thanh’s study, which included both hospitalised and non-hospitalised injuries reported by participants.16 In addition, instead of collecting data at the injured persons’ home, we obtained cost information in the hospital, thus minimising recall bias. The large burden of injury costs and the current health financing mechanism often results in catastrophic expenditure for the household. In our sample, approximately 26% of participants faced this issue. The largest proportion of catastrophic expenses was incurred by participants injured due to falls and road crashes, likely due to the severity of injury sustained. Our estimates for catastrophic payments were understandably smaller than the 84% found in the Hoang et al. study on the costs of traumatic brain injuries related to motorcycle accidents in 200810; their estimates were costs over a one year period and only included traumatic brain injuries, which we found to result in the highest costs. The proportion of 26% of catastrophic health expenditure found in this study is however substantially higher than the proportion facing catastrophic payments due to all diseases combined among a representative sample in Vietnam, estimated at about 10.5%.9 The impact of catastrophic payments can be severe, impacting on a household’s ability to pay for education and/or housing and/or other non-subsistence items, and potentially contributing to a slide into poverty for families.28 According to the 2010 Vietnam Living Standard Survey, typical households spend 10% of the nonfood expenditure on health care.34 An increased health expenditure due to injury treatment (from 10% to 40% or more) would reduce resources available for other needs. These figures highlight the important need for injury prevention, particularly for road traffic and fall-related injuries. Enforcing traffic laws and occupational health and safety regulations may help to reduce these injuries. In addition, the coverage of health insurance as a mechanism for financial risk protection needs to expand, especially for farmers and blue-collar workers. As reported in 2008, the health insurance system covers approximately 50% of the total Vietnamese population including the formally employed, the poor and children under 6 and leaves the other 50% paying for their health care costs by out-of-pocket money.8 In fact, a study on health insurance in Vietnam demonstrated that, varying by levels of facility, the average amount of saving on out-of-pocket spending was 24% among those with compulsory health insurance and 15% among those with voluntary health insurance.35 This study adds substantially to a small number of studies that have described the burden of injury and the economic costs of injury in Vietnam. The first national injury survey with data on morbidity and mortality highlighted the importance of injury as the leading cause of years of potential life lost in Vietnam.4 This national survey was large in terms of scale, but was limited to injury morbidity and mortality and did not provide any estimates on the cost of injury. In 2003, Thanh et al. reported for the first time
689
the economic burden of unintentional injuries in Vietnam.16 By sampling study subjects from a community and therefore relying on respondents’ knowledge, recall bias might limit the accuracy of information collected on cause, severity and costs of injury. In 2008, Hoang et al. studied costs of traumatic brain injury due to motorcycle accidents in Hanoi.10 By recruiting participants from the hospital, they likely increased the accuracy of information on cause and severity of injuries but there were only 35 participants included.10 The current study recruited participants from the provincial General Hospital, and data on cause and severity of injuries were collected with support from doctors’ diagnoses. Measurement errors in case identification and injury characteristics were likely reduced compared to community based studies, which rely on respondents’ knowledge. In addition, by recruiting participants from the largest trauma hospital in the province, a large proportion of hospitalised injuries with a wide range of causes were included in the study. Indeed, 892 participants in our study sample accounted for approximately 80% of the hospitalised injuries among people aged 18 years or older who had required in-patient treatment in the province during the eight month period.36 Despite improvements over prior research, there are a number of limitations in our study. First, we were limited in studying patients hospitalised for injuries, or more severe injuries, and we were unable to describe costs of non-hospitalised injuries. Second, our study was conducted in a single province, and is t therefore not fully representative of Vietnam. Compared to national figures on injury mortality and morbidity (MOH data)3 and figures on income and health expenditure (GSO data),18 Thai Binh province experiences slightly less injury cases compared to the national average. Therefore, injury costs for Vietnam, both per case and for all cases are likely larger than our estimates. Finally, the costs analysed in this study included only those incurred during the hospitalisation period, while costs would still accumulate beyond this period. In conclusion, these study results have shown that injuries lead to a significant economic burden on the family in Vietnam. Injuries due to falls and road crashes were found to have the largest costs, and as the most prevalent injuries they are causing the largest economic burden on individuals and their families. While resources are limited, injury prevention programmes need to take into account these injuries. In addition, programmes that aim to reduce the incidence and severity of falls and road crashes would likely result in significant savings for the family as well as society. Finally, to reduce the likelihood of catastrophic health payments, it is likely that the expansion of health insurance schemes would be an effective measure to protect them from impoverishment due to injuries. Conflict of interest statement H. Nguyen was funded by Atlantic Philanthropies for the study and the George Institute for Global Health, University of Sydney for student scholarship. R. Ivers and S. Jan were funded by the National Health and Medical Research Council of Australia. A. Martiniuk was funded in part by an unrestricted educational Fellowship from Merck Inc. in 2011. C. Pham was funded in part by Atlantic Philanthropies. Acknowledgements The researchers would like to thank participants and their relatives in providing information used in this study and gratefully acknowledge the huge effort made by individual doctors and nurses in Thai Binh General Hospital in collecting data.
690
H. Nguyen et al. / Injury, Int. J. Care Injured 44 (2013) 684–690
References 1. General Statistics Office of Vietnam. The 2009 Vietnam population and housing census: major findings. 2010. 2. Administration of Preventive Medicine MoH. Injury mortality statistics in 2007. Hanoi: Ministry of Health; 2008. 3. Administration of Preventive Medicine MoH. Injury mortality statistics in 2008. Hanoi: Ministry of Health; 2009. 4. Le VA, Le CL, Pham VC. Vietnam multi-center injury survey. Hanoi: Hanoi School of Public Health; 2003. 5. Peden M, McGee K, Krug E. Injury: a leading cause of the global burden of disease, 2000. Geneva: World Health Organization; 2002. 6. Zhou Y, Baker TD, Rao K, Li G. Productivity losses from injury in China. Injury Prevention Journal of the International Society for Child and Adolescent Injury Prevention 2003;9:124–7. 7. Wagstaff A, van Doorslaer E. Catastrophe and impoverishment in paying for health care: with applications to Vietnam 1993–1998. Health Economics 2003;12:921–34. 8. Ministry of Health and Health Partnership Group. Joint Annual Health Review 2008 – Health financing in Vietnam. 2009. 9. Xu K, Evans DB, Kawabata K, Zeramdini R, Klavus J, Murray CJ. Household catastrophic health expenditure: a multicountry analysis. Lancet 2003;362: 111–7. 10. Hoang HTM, Pham TL, Vo TTN, Nguyen PK, Doran CM, Hill PS. The costs of traumatic brain injury due to motorcycle accidents in Hanoi, Vietnam. Cost Effectiveness and Resource Allocation 2008;6:17–23. 11. Rice DP. Cost of illness studies: what is good about them? Injury Prevention 2000;6:177–9. 12. Meerding WJ, Mulder S, van Beeck EF. Incidence and costs of injuries in The Netherlands. European Journal of Public Health 2006;16:272–8. 13. Australian Bureau of Statistics. Injury in Australia: a snapshot, 2004–05. Canberra; 2004. 14. Lim SJ, Chung WJ, Cho WH. Economic burden of injuries in South Korea. Injury Prevention 2011;17:291–6. 15. World Health Organization. Global status report on road safety: time for action. Geneva: World Health Organization; 2009. 16. Thanh NX, Hang HM, Chuc NT, Lindholm L. The economic burden of unintentional injuries: a community-based cost analysis in Bavi, Vietnam. Scandinavian Journal of Public Health Supplement 2003;62:45–51. 17. General Statistics Office. Results of the survey on household living standards 2008. Hanoi; 2009. 18. General Statistics Office of Vietnam. Results of the survey on household living standards 2010. 2011.
19. Sethi DHS, McGee K, Peden M, Bennett S, Hyder AA, Klevens J, et al. Guidelines for conducting community surveys on injuries and violence. Geneva, Switzerland: World Health Organization; 2004. 20. Hallenbeck JL. Communication across cultures. Journal of Palliative Medicine 2004;7:477–80. 21. Association for the Advancement of Automotive Medicine. Abbreviated Injury Scale (AIS) 2005. Barrington, IL; 2005. 22. Akobundu E, Ju J, Blatt L, Mullins CD. Cost-of-illness studies: a review of current methods. Pharmacoeconomics 2006;24:869–90. 23. Corso P, Finkelstein E, Miller T, Fiebelkorn I, Zaloshnja E. Incidence and lifetime costs of injuries in the United States. Injury Prevention 2006;12:212–8. 24. Corso PS, Mercy JA, Simon TR, Finkelstein EA, Miller TR. Medical costs and productivity losses due to interpersonal and self-directed violence in the United States. American Journal of Preventive Medicine 2007;32:474–82. 25. Rice DP, MacKenzie EJ, and associates.Cost of injury in the United States: a report to Congress. San Francisco, CA: Institute for Health and Aging, University of California and Injury Prevention Center, The Johns Hopkins University; 1989. 26. Riewpaiboon A, Piyauthakit P, Chaikledkaew U. Economic burden of road traffic injuries: a micro-costing approach. Southeast Asian Journal of Tropical Medicine and Public Health 2008;39:1139–49. 27. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. 3rd ed. Oxford: Oxford University Press; 2005. 28. Xu K, Evans DB, Carrin G, Aguilar-Rivera AM, Musgrove P, Evans T. Protecting households from catastrophic health spending. Health Affairs 2007;26:972–83. 29. Gotsadze G, Zoidze A, Rukhadze N. Household catastrophic health expenditure: evidence from Georgia and its policy implications. BMC Health Services Research 2009;9:69. 30. Thuan NT, Lofgren C, Chuc NT, Janlert U, Lindholm L. Household out-of-pocket payments for illness: evidence from Vietnam. BMC Public Health 2006;6:283. 31. McNutt LA, Wu C, Xue X, Hafner JP. Estimating the relative risk in cohort studies and clinical trials of common outcomes. American Journal of Epidemiology 2003;157:940–3. 32. Zou G. A modified poisson regression approach to prospective studies with binary data. American Journal of Epidemiology 2004;159:702–6. 33. Rice DP. Cost-of-illness studies: fact or fiction? Lancet 1994;344:1519–20. 34. General Statistics Office of Vietnam. Results of the survey on household living standards 2008. 2009. 35. Sepehri A, Sarma S, Oguzoglu U. Does the financial protection of health insurance vary across providers? Vietnam’s experience. Social Science and Medicine 2011;73:559–67. 36. Thai Binh Department of Health. Report on statistics of injuries in 12-month 2010. 2011.