Journal of the Formosan Medical Association 121 (2022) 1714e1720

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.jfma-online.com

Original Article

Potential or contraindicated drugedrug

interactions with antiretroviral therapy in
real-world settings in Taiwan
Chi-Chuan Wang a,b,c,1, Hsing-Jung Li a,1, Chi-Hao Shao b,
Wang-Huei Sheng d,*

a
School of Pharmacy, National Taiwan University, Taipei, Taiwan
b
Graduate Institute of Clinical Pharmacy, College of Medicine, National Taiwan University, Taipei,
Taiwan
c
Department of Pharmacy, National Taiwan University Hospital, Taipei, Taiwan
d
Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan

Received 12 June 2020; received in revised form 8 January 2021; accepted 8 December 2021

KEYWORDS Background/Purpose: Given the complex metabolic pathway of antiretroviral therapy (ART),
Drugedrug polypharmacy may increase the risk of drugedrug interactions (DDIs). Therefore, we investi-
interaction; gated the frequency of DDIs during ART exposure to improve medical care for patients with hu-
Antiretroviral man immunodeficiency virus (HIV).
therapy; Methods: This was a nationwide cross-sectional study using claims data from the National
HIV; Health Insurance in Taiwan in 2016. Potential or contraindicated DDIs with recommended
Taiwan first-line ART (1L-ART) or protease inhibitors (PIs) were identified from the University of Liver-
pool drug interaction database. Fisher’s exact or chi-square test was used to determine the
significance of categorical variables.
Results: A total of 25,863 HIV-infected individuals were identified. Regarding 1L-ART users, pa-
tients with contraindicated DDIs accounted for 1e4%, whereas those with potential DDIs ac-
counted for 15e50%. The most frequently coprescribed medications related to potential
DDIs were diclofenac and polyvalent cation-containing antacids. Among PI users, 8e10% of
them had contraindicated DDIs while 44e50% of them had potential DDIs. The medications
related to potential DDIs with PIs were zolpidem, betamethasone, polyvalent cation-
containing antacids, and loperamide.
Conclusion: Our study showed a low prevalence of contraindicated DDIs in the HIV population;
however, more attention should be paid to a high proportion of potential DDIs. Strategies to

* Corresponding author. Department of Internal Medicine, National Taiwan University Hospital, No. 7, Zhongshan S. Rd., Zhongzheng Dist.,
Taipei City 100, Taiwan.
E-mail address: whsheng@ntu.edu.tw (W.-H. Sheng).
1
Chi-Chuan Wang and Hsing-Jung Li contributed equally as the first author.

https://doi.org/10.1016/j.jfma.2021.12.005
0929-6646/Copyright ª 2021, Formosan Medical Association. Published by Elsevier Taiwan LLC. This is an open access article under the CC
BY license (http://creativecommons.org/licenses/by/4.0/).
 Journal of the Formosan Medical Association 121 (2022) 1714e1720
avoid these DDIs should be implemented if possible. Further research that focuses on the long-
term clinical impact of potential DDIs is warranted.
Copyright ª 2021, Formosan Medical Association. Published by Elsevier Taiwan LLC. This is an
open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Introduction
With the innovative drug development of antiretroviral
therapy (ART), the life expectancy of patients with human
immunodeficiency virus (HIV) has been significantly
improved,1 and a substantial increase in the number of
elderly individuals with HIV has been observed.2 Among
these HIV-infected individuals, the complexity of aging, HIV
infection, and ART exposure may alter the risk of comorbid
conditions, such as cardiovascular diseases, metabolic
syndromes or mental disorders.3,4 Medications were,
accordingly, prescribed to address undesirable comorbid-
ities. However, the addition of concomitant medications
may contribute to pill burden and drugedrug interactions
(DDIs) in the HIV-infected population.
Attention should be paid to DDIs among HIV patients
since the interactions may alter the effectiveness and
safety of ART or other comedications, leading to detri-
mental clinical outcomes. The increased drug exposure of
HIV/non-HIV medications may lead to undesired adverse
events.5 On the other hand, the subtherapeutic drug con-
centration related to a DDI may trigger the rebound of
plasma HIV RNA and ART drug resistance.6 One of the
mechanisms prompting DDIs is drug metabolism-driven,
including chromosome P-450 (CYP) isoenzyme systems,
drug transporters, and glucuronidation enzymes.5,7,8 For
instance, protease inhibitors (PIs) are known to inhibit
CYP3A49 and are not currently recommended as first-line
treatments. In contrast, single-tablet regimens (STRs) that
would cause fewer DDIs were the recommended first-line
ART (1L-ART).10
To the best of our knowledge, the DDI profiles during ART
exposure have not been comprehensively addressed in
Taiwan. A better understanding of the DDI profiles could
help clinicians select the most appropriate ART for pa-
tients. Additionally, the costs of health care for patients
with DDIs were significantly higher than those without
DDIs.11 To improve medical care for patients with HIV and
reduce the economic burden for the government, a
detailed review of DDIs is warranted. Therefore, two major
goals in this study are listed as follows: 1) to document the
frequency of contraindicated and potential DDIs with rec-
ommended 1L-ART and PIs in real-world settings and 2) to
investigate the profiles of comorbidities and comedications
among HIV-infected patients stratified by DDI status.
Patients and methods
Data source
Data used in this study were retrieved from the Health and
Welfare Database, which contained all claims data from the
1715
National Health Insurance (NHI) program in Taiwan. The NHI
program was launched in 1995 with a coverage rate over
99.9%.12 The program covered approximately 24 million
beneficiaries in 2019.12 Data from 2016 served as the data
source. Since the identification numbers of patients in the
database were encrypted and deidentified, the study was
granted a waiver of informed consent and approved through
expedited review (IRB/REC number: 201802027RIFA) by the
Research Ethics Committee of National Taiwan University
Hospital. The International Classification of Diseases, ninth
or tenth Revision, Clinical Modification (ICD-9-CM, ICD-10-
CM) codes were used to identify diagnoses. Medications
were recognized via the Anatomical Therapeutic Chemical
(ATC) classification system.
Study population
Patients with diagnoses of HIV between January 1, 2016,
and December 31, 2016, were included in this research. HIV
patients were defined as having at least one inpatient or
two separate outpatient diagnoses ([ICD-9-CM code]: 042,
V08; [ICD-10-CM code]: B20, Z21). To verify further the
diagnosis of HIV for outpatients, at least one examination
for CD4 count or viral load ([procedure code]: 12073B,
14074B) was required between the two subsequent HIV
outpatient diagnoses.
Study design
This was a cross-sectional study that included HIV-infected
patients in 2016. The year 2016 was chosen to ensure a
comparable starting point of novel STRs; tenofovir/emtri-
citabine/efavirenz (TDF/FTC/EFV), TDF/FTC/rilpivirine
(TDF/FTC/RPV), and abacavir/lamivudine/dolutegravir
(ABC/3TC/DTG) were licensed in Taiwan in June 2010,
October 2015, and July 2015, respectively. The other STR,
elvitegravir/cobicistat/FTC/TDF, was not included in this
analysis since it was not licensed until January 2017. The
DDIs with the recommended 1L-ART (TDF/FTC/EFV, TDF/
FTC/RPV, and ABC/3TC/DTG) and PIs (atazanavir (ATV),
ATV þ ritonavir (r), darunavir (DRV) þ r, and lopinavir
(LPV)/r) were investigated separately. Atazanavir 200 mg
was used as a proxy for the “ATV alone” group, whereas
atazanavir 150 mg was used as a substitution for the
“ATV þ r” group. Since ritonavir is often used to boost other
PIs, we did not analyze it as a separate drug. Detailed
procedures are expressed as follows.
First, the antiretroviral therapy exposure period was
identified for each HIV-infected patient. The index date
was the first prescription of 1L-ART or PIs since January 1,
2016. The ART-exposed period was from the index date to
the following situations whichever came first: discontinua-
tion, death, or the end of the study (December 31, 2016).
 C.-C. Wang, H.-J. Li, C.-H. Shao et al.

Discontinuation was defined as a 14-day gap found between
the end of a day’s supply and the beginning of the next
prescription. To better outline the real-life circumstances,
patients could be included in more than one ART group as
long as they were exposed to the specific ART of interest.
Second, the comedications that could cause contra-
indicated or potential DDIs with ART of interest were recog-
nized using the University of Liverpool drug interaction
database.13 The Liverpool drug interaction database pro-
vided interactive charts for assessing the risk of DDIs between
HIV/HIV and HIV/non-HIV medications. The severity of the
interactions was categorized by icons. Only red and amber
icons, standing for contraindicated and potential DDIs,
respectively, were estimated in this research. The DDIs were
defined as having at least one-day use of systemic come-
dications within the ART-exposed period. Each person would
only be calculated once even if he/she had multiple pre-
scriptions for the same DDI pair. The frequency was esti-
mated for each drug pair for each ART of interest.
Finally, we stratified patients by their DDI status and
investigated their baseline characteristics. Patients with
potential or contraindicated DDIs were included in the
“with DDI” group, and the others were included in the
“without DDI” group. The variables of interest were age,
sex, comorbid conditions and concomitant medications.
Total numbers of comorbidities and comedications were
also summed as two independent variables. The comor-
bidities included cerebrocardiovascular diseases, metabolic
syndromes, respiratory disease, liver disease, renal failure,
psychiatric disease and malignancy. Comorbidities were
defined by at least one inpatient or two separate outpa-
tient records in 2016. The comedications were composed of
different classes of medications, including glucose-lowering
agents, cardiovascular agents, central nervous system
agents, gastrointestinal agents and anti-infectives.
Concomitant medications were recognized via at least
one prescription in 2016.
Statistical analysis
The chi-square test was used to determine if the categor-
ical variables were significant. Instead, Fisher’s exact test
was used if more than 20% of the cells in the table had an
expected value of less than five. The significant differences
between patients with or without DDIs were considered
significant with a 2-sided p-value of <0.05. A cell size less
than five is presented as “<5” per the requirement of the
Health Data Research Center under the Department of
Statistics, Ministry of Health and Welfare of Taiwan. All
statistical procedures were performed with SAS version 9.4
(SAS Institute Inc., Cary, NC, USA).
Results
A total of 25,863 HIV-infected individuals were identified;
92% of them were male, and the mean age was 39 years old.
Among patients prescribed 1L-ART, there were 5877 TDF/
FTC/EFV users, 3519 TDF/FTC/RPV users, and 2517 ABC/
3TC/DTG users. Regarding patients consuming PIs, we
identified 2314 ATV users, 656 ATV þ r users, 1291 DRV þ r
users, and 3440 LPV/r users.
Table 1 demonstrates the number of patients by
different numbers of contraindicated or potential DDIs.
Regarding 1L-ART users, patients with contraindicated DDIs
were most frequently observed among those with TDF/
FTC/RPV (4%), followed by TDF/FTC/EFV (2%) and ABC/
3TC/DTG (1%). Patients with potential DDIs were propor-
tionally highest among those who received TDF/FTC/EFV
(50%), TDF/FTC/RPV (32%) and ABC/3TC/DTG (15%). On the
other hand, patients prescribed different kinds of PIs pro-
cessed similar frequencies of contraindicated DDIs: ATV þ r
(10%), followed by DRV þ r (9%), ATV alone (9%) and LPV/r
(8%). Among those with potential DDIs, a similar trend was
found: ATV alone (50%), ATV þ r (48%), DRV þ r (48%) and
LPV/r (44%). For patients with more than three potential
DDIs, the highest frequency was found among ATV þ r users
(19%) compared to ATV users (17%), DRV þ r users (13%) and
LPV/r users (12%). The detailed frequencies of contra-
indicated or potential DDI pairs are presented in
Supplemental Tables 1e4. Note that one patient can
contribute more than one DDI pair, so the sum of all DDI
pairs may exceed the number of 1L-ART or PI users.
Table 2 presents the top three most frequently observed
comedications leading to contraindicated and potential

Table 1 Number of DDIs with recommended first-line ARTs or PIs.

N (%) First-line ARTs PIs
TDF/FTC/EFV (N Z 5877) TDF/FTC/RPV ABC/3TC/DTG ATV alone ATV þ r DRV þ r LPV/r
(N Z 3519) (N Z 2517) (N Z 2314) (N Z 656) (N Z 1291) (N Z 3440)
No. of contraindicated DDIs
0 5755 (97.92) 3387 (96.25) 2499 (99.28) 2107 (91.05) 589 (89.79) 1174 (90.94) 3160 (91.86)
1 122 (2.08) 132 (3.75) 18 (0.72) 207 (8.95) 67 (10.21) 117 (9.06) 280 (8.14)
No. of potential DDIs
0 2931 (49.87) 2383 (67.72) 2136 (84.86) 1154 (49.87) 341 (51.98) 676 (52.36) 1936 (56.28)
1 1286 (21.88) 679 (19.30) 354 (14.06) 488 (21.09) 118 (17.99) 300 (23.24) 756 (21.98)
2 685 (11.66) 286 (8.13) 20 (0.79) 286 (12.36) 73 (11.13) 141 (10.92) 328 (9.53)
3 975 (16.59) 171 (4.86) 7 (0.28) 386 (16.68) 124 (18.90) 174 (13.48) 420 (12.21)
ABC, abacavir; ARTs, antiretroviral therapies; ATV, atazanavir; ATV þ r, atazanavir þ ritonavir; DDI, drugedrug interaction; DRV þ r,
darunavir þ ritonavir; DTG, dolutegravir; EFV, efavirenz; FTC, emtricitabine; LPV/r, lopinavir/ritonavir; PIs: protease inhibitors; RPV,
rilpivirine; TDF, tenofovir disoproxil fumarate; 3TC, lamivudine.

1716
 Journal of the Formosan Medical Association 121 (2022) 1714e1720

Table 2 Among DDI drug pairs, the top 3 most frequently coprescribed medications.
N (%) First-line ARTs PIs
TDF/FTC/EFV TDF/FTC/RPV ABC/3TC/DTG ATV alone ATV þ r DRV þ r LPV/r
(N Z 5877) (N Z 3519) (N Z 2517) (N Z 2314) (N Z 656) (N Z 1291) (N Z 3440)
Among contraindicated DDI drug pairs, the top 3 most frequently used comedications
1 Midazolam Dexamethasone Phenobarbital Domperidone Domperidone Domperidone Domperidone
94 (1.60) 50 (1.42) 10 (0.40) 81 (3.50) 26 (3.96) 42 (3.25) 140 (4.07)
2 Ergotamine Omeprazole Oxcarbazepine Quetiapine Quetiapine Quetiapine Quetiapine
13 (0.22) 16 (0.45) <5 47 (2.03) 13 (1.98) 28 (2.17) 61 (1.77)
3 Triazolam Pantoprazole Phenytoin Midazolam Midazolam Midazolam Midazolam
13 (0.22) 15 (0.43) <5 24 (1.04) 10 (1.52) 19 (1.47) 46 (1.34)
Among potential DDI drug pairs, top 3 most frequently used comedications
1 Diclofenac Diclofenac Antacids Zolpidem Antacids Betamethasone Loperamide
738 (12.56) 98 (11.31) 293 (11.64) 266 (11.50) 72 (10.98) 116 (8.99) 254 (7.38)
2 Zolpidem Antacids Metformin Antacids Zolpidem Zolpidem Zolpidem
422 (7.18) 276 (7.84) 45 (1.79) 232 (10.03) 56 (8.54) 90 (6.97) 245 (7.12)
3 Betamethasone Ibuprofen Valproate Betamethasone Betamethasone Loperamide Betamethasone
397 (6.76) 201 (5.71) 27 (1.07) 145 (6.27) 49 (7.47) 75 (5.81) 197 (5.73)
ABC, abacavir; ARTs, antiretroviral therapies; ATV, atazanavir; ATV þ r, atazanavir þ ritonavir; DDI, drugedrug interaction; DRV þ r,
darunavir þ ritonavir; DTG, dolutegravir; EFV, efavirenz; FTC, emtricitabine; LPV/r, lopinavir/ritonavir; PIs: protease inhibitors; RPV,
rilpivirine; TDF, tenofovir disoproxil fumarate; 3TC, lamivudine.

Table 3 Baseline characteristics stratified by DDI status among recommended 1L-ART users.
N (%) TDF þ FTC þ EFV TDF þ FTC þ RPV ABC þ 3TC þ DTG
a a
Stratified by Without DDI With DDI p value Without DDI With DDI p value Without DDI With DDI p valuea
DDI status
N Z 2925 N Z 2952 N Z 2348 N Z 1171 N Z 2125 N Z 392
Gender (male) 2796 (95.59) 2682 (90.85) <.001 2210 (94.12) 1037 (88.56) <.001 1983 (93.32) 351 (89.54) .008
Age (50) 295 (10.09) 500 (16.94) <.001 270 (11.50) 167 (14.26) .02 253 (11.91) 94 (23.98) <.001
No. of comorbidities
0 1761 (60.21) 1170 (39.63) <.001 1255 (53.45) 512 (43.72) <.001 1072 (50.45) 154 (39.29) <.001
1 926 (31.66) 1002 (33.94) 748 (31.86) 382 (32.62) 716 (33.69) 113 (28.83)
2 201 (6.87) 509 (17.24) 257 (10.95) 176 (15.03) 263 (12.38) 73 (18.62)
3 37 (1.26) 271 (9.18) 88 (3.75) 101 (8.63) 74 (3.48) 52 (13.27)
No. of comedications
0 1634 (55.86) 567 (19.21) <.001 972 (41.40) 258 (22.03) <.001 708 (33.32) 62 (15.82) <.001
1 811 (27.73) 716 (24.25) 621 (26.45) 282 (24.08) 587 (27.62) 65 (16.58)
2 297 (10.15) 550 (18.63) 298 (12.69) 207 (17.68) 339 (15.95) 69 (17.60)
3 97 (3.32) 332 (11.25) 189 (8.05) 140 (11.96) 189 (8.89) 41 (10.46)
4 48 (1.64) 282 (9.55) 118 (5.03) 111 (9.48) 128 (6.02) 48 (12.24)
5 38 (1.30) 505 (17.11) 150 (6.39) 173 (14.77) 174 (8.19) 107 (27.3)
ABC, abacavir; ARTs, antiretroviral therapies; DDI, drugedrug interaction; DTG, dolutegravir; EFV, efavirenz; FTC, emtricitabine; RPV,
rilpivirine; TDF, tenofovir disoproxil fumarate; 1L, first-line; 3TC, lamivudine.
Statistical significance was set as P < 0.05 which are mentioned in bold.
a
Fisher’s exact or chi-square test was used to determine if there was a significant association between different DDI statuses.

DDIs. Among 1L-ART users, the most commonly coadminis-
tered medications for contraindicated DDIs were mid-
azolam (2% of TDF/FTC/EFV users) and dexamethasone (1%
of TDF/FTC/RPV users). The most frequently coprescribed
medications related to potential DDIs were diclofenac (13%
of TDF/FTC/EFV users and 11% of TDF/FTC/RPV users) and
polyvalent cation-containing antacids (12% in ABC/3TC/
DTG users). In regard to all of the protease inhibitors of
interest, the most commonly coadministered contra-
indicated medications were domperidone (3e4%), followed
by quetiapine (2%) and midazolam (1e2%). The medications
often causing potential DDIs with PIs were zolpidem,
betamethasone, polyvalent cation-containing antacids and
loperamide.
Table 3 shows the baseline characteristics stratified by
DDI status among patients receiving 1L-ART. Compared to
patients without DDIs, those with DDIs were more likely to
be female and to have more comorbidities and poly-
pharmacy regardless of which 1L-ART groups were
included. Likewise, Table 4 presents the baseline

1717
 Table 4 Baseline characteristics stratified by DDI status among PI users.
N (%) Atazanavir alone Atazanavir þ Ritonavir Darunavir þ Ritonavir Lopinavir/Ritonavir
a a a
Stratified by w/o DDI w/DDI p value w/o DDI w/DDI p value w/o DDI w/DDI p value w/o DDI w/DDI p valuea
DDI status
N Z 1128 N Z 1186 N Z 334 N Z 322 N Z 660 N Z 631 N Z 1888 N Z 1552

C.-C. Wang, H.-J. Li, C.-H. Shao et al.

Gender (male) 1068 (94.68) 1082 (91.23) <.001 317 (94.91) 300 (93.17) 0.35 639 (96.82) 602 (95.40) 0.19 1749 (92.64) 1407 (90.66) 0.04
Age (50) 218 (19.33) 383 (32.29) <.001 41 (12.28) 82 (25.47) <.001 75 (11.36) 143 (22.66) <.001 336 (17.80) 407 (26.22) <.001
No. of comorbidities
0 609 (53.99) 299 (25.21) <.001 198 (59.28) 120 (37.27) <.001 379 (57.42) 198 (31.38) <.001 904 (47.88) 444 (28.61) <.001
1 391 (34.66) 413 (34.82) 108 (32.34) 108 (33.54) 217 (32.88) 228 (36.13) 733 (38.82) 588 (37.89)
2 102 (9.04) 276 (23.27) 20 (5.99) 55 (17.08) 48 (7.27) 119 (18.86) 210 (11.12) 342 (22.04)
1718

3 26 (2.30) 198 (16.69) 8 (2.40) 39 (12.11) 16 (2.42) 86 (13.63) 41 (2.17) 178 (11.47)
No. of comedications
0 614 (54.43) 124 (10.46) <.001 151 (45.21) 39 (12.11) <.001 316 (47.88) 68 (10.78) <.001 911 (48.25) 198 (12.76) <.001
1 331 (29.34) 256 (21.59) 106 (31.74) 71 (22.05) 208 (31.52) 128 (20.29) 575 (30.46) 383 (24.68)
2 101 (8.95) 233 (19.65) 48 (14.37) 65 (20.19) 71 (10.76) 128 (20.29) 232 (12.29) 301 (19.39)
3 47 (4.17) 180 (15.18) 20 (5.99) 49 (15.22) 37 (5.61) 105 (16.64) 81 (4.29) 243 (15.66)
4 20 (1.77) 145 (12.23) 9 (2.69) 98 (30.43) 14 (2.12) 75 (11.89) 40 (2.12) 158 (10.18)
5 15 (1.33) 248 (20.91) 0 0 14 (2.12) 127 (20.13) 49 (2.60) 269 (17.33)
DDI, drugedrug interaction; PI, protease inhibitors; w/o, without; w/, with.
Statistical significance was set as P < 0.05 which are mentioned in bold.
a
Fisher’s exact or chi-square test was used to determine if there was a significant association between different DDI statuses.
 Journal of the Formosan Medical Association 121 (2022) 1714e1720
characteristics by DDI status among patients prescribed PIs.
Patients with DDIs also had a higher proportion of elderly
patients, multicomorbidities and polypharmacy than those
without DDIs. Detailed comorbidities and comedications
stratified by DDI status are provided in Supplemental Tables
5 and 6. Among the selected comorbidities, psychiatric
disease and liver disease were the most prevalent condi-
tions across all groups. Anxiolytics, hypnotic agents and
gastrointestinal agents were the most commonly prescribed
medications for all patients.
Discussion
To the best of our knowledge, this is the first nationwide
study on HIV patients in Taiwan to reveal the frequency of
contraindicated and potential DDIs with ART. Overall, the
prevalence of DDIs among patients prescribed PIs was
higher than that among patients prescribed 1L-ART. The
consequence of DDI prevalence among different ART regi-
mens was in accordance with the different metabolic pro-
files of certain ARTs, with PIs most impacted by and
involved in various metabolic enzymes and drug trans-
porters.5 Additionally, compared to patients treated with
additional boosting agents, the frequency of DDIs was
higher. That is, there was a higher occurrence of DDIs in the
“ATV þ r” group than in the “ATV alone” group. As sug-
gested by treatment guidelines, 1L-ART, which is not
composed of PIs, is a better choice to avoid DDIs.14
The results showed a low prevalence of contraindicated
DDIs among the HIV population, suggesting that physicians
are fully aware of the potential hazards therein. However,
more attention should be paid to the high risk of potential
DDIs. Strategies to avoid potential DDIs while maintaining
the safety and effectiveness of medications should be
implemented if possible, such as dose adjustment, timely
administration of medications or substitution of concomi-
tant medications with similar pharmacological effects.
Moreover, the reasons leading to the high prevalence of
potential DDIs warrant further investigation. We suspected
that patients may be hesitant to disclose their HIV/AIDS
status, which leads to insufficient medical information for
health care workers to make appropriate judgments.
Therefore, it is important to inform patients about the
severity of DDIs and to encourage them to disclose their HIV
status.
In this study, the most frequently coprescribed medi-
cations related to potential DDIs were diclofenac and
polyvalent cation-containing antacids. Given that diclofe-
nac and antacids are commonly prescribed for symptom
release, the importance of the aforementioned patient
education is again highlighted. In addition, diclofenac and
antacids are also commonly used as over-the-counter (OTC)
drugs. It is therefore important not only to educate patients
about the potential impact of DDIs but also to remind
healthcare providers actively to ask patients for OTC
products that they are taking. Patients often report pre-
scription medications but forget to report OTC products
when healthcare providers ask about their medication his-
tory. Reliance on healthcare providers is needed to survey
the use of OTC products proactively to reduce the potential
risk of DDIs between 1L ARTs and OTC drugs.
1719
Since the majority of HIV patients are relatively young,
unlike other chronic diseases, metabolic disorders and car-
diovascular diseases may not be the most imperative prob-
lems to be solved. Instead, many patients suffer from
mental disorders, which World Health Organization empha-
sizes as the goal of the ‘next 90’ for HIV patients. When we
categorized our study subjects by their DDI status, we also
found that psychiatric disease affected especially those with
DDIs, overcasting their quality of life. This finding also ex-
plains the fact that medications used for the central nervous
system, such as midazolam, quetiapine and zolpidem, were
the most commonly observed DDI comedications.
Despite the efforts made to unmask the current situation
of DDIs in the HIV-infected population in Taiwan, there are
some limitations in our study. First, we evaluated only the
interactions between HIV/non-HIV medications. Nonethe-
less, DDIs between HIV/HIV or non-HIV/non-HIV among HIV-
infected populations are worth discussing. In addition, the
proportion of DDIs was also underestimated since we did
not have information on herbal medicines and over-the-
counter drugs. Second, because of the retrospective nature
of the study, we ensured whether the two drugs were
actually taken by patients simultaneously. In addition, we
could not monitor viral loads, CD4 counts, opportunistic
infections, amusing drugs, or the clinical outcomes of DDIs
since laboratory data were not included in the database.
Finally, because of the inherent limitation of the cross-
sectional design, the causality between HIV status, ART
use, and diagnosis of comorbidities was hard to define.
Conclusions
In Taiwan, the prevalence of contraindicated DDIs was
generally low among HIV patients; however, more attention
should be paid to potential DDIs. Approaches used to pre-
vent the occurrence of contraindicated or potential
drugedrug interactions should be refined. The choice of
ART should follow the discussion among patients, doctors,
and pharmacists and consider patients’ underlying condi-
tions, including comorbidities, pill burden, and risk of DDIs.
Future research focusing on the long-term clinical impact of
potential DDIs is needed.
Author contributions
CC.W., HJ.L. and WH.S. designed the study. HJ.L. analyzed
the data. HJ.L. and CH.S. wrote and revised the paper. All
authors reviewed and contributed to the final version of the
manuscript.
Declaration of competing interest
The authors have no conflicts of interest relevant to this
article.
Acknowledgments
This study was partially funded by research grants from
GlaxoSmithKline Far East B. V. Taiwan Branch. The funder
had no role in the research design, execution, results or
 C.-C. Wang, H.-J. Li, C.-H. Shao et al.
interpretation. We thank Hsiao-Hang Lin and Elsevier Lan-
guage Editing Services for providing editorial assistance in
writing the manuscript.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jfma.2021.12.005.
References
1. Antiretroviral Therapy Cohort Collaboration. Life expectancy
of individuals on combination antiretroviral therapy in high-
income countries: a collaborative analysis of 14 cohort
studies. Lancet 2008;372(9635):293e9.
2. Greene M, Justice AC, Lampiris HW, Valcour V. Management of
human immunodeficiency virus infection in advanced age. J
Am Med Assoc 2013;309(13):1397e405.
3. Wu PY, Chen MY, Hsieh SM, Sun HY, Tsai MS, Lee KY, et al.
Comorbidities among the HIV-infected patients aged 40 years
or older in Taiwan. PLoS One 2014;9(8):e104945.
4. Tsai FJ, Cheng CF, Lai CH, Wu YC, Ho MW, Wang JH, et al.
Effect of antiretroviral therapy use and adherence on the risk
of hyperlipidemia among HIV-infected patients, in the highly
active antiretroviral therapy era. Oncotarget 2017;8(63):
106369e81.
5. Gong Y, Haque S, Chowdhury P, Cory TJ, Kodidela S,
Yallapu MM, et al. Pharmacokinetics and pharmacodynamics of
1720
cytochrome P450 inhibitors for HIV treatment. Expert Opin
Drug Metab Toxicol 2019;15(5):417e27.
6. Tang MW, Shafer RW. HIV-1 antiretroviral resistance: scientific
principles and clinical applications. Drugs 2012;72(9):e1e25.
7. Patel N, Borg P, Haubrich R, McNicholl I. Analysis of drug-drug
interactions among patients receiving antiretroviral regimens
using data from a large open-source prescription database. Am
J Health Syst Pharm 2018;75(15):1132e9.
8. Marzolini C, Elzi L, Gibbons S, Weber R, Fux C. Prevalence of
comedications and effect of potential drug-drug interactions in
the Swiss HIV Cohort Study. Antivir Ther 2010;15(3):413e23.
9. Malaty LI, Kuper JJ. Drug interactions of HIV protease in-
hibitors. Drug Saf 1999;20(2):147e69.
10. Truong WR, Schafer JJ, Short WR. Once-daily, single-tablet
regimens for the treatment of HIV-1 infection. P T 2015;
40(1):44e55.
11. Demessine L, Peyro-Saint-Paul L, Gardner EM, Ghosn J,
Parienti JJ. Risk and cost associated with drug-drug interactions
among aging HIV patients receiving combined antiretroviral
therapy in France. Open Forum Infect Dis 2019;6(3):ofz051.
12. Ministry of Health and Welfare, Taiwan, R.O.C.. Available
from: https://dep.mohw.gov.tw/DOS/cp-1735-3247-113.html.
[Accessed 4 August 2019].
13. Tita AT, Szychowski JM, Boggess K, Saade G, Longo S, Clark E,
et al. Adjunctive azithromycin prophylaxis for cesarean de-
livery. N Engl J Med 2016;375(13):1231e41.
14. Panel on Antiretroviral Guidelines for Adults and Adolescents.
Guidelines for the Use of Antiretroviral Agents in Adults and
Adolescents with HIV. Department of Health and Human Ser-
vices. Available at https://clinicalinfo.hiv.gov/sites/default/
files/guidelines/documents/AdultandAdolescentGL.pdf.