11

Methods for Burden Methods for Burden Assessment: EchinococcosisAssessment: Echinococcosis

Paul Torgerson Universität ZürichInstitut für Parasitologie

22

Highly Endemic Endemic SporadicF: Free PF: Provisionally Free

FF

PF PF

F

O O

33

44

Methods of burden assessment• Economic modelling techniques• Financial estimates

– Zoonoses– Animal and human health costs

• Non financial estimates– Levels the playing field between chronic and acute

disease– Human life has the same value in rich and poor

countries– WHO preferred instrument is the DALY

• Used to compare different disease burdens and hence drive policy

• Developed stochastic techniques to model the uncertainty of source data

55

EconomicsEconomicsCostsCosts--LivestockLivestock

Animal production lossesAnimal production lossesDecreased food conversion efficiencyDecreased food conversion efficiencyMortality or morbidityMortality or morbidityLower reproductive performanceLower reproductive performanceLower milk yieldLower milk yield

66

EconomicsEconomicsCostsCosts--Human DiseaseHuman Disease

Cost of TreatmentCost of Treatment–– SurgerySurgery–– Medical treatmentMedical treatment–– ConvalescenceConvalescence

MorbidityMorbidity–– Time off workTime off work–– Less productiveLess productive

MortalityMortality–– Economic effects of deathEconomic effects of death

77

Data SourcesData Sources

OIE reportsOIE reports–– Gross underreportingGross underreporting

Literature reportsLiterature reports–– Method of collectionMethod of collection–– BiasedBiased

Representative surveysRepresentative surveys

88

OIE DataOIE DataCountry 2003 2002 2001 2000 1999 1998 1997 1996

Albania 6

Belarus 8 11 55

Bolivia 5

Bosnia & Herzegovina 34 24 13 3

Bulgaria 639 684 628 662 691 591

Italy

Jordan 2 14 21 8

Kenya 18

Kyrgystan 470 477 566 573

Lithuania 2 4 4 4 1 4 4 8

Palestinian Auth. 10

Peru 1,326

Sudan 9

Syria 24

Tunisia 100 162 157 196 154 227 285 299

Uzbekistan 1,428 1,321 1,130 976

99

Data from literatureData from literature

CountryYears evaluated

Human incidence(annual incidence per

100,000 pop.)

Reference

Austria 1983-1992 0.21-0.67 Auer and Aspöck, 1993

Australia 1991-1994 0.23 Longbottom and Hargreaves, 1995

Corsica 1966-1970 10 Economides and Thrasou, 1999

Greece 1994 3.4 Economides and Thrasou, 1999

Italy (Entire) 1980-1984 1.92 Gabriele et al., 1997

Italy (regional) Early 2000s 9.7-Sardinia1.57- Emilia Ranagna

2.30-Sicily2.33-Apulia

Gabriele et al., 2004

Portugal 1998-2000 1.82 Seimenis, 2003

Spain 1996 0.9 Garcia, 1997

Switzerland 1984-1992 0.38(0.09- Swiss nationals

1.39- foreigners)

Eckert et al., 1995

U.K. (Powys) 1984-1990 2.3 Lloyd et al., 1998

1010

Sources of Data Sources of Data UzbekistanUzbekistan

OIE reports OIE reports –– 1999 1999 -- 1428 cases1428 cases–– 2002 2002 -- 15001500

Government figuresGovernment figures–– 2000 2000 –– 1435 cases1435 cases–– 2001 2001 -- 819819

Hospital records search and case findingHospital records search and case finding11

–– 2000 2000 –– 4636 cases4636 cases–– 2001 2001 –– 4089 cases4089 cases

1. Nazirov et al. (2002) Echinococcosis in Uzbekistan: types of problems and methods to improve traetment Uzbek Medical Journal 2-3, pp 2-5 (in Russian)

1111

Sources of Data Sources of Data JordanJordan

OIE reports OIE reports –– 19961996--20042004–– 00--21 cases per year21 cases per year

Hospital records search and case Hospital records search and case findingfinding–– 1995 1995 –– c 128 cases per yearc 128 cases per year

1212

UncertaintyUncertainty

Some costs well definedSome costs well definedOther costs ??Other costs ??–– May represent the largest lossesMay represent the largest losses

Poor or inadequate reportingPoor or inadequate reportingSample sizeSample sizeDiagnostic test efficiencyDiagnostic test efficiencyAttributable morbidityAttributable morbidityMonteMonte--Carlo techniquesCarlo techniques–– Randomly vary each Randomly vary each ““unknownunknown””

1313

Estimating unknown Data

y = 1.5476xR2 = 0.8927

0

5000

10000

15000

20000

25000

30000

0 2000 4000 6000 8000 10000 12000 14000 16000

Cost of Treatment

Per

Cap

ita G

NI

For Global Burden Estimates : Cost of treatment in each country varied as the linear predictor (mean + SEM)

1414

Estimating unknown dataEstimating unknown data

Other modelling techniquesOther modelling techniquesFrom published prevalence inFrom published prevalence in–– Dogs c 10%Dogs c 10%–– Sheep c 50%Sheep c 50%Estimate Incidence in humansEstimate Incidence in humans–– c10 cases/100,000/yearc10 cases/100,000/year–– Over estimates some (Over estimates some (egeg Muslim)Muslim)–– Under estimates Under estimates egeg ChineseChinese

1515

Hospital CostsHospital Costs

Cost of treatment and medicationCost of treatment and medication

Annual number of casesAnnual number of cases~~mean (mean (++SEM)SEM)

x x

Cost of representative sampleCost of representative sample~~mean (mean (++SEM)SEM)

1616

525.38 597.03 668.68 740.32 811.97

Cost of Treatment

743.27 846.52 949.77 1,053.02 1,156.27

Number of cases

437,891.93 544,197.65 650,503.37 756,809.09 863,114.81

Total Treatment Costs

X

=

95% CIs

1717

Other Human Health Other Human Health CCostsosts

Long term ill healthLong term ill healthDo patients fully recover?Do patients fully recover?Perhaps not?Perhaps not?Permanent decrease in Permanent decrease in quality of life.quality of life.–– How much?How much?–– Needs good studies to Needs good studies to

accurately defineaccurately defineInfected but not treatedInfected but not treated–– Sub clinical diseaseSub clinical disease

Some patients die!Some patients die!–– Capital ApproachCapital Approach–– Willingness to pay

Infected people are poorer!

Willingness to pay

1818

Echinococcosis in JordanEchinococcosis in JordanTotal CostsTotal Costs

0 2,500 ,000 5,000 ,000 7,500 ,000 10,000,000

Human health cost

Animal health cost

Total Cost

Rel

ativ

e pr

obab

ility

Cost ($US millions)

All calculations should be discounted for future values

1919

Non financial instrumentsNon financial instruments

HALYs, DALYs, QALYsHALYs, DALYs, QALYsAll measures of loss of healthAll measures of loss of healthWHO preferred measure is DALYWHO preferred measure is DALY–– Disability Adjusted Life YearDisability Adjusted Life Year–– Measures number of Measures number of ““years of full healthyears of full health””

lost due to diseaselost due to disease

2020

DALYDALY

Length of time lived with morbidityLength of time lived with morbidityDiscounted forDiscounted for–– Disability weightDisability weight–– Age of onsetAge of onset–– Losses in future years at todayLosses in future years at today’’s ratess rates

Numbers of healthy years lostNumbers of healthy years lost

2121

Morbidity costsMorbidity costsSFSF--12 results 12 results -- ChinaChina

0.00

5.00

10.00

15.00

20.00

25.00

Physical Com ponentSum m ary

Mental Com ponentSum m ary

Mea

n Sc

ore

Control

EchinococcosisPos.

2222

Country(Year)

Number of patients

Cure Morbidity Relapse Death Reference

Greece(1984-1990)

56 40 (72%) 13 (23%) 3 (5%) 0 (0%) (14)

Italy(1950-1987)

298 244 (82%) 27 (9%) 15 (5%) 12 (4%) (15)

Turkey(1992-1999)

95 32 (34%) 38 (40%) 24 (25%) 1 (1%) (16)

Turkey(1990-1995)

108 88 (81%) 19 (18%) 0 (0%) 1 (1%) (17)

Greece(1985-1990)

67 59 (86%) 4 (6%) 3 (6%) 1 (2%) (18)

Italy(1982-1994)

89 70 (79%) 17 (19%) 1 (1%) 1 (1%) (19)

Total 713 533 (75%) 118 (17%) 46 (6%) 16 (2%)

2323

Disability WeightsDisability WeightsEchinococcosisEchinococcosisDisease free liver cancer (0.200) (improve after Disease free liver cancer (0.200) (improve after surgery)surgery)Preterminal liver cancer (0.239) (Post surgical Preterminal liver cancer (0.239) (Post surgical conditions)conditions)Terminal liver cancer (0.809) (Recurrent disease, Terminal liver cancer (0.809) (Recurrent disease, multiorgans etc)multiorgans etc)Death 1Death 1Weights and duration assigned using a multinomial Weights and duration assigned using a multinomial distribution with relative proportions based on distribution with relative proportions based on published surgical studies.published surgical studies.

2424

Country Years Average age of onset/detection

Reference

China 2001-2003 35 (3)

Jordan 1994-2000 31-45 (22)

Kenya (Turkana) 1980s 21-30 (23)

Morocco 2000-2001 32 (24)

Turkey 1992-1999 44 (16)

Uruguay 1991-1992 45 (25)

Kyrgystan 1991-2000 22 (26)

2525

DALYs lost in Serchu, China county due to echinococcosis

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

< 34

,000

34,0

01 -

36,0

00

36,0

01 -

38,0

00

38,0

01 -

40,0

00

40,0

01 -

42,0

00

42,0

01 -

44,0

00

44,0

01 -

46,0

00

46,0

01 -

48,0

00

48,0

01 -

50,0

00

50,0

01 -

52,0

00

52,0

01 -

54,0

00

54,0

01 -

56,0

00

56,0

01 -

58,0

00

58,0

01 -

60,0

00

60,0

01 -

62,0

00

62,0

01 -

64,0

00

64,0

01 -

66,0

00

66,0

01 -

68,0

00

68,0

01 -

70,0

00

70,0

01 -

72,0

00

72,0

01 -

74,0

00

>74,

000

DALYS lost

Prob

abili

ty

c 0.81 DALYs lost per person

2626

Economic lossesEconomic lossesAttributable loses onlyAttributable loses onlyTrue losses are only those that are True losses are only those that are preventablepreventableCystic echinococcosis and cysticercosisCystic echinococcosis and cysticercosis–– Eliminated by veterinary public health Eliminated by veterinary public health

programmesprogrammes–– Essentially entire burden is preventableEssentially entire burden is preventable

2727

Global Burden of Global Burden of EchinococcosisEchinococcosis

US$4.1 Billion (adjusted for US$4.1 Billion (adjusted for underreporting, PPE estimate)underreporting, PPE estimate)54% Human costs54% Human costs–– >1.0 million >1.0 million DALYsDALYs–– c 200,000 cases per yearc 200,000 cases per year

46% Animal health costs 46% Animal health costs

2828

DALYsLeprosy 199,000

Onchocerciasis 484,000Dengue 616,000Chagas 667,000

Cystic echinococcosis 1,079,038Trypanosomiasis 1,525,000Schistosomiasis 1,702,000Leishmaniasis 2,090,000

Lymphatic filariasis 5,777,000TB 34,736,000

Malaria 46,486,000Japanese encephalitis 709,000

Trachoma 2,329,000Ascariasis 1,817,000Trichuriasis 1,006,000Hookworm 59,000

2929

ConclusionsConclusions

Financial estimates estimate total burden of Financial estimates estimate total burden of disease including animal health costsdisease including animal health costsPurchasing power equivalents give a better idea of Purchasing power equivalents give a better idea of disease burden in poor countriesdisease burden in poor countriesSuch results can be used to implement cost sharing Such results can be used to implement cost sharing between sectorsbetween sectorsDALYs for zoonoses can indicate priorities DALYs for zoonoses can indicate priorities compared to other diseasescompared to other diseasesResults of cost sharing can indicate the true cost Results of cost sharing can indicate the true cost benefit to health servicesbenefit to health servicesStochastic and risk analysis techniques are powerful Stochastic and risk analysis techniques are powerful tools to model uncertainty.tools to model uncertainty.

3030

Thank You

Dr Christine BudkeDr Christine BudkeNational Institutes of Health National Institutes of Health (U.S.A.) and the National (U.S.A.) and the National Science Foundation (U.S.A.) Science Foundation (U.S.A.) (Ecology of Infectious Diseases (Ecology of Infectious Diseases Programme)Programme)European UnionEuropean UnionINTASINTASSwiss National Science FundsSwiss National Science Funds

3131

Numbers of surgical cases of cystic echinococcosisKazakhstan 1984-2001

0

400

800

1984 1988 1992 1996 2001

USSRUSSR KazakhstanKazakhstan

СССР Казакхстан