int. j. epidemiol. 1997 cookson 212 9

8/13/2019 Int. J. Epidemiol. 1997 Cookson 212 9

1/8

212

International Journal of Epidemiology

International Epidemio logical Associat ion 1997

Vol. 26, No. 1

Printed in Great Britain

The seventh pandemic of cholera reached Latin

America in 1991.1 In Argentina, the first cases were

identified in February 1992 in the north2 where out-

breaks continued during the next two summers. In late

1994, the Argentine regulatory agency licensed the new

live oral cholera vaccine, CVD 103-HgR. Questions then

arose as to what segment of this population might be

immunized.

Of the cholera cases that have occurred in Argentina,

94% were in the northern provinces of Salta and Jujuy.3

The populations of Salta and Jujuy are culturally and

socioeconomically diverse.2,4 In Salta and Jujuy, chol-

era has been virtually confined to indigenous people

and migrants who live under marginal conditions and

make their livelihood as agricultural workers.57 Their

working conditions in the fields are characterized by

even fewer basic necessities than in their ramshackledhomes. In considering methods of cholera control in

Argentina, one must scrutinize the epidemiology of

cholera among these workers.

The repetitive outbreaks of cholera in a limited pro-

portion of the population of northern Argentina and the

licensing of CVD 103-HgR prompted a cost-benefit

analysis. We consider only treatment costs directly

borne by the Ministry of Health; not considered are the

* The Center for Vaccine Development and the Division of Geo-

graphic Medicine, Department of Medicine, University of Maryland

School of Medicine, Baltimore, MD 21201, USA.

** Current Address: Centers for Disease Control and Prevention

(CDC), 1600 Clifton Road, Mailstop E69, Atlanta, GA 30333, USA. FUNCEI (Central Foundation of Infectious Diseases), Buenos Aires,

Argentina. Ministerio de Salud Pblica, Salta, Argentina. Ministerio de Salud Pblica, Jujuy, Argentina.

A Cost-Benefit Analysis ofProgrammatic Use of

CVD 103-HgR Live Oral CholeraVaccine in a High-Risk PopulationSUSAN TEMPORADO COOKSON,*,** DANIEL STAMBOULIAN, JOSE DEMONTE, LUIS QUERO,

CARMEN MARTINEZ DE ARQUIZA, ALBERTO ALEMAN, ALEJANDRO LEPETIC AND

MYRON M LEVINE*

Cookson S T (The Center for Vaccine Development and the Division of Geographic Medicine, Department of Medicine,University of Maryland School of Medicine, Baltimore, MD 21201, USA), Stamboulian D, Demonte J, Quero L, Martinezde Arquiza C, Aleman A, Lepetic A and Levine M M. A cost-benefit analysis of programmatic use of CVD 103-HgR liveoral cholera vaccine in a high-risk population. International Journal of Epidemiology1997; 26: 212219.Background. Cholera spread to Latin America in 1991; subsequently, cholera vaccination was considered as an interim

intervention until long-term solutions involving improved water supplies and sanitation could be introduced. Threesuccessive summer cholera outbreaks in northern Argentina and the licensing of the new single-dose oral cholera vac-cine, CVD 103-HgR, raised questions of the cost and benefit of using this new vaccine.Methods. This study explored the potential benefits to the Argentine Ministry of Health of treatment costs averted, versusthe costs of vaccination with CVD 103-HgR in the relatively confined population of northern Argentina affected by thecholera outbreaks. Water supplies and sanitation in this area are poor but a credible infrastructure for vaccine delivery exists.Results. In our cost-benefit model of a 3-year period (19921994) with an annual incidence of 2.5 case-patients per1000 population and assumptions of vaccine efficacy of 75% and coverage of 75%, vaccination of targeted high riskgroups would prevent 1265 cases.Conclusion. Assuming a cost of US$602 per treated case and of US$1.50 per dose of vaccine, the total discountedsavings from use of vaccine in the targeted groups would be US$132 100. The projected savings would be altered lessby vaccine coverage (range 7590%) or efficacy (6085%) changes than by disease incidence changes. Our analysisunderestimated the true costs of cholera in Argentina because we included only medical expenditures; indirect losses totrade and tourism had the greatest economic impact. However, vaccination with CVD 103-HgR was still cost-beneficialin the base case.

Keywords: cost-benefit analysis, cholera vaccine, attenuated vaccines, cholera, disease outbreaks


2/8

substantial economic losses to tourism and agriculturalexports during the cholera epidemics of 19921994.

MATERIALS AND METHODS

The population at greatest risk of contracting cholera

was determined for the three cholera outbreaks in Ar-

gentina. A conservative level of vaccine coverage and

efficacy and the incidence rate in this population were

used in the base case. The actual medical cost of the

three cholera outbreaks in these areas was calculated.

An estimate of vaccine costs was determined. These ex-

penditures were used to calculate the medical costs ofthe base case with and without vaccination for a hypo-

thetical 3-year period. Sensitivity analyses were per-

formed using additional incidence rates and a range of

vaccine coverages and vaccine efficacies.

Cholera Disease Burden in Salta and Jujuy

In the summers of 1992 through 1994, cholera appeared

in northern Argentina in the provinces of Salta and

Jujuy. The first outbreak evolved from weeks 826 in

1992; the second occurred from week 50 in 1992 through

week 19 in 1993; the third extended from week 40 in

1993 through week 11 of 1994. The incidence of chol-era in Salta and Jujuy was calculated using case num-

bers and census data provided by the provincial health

ministries.4,812 Where possible, case data were ana-

lysed by age group (04, 59, 1014, and 14 years)

and sex.

Salta. The cholera outbreak in the summer of 1992

struck three of the 23 departments of Salta: San Martin,

Oran and Rivadavia.8,11 Data from Rivadavia weredeemed unreliable (all cases of diarrhoea were reported

as cholera without bacteriological confirmation)

(Carmen M de Arquiza, Direccion de Epidemiologica,

Salta, November 1994) and excluded from further ana-

lysis; in San Martin and Oran, bacteriological isolation

of Vibrio cholerae 01 was required. The second and

third cholera outbreaks, in the summers of 1993 and

1994, affected, besides San Martin and Oran, three

additional departments (General Guemes, Cerrillos and

Chicoana).911 Therefore, a total of 1819 cases of chol-

era were reported in Salta, with 89% occurring in San

Martin (N = 1011) and Oran (N = 614). Overall, indi-viduals 14 years old comprised 77% of all cases; 57%

were male. The incidence rates in San Martin and Oran

during the three successive outbreaks are shown in

Table 1; the mean yearly attack rate for San Martin for

the 3 years was 4.6 cases per 1000 population 14 years

of age; and, for Oran it was 2.0 per 1000.

Jujuy. Jujuy was affected by the cholera outbreaks in

1993 and 1994. As in Salta, cholera was limited geo-

graphically, being confined to only eight of the 16

administrative areas of the province (Yavi, San Pedro,

Santa Barbara, El Carmen, Ledesma, Humahuaca,Tilcara and Tumbaya). Age-specific data are not avail-

able for the 1993 outbreak in Jujuy. However, during

the summer 1994 outbreak, 250 of the 267 cases (94%)

were individuals 14 years of age; 69% of these were

males (Table 1).

In summary, most cholera cases in Argentina were

confined to a few northern departments (Salta) or areas

(Jujuy) where the disease clustered in agricultural

COST-BENEFIT ANALYSIS OF CHOLERA VACCINE USE 213

TABLE 1 Number of cases and incidence of cholera in two departments of Salta province and eight areas of Jujuy province during threesummer cholera outbreaks in 1992, 1993 and 1994

Summer 1992 outbreak Summer 1993 outbreak Summer 1994 outbreak

Population Cases Incidence/1000 Cases Incidence/1000 Cases Incidence/1000

Total 14 Total 14 Total 14 Total 14 Total 14 Total 14 Total 14years years years years years years years

Salta Province

San Martin 90 673 54 152 129 114 1.4 2.1 307 218 3.4 4.0 575 419 6.3 7.7

Oran 139 302 83 184 54 43 0.4 0.5 365 318 2.6 3.8 194 146 1.4 1.8

Jujuy Province

8 departmentsa 253 797 162 392 593 474b 2.3 2.9 267 250c 1.0 1.5

a The departments of Yavi, San Pedro, Santa Barbara, El Carmen, Ledesma, Humahuaca, Tilcara and Tumbaya.b Estimate based on 80% of total cases.c Actual data.


3/8

workers. Accordingly, adult agricultural workers in

Salta and Jujuy comprise the targeted high risk group

for the vaccine cost-benefit analysis; the annual incid-

ence rate was 2.7 per 1000 for the three outbreaks in

Salta and two outbreaks in Jujuy.

Assumptions of Population at Risk, Incidence Rates

and Case Loads, Vaccine Coverage and Vaccine

Efficacy over a 3-Year Hypothetical Period

The departments of San Martin and Oran in Salta and

the eight areas in Jujuy are the localities, and adults

14 years of age are the individuals, to be targeted

for possible vaccine usage. In San Martin and Oran

137 336 individuals, and in the eight areas of Jujuy

162 392 individuals, comprise the population 14 years

of age. Our cost-benefit analysis considers a 3-year

period in which, like 19921994, cholera would occurin annual outbreaks, restricted to a high-risk population.

In the cholera regions within Salta and Jujuy,

approximately 80% of the population is comprised of

permanent residents and 20% of temporary workers,

coming from Bolivia for each new harvest.4,13 We

therefore assume the 20% migrant population rep-

resents an entirely new agricultural worker cohort each

year. We estimate that there are 239 782 permanent

agricultural workers, at risk for three successive sum-

mers, who will serve as targets for vaccination. In ad-

dition, each summer harvest season would bring an

additional 59 946 temporary inhabitants (a total of179 838 for the three successive years). Thus, 100%

coverage of this population would require a total of

419 620 doses of vaccine. We also assumed only minor

side-effects from vaccination.

Base Case Model

In our base case model, the annual incidence is 2.5 per

1000 population 14 years of age, approximating the

actual incidence observed in Salta and Jujuy during

19921994 of 2.7 per 1000. We used conservative

estimates of vaccine coverage and vaccine efficacy of

75% and 75%, respectively.

Sensitivity Analysis

In our sensitivity analysis, we consider annual in-

cidence rates of 1.5, 2.0 and 3.0 per 1000 population

14 years of age; therefore, during the 3 years in this

population a total of 13492698 cholera cases would

occur (Table 2). We assume 7590% vaccine coverage

in this population. Several sources make us confident

that such levels could be achieved. In northern Argen-

tina, the Expanded Program on Immunization (EPI)

calculates that 93% of infants are immunized.14 The

agricultural workers are greatly concerned about chol-

era as an occupational hazard and therefore are stronglymotivated to co-operate with a vaccination programme.

We assume a range of vaccine efficacy from 60% to

85% based on results of volunteer challenge studies.15.16

In these controlled trials, a single dose of CVD 103-

HgR conferred 100% protection for moderate/severe

cholera following a challenge with V. cholerae of either

Inaba or Ogawa serotype or either El Tor or classical

biotype. The protection remained undiminished after

6 months (the longest interval tested).15 We assume that

the CVD 103-HgR vaccine, derived from classical bio-

type V. cholerae, confers similar levels of efficacy for

3 years. The infection-derived immunity elicited by wildtype V. cholerae 01 of classical biotype persists for at

least 3336 months in volunteers.17 Moreover, an initial

episode of cholera caused by the classical biotype con-

fers a high level (90100%) of long-lived protection

against subsequent cholera due to either the classical

or El Tor biotype.18,19 However, to account for some

degree of waning in immunity, we examined vaccine

efficacies as low as 60% for the 3-year period.

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY214

TABLE 2 The expected number of cholera cases over 3 years of passive surveillance at different incidence rates in the absenceof vaccination

Annual incidence Annual cases among population at risk 14 years old Total cases(per 1000)

Permanent Migranta 1 year 3 years(N = 239 782) (N = 59 946)

1.5 360 90 450 13492 480 120 602 1799

3 719 180 899 2698

Base case 599 150 749 2248

a This assumes that each summer during the 3 years 20% of the population represents a new cohort of migrant workers.


4/8

Estimation of Costs of Treatment of Cholera in Salta

and Jujuy

The actual medical cost of the three cholera outbreaks

for Salta and Jujuy are calculated from expenses ac-crued by the Ministries of Health of the two prov-

inces.8,9,20 The number (and duration) of hospitalized

cases and those managed as outpatients were obtained

from notification data (C Remondegui, Hospital San

Roque, Jujuy, unpublished observations, 1994).21 Es-

timates of transportation costs to the hospital were in-

cluded in managerial costs, because managers often

accompanied vehicles travelling to the areas where

cholera patients lived in order to assess the situation.

Medical and laboratory costs related to treatment of

patients with cholera were also included (Table 3).

The base cost of hospitalization for cholera wascalculated to be US$67.50 per day (Francisco Ryan,

Ex-Sub-Secretaro de Salud Pblica, Salta, personal

communication, November 1994). Managerial costs per

cholera case were obtained from Jujuy20 and adjusted

for Salta because Salta had no such data available.

Additional managerial costs in Salta consisted of two

helicopter trips each month for medical personnel to the

cholera outbreak area (Francisco Ryan, Salta, personal

communication, November 1994). These helicopters

were not used prior or subsequently in medical work.

Costs of antibiotics, rehydration solutions and medical

supplies were obtained (Gregorio Mitelman, DirectorFarmaca, Salud Pblica, Salta, personal communica-

tion, November 1994) and total costs were based on the

amount of supplies used.20 The Ministry of Health of

Jujuy calculated medical therapy for one case of chol-

era to be US$228.50. Each faecal culture cost US$0.99

(Ana G B de Bojarski, Nlida J Molina, Laboratorio

Bacteriologa, Salta, personal communication, Novem-

ber 1994). Additional laboratory tests cost US$ 1.50 per

hospitalized-day (Francisco Ryan, Salta, personal com-

munication, November 1994).

The yearly US-Argentine exchange rate has been

relatively stable at 0.99950.9905 from 1991 to 1994;22

inflation has also been stable from 1992 to 1994.22,23

However, the 1992 and 1993 summer-outbreak costs

were discounted by 5% compared to the 1994 costs.

Estimation of the Cost of Adding CVD 103-HgR

Vaccine as a Cholera Control Measure

The National Health Promotion Programme is respons-

ible for immunizing the paediatric and adult populations

(e.g. tetanus vaccine to adults). Since a reliable infra-

structure (i.e. health posts staffed by qualified nurses)

exists for delivery of vaccines, we assumed modest

administrative costs beyond the cost of the vaccine. Weassume 80 vaccine doses could be administered and

recorded per hour by one staff member of the Ministry

of Health. Assuming 10 staff mobilized for short per-

iods of time, 6000 vaccine doses can be administered

per day (7.5 h per day for vaccination and recording).

In the first year, 299 728 doses are to be given over

50 days. In the second and third years, 59 946 doses

will be administered each year, over a total of 20 days.

The salary per rural staff member is US$15.72 per day

(Carmen M de Arquiza, Salta, personal communication,

November 1994). Therefore, the total salary bill is

US$11 004 for the 3 years. Publicity and recording mat-erial costs are estimated at US$35 000 for the 3 years

and transportation costs at US$50 710. We assumed

400 reusable ice chests and 419 600 disposable cups

would be needed, at a cost of US$10 per chest and

US$0.01 per cup, this yields a total cost of US$8196 for

the 3 years. The total administrative cost is estimated

at US$104 910. For 419 620 doses to be given (100%

coverage), this is an additional US$0.25 per dose.


TABLE 3 Cost of care for patients with cholera in Salta and Jujuy, 19921994 (79, 18)

Area Outbreak Cases Days in Hospital Cost of Lab Managerial Total Discountedhospital costs antibiotics, costs costs at 5%

rehydrationfluids, misc.

Salta First 391 1050 $70 875 $165 265 $3446 $103 025 $342 611 $325 480

Second 948 1364 $92 070 $241 126 $11 826 $232 986 $555 008 $527 258Third 1018 1200 $81 000 $183 470 $10 668 $249 319 $577 502

Jujuy First 37 106 $5112 $8455 $416 $8633 $22 616 $21 485

Second 646 1855 $89 256 $147 611 $7273 $150 733 $394 873 $375 129

Third 308 884 $42 555 $70 378 $3470 $71 876 $188 279

Total $2 080 889 $2 015 133


5/8

Assuming a cost of US$1.25 for the vaccine, the total

cost is US$1.50 per dose.

Numerous studies have examined the safety of CVD

103-HgR, in children and adult vaccinees in developing

and more developed countries.2428 The vaccine has been

shown to be well tolerated with no increases in diarrhoeal

episodes or other adverse gastrointestinal reactionsamong vaccine compared with placebo recipients.2428

RESULTS

Cost Analysis

We calculate that the direct costs for management of a

cholera case in Salta are US$626 and in Jujuy US$611,

giving an average cost of US$622 per case (Table 3).

Cases in Salta have a higher cost because helicopters

were used for transporting some patients and personnel,

and antibiotic costs in the first outbreak were higher

because of the initial setting up costs for the area healthposts (Carmen M de Arquiza, Salta, personal commun-

ication, November 1994). Taking into consideration a

5% discounted cost for the first and second outbreaks,

the overall cost per cholera case is US$602.

Base Case Model

The use of CVD 103-HgR prevents 1265 cases of chol-

era over the 3-year period (Table 4). The conservative

assumption of 75% vaccine coverage, 75% vaccine effi-

cacy and a cost of US$602 per case, would result in a

total savings of US$132 100, for a break-even point

(i.e. cost/benefit ratio of 1.0) of US$1.81. Using a vac-

cine efficacy as low as 65% and vaccine coverage of

75% in the base case, vaccine use still proved to be

cost-beneficial at US$1.57/dose of vaccine.

Sensitivity Analysis

Table 2 shows the number of cases of cholera that

would occur in the target population over 3 years at

three cholera incidences in the absence of vaccination.

At the calculated cost of US$602 per case, the expend-

itures by the Ministries of Health of Salta and Jujuy

for treatment of these cases would be US$812 098,

US$1 082 998, or US$1 624 196 in the three scenarios.

The number of cases that would be prevented by

programmed use of CVD 103-HgR vaccine would

range considerably, depending on the annual incidence,the degree of vaccine coverage of the targeted popu-

lation, and the level of vaccine efficacy. The calculated

numbers of cases prevented under these various con-

ditions are shown in Table 4. With vaccination

coverage ranging from 75% to 90% and 3-year vaccine

efficacy ranging from 60% to 85%, the break-even

point (i.e. cost/benefit ratio of 1.0) was calculated. This

represents the cost per dose of vaccine such that


TABLE 4 Cholera cases prevented over 3 years by programmatic use of CVD 103-HgR live oral cholera vaccine in a target populationconsisting of 239 782 permanent residents and a total of 179 838 migrants (3 yearly cohorts of 59 946 each)

Incidence (per 1000) Vaccine Efficacy Vaccine Coverage

90% 85% 80% 75%

1.5 85% 1032 975 917 860

80% 971 917 863 80975% 911 860 809 759

70% 850 803 755 708

65% 789 745 701 658

60% 728 688 648 607

2 85% 1376 1300 1223 1147

80% 1295 1223 1151 1079

75% 1214 1147 1079 1012

70% 1133 1070 1007 944

65% 1052 994 935 877

60% 971 917 864 810

3 85% 2064 1949 1835 1720

80% 1943 1835 1727 1619

75% 1821 1720 1619 1518

70% 1700 1605 1511 1416

65% 1578 1491 1403 131560% 1457 1376 1295 1214

Base case 75% 75%

1265


6/8

vaccine cost would not exceed the costs of choleratreatment in the absence of vaccination (Table 5). We

also calculated total savings (or losses) to the Ministry

of Health from use of vaccine under the various con-

ditions if vaccine cost US$1.50 per dose.

The annual incidence of cholera had a marked effect

on the cost benefit of vaccine use, being a good invest-

ment at the high annual incidence of 3.0 per 1000.

In contrast, variations in the range of coverage of

vaccination from 75% to 90% or in vaccine efficacyfrom 60% to 85% did not substantially affect the impact

of a vaccination campaign.

DISCUSSION AND CONCLUSION

Argentina represents an interesting situation for con-

ducting a cost-benefit analysis of the new oral attenu-

ated cholera vaccine, CVD 103-HgR. First, Argentina


TABLE 5 Analyses of total savings to the Ministry of Health of Salta and Jujuy over 3 years by programmatic use of CVD 103 HgR in thetarget population of 239 782 permanent residents and 59 946 annual migrants and break-even price for vaccine (based on a total of419 620 doses required over 3 years to achieve 100% coverage of the combined permanent and migrant population at a priceof US$1.50/dose)

Incidence (per 1000) Vaccine Efficacy Vaccine Coverage

90% 85% 80% 75%

1.5 85% $8166 $42 480 $77 396 $111 710

$1.48 $1.40 $1.32 $1.23

80% $44 888 $77 396 $109 904 $142 412

$1.39 $1.32 $1.24 $1.16

75% $81 008 $111 710 $142 412 $172 512

$1.31 $1.23 $1.16 $1.09

70% $117 730 $146 024 $174 920 $203 214

$1.22 $1.15 $1.08 $1.02

65% $154 452 $180 940 $207 428 $233 314

$1.13 $1.07 $1.01 $0.94

60% $191 174 $215 254 $239 334 $264 016

$1.04 $0.99 $0.92 $0.87

2 85% $198 922 $153 170 $106 816 $61 064

$1.97 $1.87 $1.75 $1.65

80% $150 160 $106 816 $63 472 $20 128

$1.86 $1.75 $1.65 $1.55

75% $101 398 $61 064 $20 128 $20 206

$1.74 $1.65 $1.55 $1.45

70% $52 636 $14 710 $23 216 $61 142

$1.63 $1.54 $1.44 $1.35

65% $3874 $31 042 $66 560 $101 476

$1.51 $1.43 $1.34 $1.26

60% $44 888 $77 396 $109 302 $141 810

$1.39 $1.32 $1.24 $1.11

3 85% $613 098 $543 868 $475 240 $406 010

$2.96 $2.80 $2.63 $2.47

80% $540 256 $475 240 $410 224 $345 208

$2.79 $2.63 $2.48 $2.32

75% $466 812 $406 010 $345 208 $284 406$2.61 $2.47 $2.32 $2.18

70% $393 970 $336 780 $280 192 $223 002

$2.44 $2.30 $2.17 $2.03

65% $320 526 $268 152 $215 176 $162 200

$2.26 $2.14 $2.01 $1.89

60% $247 684 $198 922 $150 160 $101 398

$2.09 $1.97 $1.86 $1.74

Base case 75% 75%

$132 100

$1.81


7/8

is a country where some segments of the population live

in underprivileged conditions typical of less-developed

countries while others live in conditions typical of a

developed country. Second, despite these pockets of

poverty, the central government has instituted a sophist-

icated infrastructure for transport and treatment of chol-

era patients. This results in a high treatment cost per

patient, typical of any industrialized country setting

where a case of cholera gravis is managed. Third, theNational Ministry of Health maintains a credible infra-

structure that allows immunization of target groups of

varying ages. Therefore, the means exist to deliver CVD

103-HgR to targeted populations. Fourth, it is possible

to accurately estimate the financial burden borne by the

Ministry of Health with respect to the treatment of chol-

era. Fifth, CVD 103-HgR is conducive to large-scale use

because of its single-dose schedule and practical formu-

lation. Finally, the National Ministry of Health has been

willing to consider the purchase and implementation

of cholera vaccine if the need continues (i.e. if cholera

outbreaks continue to occur) and if a cost-benefit ana-lysis provides a convincing argument in favour of pro-

grammed use of the new cholera vaccine.

Prior cost-benefit analyses have considered the

potential use of parenteral or oral killed cholera vac-

cines.2426, 2931 These analyses involved endemic chol-

era occurring in Bangladesh where no infrastructure for

immunizing older children and adults exists and the

cost per treated case of cholera gravis is exceedingly

low. Moreover, the oral killed vaccine requires that at

least two doses be administered and length of efficacy

may be shorter than for CVD 103-HgR. These con-

ditions shift the balance against a role for vaccination.A similar cost-benefit analysis was conducted in

Santiago, Chile for evaluating the use of the Haemo-

philus influenzae type b conjugate vaccine.32 The authors

concluded that immunization represents a cost-effective

public health intervention for preventing H. influenzae

type b disease in Chilean children. In countries with

limited resources, cost-benefit analyses can provide

helpful information to public health decision makers

who must make difficult choices.

The actual benefit of vaccination was undoubtedly

underestimated in this analysis in Argentina since only

the medical costs are considered. The true overall eco-nomic costs of the cholera epidemics in northern Ar-

gentina are much greater because of marked declines in

tourism (parts of northern Argentina are popular tourist

areas) and in the sale of agricultural products. These

economic losses greatly exceed the treatment costs in-

curred.29 Less significant indirect losses due to lost

man-hours of work due to illness were also not con-

sidered. Additionally, lost productivity because of time

taken from work to receive vaccination and any time

lost due to vaccine side-effects were not considered.

Since the health clinics are located where the workers

reside and CVD 103-HgR vaccination has minimal

side-effects, we believe that these costs would not alter

our conclusions. Our base case incorporates the actual

annual incidence rate for cholera, a conservative level

for vaccine coverage (given a rate of 93% in the infant

population for BCG vaccination), and a conservativeestimate for vaccine efficacy (given results of volunteer

and field studies). Nevertheless, our base case analysis

demonstrates that CVD 103-HgR, when delivered to the

confined high-risk population of northern Argentina,

can be an important part of cholera control activities in

that region. Vaccination would prevent 1265 cases of

cholera for discounted savings to the Ministry of Health

of US$132 100.

ACKNOWLEDGEMENTS

This cost-benefit study was directly supported byFUNCEI. Considerable indirect support and unlimited

co-operation was provided by the Ministry of Health

of Argentina and of the Provinces of Salta and Jujuy.

We appreciate the editorial advice provided by Anne

Haddix of the Epidemiology Program Office and Orin

S Levine of the Division of Bacterial and Mycotic

Diseases CDC.

REFERENCES1 Tauxe R V, Blake P A. Epidemic cholera in Latin America.

JAMA 1992; 267: 138890.2 Arias L M, Aleman A, Soave L G. Provincia de Salta y el clera

perfil de una epidemia.Medicina Infant il, 1994; 1: 25459.3 Direccin de Epidemiologa, Secretara de Salud, Ministerio de

Salud y Accin Social. Clera, Argentina Febrero/1992 a

Febrero/1994. In:Boletin Epidemiolgico Nacional. Minis-

terio de Salud y Accin Social, Buenos Aires, Argentina,

1993.4 Coordinacin del Area Epidemiolgica, Direccin Provincia de

Sanidad, Ministerio de Bienestar Social.Informe Referente

al Programa de Lucha contra el Clera, Situacin Epi-

demiolgica hasta el 14/4/94. Ministerio de Bienestar

Social, Jujuy, Argentina, 1994.5 Vaamondes C, Aleman A. El Colera como Enfermedad Rela-

cionada a la Calidad de Vida y Funciones del Estado,

1994. Universidad Nacional de Salta Facultad Ciencias de

la Salud, Salta, Argentina, 1994.6 Araoz G. Perfil de Mortalidad por Desnutrition en la Provincia

de Salta, 1994. Ministerio de Salud, Pblica de la Provincia

de Salta, Secretara de Coordinacin Sanitara y Medio

Ambiente, Direccin de Epidemiolgia, Salta, Argentina,

1994.7 Direccin de Epidemiolgia, Ministerio de Salud. Comision Op-

erativa de Prevencion y Lucha contra el Clera. Ministerio

de Salud, Jujuy, Argentina, 1993.



8/8


8 Direccin de Epidemiolgia, Secretara de Coordinacin San-

itara y Medio Ambiente, Ministerio de Salud, Pblica de la

Provincia de Salta.Metodologa Provincial Empleada para

Prevencin y Control del Clera, Salta, 19911992. Min-

isterio de Salud, Salta, Argentina, 1993.9 Comision Tecnica y Logistica, Comando Operativo Clera.

Clera en la Provincia de Salta, Informe Epidemiolgico

NRO 1 . Republica Argentina,Tartagal-Salta, July 1993.10 Comision Tecnica y Logistica, Comando Operativo Clera. Cl-

era: Tercera Etapa de Laepidemia. Republica Argentina,Provincia de Salta, September 1994.

11 Direccin de Epidemiologa, Secretaria Medicina Sanitaria,

Ministerio de Salud Pblica. Clera 94/95: Proyecto de

Trabajo, Salta, 1994. Ministerio de Salud Pblica, Salta,

Argentina, 1994.12 Departamento de Bioestadistica, Ministerio de Bienestar Social.

Poblacin Estimada, Ao 1994, por Grupo Etareo y por

Area Programatica, Jujuy. Ministerio de Bienestar Social,

Jujuy, Argentina, 1994.13 Salinas G. El Clera: Determinantes para su Aparicin en la

Poblacin Rural. Universidad Nacional de Salta, Salta, Ar-

gentina, 1994.14 Mirtha Eiman Grossi. Analisis del Sistema de Vigilancia Epi-

demiologica perfil de Notificationes, 1983 a 1993. Depto

de Vigilancia Epidemiologica, Salta, Argentina, 1993.15 Tacket C, Losonsky G, Nataro J et al. Onset and duration of

protective immunity in challenged volunteers after vaccina-

tion with live oral cholera vaccine CVD 103-HgR.J Infect

Dis 1992; 166: 83741.16 Levine M M, Tacket C O. Live oral vaccines against cholera. In:

AlaAldeen D, Hormaeche C (eds).Molecular and Clinical

Aspects of Bacterial Vaccine Development. Chichester:

John Wiley, 1995, pp. 23358.17 Levine M M, Black R E, Clements M L et al. Duration of

infection-derived immunity to cholera. J Infect Dis 1981;

143: 81820.18 Clemens J D, van Loon F, Sack D A et al. Biotype as determin-

ant of natural immunising effect of cholera. Lancet1991;

337: 88384.19 Glass R I, Becker S, Huq I et al. Endemic cholera in rural Bang-

ladesh, 19661980.Am J Epidemiol 1982; 116: 5970.20 Coordinacin del Area Epidemiolgica, Direccin Provincia de

Sanidad, Ministerio de Bienestar Social. Epidemiolgica

Presupuesto 1994: Clera Enfermedades Hdricas. Minis-

terio de Bienestar Social, Jujuy, Argentina, 1994.21 Nordera J. Provincia de Salta, Clera, Parte Diario Oficial,

Comprende Informacin del Periodo 1994. Secretara de

Medicina Sanitaria y Medicina Ambiente, Ministerio de

Salud Pblica, Salta, Argentina, 1994.22International Financial Statistics. New York, International

Monetary Fund, April 1995, pp. 8486.23Monthly Bullet in of Statistics. New York, United Nations, Feb-

ruary 1995, pp. 19293.24 Simanjuntak C H, OHanley P, Punjabi N H et al. Safety,

immunogenicity, and transmissibility of single-dose live

oral cholera vaccine strain CVD 103-HgR in 24- to 59-

month-old Indonesian children. J Infect Dis 1993; 168:116976.25 Suharyono, Simanjuntak C, Witham N et al. Safety and im-

munogenicity of single-dose live oral cholera vaccine CVD

103-HgR in 59-year-old Indonesian children, Lancet

1992; 340: 68994.26 Su-Arehawaratana P, Singharaj P, Taylor D N et al. Safety and

immunogenicity of different immunization regimens of

CVD 103-HgR live oral cholera vaccine in soldiers and

civilians in Thailand. J Infect Dis 1992; 165: 104248.27 Cryz S J, Levine M M, Kaper J B, Althaus B. Randomized

double-blind placebo controlled trial to evaluate the

safety and immunogenicity of the live oral cholera vaccine

strain CVD 103-HgR in Swiss adults. Vaccine 1990; 8:

136669.28 Kotloff K L, Wasserman S S, ODonnell S, Losonsky G A,

Cryz S J, Levine M M. Safety and immunogenicity in

North Americans of a single dose of live oral cholera vac-

cine CVD 103-HgR: Results of a randomized, placebo-

controlled, double-blind crossover trial. Infect Immunity

1992; 60: 443032.29 Cvjetanovic B, Grab B, Uemura K. Dynamics of acute bacterial

diseases. Bull World Health Organ 1978; 56 (Suppl. 1):

6580.30 Mosley W, Aziz K, Rahman A et al. Report of the 196667 chol-

era vaccine trial in rural East Pakistan, Bull World Health

Organ 1972; 47: 22938.31 Holmgren J, Osek J, Svennerholm A-M. Protective oral cholera

vaccine based on a combination of cholera toxin B subunit

and inactivated cholera vibrios. In: Vibrio cholerae andCholera: Molecular to Global Perspectives, Washington,

DC: Am Soc Microbiol, 1994.32 Levine O S, Ortiz E, Contreras R et al. Cost-benefit analysis for

the use of Haemophilus influenzae type b conjugate

vaccine in Santiago, Chile. Am J Epidemiol 1993; 137:

122128.

(Revised version received July 1996)

int. j. epidemiol. 1997 cookson 212 9

Documents