Re-evaluating the burden of rabies in Africa and Asia

Darryn L. Knobel,1 Sarah Cleaveland,1 Paul G. Coleman,2 Eric M. Fèvre,1 Martin I. Meltzer,3
M. Elizabeth G. Miranda,4 Alexandra Shaw,5 Jakob Zinsstag,6 & François-Xavier Meslin7

Objective To quantify the public health and economic burden of endemic canine rabies in Africa and Asia.
Methods Data from these regions were applied to a set of linked epidemiological and economic models. The human population at
risk from endemic canine rabies was predicted using data on dog density, and human rabies deaths were estimated using a series of
probability steps to determine the likelihood of clinical rabies developing in a person after being bitten by a dog suspected of having
rabies. Model outputs on mortality and morbidity associated with rabies were used to calculate an improved disability-adjusted life
year (DALY) score for the disease. The total societal cost incurred by the disease is presented.
Findings Human mortality from endemic canine rabies was estimated to be 55 000 deaths per year (90% confidence interval (CI) =
24 000–93 000). Deaths due to rabies are responsible for 1.74 million DALYs lost each year (90% CI = 0.75–2.93). An additional
0.04 million DALYs are lost through morbidity and mortality following side-effects of nerve-tissue vaccines. The estimated annual
cost of rabies is US$ 583.5 million (90% CI = US$ 540.1–626.3 million). Patient-borne costs for post-exposure treatment form the
bulk of expenditure, accounting for nearly half the total costs of rabies.
Conclusions Rabies remains an important yet neglected disease in Africa and Asia. Disparities in the affordability and accessibility
of post-exposure treatment and risks of exposure to rabid dogs result in a skewed distribution of the disease burden across society,
with the major impact falling on those living in poor rural communities, in particular children.

Keywords Rabies/mortality/economics; Dogs; Cost of illness; Disability evaluation; Health care costs; Probability; Models, Theoretical;
Africa; Asia (source: MeSH, NLM).
Mots clés Rage (Maladie)/mortalité/économie; Chien; Coût maladie; Evaluation incapacité; Coût soins médicaux; Probabilité; Modèle
théorique; Afrique; Asie (source: MeSH, INSERM).
Palabras clave Rabia/mortalidad/economía; Perros; Costo de la enfermedad; Evaluación de la incapacidad; Costos de la atención
en salud; Probabilidad; Modelos teóricos; África; Asia (fuente: DeCS, BIREME).

Bulletin of the World Health Organization 2005;83:360-368.

Voir page 366 le résumé en français. En la página 366 figura un resumen en español. .367

Introduction These problems are not unique to rabies, and the recog-
More than 99% of all human deaths from rabies occur in the nized poor quality of much public health information from
developing world (1), and although effective and economical developing countries has prompted several investigations into
control measures are available (2, 3) , rabies remains a neglected the distribution of major infectious diseases and the mortality
disease throughout most of these countries (4, 5). A major fac- and morbidity attributable to them. Such studies are based on
tor in the low level of political commitment to rabies control estimates of occurrence extrapolated from more readily quan-
is a lack of accurate data on the true public health impact of tifiable determinants of disease, such as vector distribution or
the disease. It is widely recognized that the number of deaths host immunity (8–10). For rabies, a similar predictive approach
officially reported greatly underestimates the true incidence of has been used to estimate human deaths from rabies in the
disease. Patients may not present to medical facilities for treat- United Republic of Tanzania using a probability decision tree
ment of clinical disease; few cases receive laboratory confirma- method to determine the likelihood of clinical rabies develop-
tion; and clinical cases are often not reported by local authorities ing in a person bitten by a dog suspected to be rabid (6). Dog
to central authorities (1, 6, 7). bites are reported proportionately more frequently than human

1
Centre for Tropical Veterinary Medicine, Royal (Dick) School of Veterinary Studies, University of Edinburgh, Easter Bush, Roslin, Midlothian EH25 9RG, Scotland.
Correspondence should be sent to Dr Knobel at this address (email:d.l.knobel@sms.ed.ac.uk).
2
London School of Hygiene and Tropical Medicine, Keppel Street, London, England.
3
Centers for Disease Control and Prevention, Atlanta, GA 30333, USA.
4
Communicable Disease, Surveillance and Response, World Health Organization, Regional Office for the Western Pacific, Manila, Philippines.
5
AP Consultants, Andover SP11 7BA, England.
6
Swiss Tropical Institute, Basel, Switzerland.
7
World Health Organization, Geneva, Switzerland.
Ref. No. 03-008862
(Submitted: 16 February 2004 – Final revised version received: 3 July 2004 – Accepted: 5 July 2004)

360 Bulletin of the World Health Organization | May 2005, 83 (5)

 Research
Darryn L. Knobel et al. Burden of rabies in Africa and Asia

cases of rabies and may provide an accessible data source from rural dog-population densities (21–23), so the human popula-
which human deaths from rabies can be inferred. tion considered to be at risk was calculated using the regional
The objective of our study was to estimate the burden of rural ratios for humans to dogs. Implicit in this approach is
rabies in Africa and Asia by applying data derived from these the assumption that all human urban populations are at risk
regions to this model and to thereby present a data-driven in rabies-affected areas.
assessment of the human and economic costs of rabies in the Four basic scenarios were considered (Africa and Asia,
developing world. We define Africa as all mainland countries both rural and urban); these reflect broad differences in fac-
on the continent plus Madagascar; Asia is defined as those tors that influence rabies epidemiology and the treatment of
countries falling under the WHO-defined South-East Asia people exposed to the disease. Initial parameter estimates were
Region and Western Pacific Region, including Pakistan. Only derived following a review of the relevant literature, including
countries considered by WHO and the Office International des publications from peer-reviewed journals and grey literature
Epizooties as having endemic canine rabies were considered in sources. Parameter estimates and confidence distributions were
this analysis. The list of all countries included in the study avail- then fixed using a three-stage consensus approach (as described
in reference 9) within the WHO Burden of Rabies Working
able from http://www.vet.ed.ac.uk/ctvm/Research/Appendices/
Group. First, a workshop was held during which parameter
appendices.html.
estimates were presented and discussed and preliminary predic-
tions made using the model. Second, a comprehensive analysis
Methods using agreed upon estimates was conducted, and the model was
Human rabies deaths validated against known data. Third, each participant was sent
Full details of the methods used in the dog-bite probability the results of this analysis, along with the data and assumptions.
model have been published elsewhere (6). Briefly, the model These were reviewed and adjusted as appropriate before final
recognizes that not all bites from rabid dogs result in infection agreement was reached. The final model parameter estimates
and that not every infection leads to clinical signs and death. and distributions are presented in Table 2 (web version only,
One of the principal factors influencing the outcome of a bite available at: http://www.who.int/bulletin).
from a rabid dog is the location of the bite on the body (11, For the analysis, Asia was subdivided into three units:
12). The model uses the distribution of injuries on the body China, India and Other Asia. Initial data sources and model
together with the likelihood of the patient receiving successful validation indicated that parameter estimates for India and
treatment to predict the outcomes of bites from rabid dogs. Other Asia were similar. However, preliminary model pre-
The model thus allows the incidence of bites from suspected dictions for China overestimated the number of deaths by
one order of magnitude when these same estimates were
rabid dogs among the human population considered to be at
used compared with the 1000 deaths suggested by expert
risk to be used as a determinant of the number of human deaths
opinion. The discrepancy is possibly explained by the use of
from rabies.
post-exposure treatment in China: the country accounts for
The human population at risk from canine rabies was
two-thirds (5 million cases) of total post-exposure treatment
taken as the number of people living in areas affected by ca-
(PET) used in Asia (24), and the locally-produced tissue-culture
nine rabies where the density of the dog population exceeds vaccine is safe and relatively inexpensive (25). These factors
the threshold density at which canine rabies is capable of be- suggest that PET for rabies is more accessible and utilized more
ing maintained endemically. Dog-population densities were frequently in China than in the rest of Asia. To account for this
inferred from human densities derived from two regional in the model, the probability of a person bitten by a suspected
population density datasets (13, 14) with adjustments made rabid dog and receiving PET was assumed to be equally high
to account for population growth (15). Associated dog-popula- in both urban and rural settings in China (minimum = 95%;
tion numbers were calculated by dividing human figures by most likely = 97%; maximum = 100%). In light of the paucity of
the regional average ratio of humans to dogs, based on values data for China, the model’s predictions for this country should
given in Table 1. The threshold density for rabies persistence be treated with caution.
was taken as 4.5 dogs/km², based on predictions produced from The model was arrayed as a spreadsheet template (using
data on rabies transmission in rural Kenya (16); these data are Microsoft Excel 2000, Microsoft Inc., Redmond, WA). Uncer-
consistent with empirical observations from elsewhere in Africa tainty in parameter estimates, and inherent parameter vari-
(17–20). This density falls within the range of estimates of ability due to between-country differences, were incorporated

Table 1. Mean human:dog ratio for Africa and Asia (determined from sources available from: http://www.vet.ed.ac.uk/ctvm/Research/
Appendices/appendices.html)

Regiona
Human:dog ratio Africa Asia China
Urban Rural Urban Rural Urban Rural
b c
Mean ratio 21.2 (12.5–37.1) 7.4 (5.7–9.7) 7.5 (4.8–10.1) 14.3(0–45.0) NA NA
Mean ratio for region 12.3 (11.2–20.6) 9.5 (4.5–14.6) 48.3 (0–147.0)
a
Insufficient data were available to calculate a separate figure for India and for rural China versus urban China.
b
Figures in parentheses are 95% confidence intervals.
c
NA = not available.

Bulletin of the World Health Organization | May 2005, 83 (5) 361

 Research
Burden of rabies in Africa and Asia Darryn L. Knobel et al.

Table 4. Direct (medical) post-exposure treatment costing data and sources (available from: http://www.vet.ed.ac.uk/ctvm/Research/
Appendices/appendices.html)

Estimate
Parameter Africa Asia Source
Patient numbers
No. PETa cases per year 200 000 7 500 000 11–14, 23–27
No. (%) PET patients receiving tissue-culture vaccines 180 000 (90) 5 025 000 (67) 11–14, 23–27
No. (%) of patients receiving tissue-culture vaccine intramuscularly 180 000 (100) 4 874 250 (97) 11–14, 23–27
No. (%) of patients receiving tissue-culture vaccine intradermally 0 (0) 150 750 (3) 11–14, 23–27
No. (%) PET patients receiving nerve-tissue vaccines 20 000 (10) 2 475 000 (33) 11–14, 23–27
No. (%) PET patients receiving rabies immunoglobulin 2 000 (1) 450 000 (6) 11–14, 23–27
No. (%) PET patients receiving human rabies immunoglobulin 0 (0) 45 000 (10)
No. (%) PET patients receiving equine rabies immunoglobulin 2 000 (100) 405 000 (90)
Costs
Common costs
Material costs per injection (includes needles, syringes, swabs, etc.) US$ 0.10 E. Miranda unpublished data, 2003
Overhead costs per PET visit (includes anti-rabies clinic staff US$ 0.50 39
salaries and administration costs)
Tissue-culture vaccine costs
Intramuscular vaccination
Vaccine cost per dose US$ 10.00 WHO Procurement Services,
personal communication, 2003
Visits per patient 3 40
Injections per patient 3 Assuming standard Essen regimen
Intradermal vaccination
Vaccine cost per dose US$ 2.50 WHO Procurement Services,
personal communication, 2003
Visits per patient 3 E. Miranda unpublished data, 2003
Injections per patient 6 Assuming Thai Red Cross regimen
Nerve-tissue vaccine costs
Vaccine cost per dose US$ 0.40 7
Visits per patient 7 33
Injections per patient 7 Assuming standard regimen
Rabies immunoglobulin
Human rabies immunoglobulin cost per dose US$ 110.00 WHO Procurement Services,
personal communication, 2003
Equine rabies immunoglobulin cost per dose US$ 25.00 WHO Procurement Services,
personal communication, 2003
a
PET = post-exposure treatment.

into the model by assigning confidence distributions to input were considered: a direct DALY score derived from mortality
parameters. Parameter distributions were sampled iteratively due to the disease and an indirect DALY score, taking into ac-
(until convergence at < 1.5%) using a Monte Carlo simulation count morbidity and mortality following side-effects of nerve-
procedure (@Risk Pro 4.5, Palisade Corp., Newfield, NY). Model tissue vaccines.
predictions are reported using the means of the resulting prob-
ability distributions, with the 5th and 95th percentiles (90% Direct DALY score
confidence intervals) as the lower and upper bounds, respectively. The output of the predictive model provided the estimated
A full account of the technical details of the model, including annual number of deaths due to rabies. Age-structures and sex-
results of the sensitivity analysis, is available from: http://www. structures of rabies cases were obtained from seven reported
vet.ed.ac.uk/ctvm/Research/Appendices/appendices.html 3. studies (27–33). All parameter estimates and data sources re-
lated to the calculation of the rabies DALY score are given in
Disability-adjusted life year score Table 3 (web version only, available at: http://www.who.int/
The disability-adjusted life year (DALY) score is a standardized, bulletin). Using these parameter estimates, a DALY score for
comparative measure of the burden of disease. The DALY score rabies was determined using previously described methods (26,
for a particular condition is a composite score of the years of 34). Parameter variability was again incorporated by assign-
life lost (YLL) due to premature mortality and the years of life ing confidence distributions and using simulation software as
lived with a disability (YLD) caused by the condition (26). To described above. The 5th and 95th percentiles were used as the
calculate a DALY score for rabies the following components lower and upper bounds for the predicted scores.

362 Bulletin of the World Health Organization | May 2005, 83 (5)

 Research
Darryn L. Knobel et al. Burden of rabies in Africa and Asia

Indirect DALY burden a result of the disease. The costs due to rabies were considered
Evidence suggests that non-rabies induced morbidity and mor- under the following categories:
tality may constitute a sizeable proportion of the rabies burden • direct (medical) human costs from post-exposure treatment
in developing countries. Approximately one-third of all human • indirect (patient) costs from post-exposure treatment
rabies post-exposure treatments are carried out using crude • costs to control rabies among dogs
nerve-tissue vaccines (35), despite the occurrence of severe and • livestock losses
sometimes fatal allergic encephalomyelitic reactions (36–38). • surveillance costs.
Nerve-tissue vaccines were classified into two groups based
on differing incidence rates and clinicopathological signs of Table 4 and Table 5 give a breakdown of the costing data used
adverse reactions (37, 38). These were the Semple type (made in the economic analysis.
from phenol-treated sheep-brain or goat-brain tissue) and vac- For this analysis, direct medical costs included the cost of
cines derived from suckling-mouse brain. For the purpose of biologicals (rabies vaccines and immunoglobulin) and the cost
this preliminary analysis, disability weights for post-vaccination of their administration, including materials and staff salaries.
neurological reactions, used in the calculation of the YLDs, were Indirect costs included out-of-pocket expenses for patients,
taken as those reported for similar conditions by Murray & such as transport costs to and from rabies-treatment centres,
Lopez (39) (Table 3). The use of equivalent disability weights and loss of income while receiving treatment (40). Costs asso-
represents an admittedly crude first attempt to determine a ciated with the treatment of dog bites and the administration
YLD component of the DALY score for rabies; future attempts of antibiotics and tetanus immunizations were not included.
would benefit from a formal disability weighting procedure. Due to the erratic frequency of reporting, national numbers of
patients receiving post-exposure treatment annually were aver-
The economic burden of rabies aged over a period of 5 years (1996–2000). Countries for which
The mortality rate and DALY score provide estimates of the no reports could be found for this period were considered not to
burden of disease on human health. A second component of have treated any patients. Post-exposure treatment was catego-
the impact of disease is the economic cost incurred by society as rized by administration route (intramuscular or intradermal) on

Table 5. Indirect (patient) costs of post-exposure treatment and other costs associated with rabies. Data sources are available
from: http://www.vet.ed.ac.uk/ctvm/Research/Appendices/appendices.html

Estimate
Parameter Africa Asia Source
a
Indirect PET costs
No. of PET patient visits 680 000 32 400 000 Calculated from Table 4
Proportion of visits accompanied by an adult 0.4 0.4 2, 16
Total No. of visits (patients plus those accompanying them) 952 000 45 360 000 Calculated
Income loss
No. of working days lost per person per PET visit 0.5 0.5 15
Daily per capita Gross National Income US$ 1.87 US$ 3.50 41
Income loss per person per PET visit US$ 0.94 US$ 1.75 Calculated
Transport costs
Transport costs per person per visit US$ 2.00 US$ 3.80 40, M. Kaare, personal
communication, 2002
Dog rabies costs
Vaccination costs
No. dogs vaccinated annually 6 700 000 40 000 000 11–14, 23–27
Cost per dog vaccinated US$ 1.30 US$ 1.30 42
Population control costs
No. dogs killed annually 200 000 5 000 000 11–14, 23–27
Cost per dog killed US$ 5.00 US$ 5.00 43

Livestock losses
Total no. cattle 230 000 000 423 000 000 44
Rabies incidence rate/100 000 cattle 5 5 11–14, 23–27
Annual no. of cattle deaths from rabies 11 500 21 150 Calculated
Cost per head of cattle US$ 150.00 US$ 500.00 A. Shaw, unpublished data, 2002

Surveillance costs
No. rabies diagnostic tests per year 5 300 16 500 11–14, 23–27
Cost per test US$ 5.68 US$ 5.68 45
a
PET = post-exposure treatment.

Bulletin of the World Health Organization | May 2005, 83 (5) 363

 Research
Burden of rabies in Africa and Asia Darryn L. Knobel et al.

Table 6. Estimated human mortality caused by canine rabies in Africa and Asia

Asia
Model output India China Other Asia Africa
Urban Rural Urban Rural Urban Rural Urban Rural
Total population (millions) 284.7 732.2 459.1 816.1 295.7 525.4 294.2 498.1
Population at risk (millions) 284.7 710.4 459.1 498.3 295.7 409.1 294.2 340.1
No. bites from suspected rabid dogs (thousands) 409.4 893.4 660.1 626.7 425.2 514.5 374.3 427.8
No. of rabies deathsa 1 058 18 201 1 324 1 257 853 8 135 5 886 17 937
No. deaths/100 000 people 0.37 2.49 0.29 0.15 0.29 1.55 2.00 3.60
No. subregional deathsb 19 713 2336 9 489
(4 192–39 733) (565–5 049) (2 281–19 503)
No. regional deaths 31 539 (8 149–61 425) 23 705
(6 903–45 932)
Total no. deaths 55 270 (23 910–93 057)
Overall no. deaths/100 000 people 1.38 (0.60–2.33)
Predicted deaths in the absence of any 327 160 (166 904–525 427)
post-exposure treatment
a
Rabies deaths are the means of output probability distributions calculated independently and may therefore not sum exactly.
b
Figures in parentheses are the 5th and 95th percentiles of output probability distributions.

the basis that intradermal vaccination reduces costs by 60–80% and 9.7% in Asia); this figure was then applied to all countries
compared to the standard intramuscular regimen (41, 42). The in the region to predict the total number of dogs vaccinated.
total number of vaccine doses (i.e. injections) administered and Cost predictions were based on the use of a central-point vac-
the total number of visits made to rabies-treatment centres were cination system. Vaccination costs per dog include all com-
derived from national estimates of the proportion of patients ponents of campaign organization, public awareness efforts,
receiving treatment who were vaccinated with each schedule, and biological and material costs. Indirect costs borne by dog
after adjustment to account for patient drop-out during the owners were not included in the analysis.
course of treatment. Few published accounts dealing with Livestock losses to rabies can be significant; however, there
treatment-seeking behaviour and compliance could be found are few published estimates of rabies incidence in livestock. Sub-
so conservative estimates were made on the average number of mission of cattle specimens to central veterinary laboratories
visits per schedule: 3 instead of 5 for the Essen and Thai Red reveals an annual incidence of 0.5–2 deaths/100 000 head of
Cross regimens and 7 instead of 10–21 for the nerve-tissue cattle; this is certain to be a gross underestimate. Assuming a
vaccine schedules. The proportion of patients receiving the rate of underreporting of 10, and using the lower end of the
Zagreb schedule, based on available data, was negligible at the range of incidence rates to exclude transmission from wildlife
scale of this study. It was assumed that all children aged < 16 reservoirs, an estimated incidence of 5 deaths/100 000 cattle
years were accompanied by an adult. Assessment of transport is obtained.
costs and income loss include costs to both patients and those Insufficient data were available to enable parameter vari-
accompanying them. ability to be explicitly incorporated into the economic analysis.
Annual dog vaccination figures could not be found for all An attempt was made to model the uncertainty surrounding
countries so the average vaccination coverage of the estimated parameter estimates by inputting estimates as triangular dis-
national dog population was calculated for those countries that tributions (43), with the maxima and minima set as ± 10% of
submitted reports between 1996 and 2000 (10.3% in Africa each parameter’s values.

Table 7. Estimated disability-adjusted life year (DALY) score for rabies in Africa and Asia

DALY scorea
Component Africa Asia Total
b
Rabies deaths 747 558 (217 690–1 448 514) 994 607 (257 275–1 939 125) 1 743 015 (754 019–2 934 656)
Nerve-tissue vaccine reactions 360 (142–586) 44 525 (17 585–72 575) 44 885 (17 727–73 162)
Total 747 918 (217 954–1 449 014) 1 039 119 (302 324–1 983 646) 1 787 886 (799 615–2 984 109)
Total (assuming no post- 9 504 237 (4 848 684–15 264 050)
exposure treatment)
a
DALY scores are the means of output probability distributions calculated independently and may therefore not sum exactly.
b
Figures in parentheses are the 5th and 95th percentiles of output probability distributions.

364 Bulletin of the World Health Organization | May 2005, 83 (5)

 Research
Darryn L. Knobel et al. Burden of rabies in Africa and Asia

Results receiving post-exposure treatment is 1.25 million which is four

times lower than national estimates of 5 million; the predicted
Model outputs
number of deaths remains more than double that estimated
The results of the predicted human mortality, disability-
adjusted life year score and economic burden of rabies are by expert opinion. However, recent figures communicated
presented in Table 6, Table 7 and Table 8, respectively. The pre- by the Chinese authorities are more in line with the model’s
dicted number of human deaths remained within the reported predictions (Personal communication, 2004): more than 2000
confidence limits when the threshold dog density was cases of rabies in humans were reported in 2003, with an 80%
doubled to 9 dogs/km² (46 000; 90% confidence interval increase in the first quarter of 2004 (compared to the same
(CI) = 19 000–79 000). Setting the threshold density to period in 2003). Given the gross underreporting associated
0 (i.e. assuming that all people are at risk within areas with rabies deaths, this suggests that the model may in fact be
where canine rabies is endemic) also had little effect on the an underestimate of the true number, and the original predic-
model: in such a scenario the predicted number of deaths tion of approximately 10 000 deaths may prove more realistic.
was 67 000 (90% CI = 30 000–110 000). Detailed studies on dog ecology and rabies epidemiology and
treatment in China are necessary to resolve this.
Discussion WHO’s 1999 World survey of rabies (35) reported 1722
human rabies deaths from the study regions: 147 in Africa and
This paper provides the results of the first attempt at a quan- 1575 in Asia. The predicted figure of 55 000 deaths suggests
titative prediction of the burden of rabies in Africa and Asia.
that only 3% of human rabies deaths are recorded by central
Although we attempted to incorporate the entire range of
health authorities, a rate of underreporting of between 20
parameter variability, the final result is still likely to be an un-
times (Asia) and 160 times (Africa). This is in agreement with
derestimate of total mortality and morbidity caused by rabies in
findings from other studies using active surveillance methods
these regions. Only deaths due to canine rabies were assessed,
that found the incidence of human rabies is up to 100 times
omitting that fraction of human cases resulting from exposure
to rabid wild animals (35). In determining the number of hu- greater than what is officially recorded (6, 47).
mans at risk, only populations in areas where canine rabies is The burden of rabies is not evenly distributed across all
endemic were included. This ignores the possibility of sporadic sectors of society but is influenced by age-related and socioeco-
outbreaks in areas of low dog density or the introduction of nomic factors. The total cost (direct medical costs and indirect
the virus into a previously rabies-free population, as happened patient costs, excluding those of any accompanying people) of
on the Indonesian island of Flores in 1997 (44). an average post-exposure treatment course as determined in
The results of the modelling are consistent with those of this study is US$ 39.57 in Africa and US$ 49.41 in Asia. This
other studies that have estimated the true incidence of rabies at amounts to a substantial fraction of annual per capita gross
a national level. The recent national survey by the Association national income (GNI): 5.80% for the average person living in
for the Prevention and Control of Rabies in India (45) esti- Africa and 3.87% for someone living in Asia. Even if medical
mated a total of 18 500 human deaths from rabies in the coun- costs are fully subsidized by the government, out-of-pocket pa-
try each year, which is close to the predicted figure of 19 700 tient expenses still comprise 1.5–2% of per capita GNI. Often
in this report. The model also estimated the annual number government subsidies extend only to the provision of cheaper
of patients receiving post-exposure treatment in India as 1.07 nerve-tissue vaccines, with tissue-culture vaccines being pro-
million, which again is in good agreement with the reported vided only to those patients able to pay for them. There is there-
national estimate of 1.1 million (46). The model’s predictions fore an income-related risk factor in exposure to the side-effects
for China are less consistent: the predicted number of patients of nerve-tissue vaccines. This is compounded by occupational

Table 8. Estimated annual expenditure due to rabies

Costa
Category Africa Asia Total
b, c
PET costs 9.1 (8.2–10.0) 475.9 (435.0–520.5) 485.0 (443.4–530.1)
Direct (medical) 5.9 (5.2–6.6) 190.3 (171.4–210.5) 196.2 (176.9–216.7)
Indirect (patient) 3.2 (2.9–3.5) 285.6 (259.2–312.1) 288.7 (262.2–315.4)
Income loss 1.3 (1.2–1.4) 113.5 (104.4–122.9) 114.7 (105.7–124.1)
Transport costs 1.9 (1.7–2.1) 172.1 (154.1–190.5) 174.0 (155.9–192.5)
Dog rabies control costs 9.7 (8.8–10.6) 77.0 (71.5–82.3) 86.7 (80.7–92.8)
Vaccination costs 8.7 (7.8–9.6) 52.0 (47.1–57.0) 60.7 (55.4–66.3)
Population control costs 1.0 (0.9–1.1) 25.0 (22.5–27.5) 26.0 (23.4–28.6)
Livestock losses 1.7 (1.5–1.9) 10.5 (9.4–11.8) 12.3 (11.0–13.7)
Surveillance costs 0.03 (0.026–0.032) 0.09 (0.08–0.10) 0.12 (0.11–0.13)
Total 20.5 (19.3–21.8) 563.0 (520.0–605.8) 583.5 (540.1–626.3)
a
Costs are in millions of US$.
b
PET = post-exposure treatment.
c
Figures in parentheses are the 5th and 95th percentiles of output probability distributions.

Bulletin of the World Health Organization | May 2005, 83 (5) 365

 Research
Burden of rabies in Africa and Asia Darryn L. Knobel et al.

and socioeconomic risk factors in the initial exposure to infec- to be the most cost-effective in the medium–long term (4);
tion, further skewing the burden of rabies towards those sectors costs are typically recouped within 5–10 years, mainly through
of society least able to bear it (45, 48). The results of our study decreased expenditure on human post-exposure treatment. O
predict that there will be five times more rabies deaths in rural
areas than in urban areas. Children in particular are at a higher Acknowledgements
risk of exposure to rabid dogs. Typically, 30–50% of those re- The authors thank Dr Alex Wandeler and Dr Ursula Kayali for
ceiving post-exposure treatment are children aged < 16 years their comments on an earlier draft of this manuscript and Dr
(6, 49). Children are also more likely to suffer multiple bites Deborah Briggs for useful discussions.
and bites to the face and head, both of which carry a higher risk
of contracting rabies (31, 50). Funding: DK and SC are supported by a grant from the Well-
Rabies continues to impact human health despite the come Trust. JZ is supported by NCCR “North-South” IP-4,
existence of proven cost-effective control measures. Vaccinating which is funded by the Swiss National Science Foundation and
domestic dogs against rabies results in a significant reduction in the Swiss Development Cooperation.
the incidence of bites among the human population from dogs
suspected to be rabid, and this control strategy has been shown Competing interests: none declared.

Résumé
Réévaluation de la charge que représente la rage en Afrique et en Asie
Objectif Quantifier le fardeau économique et la charge pour de 1,74 million d’années de vie ajustées sur l’incapacité perdues
la santé publique que représente la rage canine endémique en chaque année (IC à 90% = 0,75-2,93). Quelque 0,04 million
Afrique et en Asie. d’années de vie ajustées sur l’incapacité supplémentaires sont
Méthodes Les données provenant de ces régions ont été en outre perdues du fait de la morbidité et de la mortalité dues
appliquées à un ensemble de modèles épidémiologiques et aux effets secondaires des vaccins préparés sur tissu nerveux. Le
économiques associés. On a estimé le nombre de personnes qui coût annuel estimatif de la rage est de US$ 583,5 millions (IC
seraient exposées à la rage canine endémique à partir des données à 90% = US$ 540,1 – 626,3 millions). Le coût des traitements
de la densité canine et le nombre des décès humains dus à la post-exposition supporté par les malades constitue l’essentiel des
rage au moyen d’un modèle de probabilité pas à pas utilisé pour dépenses, soit près de la moitié du coût total de la rage.
déterminer la probabilité qu’une personne mordue par un chien Conclusion La rage reste une maladie importante mais négligée
présumé enragé présente la maladie. La mortalité et la morbidité en Afrique et en Asie. Les écarts aux plans de l’accessibilité
liées à la rage fournies par les modèles ont été utilisées pour affiner financière et physique du traitement post-exposition et du risque
le calcul des années de vie ajustées sur l’incapacité (DALY) pour la d’exposition aux chiens enragés rendent inégale la distribution
maladie. Le coût social total de la maladie est présenté. de la charge de morbidité dans la société, les personnes les
Résultats On a estimé à 55 000 [intervalle de confiance (IC) à plus touchées étant les habitants des communautés rurales
90% = 24 000-93 000] le nombre annuel des décès humains dus à défavorisées, et en particulier les enfants.
la rage canine endémique. Les décès dus à la rage sont responsables

Resumen
Reevaluación de la carga de rabia en África y Asia
Objetivo Cuantificar la carga que supone la rabia canina endémica (IC) del 90% = 24 000-93 000). Las defunciones por rabia causan
en África y Asia en términos económicos y de salud pública. unas pérdidas de 1,74 millones de AVAD cada año (IC90% = 0,75-
Métodos Se aplicaron datos de esas regiones a un conjunto de 2,93). Otros 40 000 AVAD se pierden como consecuencia de la
modelos epidemiológicos y económicos relacionados. La población morbilidad y mortalidad asociadas a los efectos colaterales de las
humana expuesta al riesgo de sufrir rabia canina endémica se vacunas obtenidas con tejido nervioso. El costo anual estimado de
predijo a partir de los datos disponibles sobre la densidad de perros, la rabia asciende a US$ 583,5 millones (IC90% = US$ 540,1 - 626,3
y las defunciones por rabia humana se estimaron usando una serie millones). Los costos del tratamiento postexposición asumidos por
de pasos probabilísticos para determinar el riesgo de aparición los pacientes constituyen el grueso del gasto correspondiente, pues
de rabia clínica en una persona que hubiera sido mordida por un suponen casi la mitad del costo total de la rabia.
perro sospechoso de albergar rabia. Los resultados modelizados Conclusión La rabia sigue siendo una importante y sin embargo
sobre la mortalidad y la morbilidad asociadas a la rabia se usaron descuidada enfermedad en África y Asia. Las disparidades en la
para obtener un valor mejorado de los años de vida ajustados en asequibilidad y accesibilidad del tratamiento postexposición y en el
función de la discapacidad (AVAD) para la enfermedad. Se presenta riesgo de exposición a perros rabiosos se traducen en una distribución
el costo social total asociado a esta dolencia. asimétrica de la carga de morbilidad en la sociedad, de tal manera
Resultados La mortalidad humana por rabia canina endémica se que el impacto principal de la enfermedad recae en los habitantes de
estimó en 55 000 defunciones anuales (intervalo de confianza las comunidades rurales pobres, sobre todo en los niños.

366 Bulletin of the World Health Organization | May 2005, 83 (5)

 Research
Darryn L. Knobel et al. Burden of rabies in Africa and Asia

References
1. World Health Organization. World Survey of Rabies No. 32 for the year 16. Kitala PM, McDermott JJ, Coleman PG, Dye C. Comparison of vaccination
1996. Geneva: WHO; 1998. WHO document EMC/ZDI/98.4. strategies for the control of dog rabies in Machakos District, Kenya.
2. Bögel K, Meslin F-X. Economics of human and canine rabies elimination: Epidemiology and Infection 2002;129:215-22.
guidelines for programme orientation. Bulletin of the World Health 17. Bishop GC. Canine rabies in South Africa. In: Bingham J, Bishop GC, King AA,
Organization 1990;68:281-91. editors. Proceedings of the Third International Conference of the Southern
3. Cleaveland S, Kaare M, Tiringa P, Mlengeya T, Barrat J. A dog rabies and East African Rabies Group. Harare: Veterinary Research Laboratory;
vaccination campaign in rural Africa: impact on the incidence of dog rabies 1995. p. 104-11.
and human dog-bite injuries. Vaccine 2003;21:1965-73. 18. Brooks R. Survey of the dog population of Zimbabwe and its level of rabies
4. Meslin F-X, Fishbein DB, Matter HC. Rationale and prospects for rabies vaccination. Veterinary Record 1990;127:592-6.
elimination in developing countries. In: Rupprecht CE, Dietzschold B, 19. Cleaveland S, Dye C. Maintenance of a microparasite infecting several host
Koprowski H, editors. Lyssaviruses. Berlin: Springer Verlag; 1994. p. 1-26. species: rabies in the Serengeti. Parasitology 1995;111 Suppl:S33-47.
5. Warrell DA, Warrell MJ. Human rabies: a continuing challenge in the tropical 20. Foggin CM. Rabies and rabies-related viruses in Zimbabwe: historical,
world. Schweizerische Medizinische Wochenschrift 1995;125:879-85. virological and ecological aspects. Harare: University of Zimbabwe; 1988.
6. Cleaveland S, Fèvre EM, Kaare M, Coleman PG. Estimating human rabies 21. Childs JE, Robinson LE, Sadek R, Madden A, Miranda ME, Miranda NL.
mortality in the United Republic of Tanzania from dog bite injuries. Bulletin Density estimates of rural dog populations and an assessment of marking
of the World Health Organization 2002;80:304-10. methods during a rabies vaccination campaign in the Philippines. Preventive
7. Fekadu M. Human rabies surveillance and control in Ethiopia. In: Kitala P, Veterinary Medicine 1998;33:207-18.
Perry B, Barrat J, King AA, editors. Proceedings of the Fourth International 22. Perry BD, Brooks R, Foggin CM, Bleakley J, Johnston DH, Hill FW. A baiting
Conference of the Southern and East African Rabies Group, Nairobi, Kenya, system suitable for the delivery of oral rabies vaccine to dog populations in
4-6 March, 7-8. Lyon : Editions Fondation Mérieux ; 1997. p. 78-9. Zimbabwe. Veterinary Record 1988;123:76-9.
8. Snow RW, Craig M, Deichmann U, Marsh K. Estimating mortality, morbidity 23. Pal SK. Population ecology of free-ranging urban dogs in West Bengal, India.
and disability due to malaria among Africa’s non-pregnant population. Acta Theriologica 2001;46:69-78.
Bulletin of the World Health Organization 1999;77:624-40. 24. World Health Organization. World survey of rabies No. 34 for the year 1998.
9. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Global burden of Geneva: WHO; 1999. WHO document CSR/APH/99.6.
tuberculosis — estimated incidence, prevalence, and mortality by country. 25. Lin FT, Lina N. Developments in the production and application of rabies
JAMA 1999;282:677-86. vaccine for human use in China. Tropical Doctor 2000;30:14-6.
10. Roth F, Zinsstag J, Orkhon D, Chimed-Ochir G, Hutton G, Cosivi O, et al. 26. Murray CJL. Quantifying the burden of disease: the technical basis for
Human health benefits from livestock vaccination for brucellosis: a case disability-adjusted life years. Bulletin of the World Health Organization
study. Bulletin of the World Health Organization 2003;81:867-76. 1994;72:429-45.
11. Baltazard M, Ghodssi M. Prevention of human rabies: treatment of persons 27. Yimer E, Newayeselassie B, Teferra G, Mekonnen Y, Bogale Y, Zewde B, et al.
bitten by rabid wolves in Iran. Bulletin of the World Health Organization Situation of rabies in Ethiopia: a retrospective study 1990-2000. Ethiopian
1954;10:797-802. Journal of Health Development 2002;16:105-12.
12. Shah U, Jaswal GS. Victims of a rabid wolf in India: effect of severity and 28. Fekadu M. Rabies in Ethiopia. American Journal of Epidemiology
location of bites on development of rabies. Journal of Infectious Diseases 1982;115:266-73.
1976;134:25-9. 29. Ayalew Y. Analysis of 159 human rabies cases in Ethiopia. In: Kuwert E,
13. Deichmann U. Africa population database, version 3. Available from: Mérieux C, Koprowski H, Bögel K, editors. Rabies in the tropics. Berlin:
http://grid2.cr.usgs.gov/globalpop/ Springer-Verlag; 1985. p. 481-4.
14. Deichmann U. Asia population database. Available from: http://grid2.cr. 30. Belcher DW, Wurapa FK, Atuora DO. Endemic rabies in Ghana: epidemiology
usgs.gov/globalpop/ and control measures. American Journal of Tropical Medicine and Hygiene
15. United Nations Population Division. World population prospects: the 2002 1976;25:724-9.
revision, 2003. Available from: http://esa.un.org/unpp

Bulletin of the World Health Organization | May 2005, 83 (5) 367

 Research
Burden of rabies in Africa and Asia Darryn L. Knobel et al.

31. Kureishi A, Xu LZ, Wu H, Stiver HG. Rabies in China: recommendations for 42. Kamoltham T, Khawplod P, Wilde H. Rabies intradermal post-exposure
control. Bulletin of the World Health Organization 1992;70:443-50. vaccination of humans using reconstituted and stored vaccine. Vaccine
32. Lakhanpal U, Sharma RC. An epidemiological study of 177 cases of human 2002;20:3272-6.
rabies. International Journal of Epidemiology 1985;14:614-7. 43. Vose D. Risk analysis: a quantitative guide. 2nd ed. New York: John
33. Singh J, Jain DC, Bhatia R, Ichhpujani RL, Harit AK, Panda RC, et al. Wiley Sons; 2000.
Epidemiological characteristics of rabies in Delhi and surrounding areas, 44. Bingham, J. Rabies on Flores Island, Indonesia: is eradication possible in
1998. Indian Pediatrics 2001;38:1354-60. the near future? In: Dodet B, Meslin F-X, Heseltine E, editors. Proceedings
34. Mathers CD, Vos T, Lopez AD, Salomon J, Ezatti M. National burden of disease of the Fourth International Symposium of Rabies Control in Asia. Paris:
studies: a practical guide. 2nd edition. Geneva: World Health Organization; John Libbey Eurotext; 2001. p.148-55.
2001 (Global Programme on Evidence for Health Policy). 45. Association for the Prevention and Control of Rabies in India. Assessing
35. World Health Organization. World survey of rabies No. 35 for the year 1999. the burden of rabies in India: a national multicentric survey. Progress
Geneva: WHO; 2002. WHO document CDS/CSR/EPH/2002.10. report. Bhubaneswar, India: Association for the Prevention and Control of
36. Bahri F, Letaief A, Ernez M, Elouni J, Chekir T, Ben Ammou S, et al. Rabies; 2003.
Neurological complications in adults following rabies vaccine prepared from 46. Ichhpujani RL, Bhardwaj M, Chhabra M, Datta KK. Rabies in humans
animal brains. Presse Medicale 1996;25:491-3. in India. In: Dodet B, Meslin F-X, Heseltine E. Proceedings of the Fourth
37. Held JR, Adaros HL. Neurological disease in man following administration International Symposium of Rabies Control in Asia. Montrouge: John Libbey
of suckling mouse brain antirabies vaccine. Bulletin of the World Health Eurotext; 2001. p. 212-3.
Organization 1972;46:321-7. 47. Kitala PM, McDermott JJ, Kyule MN, Gathuma JM. Community-based active
38. Swaddiwuthipong W, Weniger BG, Wattanasri S, Warrell MJ. A high rate of surveillance for rabies in Machakos District, Kenya. Preventive Veterinary
neurological complications following Semple anti-rabies vaccine. Transactions Medicine 2000;44:73-85.
of the Royal Society of Tropical Medicine and Hygiene 1988;82:472-5. 48. Tang Q, Xiuqin Z, Zhi D. Human epidemiology and risk factors for rabies. In:
39. Murray CJL, Lopez AD. Global health statistics: a compendium of incidence, Dodet B, Meslin F-X, editors. Proceedings of the Third International
prevalence, and mortality estimates for over 200 conditions. Vol. 2. Boston Conference on Rabies Control in Asia. Paris: Elsevier; 1997. p. 130-6.
(MA): Harvard School of Public Health; 1996. 49. Dutta JK. Safety and tolerance of purified Vero rabies vaccine (Verorab):
40. Meltzer MI, Rupprecht CE. A review of the economics of the prevention an Indian experience. Journal of the Association for Prevention and Control
and control of rabies. Part 1: Global impact and rabies in humans. of Rabies in India 1999;1:26-30.
Pharmacoeconomics 1998;14:365-83. 50. Pancharoen C, Thisyakorn U, Lawtongkum W, Wilde H. Rabies exposures in
41. Pradhan S, Satapathy DM, Das BC. Pharmacoeconomics of antirabies Thai children. Wilderness and Environmental Medicine 2001;12:239-43.
vaccine treatment – the case for intradermal administration of cell culture
vaccines. Indian Journal of Internal Medicine 2001;11: 5 Suppl 1-12.

368 Bulletin of the World Health Organization | May 2005, 83 (5)

 Research
Darryn L. Knobel et al. Burden of rabies in Africa and Asia

Table 2. Parameter estimates, probability distributions and data sources used in the prediction of human deaths from rabies
from injury data on dog bites (data sources available from: http://www.vet.ed.ac.uk/ctvm/Research/Appendices/appendices.html)

Parameter estimates
Param- Description Probability Africa Asia Source
eter distribution of data
Urban Rural Urban Rural
Annual incidence of Trigen:a 1–6
suspect bites from Practical minimum 6 6 50 15
rabid dogs per Most likely 100 100 120 100
100 000 humans Practical maximum 227 227 250 250
Bottom and top percentiles 0%; 95% 0%; 95% 5%; 95% 5%; 95%
P10 Probability of an Trigen: 2, 7
individual bitten by a Practical minimum 0.80 0.55 0.95 0.70
dog suspected to be Most likely 0.85 0.60 .097 0.75
rabid receiving Practical maximum 0.90 0.60 1.00 0.80
successful post- Bottom and top percentiles 5%; 95% 5%; 95% 10%; 100% 10%; 90%
exposure treatment
P1 Probability of a Beta:b India Other Asia 2, 3, 8–15
suspected rabid dog No. of suspect dogs examined 9 285 5 863 59 588
being confirmed No. confirmed rabid 5 291 2 906 22 923
rabid on laboratory P1 0.64 0.50 0.38
diagnosis
P2 Probability of a bite Point probability 0.07 2, 16, 17,
to the head or neck E. Miranda
unpublished
data, 2003
P3 Probability of a bite to Point probability 0.38 2, 16, 17,
the upper extremity E. Miranda
(arm or hand) unpublished
data, 2003
P4 Probability of a bite Point probability 0.06 2, 16, 17,
injury to the trunk of E. Miranda
the body unpublished
data 2003
P5 Probability of a bite to Point probability 0.49 2, 16, 17,
the lower extremity E. Miranda
(leg or foot) unpublished
data, 2003
P6 Probability of Triangular: 18–21
developing rabies Minimum 0.30
following a bite to the Most likely 0.45
head by a rabid dog Maximum 0.60
P7 Probability of Triangular: 18–21
developing rabies Minimum 0.15
following a bite to an Most likely 0.28
upper extremity by a Maximum 0.40
rabid dog
P8 Probability of Triangular: 18–21
developing rabies Minimum 0
following a bite to the Most likely 0.05
trunk by a rabid dog Maximum 0.10
P9 Probability of Triangular: 18–21
developing rabies Minimum 0
following a bite to a Most likely 0.05
lower extremity by Maximum 0.10
a rabid dog
a
The Trigen distribution avoids the use of absolute maxima and minima by allowing the specification of a likely range for the parameter together with an estimation
of the probability that the parameter will fall outside this range (top and bottom percentiles).
b
The Beta distribution is a binomial process allowing estimation of the probability of success p, given s successes from n trials. Assuming a non-informative Uniform
(0, 1) prior, the Beta distribution takes the form p = Beta (s + 1, n – s + 1). See reference 22 for a discussion of commonly used probability distributions.

Bulletin of the World Health Organization | May 2005, 83 (5) A

 Research
Burden of rabies in Africa and Asia Darryn L. Knobel et al.

Table 3. Parameter estimates and data sources used to calculate the disability-adjusted life year (DALY) score for rabies (data
sources available from: http://www.vet.ed.ac.uk/ctvm/Research/Appendices/appendices.html)

Estimate
Parameter Africa Asia Source
a
No. human rabies deaths per year 23 788 (7 280–44 112) 30 942 (6 017–61 657) Model output (Table 3)
b
No. PET cases per year 200 000 7 500 000 12–14, 23–27
No. (%) of PET patients receiving nerve-tissue vaccines 20 000 (10) 2 475 000 (33) 12–14, 23–27
c
No. (%) of NTV patients receiving Semple-type 16 000 (80) 1 980 000 (80) 12–14, 23–27
nerve-tissue vaccine
No. (%) of NTV patients receiving suckling-mouse brain 4 000 (20) 495 000 (20) 12–14, 23–27
nerve tissue vaccine
Rate of neurological complications per 100 patients Triangular: 16, 28–31
receiving Semple vaccine Minimum = 0.035; Most likely = 0.40;
Maximum = 0.83
Case–fatality rate for cases of Semple neurological 0.17 30, 32, 33
complications
Disability weight for Semple neurological complications 0.613 34 (disability weight
for an episode of
bacterial meningitis)
Disability duration for Semple neurological complications Triangular: 35
Minimum = 1 day; Most likely = 8 days;
Maximum = 1 year
Rate of neurological complications per 100 patients Triangular: 29, 31, 35, 36
receiving suckling-mouse brain vaccine Minimum = 0.013; Most likely = 0.03;
Maximum = 0.08
Case–fatality rate for cases of neurological complications 0.22 29
from suckling-mouse brain vaccine
Disability weight for suckling-mouse brain neurological 0.725 34 (disability weight
complications for an injured spinal
cord)
Disability duration for suckling-mouse brain neurological 200 days 37
complications
DALY formula parameters
Discount rate r 0.03 38
Age-weighting correction constant C 0.1658 38
Age-weighting function constant 0.04 38
a
Values in parentheses are 90% confidence intervals.
b
PET = post-exposure treatment.
c
NTV = nerve-tissue vaccine.

B Bulletin of the World Health Organization | May 2005, 83 (5)