2015 Article 1098
2015 Article 1098
2015 Article 1098
Abstract
Background: We report an economic analysis of Human Immunodeficiency Virus (HIV) care and treatment in Indonesia
to assess the options and limitations of costs reduction, improving access, and scaling up services.
Methods: We calculated the cost of providing HIV care and treatment in a main referral hospital in West Java, Indonesia
from 2008 to 2010, differentiated by initiation of treatment at different CD4 cell count levels (0–50, 50–100, 100–150,
150–200, and >200 cells/mm3); time of treatment; HIV care and opportunistic infections cost components; and the costs
of patients for seeking and undergoing care.
Discussion: Before antiretroviral treatment (ART) initiation, costs were dominated by laboratory tests (>65 %), and
after initiation, by antiretroviral drugs (≥60 %). Average treatment costs per patient decreased with time on
treatment (e.g. from US$580 per patient in the first 6 month to US$473 per patient in months 19–24 for those
with CD4 cell counts under 50 cells/mm3). Higher CD4 cell counts at initiation resulted in lower laboratory and
opportunistic infection treatment costs. Transportation cost dominated the costs of patients for seeking and
undergoing care (>40 %).
Conclusions: Costs of providing ART are highest during the early phase of treatment. Costs reductions can potentially
be realized by early treatment initiation and applying alternative laboratory tests with caution. Scaling up ART at the
community level in certain high prevalence settings may improve early uptake, adherence, and reduce transportation
costs.
Keywords: Cost Analysis, Antiretroviral Therapy, HIV, Indonesia
© 2015 Siregar et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Siregar et al. BMC Health Services Research (2015) 15:440 Page 2 of 12
these activities were obtained from the actual budgetary collected information including (but not limited to) clinic
or governmental records. Market prices were used to esti- service fee, travel costs, travelling time, the average num-
mate other capital costs, including equipment, furniture, ber of daily working hours, and monthly expenditures.
and start-up costs (e.g., renovation costs, if applicable). Based on this information, first we calculated the amount
Capital costs were subsequently annualized on the basis of of productive time per patient (in minutes), basically total
the lifetime of the capital items, using a 3 % discount rate time spent at work per month in minutes. Second, we cal-
[17]. We omitted the cost of utilities (i.e., water and elec- culated patient productivity per minute (i.e. monthly ex-
tricity). This result was then multiplied by the proportion penditure per minute as a proxy for income/productivity
of time allocated by the clinic to deliver ART, calculated per minute). Third, we estimated the total productivity
through a separate time motion study in which we loss per patient as the amount of minutes spent to
observed (in a week time, within the clinic) all clinical undergo the treatment (i.e. time spent for traveling,
activities and calculated the amount of time spent on queueing/waiting, and treatment) times patient productiv-
ART-related duties per week by the clinic staff. The ity per minute. Patients did not have to pay for ARV, ARV
total outpatient visit cost was then divided by the monitoring, other lab tests, or OI medication/treatment.
number of total outpatient visits to obtain the unit To avoid double counting in calculating treatment costs
cost per outpatient visit. Patients registered as out- from the societal perspective, we exempted the clinic
patient were never also registered as inpatient in the service fee from this specific calculation (but it is
same period. For example, a patient who is registered included in the patient cost calculation).
as inpatient in a certain month may be registered as All costs were measured in Rupiah, and converted to
outpatient in the next month, but never at the same US$ using the 2010 exchange rate [21]. Both the
month. The details of the outpatient cost are pre- utilization and cost data were analyzed using Microsoft
sented in the Appendix: Table 6). Excel 2007. We report the costs from both health care
The OI treatment costs were calculated based on the system and patient perspectives, and performed statistical
medical resources consumed by OI treatment (e.g., drugs tests to find out the significance of any costs differences
and equipment), excluding hospitalization. Medical between period and CD4 cell count level.
records and the physician’s patient database were used Secondary data related to patients (e.g. ARV and OI
to estimate resource utilization, and the official hospital drugs intake per patient) were taken from the clinic’s
prices issued in 2011 were used to calculate the unit patients database (in the form of an Excel file). The
costs of drugs and equipment. We obtained the unit cost database was anonymized prior to analysis (we utilize
of OI treatment by dividing the total cost of OI treatment the patients hospital ID number during analysis), and
for each CD4 cell count group by its population. We were none of the patients personal identity was published
unable to retrieve data regarding the specific OIs that in any part of the study. On the event of patient
drugs and equipment were used to treat. The average unit costs (primary) data collection, all patients were asked
cost of OI treatment is presented in Appendix: Table 7). to fill in written informed consent forms prior to
Because the data was limited, we did not perform micro- participating in the survey and the survey was
costing when calculating the costs of hospitalization, ARVs, anonymous (no patient names were collected). The
or laboratory tests. We used the World Health Organiza- survey was conducted by a group of enumerators and
tion’s Choosing Interventions that are Cost Effective authors only received the results. The study was
(WHO-CHOICE) estimates [19] to estimate the per day approved by the Padjadjaran University Indonesia,
inpatient hotel cost, which we then used to calculate the Medical Faculty ethical committee.
total hospitalization cost. The WHO-CHOICE estimates
for inpatient cost include items such as personnel, capital,
and food costs, and exclude drugs and diagnostic test Results
costs. The prices of ARV drugs (except for Tenofovir) Patient characteristics
issued by Kimia Farma (a national pharmaceutical Patient characteristics are presented in Table 1. The major-
corporation) were used as the unit costs of ARVs, ity of patients within the CD4 cell count of 0 – 150 cells/
while the price of Tenofovir was based on Bender et mm3 group are males (151 male, 34 female), while females
al. [20]. The unit costs of laboratory tests (CD4 cell dominate the >150 cells/mm3 group (11 male, 16 female).
count, viral load, and other laboratory tests) were de- In average, there are more males across the CD4 cell count
rived from the 2011 official hospital price for each groups (72 %). The average age of patients across the
test. The summary of all unit costs used is presented groups is 30 years old, and the majority are married,
in the Appendix: Table 8). employed, and have experience with injecting drug use.
The patient costs were estimated by conducting a survey The highest education level attained by patients was the
among 41 patients undergoing ART at the hospital. We secondary level (high school).
Siregar et al. BMC Health Services Research (2015) 15:440 Page 4 of 12
Resource utilization and costs of providing ART count < 50 cells/mm3, 50 – 100 cells/mm3, and those
Table 2 presents the resources used to provide ART. with >200 cells/mm3. The average 2 year treatment costs
Hospitalization occurred only before ART and up to difference between the patients with CD4 cell count < 50
6 months after treatment was initiated; the duration cells/mm3 and the groups with higher CD4 cell count is
ranged from 3 to 20 days. The switch to second line also statistically significant, except with patients with CD4
ART occurred in 5 % of patients with a CD4 cell count cell count between 100 – 150 cells/mm3. The statistical
of 0 − 50 cells/mm3 and 15 % of those with 50 − 100 significance test is summarized in Appendix: Table 9.
cells/mm3. Few patients with CD4 cell counts >200
cells/mm3 were hospitalized and received OI treatment.
Details regarding unit costs per item are summarized in Patient costs per visit
the Appendix: Table 8). Table 4 presents patient costs. The average patient costs
Table 3 details the costs associated with providing ART. per visit are US$10 and US$11, for patients with CD4 cell
Before ART initiation, costs were mainly dominated by la- counts below and above 200 cells/mm3, respectively.
boratory tests (including the CD4, viral load, and routine Transportation cost and the clinic fee dominated the
laboratory tests). After the initiation of ART, costs were costs, while productivity loss accounted for less than 25 %
dominated by ARV, regardless of patients’ CD4 levels. of the total cost per visit. Per visit, almost all patients
Both total costs and per patient average costs decreased spent US$2 for the registration fee and US$5 − US$6 for
over time after ART initiation. The one anomaly was transportation. The mean time to reach the clinic was
the OI drugs/treatment cost for patients with a CD4 approximately 1 h (most patients lived < 20 km away) and
level of 50 − 100 cells/mm3, which increased from the average time spent in the clinic was approximately 100
US$725 in 1–6 months to US$2099 in 7–12 months. A min. There were no major differences in patient costs
relatively high CD4 cell count at treatment initiation between patients with CD4 cell counts less than or greater
relates to relatively low costs of ARVs, laboratory tests, than 200 cells/mm3 (we did not perform significance test
and OI drugs/treatment. Figure 1 shows the average for this due to our small sample size). Most patients in
costs per patient for different CD4 cell count levels and our survey travelled using either public transport (54 %)
over time. The highest average costs for 24 months of or motorcycle (42 %).
ART per patient were for patients with a CD4 cell
count <50 cells/mm3. The distribution of cost is provided Discussion
in the Appendix: Figure 2). The average costs difference To our knowledge, this study is the first in Indonesia
between patients undergoing the first 6 months of treat- and among the few in Asia [22, 23] to estimate the cost
ment and the 24 months of treatment is statistically of providing ART. The overall cost profile shows that
significant within the group of patients with CD4 cell the total costs and average costs per patient are
Siregar et al. BMC Health Services Research (2015) 15:440 Page 5 of 12
Table 2 Resource utilization of patients on ART by CD4 cell count at the start of ART, per specified period
CD4 cell Item Period Average
count 1–24 months
Before ART 1–6 months 7–12 months 13–18 months 19–24 months
0–50 Number of patients 96 96 95 84 61
a
% hospitalized 14 % 22 % - - -
Average days of hospitalizationb 6 (4–8) 15 (10–20) - - - 0.2c
Number of outpatient visit 95 47 75 63 41 3c
% of patients treated for OIa 2% 63 % 27 % 18 % 8% 29 %
nd a
% switched to 2 line ARV - - 1% 2% 3% 2%
Number of CD4 tests 95 47 75 63 41 3c
Number of viral load tests - 6 16 9 4 0.4c
Number of routine lab tests 87 45 74 62 40 3c
50–100 Number of patients 33 33 33 32 25
% hospitalizeda 3% 12 % - - -
Average days of hospitalizationb 3 6 (5–7) - - - 0.2c
Number of outpatient visit 36 13 29 20 10 2c
a
% of patients treated for OI 3% 36 % 21 % 13 % 4% 19 %
% switched to 2nd line ARVa - 3% 3% 6% 12 % 7%
Number of CD4 tests 36 13 29 20 10 2c
Number of viral load tests - 5 6 4 - 0.5c
Number of routine lab tests 30 14 28 20 9 2c
100–150 Number of patients 22 22 22 18 13
a
% hospitalized - 14 % - - -
Average days of hospitalizationb - 7 - - - 0.3c
Number of outpatient visit 29 8 18 13 7 2c
% of patients treated for OIa - 45 % 18 % 11 % 15 % 22 %
nd a
% switched to 2 line ARV - - - - -
Number of CD4 tests 29 8 18 13 7 2c
Number of viral load tests 1 1 4 3 - 0.4c
Number of routine lab tests 21 8 17 13 7 2c
150–200 Number of patients 16 16 13 13 11
% hospitalizeda 12 % 6% - - -
Average days of hospitalizationb 5 3 - - - 0.2c
Number of outpatient visit 26 9 9 11 7 3c
a
% of patients treated for OI - 47 % 14 % 21 % - 27 %
% switched to 2nd line ARVa - - - - -
Number of CD4 tests 26 9 9 11 7 3c
Number of viral load tests 1 4 4 1 - 0.6c
Number of routine lab tests 16 8 9 11 7 3c
>200 Number of patients 10 10 10 7 4
a
% hospitalized - - - - -
Average days of hospitalizationb - - - - -
Number of outpatient visit 17 5 6 4 4 3c
% of patients treated for OIa - 10 % - - - 10 %
nd a
% switched to 2 line ARV - - - - -
Siregar et al. BMC Health Services Research (2015) 15:440 Page 6 of 12
Table 2 Resource utilization of patients on ART by CD4 cell count at the start of ART, per specified period (Continued)
Number of CD4 tests 17 5 6 4 4 3c
Number of viral load tests - - - - -
Number of routine lab tests 9 5 6 4 4 3c
a b c
for the whole sample within the indicated period, 95 CI%, per patient
reduced when patients’ CD4 levels are higher at the countries [33], which showed that ART can be
time of clinic enrollment and ART initiation. In most delivered safely without routine laboratory
instances, hospitalization, OI treatment, and ART aver- monitoring for toxicity. A study within the same
age costs per patient decrease with longer use of ART. hospital clinic with our study demonstrated that the
During the early phase of the treatment, the highest total lymphocyte count (TLC) is a good alternative
costs are the costs of hospitalization, OI treatment, and for CD4 cell count as it is much cheaper and easier
ART initiation; these costs decrease over time as a re- to implement in rural settings. Combining TLC test
sult of patients’ improved health. This trend is compar- results with an algorithm of simple patient
able with the results of a study conducted in Southern characteristics could save US$14 per patient
Africa [24]. compared with the current scenario [11]. Also, De
These findings lead to several observations in response Jong et al.[12] in Indonesia (study conducted at the
to the research questions. Regarding the cost of treat- same clinic as our study) and Kumarasamy et al.
ment and potential cost reduction, the study confirms [34] in India found that TLC may reduce the need
the hypothesis that the following measures have poten- for routine CD4 measurements during ART
tial to reduce cost in ART delivery: (excepting the first year of treatment). In more
recent studies, however, it is found that TLC may
1) Early ART initiation. Most cost items are lower not be a reliable predictor for CD4 cell count in
when patients’ CD4 cell counts are higher at ART HIV-infected individuals in certain settings [35, 36],
initiation and especially the total cost share of while it is reliable in others [37–41]. Therefore,
hospitalization and OI treatment was reduced. CD4 caution is needed in applying TLC test as a
cell counts can predict the likelihood of OIs; replacement for CD4 test as it seems like the success
patients with CD4 cell counts >200 cells/mm3appear of its application is varied, depending much on the
to be at lower risk for the majority of OIs, compared specific settings and population in which it is applied.
with patients with <200 cells/mm3 [25–28] and this In Indonesia, another method to reduce laboratory
explains our study results. Early treatment is also costs is proposed by Indrati et al. [13], who found that
cost effective in resource-limited settings as well as a dual-test or single rapid-test algorithm (instead of a
on the global epidemic setting [9, 29, 30]. Additionally, serial three-test algorithm) may be just as accurate
the NIAID START Trial shows that early ART and more cost-effective, although the single rapid-test
protects the health of PLHIV [31]. WHO is now should be interpreted carefully. Although these alter-
preparing for a new ARV provision guideline in which native laboratory testing methods may lead to costs
it may recommend an even earlier ART initiation reductions, more research is needed to determine the
compared to its 2013 guideline [32]. Thus, our potential cost savings.
recommendation is strengthened by the WHO
discussion and it warrants further attention. The costs Regarding our second research question about scaling
implication of prolonged treatment, however, should up ART, the study indicates that although increased
be further studied to determine whether the costs re- ART coverage may cause a large increase in health ex-
duction and long term costs saving resulting from early penditure in the short run [23, 42], it could potentially
treatment offset the costs of prolonged treatment. save costs in the long run. By reaching more people in
2) Alternative diagnostics. Before ART initiation, costs need of ART, assuming that these are detected at earlier
are dominated by laboratory tests (CD4, viral load, stages, costs related to opportunistic infections and
and routine laboratory tests) followed by outpatient hospitalization may be avoided as has been shown in our
visits and hospitalization. After ART initiation, costs study. Importantly, providing ART can also act as HIV/
were dominated by ARV drugs followed by AIDS prevention [43, 44] because ART treatment re-
laboratory tests. Although costs reductions in duces transmission rates. Universal voluntary HIV test-
hospitalization and ARV use might be difficult to ing and early ART could therefore have a major effect
realize, reductions are possible for laboratory test on the HIV/AIDS epidemic and could be cost saving
costs, as is found in the DART trial in four African [29, 45]. The costs of treatment of new HIV infections
Siregar et al. BMC Health Services Research (2015) 15:440 Page 7 of 12
Table 3 Health care costs of patients on ART by CD4 cell count at the start of ART, per specified period (US$a)
CD4 cell Item Period Average
count 1–24 months
Before ART 1–6 months 7–12 months 13–18 months 19–24 months
per patient
0–50 Number of patients 96 96 95 84 61
Hospitalization 1483 (10 %) 6161 (9 %) - - - 64
Outpatient visits 1110 (8 %) 6707 (10 %) 6462 (12 %) 5247 (11 %) 3517 (11 %) 258
OI treatment 86 (1 %) 10867 (16 %) 2163 (4 %) 564 (1 %) 1538 (5 %) 168
ARV drugs - 40012 (60 %) 41281 (78 %) 37724 (81 %) 26915 (81 %) 1742
CD4 test 1254 (9 %) 620 (1 %) 990 (2 %) 832 (2 %) 541 (2 %) 36
Viral load test - 396 (1 %) 1056 (2 %) 594 (1 %) 264 (1 %) 27
Routine lab test 10452 (73 %) 1488 (2 %) 1289 (2 %) 1548 (3 %) 363 (1 %) 53
c c c
Total costs 14377 (100 %) 66205 (100 %) 53196 (100 %) 46472 (100 % 33115 (100 %) 2346
Average costs per patientb 150 (139 – 160) 690 (593–787) 560 (521–599) 553 (515–592) 543 (480–606)
50–100 Number of patients 33 33 33 32 25
Hospitalization 57 (1 %) 418 (2 %) - - - 13
Outpatient visits 421 (9 %) 2337 (12 %) 2232 (11 %) 1951 (11 %) 1110 (11 %) 244
OI treatment 12 (0.3 %) 725 (4 %) 2099 (10 %) 13 (0.1 %) 9 (0.1 %) 86
ARV drugs - 14694 (77 %) 15360 (73 %) 14763 (83 %) 9058 (87 %) 1734
CD4 test 475 (10 %) 172 (1 %) 383 (2 %) 264 (1 %) 132 (1 %) 30
Viral load test - 330 (2 %) 396 (2 %) 264 (1 %) - 30
Routine lab test 3608 (79 %) 478 (2 %) 494 (2 %) 508 (3 %) 104 (1 %) 49
Total costs 4570 (100 %) 19138 (100 %) 20947 (100 %) 17749 (100 %) 10405 (100 %)c 2186
b
Average costs per patient 138 (126–151) 580 (567–593) 635 (622–648) 555 (542–568) 473 (460–486)
100–150 Number of patients 22 22 22 18 13
Hospitalization - 418 (4 %) - - - 19
Outpatient visits 339 (10 %) 1566 (14 %) 1391 (13 %) 1157 (13 %) 771 (14 %) 258
OI treatment - 815 (7 %) 77 (0.7 %) 29 (0.3 %) 2 (0.04 %) 42
ARV drugs - 8147 (72 %) 8587 (79 %) 6858 (79 %) 4661 (83 %) 1500
CD4 test 383 (11 %) 106 (1 %) 238 (2 %) 172 (2 %) 92 (2 %) 32
Viral load test 66 (2 %) 66 (1 %) 264 (2 %) 198 (2 %) - 26
Routine lab test 2547 (76 %) 222 (2 %) 293 (3 %) 317 (4 %) 63 (1 %) 46
Total costs 3333 (100 %) 11329 (100 %) 10839 (100 %) 8722 (100 %) 5585 (100 %) 1922
Average costs per patientb 159 (144–173) 515 (500–529) 493 (478–507) 485 (470–499) 430 (415–444)
150–200 Number of patients 17 17 14 14 11
Hospitalization 190 (7 %) 57 (1 %) - - - 3
Outpatient visits 304 (11 %) 1133 (14 %) 982 (15 %) 935 (14 %) 654 (14 %) 263
OI treatment - 66 (1 %) 3 (0.05 %) 5 (0.1 %) - 4
ARV drugs - 6188 (77 %) 4865 (76 %) 5162 (77 %) 3717 (82 %) 1418
CD4 test 343 (12 %) 119 (1 %) 119 (2 %) 145 (2 %) 92 (2 %) 34
Viral load test 66 (2 %) 264 (3 %) 264 (4 %) 66 (1 %) - 39
Routine lab test 1969 (69 %) 200 (2 %) 195 (3 %) 362 (5 %) 63 (1 %) 57
Total costs 2870 (100 %) 8020 (100 %) 6421 (100 %) 6670 (100 %) 4522 (100 %) 1818
Average costs per patientb 169 (136–201) 472 (445–499) 459 (432–485) 476 (450–503) 411 (384–438)
>200 Number of patients 10 10 10 7 4
Hospitalization - - - - -
Siregar et al. BMC Health Services Research (2015) 15:440 Page 8 of 12
Table 3 Health care costs of patients on ART by CD4 cell count at the start of ART, per specified period (US$a) (Continued)
Outpatient visits 199 (13 %) 701 (15 %) 608 (13 %) 363 (12 %) 257 (20 %) 247
OI treatment - 1 (0.02 %) - - - 0.1
ARV drugs - 3637 (79 %) 3797 (83 %) 2715 (84 %) 938 (73 %) 1366
CD4 test 224 (15 %) 66 (1 %) 79 (2 %) 53 (2 %) 53 (4 %) 35
Viral load test - - - - -
Routine lab test 1119 (73 %) 223 (5 %) 92 (2 %) 68 (2 %) 35 (3 %) 50
c
Total costs 1541 (100 %) 4623 (100 %) 4572 (100 %) 3218 (100 %) 1281 (100 %) 1699
Average costs per patientb 154 (114–195) 462 (437–487) 457 (432–482) 460 (435–485) 320 (295–345)
a
except for number of patients, b95 % CI, cdifference is significant between first 6 months and the period measured, 95 % CI
will be averted and may potentially free resources to pre- community clinics, reducing the hospital burden. Patients
vent even more infections [14]. Considering these find- that initiate ART at >200 cell/mm3 could also obtain ART
ings, we suggest further study regarding advantages (e.g., at the community health care center, as our analysis sug-
health benefits of early treatment) and disadvantages gests that hospitalization and OIs are rare in this popula-
(e.g., budget impact) of scaling up ART in Indonesia tion. In addition, patients mostly utilize first line ARV, and
from both the short run and long run perspectives. no patients with CD4 cell counts >100 cells/mm3 switch to
In terms of the location for scaling up ART, we cannot second line ARV (Table 2), indicating a low rate of treat-
draw strong conclusions on the basis of the costing analysis ment failure within this group [51]. As such, the ARV dis-
in the hospital only. Yet, there seem to be advantages in tribution (in terms of medicine type) in community/
scaling up ART at the community level as this may poten- primary health care centres for patients with CD4 cell
tially increase early detection and reduce the burden in hos- counts >100 cells/mm3 might not be too complex, as most
pital clinics [46]. Also, the shorter waiting and travel time patients are likely to require only first line ARV. Currently,
to the clinic may lead to lower patient’s costs and better up- there are only two primary health care centers in Bandung
take and adherence of ART [46–49]. In this scenario, the that provide ART, which presents considerable potential to
hospital and community health centres will have different increase the service to other community clinics.
roles (Table 5). The hospital will be a referral centre for However, providing ART at all community health clinics
complicated AIDS cases and treatment of OIs just as in Indonesia at this stage seems inefficient due to the low
current practice [46, 50]. Because patients become relatively HIV prevalence in the general population which will result
stable over time (indicated by decrease in hospitalization in a low patient load per clinic for which all community
and OI treatment over time) they could continue ART at staff will require training [52]. Therefore, providing HIV
Fig. 1 Average service costs per patient per specified period, health care system perspective (US$). This figure presents the average service costs
per patient taking ART. The average costs are separated into specific periods, namely before ART, 1–6 months, 7–12 months, 13–18 months, and
19–24 months within ART. These costs are further separated into CD4 cell count group, namely 0–50, 50–100, 100–150, 150 - 200, and >200 cells/
mm3. The figure shows how the average costs per patient in different CD4 cell count groups relatively decrease after the start of ART
Siregar et al. BMC Health Services Research (2015) 15:440 Page 9 of 12
services through clinics in certain high prevalence settings results. Cost structures and levels as well as patient popula-
such as prisons or cities may be preferable [53, 54], tions are likely to vary between clinics, and specific costing
although this strategy requires further study. studies for other settings (e.g., other hospitals, community/
Regarding our third research question about patients’ primary health centers, and prisons) should be considered.
financial burden, the study shows that patient costs per Caution should be exercised when interpreting our result
visit are US$10 and US$11, for patients with CD4 cell in other resource limited settings. Second, we may have
counts below and above 200 cells/mm3. This relates to overestimated the total patient costs of seeking and under-
approximately 14 and 7 % of their monthly expenditure going care as this was based on assumptions regard-
(a proxy of monthly income), respectively. Especially for ing patients’ labour productivity losses, and not on
patients with CD4 cell counts <200 cells/mm3 these empirical data collection on these losses per se. Third,
costs could be a barrier, as it exceeds 10 % of their although we have conducted a time motion study to
monthly expenditure and can be considered to be cata- control for inefficiency in personnel performance and
strophic for a household economy [55]. equipment use related to ART delivery in the clinic,
Transportation comprises the highest proportion of discrepancies may still exist, and we may have over-
costs: 62 and 43 % for patients with CD4 cell counts below or undervalued the total costs. Fourth, we did not
and above 200 cells/mm3, respectively, and this is compar- perform any comparison between WHO CHOICE es-
able with the finding of Riyarto et al. [16] in Indonesia. A timates (that we used for calculating inpatient cost)
study by Haroen et al. [49] in Bandung, Indonesia, and and any local data. Although this is an important as-
international studies by Portelli et al. [47], Brinkhof et al. pect, currently there is very limited local data avail-
[48], and Posse et al. [15] have shown that transportation able to do this comparison. Fifth, it is important to
costs are a common reason why patients cease ARV. This note that the unit costs and prices data that we used
information provides another reason to scale up ART at are from year 2010–2012, depending on availability.
community level, as it likely reduces transportation costs As such, we believe these unit costs have changed
for patients and may increase the uptake of ART, espe- overtime and our results should be interpreted with
cially of patients with CD4 cell counts <200 cells/mm3. this note in mind (e.g. ARV drugs prices may have
decreased since government are producing more ARV
Study limitations locally [2], patients monthly expenditure has been ris-
Our results should be interpreted with some caution. First, ing due to inflation).
this study has evaluated a contextualized ART service deliv-
ery model, which may hamper the generalizability of its
Conclusion
Table 5 Recommendation on role of clinics in delivering ARTa
Three main conclusions can be derived from our study.
Costs Items Type of Clinic First, we show that the costs of providing ART are highest
Hospital Community/Primary Health Care Centre during the early phase of treatment, and will decrease and
Hospitalization + - stabilize as treatment progresses. Second, our findings sug-
Outpatient visits +/- + gest that costs reduction can be potentially realized by early
OI treatment +b + treatment initiation (which may reduce hospitalization, OI
ARV drugs +/- +
drug/treatment costs, and patient mortality) and by apply-
ing alternative laboratory tests with caution. Third, scaling
CD4 test + -
up ART at the community level in certain high prevalence
Viral load test +c - settings has potential to save costs and improve uptake and
Routine lab test + +/- adherence. However, provision of ART at all community
ART antiretroviral treatment clinics seems inefficient due to the low prevalence in the
a
‘ + ’ and ‘-’ denote respectively a role of high and low importance for the
clinic in the specified activities in HIV/AIDS control
general population and options to select certain clinics in
b
for severe cases, c if necessary high prevalence areas need further investigation.
Siregar et al. BMC Health Services Research (2015) 15:440 Page 10 of 12
Table 7 Average unit costs of OI treatment by CD4 cell count at the beginning of ART (US$)
CD4 cell Period
count
Before ARTa 1–6 monthsa 7–12 monthsa 13–18 monthsa 19–24 monthsa
0–50 1 (0–2) 113 (48–179) 23 (3–42) 7 (0–16) 25 (0–60)
50–100 0.4 (0–1) 22 (1–43) 64 (0–170) 0.4 (0–1) 0.3 (0–1)
100–150 - 37 (0–94) 3 (0.1–7) 2 (0–3) 0.2 (0–0.4)
150–200 - 4 (0–9) 0.2 (0–1) 0.4 (0–1) -
>200 - 0.1 (0–0.3) - - -
a
95 % CI
Fig. 2 Appendix (Figure A). HIV treatment total costs per patient on ART, over 24 months. This figure shows the scatter diagram of total costs of
patients undertaking HIV treatment over 24 months. The patients within the group of 0–50 cells/mm3 CD4 cell count have the largest variance in
costs (e.g. patients with highest or lowest average treatment costs are within this group)
14. Stover J, Bertozzi S, Gutierrez J-P, Walker N, Stanecki K a, Greener R, et al. 37. Chen J, Li W, Huang X, Guo C, Zou R, Yang Q, et al. Evaluating Total
The global impact of scaling up HIV/AIDS prevention programs in low- and Lymphocyte Count as a Surrogate Marker for CD4 Cell Count in the
middle-income countries. Science. 2006;311:1474–6. Management of HIV-Infected Patients in Resource-Limited Settings: A
15. Posse M, Meheus F, van Asten H, van der Ven A, Baltussen R. Barriers to Study from China. PLoS One. 2013;8:e69704.
access to antiretroviral treatment in developing countries: a review. Trop 38. Jain S, Singh A, Singh R, Bajaj J, Damle A. Evaluation of total lymphocyte
Med Int Heal. 2008;13:904–13. count (TLC) as a surrogate marker for CD4 count in HIV-positive patients for
16. Riyarto S, Hidayat B, Johns B, Probandari A, Mahendradhata Y, Utarini A, et resource-limited settings. Muller J Med Sci Res. 2015;6:23.
al. The financial burden of HIV care, including antiretroviral therapy, on 39. Obirikorang C, Quaye L, Acheampong I. Total lymphocyte count as a surrogate
patients in three sites in Indonesia. Health Policy Plan. 2010;25:272–82. marker for CD4 count in resource-limited settings. BMC Infect Dis. 2012;12:128.
17. Drummond MF, Sculpher MJ, Torrance GW. Methods for the Economic 40. Daud MY, Qazi RA, Bashir N. Total lymphocyte count as a susbtitute to CD4
Evaluation of Health Care Programmes. New York: Oxford University count in management of HIV infected individuals in resource lmited society.
Press; 2005. J Ayub Med Coll. 2015;27:29–31.
18. Ministry of Finance (MoF) of Indonesia. Circular Letter No. SE-03/PB/2007 on 41. Karanth SS, Rau NR, Gupta A, Kamath A, Shanbhogue V, Pruthvi BC. Utility of
the Main Salary Adjustment for Government Official, Judge of General, State total lymphocyte count as a surrogate for absolute CD4 count in the adult
Administrative, and Religious Courts, National Army, and National Police. 2007. Indian HIV population: A prospective study. Avicenna J Med. 2014;4:1–4.
19. Adam T, Evans DB, Murray CJL. Econometric estimation of country-specific 42. Afriandi I, Aditama TY, Mustikawati D, Oktavia M, Alisjahbana B, Riono P. HIV
hospital costs. Cost Eff Resour Alloc. 2003;10:1–10. and injecting drug use in Indonesia: epidemiology and national response.
20. Bender MA, Kumarasamy N, Mayer KH, Wang B, Walensky RP, Flanigan T, et Acta Med Indones. 2009;41 Suppl 1:75–8.
al. Cost-effectiveness of tenofovir as first-line antiretroviral therapy in India. 43. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy
Clin Infect Dis Off Publ Infect Dis Soc Am. 2010;50:416–25. N, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl
21. The World Bank: Official Exchange Rate (LCU per US$, period average) 2010. J Med. 2011;365:493–505.
accessed 2011, http://data.worldbank.org/indicator/PA.NUS.FCRF . 44. Mathers BM, Degenhardt L, Ali H, Wiessing L, Hickman M, Mattick RP, et al.
22. Paton NI, Chapman CAT, Sangeetha S, Mandalia S, Bellamy R, Beck EJ. Cost HIV prevention, treatment, and care services for people who inject drugs: a
and cost-effectiveness of antiretroviral therapy for HIV infection in systematic review of global, regional, and national coverage. Lancet.
Singapore. Int J STD AIDS. 2006;17:699–705. 2010;375:1014–28.
23. Kitajima T, Kobayashi Y, Chaipah W, Sato H, Chadbunchachai W, Thuennadee 45. De Cock KM, Gilks CF, Lo Y-R, Guerma T. Can antiretroviral therapy eliminate
R. Costs of medical services for patients with HIV/AIDS in Khon Kaen, Thailand. HIV transmission? Lancet. 2009;373:7–9.
AIDS. 2003;17:2375–81. 46. Siregar AYM, Komarudin D, Wisaksana R, van Crevel R, Baltussen R. Costs
24. Leisegang R, Cleary S, Hislop M, Davidse A, Regensberg L, Little F, et al. Early and outcomes of VCT delivery models in the context of scaling up services
and late direct costs in a Southern African antiretroviral treatment in Indonesia. Trop Med Int Heal. 2011;16:193–9.
programme: a retrospective cohort analysis. PLoS Med. 2009;6:e1000189. 47. Portelli MS, Tenni B, Kounnavong S, Chanthivilay P: Barriers to and
25. Masur H, Ognibene FP, Yarchoan R, Shelhamer JH, Baird BF, Travis W, et al. Facilitators of Adherence to Antiretroviral Therapy Among People Living
CD4 counts as predictors of opportunistic pneumonias in human With HIV in Lao PDR: A Qualitative Study. Asia Pac J Public Health 2012, 53.
immunodeficiency virus (HIV) infection. Ann Intern Med. 1989;111:223–31. 48. Brinkhof MWG, Pujades-Rodriguez M, Egger M. Mortality of patients lost to
26. Phair J, Muñoz A, Detels R, Kaslow R, Rinaldo C, Saah A. The risk of Pneumocystis follow-up in antiretroviral treatment programmes in resource-limited
carinii pneumonia among men infected with human immunodeficiency virus settings: systematic review and meta-analysis. PLoS One. 2009;4:e5790.
type 1. Multicenter AIDS Cohort Study Group. N Engl J Med. 1990;322:161–5. 49. Haroen H, Handayani M, Puspitasari TS, Pinxten L, Van Den Bome B, Halfens
R. Lost to follow up from ARV treatment: reasons and true outcome of patients
27. Miller V, Mocroft A, Reiss P, Katlama C, Papadopoulos AI, Katzenstein T, et al.
in Bandung, Indonesia. Indones: One Results IMPACT Progr Bandung; 2011.
Relations among CD4 lymphocyte count nadir, antiretroviral therapy, and
50. Wisaksana R, Alisjahbana B, van Crevel R, Kesumah N, Sudjana P, Sumantri R.
HIV-1 disease progression: results from the EuroSIDA study. Ann Intern Med.
Challenges in delivering HIV-care in Indonesia: experience from a referral
1999;130:570–7.
hospital. Acta Med Indones. 2009;41 Suppl 1:45–51.
28. Kovacs JA, Masur H. Prophylaxis against opportunistic infections in patients
51. World Health Organization HIV Department: Prioritizing Second-Line
with human immunodeficiency virus infection. N Engl J Med.
Antiretroviral Drugs for Adults and Adolescents: A Public Health Approach,
2000;342:1416–29.
Report of a WHO Working Group Meeting. Geneva; 2007(May).
29. Granich RM, Gilks CF, Dye C, De Cock KM, Williams BG. Universal voluntary HIV
52. Curran J, Debas H, Arya M, Knobler S, Pray L. Scaling Up Treatment for the
testing with immediate antiretroviral therapy as a strategy for elimination of
Global AIDS Pandemic: Challenges and Opportunities. Washington: The
HIV transmission: a mathematical model. Lancet. 2009;373:48–57.
National Academies Press; 2005.
30. Johansson KA, Robberstad B, Norheim OF. Further benefits by early start of
53. Nelwan EJ, Van Crevel R, Alisjahbana B, Indrati AK, Dwiyana RF. Human
HIV treatment in low income countries: survival estimates of early versus
immunodeficiency virus, hepatitis B and hepatitis C in an Indonesian
deferred antiretroviral therapy. AIDS Res Ther. 2010;7:3.
prison: prevalence, risk factors and implications of HIV screening. Trop
31. NIAID START Trial Confirms that Immediate Treatment of HIV with
Med Int Heal. 2010;15:1491–8.
Antiretroviral Drugs (ARVs) Protects the Health of People Living with HIV
54. World Health Organization. Scaling up HIV/AIDS care: service delivery &
[http://www.who.int/hiv/mediacentre/news/niaid_start/en/#]. Accessed date
human resources perspectives. 2004.
7 September 2015
55. Russell S. The economic burden of illness for households in developing
32. World Health Organization. Consolidated Guidelines on the Use of Antiretroviral countries: a review of studies focusing on malaria, tuberculosis, and human
Drugs for Treating and Preventing HIV Infection: Recommendations for a Public immunodeficiency virus/acquired immunodeficiency syndrome. Am J Trop
Health Approach, June 2013. Geneva: World Health Organization; 2013. Med Hyg. 2004;71(2 Suppl):147–55.
33. Mugyenyi P, Walker AS, Hakim J, Munderi P, Gibb DM, Kityo C, et al. Routine
versus clinically driven laboratory monitoring of HIV antiretroviral therapy in
Africa (DART): a randomised non-inferiority trial. Lancet. 2010;375:123–31.
34. Kumarasamy N, Mahajan AP, Flanigan TP, Hemalatha R, Mayer KH, Carpenter
CCJ, et al. Total lymphocyte count (TLC) is a useful tool for the timing of
opportunistic infection prophylaxis in India and other resource-constrained
countries. J Acquir Immune Defic Syndr. 2002;31:378–83.
35. Wondimeneh Y a, Ferede GE, Yismaw GW, Muluye DF. Total Lymphocyte
Count as surrogate marker for CD4 Cell Count in HIV-Infected Individuals in
Gondar University Hospital, North West Ethiopia. AIDS Res Ther. 2012;9:21.
36. Abdollahi A, Saffar H, Shoar S, Jafari S. Is total lymphocyte count a predictor
for CD4 cell count in initiation antiretroviral therapy in HIV-infected patients?
Niger Med J. 2014;55:289.