22.10.2014 Views

Child Drowning

Child Drowning

Child Drowning

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

data from multiple sources including national death registries, coronial investigations and other national<br />

mortality reporting systems. The differences between GBD figures and the member reported figures<br />

range from +50 per cent (Australia) to +300 per cent (United Kingdom). 17<br />

The GBD estimates have many uncertainties built into them. Normally, estimates with such an amount<br />

of uncertainty are stated with confidence intervals (e.g. 95% confidence intervals) or a range of<br />

likelihood scenarios (e.g. low, medium and high ranges). Listing a single, precise number can suggest<br />

precision in an estimate that is not there, especially to non-technically trained policy-makers.<br />

Table 4 below shows for children aged 0-14, direct comparisons for the GBD 2004 estimates of fatal<br />

child drowning with the estimated numbers of drowning deaths in three countries surveyed.<br />

Table 4: WHO Global Burden of Disease estimates and survey estimates of fatal drowning among<br />

children 0-14 years old<br />

Country<br />

GBD 2004 estimated<br />

drowning deaths<br />

<strong>Drowning</strong> deaths<br />

estimated in the<br />

country survey<br />

Year of<br />

survey<br />

Percentage<br />

difference<br />

Bangladesh 3,828 16,452 2002 +330%<br />

Thailand 1,322 2,093 2003 +58%<br />

Cambodia 1,074 1,871 2006 +74%<br />

Source: Data from countries included in surveys (Bangladesh Health and Injury Survey, Cambodia Accident and Injury Survey, Thailand National<br />

Injury Survey) and World Health Organization Global Burden of Disease website. The comparison is between children aged 0–14, as GBD does<br />

not include a category for children aged 0- 17 years.<br />

There is a substantial difference between the numbers of fatal child drowning deaths estimated by the<br />

GBD 2004 and by the surveys done in the three countries. Apart from regular seasonal flooding, there<br />

were no aquatic disasters or other factors that would increase the rate of child drowning in the three<br />

countries during the period. National records do not show any significant differences in annual<br />

reporting during the same period. The most likely explanation for the large difference is the virtual<br />

impossibility of adjusting for all the errors, biases, undercounts, and incompleteness in the raw data<br />

used in the estimation process. Many of the factors involved are simply unknowable. There are no<br />

means of validating the results against reliable drowning data from other LMICs as such reliable data<br />

does not exist. Hence the estimates, even when adjusted for the known factor of not counting drowning<br />

from transport, natural cataclysms and intentional injury, may have substantial inaccuracies built in.<br />

The GBD 2008 update reported even lower drowning numbers in Bangladesh (-32.8%), Cambodia<br />

(25.9%), and an increase in Thailand (+21.3%) resulting in a net reduction in drowning deaths compared<br />

to 2004. The increase in Thailand is not related to the Indian Ocean tsunami (Dec 26, 2004). It was<br />

during the 2004 GBD time frame and this was a cataclysm of nature; thus these drowning deaths are<br />

excluded from GBD estimates. There were no other changes in environmental conditions or risk factors<br />

for drowning in the three countries between 2005 and 2008, further increasing the disparity between<br />

the GBD estimates and the country surveys.<br />

Other corroborating evidence of the potential for large undercounts is available from India. The GBD<br />

2004 estimate for India was 21,785 drowning deaths in children 0-14 years of age. The Million Death<br />

Study (MDS), a nationally representative verbal autopsy study in India, covered more than 1.1 million<br />

17 Franklin R.C. and J. Scarr (2011). ‘ILS <strong>Drowning</strong> Data and Research Survey’, International Life Saving Federation, Leuven; in<br />

press.<br />

25

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!