Home >News > ProMED翻訳情報(373回) ~インドの調査に関する意見(2題)~

ProMED翻訳情報(373回) ~インドの調査に関する意見(2題)~


[1]Date: Thu 23 Apr 2015    From: Louise Taylor [edited]


As the moderator invited informed comment, we would like to point out some facts, and some errors, relating to the previous post.
The Sudarshan (National Multicentric Rabies Survey) 2007 study did identify 235 deaths, but from 22 hospitals affiliated to medical schools. These deaths were subsequently used as index cases for community searches, estimated to cover a population of 10.8 million in India (less than 1 per cent of the total human population). These findings were then extrapolated to the whole country, as summarized in
the paper’s abstract: “Results: The annual incidence of human rabies was estimated to be 17 137 (95 per cent CI 14 109 – 20 165). Based on expert group advice, an additional 20 per cent was added to this to include paralytic/ atypical forms of rabies, providing an estimate of 20 565 or about 2 per 100 000 population.” [Sudarshan et al, 2007, International Journal of Infectious Diseases].
2007年に行われたSudarshan(国内多拠点狂犬病調査)では235人の死者が明らかになりましたが、回答は医学部付属の22病院からでした。これらの死者数は地域調査の指標にされ、インドにおける1080万人(全体人口の1%未満)の人口に及ぶと推定されました。そこで、このデータは、以下の論文要旨にまとめられているようにインド全体の推定基準とされた。「結果:狂犬病による年間死者数は17,137人[95% CI(信頼限界)で14,109-20,165人]と推定された。専門家グループからの助言を基に、麻痺型/無症状型の狂犬病を含めると、さらに20%がこれに追加される。したがって、20,565人または10万人当たり2人の死者と推定される。」[Sudarshan et al, 2007, International Journal of Infectious Diseases]
>From the abstract of the Suraweera (verbal autopsy) 2012 study conducted in India: “Conclusions and Significance: Rabies remains an avoidable cause of death in India. As verbal autopsy is not likely to identify atypical or paralytic forms of rabies, our figure of 12 700 deaths due to classic and clinically identifiable furious rabies underestimates the total number of deaths due to this virus” [Suraweera et al, 2012, PLOS Neglected Tropical Diseases].
インドで行われた2012年のSuraweera(口頭分析)調査の要旨からでは、「結論と意義:狂犬病は避けれらない死因としてインドに存在する。口頭剖検では非特異的、麻痺性などの狂犬病との区別はつけられず、12,700人という我々が出した数字は、典型的な臨床的に判断が可能である狂騒型の狂犬病の件数であり、狂犬病ウイルスによる死者数全体としては少なく見積もられている。」[Suraweera et al, 2012, PLOS Neglected Tropical Diseases]
The most recent Global Burden of Disease study — utilizing largely verbal autopsy data (which as mentioned above will miss paralytic forms of rabies) estimated 12 349 (95 per cent CI 8196 – 15 399) deaths from rabies in India in 2013 [GBD 2013 authors. 2015, Lancet].
最新の世界疾病負担研究、膨大な口頭分析(上記でも述べたように、麻痺型狂犬病は見逃してしまう)のデータを活用したもの、その研究結果は2013年にインドで12,349人(95%CIで8,196-15,399人)の狂犬病死者があったと推測している。[GBD 2013 authors. 2015, Lancet]
The recently published Partners for Rabies Prevention study, estimated 20 847 (95 per cent CI 7000 – 55 000) deaths in India [Hampson et al, 2015, PLOS Neglected Tropical Diseases].
さらに、最近公開されたPartners for Rabies Preventionの研究では、20,847人(95%CIでは7,000-55,000人)の死者がインドで発生していると推測された。[Hampson et al, 2015, PLOS Neglected Tropical Diseases]

These estimates of human deaths from rabies in India vary across the studies, as is to be expected from the divergent methodologies, but none of the 95 per cent confidence intervals fall as low as the number of human deaths (generally between 200 and 350 deaths per year) reported by India’s Central Bureau of Health Intelligence (CBHI). The
CBHI figures are not the human deaths experienced by India, but the human deaths officially reported in India (where human rabies is not a notifiable disease).
This discrepancy is entirely consistent with other active surveillance studies of human rabies deaths that have shown that passive surveillance records grossly underestimate human rabies incidence, sometimes by a factor of 100 [Cleaveland et al. 2002, Bulletin WHO]. Whilst 1 or 2 cases in urban areas may make headline news, in many countries, deaths are likely concentrated amongst the rural poor who die outside of the health system, and with no media interest.
この矛盾は、受動的調査の記録が時に100分の1程度にヒト狂犬病症例を著しく低く見積もることを示した他のヒト狂犬病死者数の能動的調査と完全に一致している[Cleaveland et al. 2002, Bulletin WHO]。多くの国では、都会で発生した1例、2例はニュース速報になる一方で、狂犬病による死亡は医療組織と無関係に亡くなる地方の貧民の間に集中しており、報道機関が注目することもない。


If passive surveillance records were thought to be adequate to answer the question of how many people die of human rabies in canine rabies endemic countries, there would be no need for the detailed and time consuming research activities described above. Such studies attempt to quantify the extent of underreporting (of rabies and other diseases) to help decision makers prioritize disease control efforts and save lives. The data inputs, assumptions and methods utilized in each of the studies are explained in detail in the respective papers and should not be summarised as “abstract projections”. We invite any scientific analysis of the data and the robust methodologies applied. Peer-reviewed scientific evidence, especially new active surveillance studies of rabies from India and elsewhere, will be of value to the whole rabies community.

Louise Taylor, PRP Coordinator and Newsletter Editor, Global Alliance for Rabies Control,
Louis Nel, Global Alliance for Rabies Control,
Katie Hampson, University of Glasgow,
[2]Date: Thu 23 Apr 2015  From: Merritt Clifton [edited]
What GARC [Global Alliance for Rabies Control] is saying here, in a great many more words, is that they and others took the actual verifiable body count from the 22 institutions most likely to actually be handling human rabies deaths (because most Indian hospitals don’t knowingly admit rabies cases), and then multiplied them up to get a number similar to the number already claimed for decades, which was for them the politically most convenient number to use.

Meanwhile, the actual verifiable [Indian] body count since 2003, reported by the Central Bureau of Health Intelligence, has been 235, 274, 361, 221, 244, 260, 162, 223, 212, and 138, total of 2330, average of 233, median of 229, within a range of 138-361, with 4 of the 5 lowest annual counts coming in the past 4 years.
GARC(狂犬病制御国際連合)がここに述べていることは、文字通り、彼らと他の人達は人の狂犬病死者を実際に扱っていると思われる22 病院から、実際に確定することのできる死者数を数え、(ほとんどのインドの病院は狂犬病とわかって入院させないため)、次いで、何十年間も報告されてきた数字に近くなるまで乗算したということです。その数字は彼らにとって政治的にもっとも利便性のある数字でした。

一方、CBHIにより報告された2003年以来の実際に確定できるインドの狂犬病死者数は235, 274, 361, 221, 244, 260, 162, 223, 212および 138で、合計2,330、年平均233、中央値229、138〜361の範囲で、過去4年間が最も年間死者数の低い5年のうちの4年に当たる。
If the annual death tolls were actually close to 100 times higher, would not the numbers of cases seen by the 22 institutions most likely to be handling rabies deaths at least occasionally spike higher?
If under-reporting from remote villages really was causing such an extraordinary discrepancy in data back in 2003, when India was already among the most internet-wired and savvy nations in the world, would that condition really not have been remedied at all in the ensuing decade, during which India has become the call center and tech support desk for practically the whole of the English-speaking world?
If 17 000 to 21 000 Indians per year really were dying of rabies, could we not expect to see a rise in the reported numbers of cases, as has occurred with a variety of other diseases due to improved reporting, instead of seeing a pronounced decline from low to lower?

If human rabies deaths really were occurring at approximately the same rate as deaths by house fire (National Health Profile table, would they not receive approximately the same sort of media coverage, i.e. usually local notice only, rather than banner headlines throughout an entire community of millions, if a rabies case is even suspected?

もし、ヒトの狂犬病死亡例が本当に、家屋火災による死亡(国家保健白書 表3.1.3.1)とほぼ同じ率で起きているのなら、1例の狂犬病事例が疑われても、通常は狂犬病の症例は地元住民への通知だけで、何百万という地域全体へのトップ記事報道のような火災と同等の報道が行われないのでしょうか?


Would government hospitals equipped to handle human rabies cases not be as common as fire departments?

The phrase “writ of habeas corpus” applies here, but in a forensic rather than legal contest. A “writ of habeas corpus” is literally a document requiring the jailer to produce the body.
Where 17 000 or 20 000 or 21 000 bodies are claimed to exist, but fewer than 250 can be found, the burden of proof is on the persons making the claim to find them. Exercises in math based on presumptive projection are no substitute for forensic evidence in bringing a murder charge, and should not be substituted for forensic evidence in allocating resources in disease control, either.

17,000や20,000 または21,000人の狂犬病死者がいるとされているが、実際には250人以下しか見つかっていない。証明の苦労はこれらの見逃されている人々を見つける(という主張を行う)ことだ。予測に基づいた数理的な課題は殺人の代償をもたらす法廷の代わりにはならないし、疾病制御で物資を分配することにおいて法廷の証拠の代わりにされるべきではない。
The consistently low numbers of verifiable rabies cases in India over the past decade suggests that rabies is close to eradication, if resources are allocated to vaccinating it out of existence in the places where it verifiably still occurs.
As a postscript, let me call your attention to the numbers below, posted by ProMED 2 days ago as Ebola update (60). What you may notice 1st of all is that Guinea, Liberia, and Sierra Leone are all much less economically, technologically, and educationally advanced than India. Yet they are able to produce reasonably credible and frequently
updated body counts. Second, the death tolls in Liberia and Sierra Leone, if projected over a year’s time, end up in about the same range as the GARC-projected rabies death toll in India. The Ebola death toll in western Africa produced global concern even when a fraction as high. Yet we are to believe that deaths on this order occurring in India, each and every year, are somehow eluding everyone’s notice?


(Data up to 12 Apr 2015 for Guinea and Sierra Leone and 11 Apr 2015 for Liberia)

Guinea cumulative cases (deaths):
Confirmed cases 3117 (1932) [106 in past 21 days]
Probable cases 414 (414)
Suspected 17 (not available)
Total cases 3548 (2346) [106 in past 21 days]
ギニア 累積症例(死者数):

確定症例 3117(1932)[過去21日間で106人]

推定例 414(414)

疑い例 17(データなし)

総計 3548症例(2346死者数)[過去21日間で106人]
Liberia cumulative cases (deaths):
Confirmed cases 3151 (not available) [0 in past 21 days]
Probable cases 1879 (not available)
Suspected 5012 (not available)
Total cases 10 042 (4486) [0 in past 21 days]
リベリア 累積症例(死者数):

確定症例 3151(データなし)[過去21日間で0人]

推定例 1879(データなし)

疑い例 5012(データなし)

総計 10042症例(4486死者数)[過去21日間で0人]
Sierra Leone cumulative cases (deaths):
Confirmed cases 8563 (3491) [43 in past 21 days]
Probable cases 287 (208)
Suspected 3351 (158)
Total cases 12 201 (3857) [43 in past 21 days]
シエラレオネ 累積症例(死者数):

確定症例 8563 (3491)[過去21日間で43人]

推定例 287(208)

疑い例 3351(158)

総計 12201症例(3857死者数)[過去21日間で43人]
All countries
Confirmed cases 14 831 (not available) [149 in past 21 days]
Probable cases 2580 (not available)
Suspected cases 8380 (not available)
Total 25 791 (10 604 deaths) [149 in past 21 days]
3か国 累積症例(死者数):

確定症例 14831(データなし)[過去21日間で149人]

推定例 2580 (データなし)

疑い例 8380(データなし)

総計 25791症例(10604死者数)[過去21日間で149人]

Merritt Clifton  Editor, Animals 24-7  Greenbank, Washington 98253  USA
[Thanks to Merritt and Louise & her colleagues. However, I cannot but echo Merritt’s last statement. If the scope of the rabies problem in India were to be accurately assessed, it would be seen to be a series of relatively isolated outbreaks, vulnerable to eradication by targeting the areas for saturation vaccination of dogs. Then canine rabies could be eliminated from India altogether within just a few years of intensive effort.
Probably the same could be actively considered in Nepal, Pakistan, and Bangladesh. India, though, has probably been within striking distance of eradicating rabies for 20-30 years now — if the strikers were taking accurate aim at the problem — instead of wasting effort with just retroactive village vaccination following a human case. One has to be proactive in areas, not just villages, with a recognized problem. Rabies should be relatively easily eradicable in Nepal, as well, but in Nepal (as in the US, Canada, & Europe) the wildlife vectors appear to be at least as problematic as the canine vector, and oral vaccinating foxes et al amid the rugged terrain will not be as simple as vaccinating urban street dogs. – Mod.MHJ]