-
Unit Costs for Delivery of AntiretroviralTreatment and
Prevention of Mother-to-Child Transmission of HIVA Systematic
Review for Low- and Middle-Income Countries
Omar Galarraga,1,2 Veronika J. Wirtz,3 Alejandro Figueroa-Lara,3
Yared Santa-Ana-Tellez,2
Ibrahima Coulibaly,4 Kirsi Viisainen,4 Antonieta Medina-Lara5
and Eline L. Korenromp4,6
1 Health Services, Policy and Practice (Health Economics), Brown
University, Providence, Rhode Island,
USA
2 Center for Evaluation Research and Surveys, Division of Health
Economics, National Institute of Public
Health (INSP)/Mexican School of Public Health, Cuernavaca,
Mexico3 Center for Health Systems Research, National Institute of
Public Health, Cuernavaca, Mexico
4 The Global Fund to Fight AIDS, Tuberculosis and Malaria,
Geneva, Switzerland
5 Center for Research on Health and Social Care Management
(CERGAS), Bocconi University, Milan, Italy
6 Department of Public Health, Erasmus MC, University Medical
Center Rotterdam, Rotterdam,
the Netherlands
Contents
Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . 5801. Literature Review . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
581
1.1 Search. . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . 5811.1.1 Inclusion Criteria . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . 5811.1.2 Data Extraction.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . 5821.1.3
Standardization of Data . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
582
1.2 Data Aggregation and Sensitivity Analysis. . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. 5831.3 Results . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . 583
2. Antiretroviral Treatment . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . 5832.1 Antiretroviral Medicines . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . 5892.2 Laboratory . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5892.3 Personnel . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . 5892.4 Programme-Level Costs . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . 5932.5 Paediatric
Antiretroviral Treatment . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
3. Prevention of Mother-to-Child Transmission . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . 5934. Sensitivity Analyses . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . 5945. Discussion . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . 594
5.1 Limitations . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . 5955.2 Future Work. . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . 596
6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . 597
REVIEWARTICLE Pharmacoeconomics 2011; 29 (7):
579-5991170-7690/11/0007-0579/$49.95/0 2011 Adis Data Information
BV. All rights reserved.
-
Abstract As antiretroviral treatment (ART) for HIV/AIDS is
scaled up globally,information on per-person costs is critical to
improve efficiency in servicedelivery and to maximize coverage and
health impact. The objective of thisstudy was to review studies on
unit costs for delivery of adult and paediatricART per
patient-year, and prevention of mother-to-child transmission(PMTCT)
interventions per mother-infant pair screened or treated, in
low-and middle-income countries. A systematic review was conducted
of English,French and Spanish publications from 2001 to 2009,
reporting empiricalcosting that accounted for at least
antiretroviral (ARV)medicines, laboratorytesting and personnel.
Expenditures were analysed by country-income leveland cost
component. All costs were standardized to $US, year 2009
values.Several sensitivity analyses were conducted.
Analyses covered 29 eligible, comprehensive, costing studies. In
the basecase, in low-income countries (LIC), median ART cost per
patient-year was$US792 (mean: 839, range: 6821089); for
lower-middle-income countries(LMIC), the median was $US932 (mean:
1246, range: 1563904); and, forupper-middle-income countries
(UMIC), the median was $US1454 (mean:2783, range: 12305667). ARV
drugs were the largest component of overallART costs in all
settings (64%, 50% and 47% in LIC, LMIC and UMIC,respectively). Of
26 ART studies, 14 reported the drug regimes used, and onlyone
study explicitly reported second-line treatment costs. The second
costdriver was laboratory cost in LIC and LMIC (14% and 20%), and
personnelcosts in UMIC (26%). Two ART studies specified the types
of laboratorytests costed, and three studies specifically included
above facility-level per-sonnel costs. Three studies reported
detailed PMTCT costs, and three studiesreported on paediatric
ART.
There is a paucity of data on the full unit costs for delivery
of ART andPMTCT, particularly for LIC and middle-income countries.
Heterogeneityin activities costed, and insufficient detail
regarding components included inthe costing, hampers
standardization of unit cost measures. Evaluation ofprogramme-level
unit costs would benefit from international guidance on
stan-dardized costing methods, and expenditure categories and
definitions. Futurework should help elucidate the sources of the
large variations in delivery unitcosts across settings with similar
income and epidemiological characteristics.
As antiretroviral treatment (ART) for HIV/AIDS is scaled up
globally, information on per-person costs is critical to improve
efficiency inservice delivery and to maximize coverage andhealth
impact. Cost evaluations support programmeplanning and budgeting,
can help to ensure pro-gramme sustainability, and are a
pre-requisite toidentifying opportunities for efficiency
gains.[1,2]
As global health institutions strive to ensure valuefor money,
they are committed to supportingcountries in the measurement of
per-person de-livery costs of key services.[3-8] Since 2008,
many
national HIV/AIDS programmes have made im-portant advances in
complying with the bi-annualroutine reporting of nationally
aggregated expen-ditures as stated in the 2001 Declaration of
Com-mitment on HIV/AIDS (UNGASS) to UNAIDS(the Joint United Nations
Programme on HIV/AIDS).[9] UNAIDS summaries of UNGASS datatry to
standardize reported programme expendi-tures for ART and PMTCT as
per-person costs,[10]
but do not request a comprehensive breakdown ofthe cost
components included in each service area.As of 2011, most HIV/AIDS
programmes do not
580 Galarraga et al.
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
routinely assess their cost per person or per unitof service
delivery, nor do they breakdown ex-penditure by cost components
using standardizedmethods.
For antiretroviral treatment (ART), scale-upof treatment access
in low- and middle-incomestudies started in 20045, when funding
increasedsubstantially with new donor funding from theGlobal Fund
to Fight AIDS, Tuberculosis andMalaria, and the US Presidents
Emergency Planfor AIDS Relief (PEPFAR), among others. As atthe end
of 2009, an estimated 5.2 million HIV-infected people were
receiving ART globally.With the WHOs revised 2010 guidelines on
HIVtreatment in resource-limited settings, the totalimmediate need
is now estimated at 15 millioneligible people worldwide.[11] The
ongoing scale-up and progress toward universal access willdepend on
both total available funding and thedelivery cost per patient-year
in countries withhigh HIV prevalence.
Previous reviews of facility-level or per-patientART costs found
only a very small number ofstudies from low- andmiddle-income
countries[12,13]
and estimated the cost of antiretroviral (ARV)medicines from
manufacturers procurement pricelists, as an average per
country-income group.[13]
For prevention of mother-to-child transmission(PMTCT) of HIV,
(unit) delivery costs have notbeen reviewed since 2000.[14] Themost
recent reviewof ART costs[15] included more studies from low-and
middle-income countries and presented a scor-ing system to rate the
methodological quality ofstudies without attempting a quantitative
meta-analysis of cost results.
To complement these reviews, we conducted asystematic literature
review and a quantitativeanalysis of per-patient expenditure for
adult andpaediatric ART, as well as ARV prophylaxis usedfor PMTCT,
using explicit guidelines[16] for sys-tematic reviews, and
standardizing data from eli-gible studies into common service
delivery unitsand cost component categories. Per-patient costsare
presented as median, arithmetic average andranges across eligible
study sites, studies, countriesand country-income groupings,
separately for themost important standardized cost components.We
discuss results in the light of information re-
quirements anticipated from national AIDS pro-grammes from major
global financiers of HIV/AIDS programme scale-up.
1. Literature Review
1.1 Search
The literature search followed the PRISMA(PreferredReporting
Items for Systematic ReviewsandMeta-Analyses) guidelines (figure
1).[16] Orig-inal articles published in English, French andSpanish
from January 2001 to October 2009were searched using PubMed,
EconLit and theCochrane Library, and grey literature throughGoogle
Scholar and POPLINE. In addition,websites of international donor
organizations suchas UNAIDS, WHO, World Bank, PEPFAR andthe United
Nations Childrens Fund (UNICEF)were searched. Abstracts from the
InternationalAIDS Conferences 20018, International HealthEconomics
Association 20068, American SocietyforHealth Economists 20068,
International AIDSEconomics Network 2008 and PEPFAR Im-plementers
meetings 20089 were also screened forcontextual information to the
published data.References identified as relevant served to
trackadditional studies of interest. Annex A details thesearch
terms used on electronic databases; annexB shows the search
commands and number ofstudies found (all annexes are available as
Sup-plemental Digital Content only;
http://links.adisonline.com/PCZ/A117).
1.1.1 Inclusion Criteria
The following inclusion criteria were applied.A list of excluded
studies is available in Annex D. Country-specific, empirical
measurement of
actual expenditure, from a micro-costing oringredients approach
(if the study used solely astep-down approach, it was
excluded).
Study (or study data point) included at least25 patients.
Low- and middle-income countries, accord-ing to the July 2009
World Bank list of econ-omies,[17] with both publication and
datacollection between January 2001 and October2009.
Antiretroviral Treatment Unit Costs 581
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
Costing included at least ARVs, laboratoryexpenses and
personnel, as three individuallyreported expenditure components, or
as ex-plicitly named components of a comprehen-sive but unbundled
overall expenditure.
Economic costs from the healthcare providerperspective;
financial costs included if econom-ic costs were not available.
Economic costsaccruing to patients, such as waiting
and/ortransportation time and productivity losseswere not
considered because of the inherentdifficulties associated with
assigning opportu-nity costs in the absence of accurate data
onwages and time costs.
1.1.2 Data Extraction
Eligible articles were reviewed for inclusion byat least two
independent reviewers (OG and VWor AML), using a pre-determined
data extractionform (see Annex C). When the decision was indoubt, a
third or fourth reviewer (YST or AFL)analysed the article and
inclusion was solved byconsensus. If the information contained in
thefull-text article was incomplete, the authors werecontacted to
obtain missing data.
1.1.3 Standardization of Data
ART delivery unit cost data were standardizedas estimates per
patient-year, separately for first-and second-line ART, or weighed
by the local mixof patients. Paediatric ARTwas distinguished
fromadult ART, using an age cut-off of 15 years. Com-ponents of
total costs were categorized as follows.
ARVs, including, if available, transport of drugsto point of
care, and storage.
Personnel involved in delivering ART, includ-ing education and
training, as well as supra-facility-level human resources, if
specified (suchas personnel in distribution centres or
district/provincial/national programme management,or monitoring and
evaluation).
Laboratory tests, including CD4 cell countsand viral load, as
well as other tests, if specificallymentioned. Overhead costs,
including adminis-tration and utilities (electricity, water,
phone,rent, maintenance costs, cleaning, security).
Other components considered and reported aspart of delivery unit
costs, such as concomi-tant medications (for opportunistic
infectionsor prophylaxis), other supplies (e.g. cotton, sy-ringes,
gloves), staff transport for home visits
Records after removalof duplicates
(n = 17)
Articles screened(n = 150)
Publicationsidentified through
PubMed(n = 103)
Full-text articleseligible for further analysis
(n = 139)
Publications identifiedthrough EconLit
(n = 30)Additional recordsidentified through
POPLINE and othersources1
(n = 34)
Editorials andcommentaries
(n = 11)
Full-text articles eligiblefor in-depth analysis
(n = 29)
Full-text articlesexcluded because theydid not fulfil
inclusion
criteria2 (n = 110)
Fig. 1. Literature review on antiretroviral therapy and
prevention of mother-to-child transmission per person-year delivery
unit costs: flowchartof search and review process. 1 Seven studies
included for in-depth analysis were from the grey literature; 2 see
Annex D in the SupplementalDigital Content 1,
http://links.adisonline.com/PCZ/A117, for a list of excluded
studies.
582 Galarraga et al.
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
(i.e. trained field officers visit patients homesperiodically to
deliver drugs, clinically moni-tor participants with a checklist on
signs andsymptoms of drug toxicity or disease progres-sion and to
provide adherence support) anddepreciated capital costs (medical
and non-medical equipment), as well as other recurrentcosts when
those breakdowns were available.For PMTCT, reported costs were
either per
HIV-infected pregnant woman receiving ARVprophylaxis, or per
pregnant woman initiallytested and counselled for HIV. In either
case,costs covered the woman-neonate pair through-out the
pregnancy/birth period. Cost componentswere categorized as follows.
ARVs given before, during and/or shortly
after birth to mother and/or child to preventtransmission
(including transport cost of ARVsto point of care and storage);
personnel in-volved in delivering PMTCT, including educa-tion and
training, as well as supra-facility-levelpersonnel, if
specified.
Laboratory tests (including HIV testing, CD4cell counts and
viral load).
Overhead costs, including administration andutilities
(electricity, water, phone, rent, main-tenance costs, cleaning,
security).
Other components, as listed by specific studies.All costs are
presented as $US, October 2009
values, after foreign currency conversion usingaverage annual
exchange rates provided by orig-inal study authors or from the
InternationalMonetary Fund (IMF)[18] (except for a study ofART in
Lesotho,[19] for which rates were directlyfrom a Central Bank[20]
source, for lack of therelevant exchange rate from the IMF).
Onceconverted into $US, costs were adjusted for in-flation using
the US Consumer Price Index.[21]
1.2 Data Aggregation and Sensitivity Analysis
The analysis focused on economic costs fromthe perspective of
the provider. Specified econ-omic costs to the patient were
included only if amonetary transaction took place so that the
ser-vice would occur (e.g. a co-payment or re-cuperation fee paid
by the patient), in which casethey were added to the overall
costs.
Unit costs and unit cost components weresummarized as mean,
median and range acrosscountries studied (after aggregation across
mul-tiple sites, or data points within each country)and
country-income groups: low-income coun-tries (LIC),
lower-middle-income countries (LMIC)and upper-middle-income
countries (UMIC), ac-cording to the World Bank 2009 country
classi-fication.[17] This method of aggregation preventedcountries
with relatively large numbers of datapoints from skewing the
results. In sensitivity anal-yses, we explored the robustness of
results againstalternative ways of aggregating results across
sites,countries and country-income groupings, and ro-bustness of
different study inclusion criteria.
1.3 Results
Of 574 abstracts retrieved, 150 full-text studieswere assessed
for eligibility; of those, 29 were in-cluded in the analysis: 26 on
ART (23 on adultART, 2 on adult and paediatric ART[22,23] and 1on
paediatric ART only)[24] (table I). Table II pro-vides estimates of
ART cost per patient per yearaccording to country income level
(base-case andsensitivity analyses), and figure 2 summarizes
themost relevant findings. Three studies for PMTCTare detailed in
table III (expanded versions of thetables, with additional
information and annexes,are available in the Supplemental Digital
Con-tent). The median year of reported cost data was2004 for ART
and 2005 for PMTCT.
2. Antiretroviral Treatment
Six studies from four LIC were included(Benin, Ethiopia, Haiti
and Uganda); the mediancost per patient-year of ART was $US792
(mean:839, range: 6821089) [table II].[23,25-29] Studiesfor LMIC
were found for India,[30,31] Lesotho,[19]
Morocco,[32] Nigeria[33,34] and Thailand;[35] themedian cost of
ART was $US932 (mean: 1246,range: 1563904) [table II]. In UMIC
Brazil,[24,36]
Mexico[37-39] and SouthAfrica[22,40-46] themedianART cost was
$US1454 (mean: 2783, range:12305667) [table II].
ART unit costs and their components variedconsiderably between
both countries and studies
Antiretroviral Treatment Unit Costs 583
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
Table I. Delivery unit costs ($US, year 2009 values) of
antiretroviral treatment (ART), per patient-yeara
Study, year of
publication
(country)
City/setting Scope ART unitcosts
Cost components (% of total cost) ART regimenb
facilitiesc
(provider type)
patients ARV
drugs
personnel laboratory tests overheads
Low-income countries
Hounton et al.,[25]
2008d,e (Benin)
Cotonou/U 1 (Nat/NGO/D) 122outpts
681.9 500.5 (73.4) 51.4 (7.5) 99.5 (14.6) 21.1 (3.1) NA
Bikilla et al.,[26]
2009e,f (Ethiopia)
Arba Minch/U 1 (Nat) 209outpts
262.8 205 (78) 37.8 (14.4) 15.7 (6) INU (d4T/3TC +NVP)or (AZT +
3TC+NVP)or (d4T/3TC +EFV)or (AZT + 3TC+EFV).All regimens are
first-line
Kombe et al.,[27]
2004g (Ethiopia)
Nazareth, Addis,
Bahir Dar/U6 (Nat) NA 804.1 548.3 (68.2) 14.2 (1.8) Dr,
nurse, counsellor,
pharmacist, lab
technician
241.5 (30) FBC,
blood chemistry,
blood sugar,
CD4 count, viral
load
Not
included
AZT +d4T +NVPfirst-line
PEPFAR ART
Costing Project
Team,[23] 2009h,i
(Ethiopia)
U 9 (Nat) NA
outpts
741.7 (new
adult
patients)
428.3 (57.7) 21.8 (2.9) 191.1 (25.8) 59.9 (8) NA
610.8
(established
adult
patients)
428.3 (70.1) 12 (2) 80.8 (13.2) 33.7 (5.5)
960.8 (new
paediatric
patients)
606.7 (63.1) 21.8 (2.2) 191.1 (19.9) 62.7 (6.5)
933.4
(established
paediatric
patients)
606.7 (65) 12.3 (1.3) 94.5 (10.1) 39.4 (4.2)
Koenig et al.,[28]
2008e (Haiti)
Port-au-Prince/U 1 (NGO) 218in/outpts
1088.7 402.6 (37) INU 147.4 (13.5) 129.6
(11.9)
NA
Jaffar et al.,[29]
2009e (Uganda)
Jinga/R 1 (NGO) NAoutpts
789 416.6 (52.8) 181.5 (23) INU INU NA
833.8 417.7 (50.1) 242.6 (29.1)
Continued next page
584Galarraga
etal.
2011AdisData
Inform
atio
nBV.Allrig
htsreserved.
Pharm
acoeconomics2011;29(7)
-
Table I. Contd
Study, year of
publication
(country)
City/setting Scope ART unitcosts
Cost components (% of total cost) ART regimenb
facilitiesc
(provider type)
patients ARV
drugs
personnel laboratory tests overheads
Lower-middle-income countries
Gupta et al.,[30]
2009e,j,k (India)
Chennai/U 1 (Nat) 2606 96.7 46.2 (47.8) 7.1 (7.3) 18.9 (19.5)
Notincluded
d4T +3TC +NVPFirst-line
Imphal/U 1 (Nat) 226 377.2 130.3 (34.5) 33.3 (8.8) 97.8
(25.9)
Ahmedabad/U 1 (Nat) 1210 133.2 60.7 (45.5) 8.5 (6.4) 42.4
(31.8)
New Delhi/U 2 (Nat) 1728 120 53.4 (44.5) 11.4 (9.5) 20.3
(16.9)
Trivandrum/U 1 (Nat) 498 217.9 118.3 (54.2) 12.8 (5.9) 60.3
(27.6)
Thrissur/U 1 (Nat) 308 155.5 101.4 (65.1) 8.7 (5.6) 0 (0)
John et al.,[31]
2006h,k (India)
Bangalore/U 1 (NGO) 25in/outpts
461.8 257.7 (55.8) INU project
coordinator, medical
officer, nurses,
laboratory
technician,
counsellors
17.4 (3.8) INU AZT/d4T + 3TC+NVPFirst-line
Cleary et al.,[19]
2007e,l (Lesotho)
Maseru/U 1 (Nat/NGO) 271 inpts 161.7 127.4 (78.8) INU 15.3 (9.5)
INU 3TC+ d4T +NVPFirst-line
Loubiere et al.,[32]
2008e,m (Morocco)
Casa Blanca/U 1 (Nat) 167in/outpts
1076.9 INU INU INU INU NA
Kombe et al.,[33]
2004e,n (Nigeria)
Lagos/U;Abuja/U
5 (Nat) NA 841 420.5 (50) 185 (22) 193.4 (23) 33.6 (4) d4T +3TC
+NVPFirst-line
PHR,[34] 2004g,k,o
(Nigeria)
Anambra/U;Bauchi/R;Edo/U;Federal Capital
Territory/U;Kano/U;Lagos/U;Nassarawa/U;Rivers/U
66
(Nat/FBO/NGO)NA
in/outpts1023.3 362.1 (35.4) 391.2 (38.2) 237.5 (23.2) 18.6
(1.8) 3TC+ d4T +NVP
First-line
Continued next page
Antiretro
viral
Treatm
entUnitCosts
585
2011AdisData
Inform
atio
nBV.Allrig
htsreserved.
Pharm
acoeconomics2011;29(7)
-
Table I. Contd
Study, year of
publication
(country)
City/setting Scope ART unitcosts
Cost components (% of total cost) ART regimenb
facilitiesc
(provider type)
patients ARV
drugs
personnel laboratory tests overheads
Kitajima et al.,[35]
2003e,p (Thailand)
Khon Kaen/R 2 (Nat) 106outpts
3903.8 3415.1 (87.5) INU 462 (11.8) 26.7 (0.7) NA
Upper-middle-income countries
Acurcio et al.,[36]
2006e,g,q (Brazil)
Belo Horizonte/U 2 (Nat) 157adherent
patients
1454 1290.3 (88.7) 6.2 (0.4) infectious
diseases specialist
and other specialists
45 (3)
lymphocyte
CD4/CD8, viralload test, other
tests
Not
included
NA
40 non-
adherent
patients
1340.1 1290.3 (96.3) 8.7 (0.6) infectious
diseases specialist
and other specialists
39.9 (2.9)
lymphocyte
CD4/CD8, viralload test, other
tests
Marques et al.,[24]
2007i (Brazil)
Sao Paulo/U 1 (Nat) 140 outptchildren
2038.8 INU INU INU INU RTV+NFV +3CT +AZT +DDL + IDV
Aracena-Genao
et al.,[37] 2008e,r
(Mexico)
Mexico City/U 1 (Nat) 80 7688 INU INU INU INU NA
Bautista-Arredondo
et al.,[38] 2003;
Bautista-Arredondo
et al.,[39] 2008e,s
(Mexico)
Mexico City/U; 11 (Nat) 902 6651.3 firstyear
6024.4 (90.6) INU 223.4 (3.4) INU NA
Cuernavaca/U; 4256.8second year
3809.4 (89.5) 165.8 (3.9)
Guadalajara/U 4682.9 thirdyear after
ART
initiation
3960.7 (84.6) 254.6 (5.4)
Cleary et al.,[40]
2006e,t
(South Africa)
Cape Town/U 3 (Nat) 1729 987.3 338.8 (34.3) INU 96 (9.7) INU AZT
+3TC +NVP/EFV first-line
1774.3 1108.3 (62.5) 112 (6.3) AZT +DDL +LPV/RTV second-line
Continued next page
586Galarraga
etal.
2011AdisData
Inform
atio
nBV.Allrig
htsreserved.
Pharm
acoeconomics2011;29(7)
-
Table I. Contd
Study, year of
publication
(country)
City/setting Scope ART unitcosts
Cost components (% of total cost) ART regimenb
facilitiesc
(provider type)
patients ARV
drugs
personnel laboratory tests overheads
Deghaye et al.,[41]
2006e,u
(South Africa)
Durban/U 2 (Nat) 41 1022.3 411 (40.2) 330 (32.3) 218.8 (21.4)
16.4 (1.6) d4T +3TC +EFVfirst-line
Harling et al.,[42]
2007e,v
(South Africa)
Cape Town/U 1 (Nat/D) NT 521.6 INU INU Dr, nurse,counsellor,
pharmacist
INU INU NT
Harling et al.,[43]
2007e,u
(South Africa)
Cape Town/U 1 (Nat/NGO) 172outpts
1296.5 535.9 (41.3) 203.3 (15.7) 341 (26.3) 25.2 (1.9) d4T +3TC
+EFVfirst-line
129 inpts 1834.3 535.9 (29.2) 545.5 (29.7) 442.1 (24.1) 212.7
(11.6)
Kevany et al.,[44]
2009w
(South Africa)
Cape Town/U 1 (Nat) 48 outpts 1795.8 103.9 (5.8) 988.6
(55.1)consultant, medical
officer, nursing
sister, medical ward
staff
478.1 (26.6) Not
included
3TC+ d4T +NVPfirst-line
25 inpts 6579.5 271.7 (4.1) 2097.8 (31.9)
consultant, medical
officer, nursing
sister, medical ward
staff
847.6 (12.9)
Martinson et al.,[45]
2009e,v
(South Africa)
Soweto/U 1 (D) 591 2853.6 (newART
patients)
958 (33.6) 435.5 (15.3) primary
health care nurse,
pharmacist, medical
officer
1174.4 (41.2) 190.5 (6.7) d4T +3TC +NVPfirst-line
1770.4
(established-
ART
patients)
958 (54.1) 435.5 (24.6) primary
health care nurse,
pharmacist, medical
officer
114.5 (6.5) 190.5
(10.8)
Rosen et al.,[46]
2008e (South Africa)
Gauteng/U 2 (Nat/NGO) 200 1122.5(patient
responding)
INU INU INU INU 3TC+ d4T +EFVfirst-line
1210.3
(patient not
responding)
Continued next page
Antiretro
viral
Treatm
entUnitCosts
587
2011AdisData
Inform
atio
nBV.Allrig
htsreserved.
Pharm
acoeconomics2011;29(7)
-
Table I. Contd
Study, year of
publication
(country)
City/setting Scope ART unitcosts
Cost components (% of total cost) ART regimenb
facilitiesc
(provider type)
patients ARV
drugs
personnel laboratory tests overheads
Eastern Cape/U 1 (NGO) 100 1241.1(patient
responding)
1177.3
(patient not
responding)
Mpumalanga/R 1 (NGO) 100 1229.4(patient
responding)
1182.7
(patient not
responding)
Vella et al.,[22]
2008e,i
(South Africa)
KwaZulu-Natal/U 32 (Nat) 2835 1068.1 500 (46.8) 300 (28.1) 236
(22.1) 16.6 (1.5) NA
a Dollar amounts and proportions of overall costs may not add to
100% due to rounding and because some costs do not fit into the
classifications used. For example, transportationcosts to visit
patients homes, non-ARV medicines, opportunistic infections
medicines and recurrent costs such as cotton, syringes, gloves,
etc. are in the Other category, which
has been excluded from the table (but could be estimated by
subtracting the current total from 100%). ARV medicine costs
include only the costs of ARVs, mostly first-linetreatment unless
otherwise noted in the Regimen column. Personnel costs include
human resource expenses incurred in service delivery, including
health workers doing patient
care and, in some studies, administrative personnel. Laboratory
costs may include HIV testing, CD4 cell count and viral load
measurement. Overheads may include capital costs
(e.g. medical equipment, computers) and/or recurrent costs
(utilities: electricity, water, phone, rent, maintenance costs,
storage, cleaning or security). If the study reviewed
provideddetails about what was included in each category, those
details have been incorporated in the table. Costs were converted
into $US, year 2009 values, using average exchange
rates during the study period (see sources in theMethods
section). For further details see the Supplemental Digital Content
(http://links.adisonline.com/PCZ/A117) versions of the table.
b Refers to the ARV line costed by the study.
c Refers to the number of facilities in which the costs were
collected; whether the programme was national or local; and whether
the provider was different from public/local government(i.e. NGO,
FBO or D).
d International dollars converted into $US using the purchase
power parity conversion factor provided by the UN Statistical
Office.[47]
e The study reports the average costs.
f Outpts only (includes patients aged 15 years).
g The costs collected are financial costs.
h The study reports the median costs.
i Includes children (
-
(such as South Africa[22,40-46]) and between siteswithin one
study (e.g. Gupta et al.,[30]) [figure 2].While variations between
studies may in partreflect differences in costing methods and
defini-tions, variations within studies point to the ex-istence of
real differences in actual cost betweensites and/or patient
populations.
2.1 Antiretroviral Medicines
In all three income groups, ARV medicineswere the largest cost
component. In LIC, theyrepresented 64% of the overall costs, at a
median$US428 per patient-year (mean: 456 [62%], range:205607
[3778%]) [summary of table I is availablein the Supplemental
Digital Content only]. InLMIC, ARVmedicines represented 50% of
overallcosts (median: $US127, mean: 463 [54%], range:463415
[3588%]) [see the Supplemental digitalContent]. In UMIC, ARV
medicines covered47% of costs: median $US958 per patient-year,mean:
1473 (53%), range: 1046024 (496%)[table I].
2.2 Laboratory
In LIC and LMIC, laboratory costs werethe second most important
cost component. InLIC they accounted for 14% of overall
costs(median: $US123; mean: 133 [17%], range: 16242[630%]) [see
table I in the Supplemental DigitalContent]. In LMIC, laboratory
costs correspondedto 20% of overall costs (median: $US42; mean:106
[18%], range:
-
were the second most important cost component,at 26% of overall
cost (median: $US383, mean:535 [24%], range: 62098 [0.455%]) [table
I].
Most (23 of 26) studies included the costof personnel only at
the facility level. While typesof services costed varied, only
seven studies re-
Table II. Base-case and sensitivity analyses: delivery unit cost
($US, year 2009 values) of antiretroviral treatment (ART) per
patient per year,
by country income levela
Low
income
Lower-middle
income
Upper-middle
income
All
Base case
Mean 839 1246 2783 1494
Median 792 932 1454 1005
SD 175 1546 2500 1628
Range 6821089 1563904 12305667 1565667
Countries 4 5 3 12
Studies (n) 6 7 13 26
Data points/sites (N) 10 12 24 461. Weighing all study sites,
instead of all countries equally
Mean 770 714 2366 1588
Median 797 298 1397 1073
SD 226 1067 2034 1754
Range 2631089 973904 5227688 977688
Countries 4 5 3 12
Studies (n) 6 7 13 26
Data points/sites (N) 10 12 24 462. Adding studies reporting
total unit cost without ARV/laboratory/human resource breakdownMean
713 1314 Unchanged from
base case
1422
Median 773 932 932
SD 319 1637 1639
Range 2111089 1624212 1625667
Countries 5 5 13
Studies (n) 7 9 29
Data points/sites (N) 11 14 483. Country income level according
to costing yearb
Mean 565 1683 3068 1588
Median 646 1454 1241 1073
SD 346 886 2537 1754
Range 971089 5223904 10687688 977688
Countries 7 4 2 12
Studies (n) 11 9 6 26
Data points/sites (N) 20 13 13 46a The base case and scenario 2
aggregate total ART unit costs per person per year by first taking
the median within each country. Sensitivity
analyses are as follows: 1. All study sitesweighted equally,
regardless of country or number of sites per country. 2.
Expandeddataset to include
studies that did not include a breakdown of cost components, but
had data on total ART costs. 3. Use World Bank gross national
income per
capita data at the time of the study costing, instead of 2009
gross national income, to classify countries according to income
level. Baseline
estimates based on more than 14765 patients from 26 studies,
plus additional studies that did not disaggregate costs by cost
component.
Twelve studies specifically reported costs for first-line
treatment; one study[40] specifically included second-line
treatment but those specific
costs were excluded. See table I, text and Supplemental Digital
Content, 1 (http://links.adisonline.com/PCZ/A117) for further
details.
b The total number of countries in scenario 3 adds to 12 only
because, given the timing of the costing, and the countrys economic
development,
South Africa is in both the lower-middle-income and the
upper-middle-income categories.
ARV= antiretroviral.
590 Galarraga et al.
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
8000
7000
6000
(
$
U
S
,
y
e
a
r
2
0
0
9
v
a
l
u
e
s
)
5000
4000
3000
2000
1000
K
o
m
b
e
e
t
a
l
.
,
2
0
0
4
B
i
k
i
l
l
a
e
t
a
l
.
,
2
0
0
9
P
E
P
F
A
R
t
e
a
m
,
2
0
0
9
J
a
f
f
a
r
e
t
a
l
.
,
2
0
0
9
K
o
e
n
i
g
e
t
a
l
.
,
2
0
0
8
H
o
u
n
t
o
n
e
t
a
l
.
,
2
0
0
8
J
o
h
n
e
t
a
l
.
,
2
0
0
6
G
u
p
t
a
e
t
a
l
.
,
2
0
0
9
C
l
e
a
r
y
e
t
a
l
.
,
2
0
0
7
P
H
R
,
2
0
0
4
K
o
m
b
e
e
t
a
l
.
,
2
0
0
4
K
i
t
a
j
i
m
a
e
t
a
l
.
,
2
0
0
3
L
o
u
b
i
e
r
e
e
t
a
l
.
,
2
0
0
8
C
l
e
a
r
y
e
t
a
l
.
,
2
0
0
6
D
e
g
h
a
y
e
e
t
a
l
.
,
2
0
0
6
H
a
r
l
i
n
g
e
t
a
l
.
,
2
0
0
7
H
a
r
l
i
n
g
e
t
a
l
.
,
2
0
0
7
V
e
l
l
a
e
t
a
l
.
,
2
0
0
8
R
o
s
e
n
,
2
0
0
8
K
e
v
a
n
y
e
t
a
l
.
,
2
0
0
9
M
a
r
t
i
n
s
o
n
e
t
a
l
.
,
2
0
0
9
A
c
u
r
c
i
o
e
t
a
l
.
,
2
0
0
6
M
a
r
q
u
e
s
e
t
a
l
.
,
2
0
0
7
B
a
u
t
i
s
t
a
-
A
r
r
e
d
o
n
d
o
e
t
a
l
.
,
2
0
0
3
A
r
a
c
e
n
a
-
G
e
n
a
o
,
2
0
0
8
Ethiopia* *
UgandaHaiti
Benin India
ARV drugsLaboratory testsPersonnelOther costsCosts not
unbundled
Lesotho
Nigeria Morocco
Thailand
South Africa
*
*
Brazil
Mexico
0
Fig. 2. Unit costs for delivery of antiretroviral treatment
(ART), per patient-year, by cost component, country and
site.[19,22-31,33-38,40-46] Countries are ranked from left to right
accordingto increasing per capita gross national income (GNI) and
then by year of study publication. Studies including (or limited
to) paediatric ART are indicated with an asterisk (*) over the
bar.Other costs refer to components of ART other than
antiretroviral (ARV) medicines, laboratory and personnel costs;
other costs are not comparable across studies because they donot
contain the same elements. Unbundled costs refer to the overall
cost of ART in case the study did not report the separate cost
components but these still included at least the threemain
components under review. Each bar represents a separate data point
within each country or study (e.g. different region, health
facility/site, patient type, delivery modality,temporality, etc.).
GNI per capita (year 2009 values): Ethiopia $US280, Uganda $US420,
Haiti $US660, Benin $US690, India $US1070, Lesotho $US1080, Nigeria
$US1160, Thailand$US2840, South Africa $US5820, Brazil $US7350,
Mexico $US9980. See table I for additional details. PEPFAR =US
Presidents Emergency Plan for AIDS Relief; PHR =The Partnersfor
Health Reformplus.
Antiretro
viral
Treatm
entUnitCosts
591
2011AdisData
Inform
atio
nBV.Allrig
htsreserved.
Pharm
acoeconomics2011;29(7)
-
Table III. Delivery unit cost ($US, year 2009 valuesa) of
prevention of mother-to-child transmission (PMTCT) of HIV
Study, year of
publication
(country)
City/settingb
(costing yearc)
Scope Provider
typefPMTCT unit cost Cost (% of total cost) Regimeng
facilitiesd patientse ARV
drugs
personnel laboratory
tests
overheads
Dandona
et al.,[48] 2008
(India)
Andhra Pradesh/U(20056)
16 1212 Nat 2.4 (post-test
counselling)
INU 1.1 (47.3) INU 0.35 (14.7) NVP
251.1h (mother-
neonate pair
receiving NVP)
118.7 (47.3) 36.9 (14.7)
Desmond
et al.,[49] 2004
(South Africa)
Durban/U, Paarl/U,Siloam/R, Frankfort/R(2002)
9 709 (pre-test
counselling)
Nat/D 42.4 INU INU INU Notincluded
NVP +CTX
570 testing 32.5
532 (post-test
counselling)
23.8
117 (mother-
neonate pair
receiving NVP)
208.8i
63 (mother-
neonate pair
follow-up)
327.2j
McMennamin
et al.,[50] 2007
(Rwanda)
Gicumbi/R (2005) 6 NA Nat/FBO 6.42 specificintervention
INU INU INU INU Not reported
a Costs were converted into $US, year 2009 values, using average
exchange rates during the study period.
b Refers to the city and the type of location.
c Refers to the year for which the study reported the costs.
d Refers to the number of facilities in which the costs were
collected.
e Refers to the number of patients studied to cost the ART
delivery.
f Refers to the type of ART provider.
g ARV line costed by the study.
h The mother was given one tablet of NVP 200mg at the onset of
labour and the neonate was given NVP 2mg/kg bodyweight within 72
hours of birth.
i This cost includes re-issue of NVP to mothers, and provision
of NVP suspension to the child and any formula milk for the
duration of the hospital stay.
j PMTCT activities delivered after the mother and infant are
discharged from hospital after delivery, including CTX and the
provision of formula feeding. It does not include infant HIV
testing.
ARV = antiretroviral; CTX = cotrimoxazole; D =donor; FBO =
faith-based organization; INU = included but not unbundled; NA= not
available in the study; Nat =National (if Nat notmentioned it means
that the provider is local); NGO =non-government organization; NVP
= nevirapine; R = rural; U =urban.
592Galarraga
etal.
2011AdisData
Inform
atio
nBV.Allrig
htsreserved.
Pharm
acoeconomics2011;29(7)
-
ported details of the types of personnel
in-cluded.[27,31,36,40,42,44,45] Of these, three
includedabove-facility-level personnel such as humanresources at
distribution centres or district/provincial/national programme
management, ormonitoring and evaluation in the personnel
costcomponent[30,31,34] (see section 2.4).
2.4 Programme-Level Costs
Three studies included programme-level costsassociated with ART
delivery.[30,31,34] Across twourban sites in India,[30]
programme-level costs onmedical and administrative personnel other
thanthose directly involved in ART delivery to patientswere
reported to be about 16% of per patient-yearrecurrent costs during
the first year of the pro-gramme, falling to about 7% in the second
year.Another study in India[31] included a project co-ordinator
working across different health facilitiesunder personnel cost. In
urban Nigeria,[34] reportedabove-facility expenses included
transport of drugsfrom federal central facilities to the ARV
centres.
2.5 Paediatric Antiretroviral Treatment
We identified three studies reporting costs ofART for patients
aged 60% of totalcost), and laboratory tests $US191 (20%) in new
or$US94 (10%) in established patients.
3. Prevention of Mother-to-ChildTransmission
Three eligible PMTCT costing studies wereanalysed. The cost per
mother-neonate pair re-ceiving nevirapine around child delivery
was$US251.1 in India,[48] $US208.8 in SouthAfrica[49]
(both middle-income countries), but only $US6.4in Rwanda,[50] an
LIC (table III). The large dif-ference in costs suggests that these
studies includeddifferent packages of PMTCT services and/or
different cost components (see Discussionsection).
The SouthAfrican study[49] presented a thoroughcosting in
already existing facilities, focusing onthe added financial costs
utilized specifically bythe PMTCT programme. Across four
facilities,the average unit cost per HIV-infected mother-neonate
pair receiving prophylactic nevirapineservices was $US208.8. This
cost included nevi-rapine to mothers at delivery, as well as
provisionof nevirapine suspension to the infant and for-mula milk
for the duration of the hospital stay.Additional costs estimated in
the study were$US42.4 for the pre-test counselling and $US32.5for
testing costs. The study reported a substan-tially higher cost for
post-discharge follow-upcare, at $US327.2 per HIV-infected
mother/infantpair covering cotrimoxazole and provision of in-fant
formula after birth for the duration of thehospital stay (table
III). The unit costs varied ac-cording to the denominator unit used
(pregnantwomen screened, or HIV-infected women andneonate pair
receiving ARV prophylaxis) as wellas between the four locations
according to theireconomic and epidemiological differences, e.g.
ahigh HIV prevalence and resource-poor setting(Frankfort, Free
State) compared with a lowHIVprevalence and better resourced
setting (Paarl,Western Cape).
In Andhra Pradesh, India,[48] across 16 PMTCTcentres, 1212
HIV-infected pregnant women re-ceived PMTCT, of 125 073 counselled
and testedpregnant women. The average economic cost perHIV-infected
woman receiving nevirapine aroundchild delivery was $US251. Average
unit cost,expressed per pregnant woman counselled andtested,
irrespective of HIV status, was $US2.40.Given the low price of
nevirapine, personnel wasthe major cost component, at $US119 or 47%
ofoverall cost per HIV-infected woman, followedby $US37 (15%) on
overheads.
In Rwanda,[50] the PMTCT unit cost per spe-cific package was
much lower, at $US6.42. Themarked contrast with India and South
Africa is
Antiretroviral Treatment Unit Costs 593
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
in part explained by Rwandas much lower per-capita income level
and health worker salaries.Equally important, this unit cost was
achieved ina large-scale programme in a high HIV-prevalencesetting,
with the six health centres covering acatchment population of 148
151 individuals, witha high PMTCT uptake rate of 71.6%, which
likelyachieved significant economies of scale. How-ever, the study
does not specify number of preg-nant women screened for PMTCT, the
number ofHIV-infected women or the number of mother-infant pairs
given nevirapine.
4. Sensitivity Analyses
A first sensitivity analysis assessed the effect ofaggregating
unit cost within country-income groupsby weighing all data points
and sites across stud-ies and countries equally, instead of the
base-casemodel (presented above), which weighed allcountries
equally, irrespective of their number ofdata points. In this
variant model, median unitcosts for ART were nearly unchanged for
LICand UMIC: in LIC, the median was $US797(vs $US792 in the
base-case model) and 1397in UMIC (vs $US1454 in the base-case
model;table II). However, for LMIC, this alternativeaggregation
method resulted in a 3-fold reducedmedian cost estimate per
patient-year: $US298compared with $US932 in the base case.
Thisdifference reflects an over-representation of studysites and
data points from India among LMIC,with exceptionally low unit costs
reflecting Indiasuniquely low ARV prices due to the
countrysimportant generic ARV industry and resultingstrong
pharmaceutical negotiation capacity.
A second sensitivity analysis, to further check
onrepresentativeness, expanded the data set to includea number of
studies that did not include a break-down of cost components, but
had data on totalART costs. With the addition of one study
fromKenya,[51] the LICmedian cost per patient per yearbecame $US773
(mean: 713, range: 2111089), sim-ilar to the original estimate of
$US792. AmongLMIC, adding data points from Thailand andIndia[52-54]
did not alter the median cost per patientper year from the default
of $US932 (but the meanincreased to 1314, range: 1624212).
Third, we replaced the country-income classi-fication according
to each countrys 2009 incomelevel by a classification using each
countrys in-come level in the year of cost data collection(while in
both cases applying World Bank 2009income thresholds). This shifted
five countriesthat had grown richer between the costing studyyear
and 2009 (Brazil, India, Lesotho, Nigeriaand South Africa) into a
lower income category.This country regrouping reduced the medianART
per-patient cost among LIC (from $US792to $US646), increased the
estimate for LMIC(from $US932 to $US1454) and decreased theestimate
for UMIC (from $US1454 to $US1241).
5. Discussion
Empirical data on ART delivery unit costs inlow- and
middle-income countries has increasedsignificantly in recent years,
as access to ART hasexpanded. The current review identifies
severalmore good-quality costing studies of both out-patient and
home-based ART than earlier re-views,[12,13,15] and provides
separate cost estimatesfor the components ARVs, personnel and
labor-atory costs.
The results have several implications forprogramme planning and
budgeting, and for seek-ing efficiency improvement. For ART, the
key costdrivers within health facilities are ARV procure-ment,
laboratory tests and personnel. ARV priceshave declined
substantially over the last decade,with average declines of 1239%
over 20069 forthe first-line regimens most commonly used inlow- and
middle-income countries.[55] Lackingspecification of regimens and
drug sources in sitesstudied, it was not possible to adjust cost
estimatesfor price declines. Only six of the 24 ART costingstudies
reported whether the programme usedgeneric or innovator drugs (see
the expanded ver-sions of table I in the Supplemental Digital
Con-tent). Of 14 studies that specified ARV regimens,the majority
used lamivudine (3TC)+ nevirapine+stavudine (d4T) as the most
common first-lineregimen, the regimen for which recent price
de-clines have been smallest: 912% per year.[55] Thepredominance of
this regimen may explain why,over 20049, the average proportion of
overall
594 Galarraga et al.
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
ART cost covered by ARV drugs was relativelystable, at 4852%
every year, rather than declining.
The observed large variations in laboratorycosts between
countries and sites indicate possibleopportunities for efficiency
gains. Variations maypartly relate to patients differing clinical
stages andresponses to treatment,[46] with laboratory costsbeing
higher during the first 6 months after ARTinitiation.[23]
Furthermore, laboratory costs tend tobe lower in established
programmes or in hospitals,for example, $US15.7 or 6% of overall
cost inEthiopia in 20045,[26] compared with programmesor clinics
that are starting up and therefore invest-ing more in laboratory
equipment.[27]
Available data (table I) were limited in thespecification of
laboratory tests performed, test-specific costs and precise
laboratory testingpractices and frequencies. We could therefore
notanalyse whether sites with routine, systematicCD4 and viral load
monitoring were more ex-pensive than sites with selective viral
load mon-itoring, sites with only (or less frequent) CD4testing, or
sites with just clinical monitoring, andcannot estimate the
possible savings from shiftingtesting patterns. In the DART
(Development ofAntiRetroviral Therapy in Africa) trial of
clinicallydriven versus laboratory-guided ART in Ugandaand
Zimbabwe, haematology and biochemistryadded little benefit.
Twelve-weekly CD4 monitor-ing improved clinical outcomes from the
2nd yearfrom treatment initiation onwards. However, thecost of CD4
testing would have to drop to $US3.8(from current $US175) in order
to make this mon-itoring practice cost effective.[56-58]
Another approach to improving the efficiencyof ART programmes
may be personnel taskshifting.[59-61] In a non-Government
Organiza-tion-led programme in rural Uganda, costs to theprovider
were similar for clinic-based and home-based ART ($US834 and $US789
per patient-year, respectively), at comparable patient retentionand
clinical outcomes.[29] The study suggests anopportunity to increase
ART access in settingswith restrained health system capacity. The
mark-edly higher cost of inpatient versus outpatientART (two
studies in South Africa[43,44] [table I])further demonstrates the
financial importance ofpreventing hospitalizations.
5.1 Limitations
This review reflects the state of evidence inthe first decade of
global ART scale-up. Giventhe small number of studies and settings,
the avail-able data can hardly claim global representative-ness.
Despite the recent increase in good-qualitystudies (as defined in
our inclusion criteria),costing information from South Asia (2
studies),Latin America and the Caribbean (5) and EastAsia and the
Pacific (1) remains extremely lim-ited. Furthermore, in sub-Saharan
Africa, 8 ofthe 16 studies were conducted in South Africa,which,
with its relatively high income level, is thecountry least
representative of the region. Thelack of regional
representativeness is one reasonthat our first sensitivity analysis
found highermedian ART costs in LIC than in LMIC, a resultthat
seems contrary to intuition and to observedpatterns in ARV
prices.[55] Despite the inclusionof studies from the grey
literature (of which sevenwere eligible[19,22,23,27,33,34,38]), we
cannot excludethe possibility of publication bias in
availablecosting studies. Several recent, well executedART and
PMTCT cost studies that are re-cognized for influencing donor
funding policieshave not been published in books,
peer-reviewedjournals or made otherwise publically available,so
were unfortunately not included in our meta-analyses.
Innovative monitoring and evaluation strategiesare needed to
address the gaps in good-qualityevidence, especially concerning
above-facilityand programme-level costs. We intended toanalyse all
relevant cost components, includingprogramme-level (above
facility-level) activitiesand expenses, including managerial
overheads,administration, monitoring and evaluation andtraining
borne at district, province and nationallevels. However, only three
ART studies[30,31,34]
provided such programmatic costs, without suf-ficient
description of the relevant activities orcost items to analyse the
determinants of thesepotentially significant contributions to
overalldelivery cost.
Quality and completeness of available cost-ing studies varied.
Many did not explicitly re-port all cost components, missing
information
Antiretroviral Treatment Unit Costs 595
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
on such important cost determinants as ARVdrugs, regimens, or
type and frequency of la-boratory testing used. Most studies also
lackeddetails about variation in costs between differenttypes of
facilities and patient populations stud-ied (in terms of, for
example, age and CD4 cellcount at treatment initiation).[23] Basic
infor-mation on cost components was lacking in manystudies; often
either capital costs (such as thoserelated to setting up laboratory
equipment) orselected recurrent costs (such as utilities)
wereomitted. In particular, definitions of overheadcosts varied,
with some studies not even includ-ing any overheads. Only three
studies[33,34,36]
mentioned explicitly that their data collectioninstruments had
been validated or piloted be-fore application. The best studies
reviewed useda micro-costing approach, where expenditureson each
component of providing ART orPMTCT delivery were separately listed
and val-ued. Even when an ingredients approach wasutilized, most
studies in the last decade did notprovide the type of ART used (not
even gener-ic vs patent, or first- vs second-line). In manystudies,
sample sizes appear to have been drivenby clinical outcomes, rather
than by costing-related statistical considerations. Sample
sizesvaried between 209 (median for LIC, range122218) and 430
(median for UMIC, range222835) patients, which may be
consideredappropriate for an overall estimate from onesingle
locality. However, convenience samplinglimits the precision and
power to evaluate costsfor relevant strata of sites, patient age
groups,treatment regimens and modality and phasesof treatment.
Heterogeneity in unit denominators, notablythe calculation of
patient-years, further compli-cates comparisons among studies. For
illustra-tion, a study in India[30] reported 1094 clients whoever
started ART (as average across sites),939 clients at the end of
study and 503 clientswho had been on therapy for the entire
studyperiod. These authors selected the corresponding9460
client-months of ART as an appropriatedenominator unit; however,
the large differencesbetween patients starting and patients
retainedillustrates the critical influence that varying de-
nominator definitions will have on unit cost esti-mates,
especially in programmes with low patientretention and/or rapid
scale-up, with many newpatients initiating ART.
5.2 Future Work
The limitations in availability and compar-ability of existing
cost data described abovehighlight an urgent need for development
anddissemination of standardized cost-measurementmethodologies and
for capacity building. Await-ing such guidance and improved data
quality andstandardization, any cost comparisons betweensites and
studies should be carried out with ex-treme caution and attention
to methodologicaldifferences; interpretations on relative
efficiencywill typically be limited to within-study determi-nants.
Future strategic use of cost data for im-proving the efficiency of
ART delivery will, inparticular, benefit from more careful costing
andreporting of ARV drug sources and regimens,laboratory costs by
type and frequency of tests,and staff costs by type of personnel,
includingthose above the health-facility level.
Future assessments should link costs to healthoutcomes,
including cost differentials associatedwith clinical response,[46]
drug toxicity and resis-tance.[44] This is ever more important in
view ofthe WHO revised 2010 guidelines for adult ARTin
resource-limited settings, which recommendedan expanded access to
ART, starting at CD4 cellcount
-
second, to exchange experiences in increasingservice efficiency.
Multi-country exercises shouldhelp to improve and standardize
costing meth-odologies and tools for comprehensive data
col-lection. Economies of scale, as apparent in somesettings,[30]
are worth investigating to establishmore clearly if these are
achieved merely by allo-cating fixed costs over a larger number of
pa-tients, or by improved technical efficiency (i.e.transformation
of inputs to outputs) associatedwith programme maturation and
learning.
Whenever possible (conditional on the avail-ability of data on
patient wages and other oppor-tunity costs), both financial and
economic costsshould be collected. Financial costs are relevantfor
budgeting, while economic costs (includingopportunity costs,
productivity losses, transportand out-of-pocket expenditure by
patients) willdetermine cost effectiveness and prioritization
ofresource allocations. For example, if an ART ser-vice package
includes in-kind contributions suchas food subsidies or other
resources donated fromexternal funding sources, these will
influence fu-ture planning and sustainability. Ignoring
suchexternal assistance would distort the picture ofoverall costs
and incentive structures in which anART programme operates.
6. Conclusions
There is a paucity of information about unitcosts for delivery
of ART and PMTCT in differ-ent HIV/AIDS programmes, particularly in
LIC.Future evaluations of programme-level ART andPMTCT unit costs
will benefit from internation-al guidance on standardized
expenditure defini-tions and categories, standardized formats
forspecifying ARV regimen mixes, laboratory test-ing practices
(including type of tests and frequency)and human resource
disaggregation (facility levelvs above-facility and programme
level); as well asstandardized service unit denominators
(possiblyincluding a component of service quality or pa-tient
retention). The large differences in ARTunit costs observed in
settings with similar epi-demiological and economic characteristics
deserveadditional assessments focusing on cost deter-minants and
opportunities for efficiency gain
in programme implementation and scale-up. Toscale-up ART and
PMTCT to global universalaccess, innovative options are needed to
containcosts while maintaining or improving quality andhealth
gain.
Acknowledgements
The authors thank Daniel Acuna, who conducted the ini-tial
bibliographic search, as well as Lazarus Muchabaiwa andJesse Kigozi
for their research assistance. The authors alsoacknowledge Sergio
Bautista-Arredondo, Lisa DeMaria,Steven Forsythe, Lori Bollinger,
Megan OBrien and variousparticipants at the first Latin America
& the Caribbean Con-ference on Global Health in Cuernavaca,
Mexico and at theXVIII International AIDS Conference in Vienna,
Austria, forhelpful comments. The research was partially funded by
theGlobal Fund to Fight AIDS, Tuberculosis and Malaria(PO#2008301)
through the National Institute of Public Health(INSP)/Consortium
for Research on HIV/AIDS and Tu-berculosis (CISIDAT[63]).
Additional funding for OmarGalarraga was provided by the US
National Institutes ofHealth (NIH)/Fogarty International Center
(K01-TW008016-02)through the Institute of Business
andEconomicResearch (IBER)at theUniversity of California, Berkeley,
CA,USA. The opinionsexpressed in the review do not reflect the
views of any of thefunding or other organizations that supported or
facilitated thestudy. The authors are solely responsible for the
contents; theyreport no conflicts of interest.
References1. WHO/UNAIDS/UNICEF. Towards universal access:
scal-
ing up priority HIV/AIDS interventions in the health sec-tor.
Progress report. Geneva: WHO, 2009
2. Beck EJ, Santas XM, Delay PR. Why and how to monitorthe cost
and evaluate the cost-effectiveness of HIV servicesin countries.
Aids 2008 Jul; 22 (Suppl. 1): S75-85
3. The Global Fund to Fight AIDS, Tuberculosis and MalariaThird
Replenishment 20112013. Improving value formoney in Global
Fund-supported programs. Geneva: TheGlobal Fund to Fight AIDS,
Tuberculosis and Malaria,2010 Mar [online]. Available from URL:
http://www.theglobalfund.org/documents/replenishment/2010/Improving%20Value%20for%20Money%20in%20Global%20Fund%20Supported%20Programs.pdf
[Accessed 2010 Mar 28]
4. JointUnitedNations Programme onHIV/AIDS (UNAIDS).What
countries need: investments needed for 2010 targets.Geneva: UNAIDS,
2009 [online]. Available from
URL:http://data.unaids.org/pub/Report/2009/jc1681_what_countries_need_en.pdf
[Accessed 2010 Jun 15]
5. UNAIDS and The World Bank Global HIV/AIDS pro-gram. The
global economic crisis and HIV preventionand treatment programmes:
vulnerabilities and impact.Washington, DC: The World Bank, 2009
6. Tom Lantos and Henry J. Hyde United States Global Lea-dership
Against HIV/AIDS, Tuberculosis, and MalariaReauthorization Act of
2008, HR5501, 110th Cong
Antiretroviral Treatment Unit Costs 597
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
7. UK Department for International Development.
Achievinguniversal access: the UKs strategy for halting and
reversingthe spread of HIV in the developing world.
Monitoringperformance and evaluating impact. 2008 [online].
Avail-able from URL:
http://www.aidsportal.org/repos/achieving-universal-access.pdf
[Accessed 2011 Feb 17]
8. The Global Fund to Fight AIDS, Tuberculosis andMalaria.The
Global Fund 2010: innovation and impact. Geneva:The Global Fund to
Fight AIDS, Tuberculosis and Ma-laria, 2010Mar [online]. Available
fromURL:
http://www.theglobalfund.org/documents/replenishment/2010/Global_Fund_2010_Innovation_and_Impact_en.pdf
[Accessed 2010Jul 10]
9. Izazola-Licea JA, Wiegelmann J, Aran C, et al. Financingthe
response to HIV in low-income and middle-income coun-tries. J
Acquir Immune Defic Syndr 2009 Dec; 52 Suppl. 2:S119-26
10. UNAIDS. Country factsheets. Geneva: UNAIDS, 2010
Dec[online]. Available from URL:
http://www.unaids.org/en/dataanalysis/tools/aidsinfo/countryfactsheets/
[Accessed 2011Feb 19]
11. UNAIDS. Report on the global AIDS epidemic. Geneva:Joint
United Nations Programme on HIV/AIDS (UNAIDS),2010
http://www.unaids.org/globalreport/Global_report.htm [Accessed 2011
Feb 17]
12. Beck EJ, Miners AH, Tolley K. The cost of HIV treatmentand
care: a global review. Pharmacoeconomics 2001 Jan;19 (1): 13-39
13. Bollinger L, Stover J. Financial resources required to
achieveuniversal access to HIV prevention, treatment, care
andsupport: methodology for care and treatment interven-tions. 2007
[online]. Available from URL:
http://data.unaids.org/pub/Report/2007/20070925_annex_iii_treatment_care_methodology_en.pdf
[Accessed 2010 Apr 26]
14. Ades AE, Ratcliffe J, Gibb DM, et al. Economic issues in
theprevention of vertical transmission of HIV. Pharmacoeco-nomics
2000 Jul; 18 (1): 9-22
15. Beck EJ, Harling G, Gerbase S, et al. The cost of
treatmentand care for people living with HIV infection:
implicationsof published studies, 19992008. Curr Opin HIV AIDS2010
May; 5 (3): 215-24
16. PRISMA. Preferred Reporting Items for Systematic Re-views
and Meta-Analyses [online]. Available from
URL:http://www.prisma-statement.org/statement.htm [Accessed2009 Feb
17]
17. World Bank. World Bank list of economies (July
2009)[online]. Available from URL:
http://siteresources.worldbank.org/DATASTATISTICS/Resources/CLASS.XLS[Accessed
2009 Jul 17]
18. International Monetary Fund. International exchange
rates.Washington, DC: IMF, 2009
19. Cleary S, Tshehlo R, Jouquet G, et al. Ensuring access
tofree HIV/AIDS care and treatment in Lesotho. Capetownand Maseru:
University of Cape Town, Ministry of Health& Social Welfare,
Kingdom of Lesotho, Doctors WithoutBorders/Medecins Sans
Frontie`res, Scott Hospital HealthService Area, 2007
20. Lesotho Central Bank. Official exchange rates. 2009
[on-line]. Available from URL:
http://www.centralbank.org.ls/statistics/2006%20Annual%20Report%20Tables1(final).pdf[Accessed
2009 Dec 15]
21. US Bureau of Labor Statistics. Consumer price index:
his-torical dataset. Washington, DC: Bureau of Labor
Statistics,2009
22. Vella V, Govender T, Dlamini S, et al. Evaluation of
anti-retroviral therapy against HIV/AIDS in KwaZulu-Natal,South
Africa. 2008 Oct [online]. Available from
URL:http://www.kznhealth.gov.za/italian/arv.pdf [Accessed2009 Dec
12]
23. PEPFAR ART Costing Project Team. The costs of com-prehensive
HIV treatment in Ethiopia: report of a coststudy of
PEPFAR-supported HIV treatment programs inEthiopia. Atlanta (GA):
CDC and PEPFAR, 2009
24. Marques HH, Couttolenc BF, Latorre Mdo R, et al. Costsof
care provided in a university hospital for children ex-posed to or
infected with the HIV/AIDS. Cad Saude Pub-lica 2007; 23 Suppl. 3:
S402-13
25. Hounton SH, Akonde A, Zannou DM, et al. Costing uni-versal
access of highly active antiretroviral therapy inBenin. AIDS Care
2008 May; 20 (5): 582-7
26. Bikilla AD, Jerene D, Robberstad B, et al. Cost estimates
ofHIV care and treatment with and without anti-retroviraltherapy at
Arba Minch Hospital in southern Ethiopia.Cost Eff Resour Alloc
2009; 7: 6
27. Kombe G, David G, Raj G, et al. Human and financial
re-source requirements for scaling up HIV/AIDS services inEthiopia.
Bethesda (MD): The Partners for Health Re-formplus Project, Abt
Associates, 2004
28. Koenig SP, Riviere C, Leger P, et al. The cost of
antiretro-viral therapy in Haiti. Cost Eff Resour Alloc 2008; 6:
3
29. Jaffar S, Amuron B, Foster S, et al. Rates of
virologicalfailure in patients treated in a home-based versus a
facility-based HIV-care model in Jinja, southeast Uganda: a
cluster-randomised equivalence trial. Lancet 2009 Dec 19; 374
(9707):2080-9
30. Gupta I, TrivediM,Kandamuthan S.Recurrent costs of
Indiasfree ART program. In: Haacker M, Claeson M, editors. HIVand
AIDS in South Asia: an economic development risk.Washington, DC:
World Bank, 2009: xxvi, 244
31. John KR, Rajagopalan N, Madhuri KV. Brief commu-nication:
economic comparison of opportunistic infectionmanagement with
antiretroviral treatment in people livingwith HIV/AIDS presenting
at an NGO clinic in Bangalore,India. MedGenMed 2006; 8 (4): 24
32. Loubiere S, el Filal KM, Sodqi M, et al. When to
initiatehighly active antiretroviral therapy in low-resource
set-tings: the Moroccan experience. Antivir Ther 2008; 13
(2):241-51
33. Kombe G, Galaty D, Nwagbara C. Scaling-up
antiretroviraltreatment in the public sector in Nigeria: a
comprehensiveanalysis of resource requirements. Washington, DC:
Part-ners for Health Reform plus, Abt Associates, 2004
34. PHR. Nigeria: rapid assessment of HIV/AIDS care in thepublic
and private sectors. Bethesda (MD): The Partnersfor Health
Reformplus (PHR) Project, DELIVER,POLICY Project, Abt Associates,
2004
35. Kitajima T, Kobayashi Y, Chaipah W, et al. Costs of med-ical
services for patients with HIV/AIDS in Khon Kaen,Thailand. AIDS
2003 Nov 7; 17 (16): 2375-81
36. Acurcio Fde A, Puig-Junoy J, Bonolo Pde F, et al.
Cost-effectiveness of initial adherence to antiretroviral
therapy
598 Galarraga et al.
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
among HIV infected patients in Belo Horizonte, Brazil.Rev Esp
Salud Publica 2006 Jan-Feb; 80 (1): 41-54
37. Aracena-Genao B, Navarro JO, Lamadrid-Figueroa H,et al.
Costs and benefits of HAART for patients with HIVin a public
hospital in Mexico. AIDS 2008 Jul; 22 Suppl. 1:S141-8
38. Bautista-Arredondo S, Dmytraczenko T, Kombe G, et al.Costing
ofHIV/AIDS treatment inMexico. Bethesda (MD):Partners for Health
Reformplus Project, Abt AssociatesInc., 2003. Report No. 020
39. Bautista-Arredondo S, Dmytraczenko T, Kombe G, et al.Costing
of scaling up HIV/AIDS treatment in Mexico.Salud Publica Mex 2008;
50 Suppl. 4: S437-44
40. Cleary SM, McIntyre D, Boulle AM. The cost-effectivenessof
antiretroviral treatment in Khayelitsha, South Africa:a primary
data analysis. Cost Eff Resour Alloc 2006; 4: 20
41. Deghaye N, Pawinski RA, Desmond C. Financial and eco-nomic
costs of scaling up the provision of HAART to HIV-infected health
care workers in KwaZulu-Natal. S AfrMedJ 2006 Feb; 96 (2):
140-3
42. Harling G, Bekker LG, Wood R. Cost of a dedicated ARTclinic.
S Afr Med J 2007 Aug; 97 (8): 593-6
43. Harling G, Wood R. The evolving cost of HIV in SouthAfrica:
changes in health care cost with duration on anti-retroviral
therapy for public sector patients. J Acquir Im-mune Defic Syndr
2007 Jul 1; 45 (3): 348-54
44. Kevany S, Meintjes G, Rebe K, et al. Clinical and
financialburdens of secondary level care in a public sector
antire-troviral roll-out setting (G. F. Jooste Hospital). S Afr
MedJ 2009 May; 99 (5): 320-5
45. Martinson N, Mohapi L, Bakos D, et al. Costs of
providingcare for HIV-infected adults in an urban HIV clinic
inSoweto, South Africa. J Acquir Immune Defic Syndr2009 Mar 1; 50
(3): 327-30
46. Rosen S, Long L, Sanne I. The outcomes and outpatient
costsof different models of antiretroviral treatment delivery
inSouth Africa. TropMed Int Health 2008 Aug; 13 (8): 1005-15
47. United Nations. Purchasing power parities (PPP)
conversionfactor, local currency unit to international dollar.
MillenniumDevelopment Goals indicators 2010 [online]. Available
fromURL:
http://unstats.un.org/unsd/mdg/Metadata.aspx?IndicatorId=0&SeriesId=699
[Accessed 2010 Feb 17]
48. Dandona L, Kumar SG, Ramesh YK, et al. Outputs, costand
efficiency of public sector centres for prevention ofmother to
child transmission of HIV in Andhra Pradesh,India. BMC Health Serv
Res 2008; 8: 26
49. Desmond C, Franklin L, Steinberg M. The prevention
ofmother-to-child HIV transmission: costing the service in
foursites in South Africa. Durban: Health Systems Trust, 2004
50. McMennamin T, Fellow A, Fritsche G. Cost and revenueanalysis
in six Rwandan health centers: 2005 cost and rev-enues. Boston
(MA): Management Sciences for Health(MSH), 2007
51. Guinness L, Arthur G, Bhatt SM, et al. Costs of hospitalcare
for HIV-positive and HIV-negative patients at
Kenyatta National Hospital, Nairobi, Kenya. AIDS 2002Apr 12; 16
(6): 901-8
52. Calmy A, Klement E, Teck R, et al. Simplifying and adapt-ing
antiretroviral treatment in resource-poor settings:a necessary step
to scaling-up. AIDS 2004 Dec 3; 18 (18):2353-60
53. Ford N, Wilson D, Costa Chaves G, et al. Sustaining accessto
antiretroviral therapy in the less-developed world: les-sons from
Brazil and Thailand. AIDS 2007 Jul; 21 Suppl. 4:S21-9
54. Over M, Marseille E, Sudhakar K, et al.
Antiretroviraltherapy and HIV prevention in India: modeling costs
andconsequences of policy options. Sex Transm Dis 2006 Oct;33 (10
Suppl.): S145-52
55. WHO Global Price Reporting Mechanism (GPRM).Transaction
prices for antiretroviral medicines and HIVdiagnostics from 2008 to
March 2010 [online]. Availablefrom URL:
http://www.who.int/hiv/pub/amds/GPRMsummary_report_may2010.pdf
[Accessed 2010 Jul 28]
56. Mugyenyi P, Walker AS, Hakim J, et al. Routine
versusclinically driven laboratory monitoring of HIV
antiretro-viral therapy in Africa (DART): a randomised
non-inferiority trial. Lancet 2010 Jan 9; 375 (9709): 123-31
57. Medina-Lara A, Kigozi J, Amurwon J, et al. Cost
effec-tiveness analysis of routine laboratory or clinically
drivenstrategies for monitoring anti-retroviral therapy in
Ugandaand Zimbabwe (DART Trial). 5th IAS Conference on
HIVPathogenesis, Treatment and Prevention; 2009 Jul 1922;Cape
Town
58. Alcorn K. DART study shows HIV treatment without
labmonitoring safe, effective in Africa. London: NAM Pub-lications,
2009 Jul 21 [online]. Available from URL:
http://www.aidsmap.com/page/1435211/ [Accessed 2011 Mar 15]
59. Majchrowicz M. Beyond antiretroviral access:
low-costlaboratory tests needed for the developing world. AIDS2003;
17 Suppl. 4: S13-5
60. Schneider H, Blaauw D, Gilson L, et al. Health systems
andaccess to antiretroviral drugs for HIV in Southern
Africa:service delivery and human resources challenges.
ReprodHealth Matters 2006 May; 14 (27): 12-23
61. Hirschhorn LR, Oguda L, Fullem A, et al. Estimating
healthworkforce needs for antiretroviral therapy in
resource-limited settings. Hum Resour Health 2006; 4: 1
62. WHO. Antiretroviral therapy for HIV infection in adultsand
adolescents: recommendations for a public health ap-proach. 2010
version. Geneva: WHO, 2010 [online]. Avail-able from URL:
http://whqlibdoc.who.int/publications/2010/9789241599764_eng.pdf
[Accessed 2011 Feb 19]
63. Consorcio de Investigacion sobre VH/SIDA/TB
[online].Available fromURL: http://www.cisidat.org.mx [Accessed2011
Feb 17 ]
Correspondence: Dr Veronika J. Wirtz, Av. Universidad
655,Cuernavaca, CP 62100, Mexico.E-mail: [email protected]
Antiretroviral Treatment Unit Costs 599
2011 Adis Data Information BV. All rights reserved.
Pharmacoeconomics 2011; 29 (7)
-
Copyright of PharmacoEconomics is the property of ADIS
International Limited and its content may not becopied or emailed
to multiple sites or posted to a listserv without the copyright
holder's express writtenpermission. However, users may print,
download, or email articles for individual use.