Ethics statement

Ethics approval was granted for the TAHOD study design, methods and consent procedures by the University of New South Wales Human Research Ethics Committee (HC17825). Site specific study governance was granted by site-relevant institutional review boards: Ministry of Health National Ethics Committee for Health Research (Cambodia), Ethical Committee of Beijing Ditan Hospital Affiliated to Capital Medical University (China), Research Ethics Committee Kowloon Central / Kowloon East, Hospital Authority IRB (China), Institutional Review Board Of YRG CARE (India), Institutional Ethics Committee Rao Nursing Home (India), Kerti Praja Foundation IRB (Indonesia), Committee of Medical Research Ethics, Faculty of Medicine University of Indonesia (Indonesia), National Center for Global Health and Medicine Human Research Ethics Committee (Japan), Medical Research & Ethics Committee, Ministry of Health (for Sungai Buloh Hospital and Hospital Raja Perempuan Zainab II, Malaysia), Medical Ethics Committee, University Malaya Medical Centre (Malaysia), Research Institute for Tropical Medicine, Department of Health (Philippines), National Healthcare Group IRB, Domain Specific Review Board (Singapore), Severance Hospital Yonsei University College of Medicine Institutional Review Board (South Korea), Institutional Review Board of Taipei Veterans General Hospital (Taiwan), The Internal Ethical Committee for Research in Human Subject, Chiangrai Prachanukroh Hospital (Thailand), Institutional Review Board Faculty of Medicine, Chulalongkorn University (Thailand), Committee on Human Rights Related to Research Involving Human Subjects Faculty of Medicine Ramathibodi Hospital, Mahidol University (Thailand), Research Ethics of the Faculty of Medicine, Chiang Mai University (Thailand), Siriraj Institutional Review Board, Mahidol University (Thailand), Ministry of Health, Hanoi School of Public Health IRB (Vietnam), and National Hospital of Tropical Diseases IRB (Vietnam).

Written informed consent was not sought in TAHOD unless required by a site’s local institutional review board. The need for written consent was waived by the following ethics committees: Ministry of Health National Ethics Committee for Health Research (Cambodia), Ethical Committee of Beijing Ditan Hospital Affiliated to Capital Medical University (China), Research Ethics Committee Kowloon Central / Kowloon East, Hospital Authority IRB (China), Institutional Review Board Of YRG CARE (India), Institutional Ethics Committee Rao Nursing Home (India), Kerti Praja Foundation IRB (Indonesia), Committee of Medical Research Ethics, Faculty of Medicine University of Indonesia (Indonesia), Medical Research & Ethics Committee, Ministry of Health (for Sungai Buloh Hospital, Malaysia), Medical Ethics Committee, University Malaya Medical Centre (Malaysia), The Internal Ethical Committee for Research in Human Subject, Chiangrai Prachanukroh Hospital (Thailand), Ministry of Health, Hanoi School of Public Health IRB (Vietnam), and National Hospital of Tropical Diseases IRB (Vietnam).

Study population

We used individual, de-identified patient data from Thai sites contributing to the TREAT Asia HIV Observational Database (TAHOD). TAHOD involves 21 HIV clinics in the Asia-Pacific region and is part of the International Epidemiology Databases to Evaluate AIDS collaboration [9]. Ethics approval was granted by the University of New South Wales Human Research Ethics Committee (HC17825). Clinics follow local guidelines and regulations regarding patient consent and ethics review (see Ethics statement in Acknowledgements). The study population was selected from patients enrolled at Ramathibodi Hospital, Bangkok; HIV-NAT Research Collaboration/Thai Red Cross AIDS Research Centre, Bangkok; Research Institute for Health Sciences, Chiang Mai; or Chiangrai Prachanukroh Hospital, Chiang Rai. Included patients were required to have documentation of ≥1 visit between 1 January 2013 and 1 September 2019 and, at their last clinic visit, be aged 35–75 years, have no history of ASCVD, have been using ART for at least 6 months, and have a CD4 cell count >100 cells/mm³. S1 Table in S1 File further characterizes the PLHIV in our study population.

Model structure, parameterization and validation

We developed a microsimulation model that randomly selected (with replacement) 100,000 patients from our study population (N = 1,379). The model simulated their probability of ASCVD over time. Direct medical costs and outcomes were assigned in annual cycles over a lifetime horizon and discounted at 3% per year [10]. We assumed the Thai healthcare sector perspective. Primary ASCVD risk was calculated using the reduced Data-collection on Adverse Effects of Anti-HIV Drugs (D:A:D) CVD risk equation [11]. Background mortality rates were based on those of the Asian population on ART [12]. Recurrent event rates were modelled as per our earlier analysis evaluating pravastatin and pitavastatin expansion [8]. Further detail on the model is provided in the Supplementary Material, including S1 Fig in S1 File which presents a schematic of the model structure. Model parameters are shown in Table 1. We calibrated our model using a goodness-of-fit approach based on the observed rates of all-cause and cardiovascular death among TAHOD participants between 2009 and 2019. Fig 1 shows that our calibrated model estimates provided an accurate reflection of the observed data.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Observed versus modelled probability of all-cause and cardiovascular death over time.

Observed data is from Thai sites in the TREAT Asia HIV Observational Database (TAHOD). Shaded area is 95% confidence interval for observed data.

https://doi.org/10.1371/journal.pone.0256926.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Key model parameters.

https://doi.org/10.1371/journal.pone.0256926.t001

Treatment strategies

Current Thai guidelines recommend PLHIV with an LDL-C level ≥190mg/dL, those aged ≥40 years with diabetes, and those aged ≥50 years with chronic kidney disease and an LDL-C level ≥100mg/dL should be started on a statin regardless of ASCVD risk score [6,7]. For those who do not meet these criteria, statin therapy is recommended if their 10-year ASCVD risk is ≥10%. Individuals with an ASCVD score <10% may be considered for statin therapy if there is evidence of subclinical atherosclerosis but such information is rarely available in routine practice in Thailand. Consistent with Thai guidelines, we assumed all individuals with an LDL-C level ≥190mg/dL would be prescribed high intensity statin therapy, and those with a statin indication associated with diabetes or chronic kidney disease would be prescribed low/moderate intensity statin therapy. We evaluated three different treatment strategies for individuals who currently require an ASCVD risk score to determine statin eligibility: 1) treating those with a 10-year ASCVD risk ≥10% with a low/moderate intensity statin (control group); 2) treating those with a 10-year ASCVD risk ≥7.5% with a low/moderate intensity statin; and 3) treating those with a 10-year ASCVD risk ≥5% with a low/moderate intensity statin.

We modelled the effectiveness of statins through the simulated change in LDL-C. We assumed fully adherent individuals using high intensity and low/moderate intensity statins would achieve LDL-C reductions of 55% and 40%, respectively [3,13]. Statin adherence was assumed to be 86.7% in the first year of statin use, 72.6% in the second, 61.1% in the third, 56.6% in the fourth, and 58.1% in all years thereafter [14]. We assumed PLHIV would only accrue the cost of statin use, exhibit side effects of statins, and benefit from statin LDL-C cholesterol reduction while they were using a statin and hence these parameters were adjusted in line with the decline in adherence over time.

We assumed statin therapy only reduced ASCVD risk by improving LDL-C levels. In scenario analyses, we assumed additional ASCVD preventative efficacy to account for the possibility that the anti-inflammatory properties of statins provide additional benefit in PLHIV [5]. We assumed statins do not prevent non-ASCVD events as current evidence suggests little or no benefit for such outcomes [33]. We modelled adverse events related to statin use (hemorrhagic stroke, diabetes, and myopathy) based on rates observed in the general population (see Table 1 and Supplementary Material in S1 File). We did not account for differences in statin type or dose as current evidence suggests these factors have little impact on the type or frequency of adverse events observed [34].

Cost and quality-of-life estimates

Health-related costs and quality-of-life adjustments were assigned to clinical events and health states in annual cycles. We included medical costs regardless of who paid for them. Cost estimates from earlier years were inflated to 2018 Thai Baht equivalents [35]. The cost of HIV management and rates of second-line ART use were based on published literature [15,16]. Statin costs were estimated based on unit costs [25] and a survey of statin use among PLHIV on ART at the Thai TAHOD sites contributing to our study population. S4 Table in S1 File provides a breakdown of how we arrived at the average annual costs for low/moderate and high intensity statin use (2,301 Baht [$US74] and 9,486 Baht [$US304], respectively). Other costs were based on published estimates for the general population (see Table 1). Quality-of-life adjustments were largely based on data from the 2017 Global Burden of Disease study [31]. Since patients using ART already take at least one daily pill, our base-case model assumed the inconvenience of taking a daily statin (pill burden) was not associated with a quality-of-life decrement. Earlier studies among the HIV and general populations have also assumed regular statin use is not associated with a pill burden [8,32,36].

Outcomes

The primary outcome was the incremental cost-effectiveness ratio (ICER). The threshold for an intervention being considered cost-effective (willingness-to-pay threshold) was as an ICER below 160,000 Baht ($US5,315). This is consistent with recommendations from the Thai Ministry of Public Health [37].

Scenario analyses

In addition to our base-case analyses, we investigated the following scenarios:

1. Assuming statins reduce ASCVD event probability an additional 15% to account for the possibility that their anti-inflammatory properties provide additional benefit among PLHIV;
2. As above but assuming statins reduce ASCVD event probability an additional 30%;
3. Using the Rama-EGAT equation to calculate T1MI and ischemic stroke risk. The Rama-EGAT equation [38] has been validated in the general Thai population [39] and is a reasonable alternative to the D:A:D equation.

Sensitivity analyses

We used sensitivity analyses to evaluate whether our results were impacted by uncertainty in key input parameters. In deterministic sensitivity analyses, we varied one or two input parameters at a time. In probabilistic sensitivity analyses we varied multiple input parameters simultaneously across prespecified distributions over 1,000 iterations. We used beta distributions for utilities and event probabilities, and log-normal distributions for hazard ratios, safety and efficacy measures, and costs.

Software

Data management and statistical analysis was conducted using SAS 9.4 (SAS Institute Inc, Cary, North Carolina). Modelling was undertaken in TreeAge Pro 2020 Version R1.0 (TreeAge Software, Williamstown, Massachusetts).

Base-case analysis

Modelled incidence rates for T1MI, ischemic stroke and fatal CVD among the ≥10% ASCVD risk threshold group were 6.91, 3.06 and 3.58 per 1,000 person-years, respectively. Individuals in this group were projected to accumulate a discounted average of 22.268 QALYs, 22.587 life-years and 1,446,067 Baht ($US46,407) in direct medical costs (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Incremental cost-effectiveness of different statin initiation thresholds for primary prevention of ASCVD among PLHIV.

https://doi.org/10.1371/journal.pone.0256926.t002

Compared with the ≥10% ASCVD risk threshold group, reductions in the incidence of T1MI, ischemic stroke and fatal CVD were projected to be 0.9%, 1.3% and 0.7%, respectively, in the ≥7.5% ASCVD risk threshold group. These reductions contributed to the accumulation of 0.015 additional QALYs at an incremental cost of 3,539 Baht ($US113), giving an ICER of 239,000 Baht ($US7,670)/QALY gained. Compared with the ≥7.5% ASCVD risk threshold group, reductions in the incidence of T1MI, ischemic stroke and fatal CVD were projected to be 0.4%, 0.8% and 0.8%, respectively, in the ≥5% ASCVD risk threshold group. These reductions contributed to the accumulation of 0.015 additional QALYs at an incremental cost of 5,381 Baht ($US172), giving an ICER of 349,000 Baht ($US11,200)/QALY gained (Table 2).

Scenario analyses

Our scenario analyses results are shown in Table 2. Assuming statins reduce ASCVD risk an additional 15% due to their anti-inflammatory effects (Scenario 1) reduced the ICERs for both the ≥7.5% ASCVD risk threshold versus the ≥10% ASCVD risk threshold (231,000 Baht [$US7,413]/QALY gained), and for the ≥5% ASCVD risk threshold versus the ≥7.5% ASCVD risk threshold (299,000 Baht [$US9,595]/QALY gained). Assuming statins reduce ASCVD risk an additional 30% due to their anti-inflammatory effects (Scenario 2) further reduced the respective ICERs: 177,000 Baht ($US5,680)/QALY gained and 256,000 Baht ($US8,215)/QALY gained. When replacing the D:A:D equation with the Rama-EGAT equation to estimate ASCVD risk (Scenario 3), our model predicted higher incidence rates of T1MI, ischemic stroke and fatal CVD. The ICERs were 272,000 Baht ($US8,729)/QALY gained for the ≥7.5% ASCVD risk threshold versus the ≥10% ASCVD risk threshold, and 424,000 Baht ($US13,607)/QALY gained for the ≥5% ASCVD risk threshold versus the ≥7.5% ASCVD risk threshold.

Sensitivity analyses

One-way sensitivity analysis results are presented in Fig 2 (≥7.5% versus ≥10% ASCVD risk threshold) and S2 Fig in S1 File (≥5% versus ≥7.5% ASCVD risk threshold). Our findings were sensitive to changes in the annual cost of low/moderate intensity statin therapy, the pill burden associated with daily statin use, and the discount rate.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Tornado plot showing impact of changes in model parameters on the incremental cost-effectiveness ratio for statin initiation threshold ≥7.5% ASCVD risk versus ≥10% ASCVD risk.

Shading of bars indicates directionality: Lighter bars represent the smaller values in the sensitivity range and darker bars indicate the larger values. Directionality also indicated by the order of values shown in the text description. Cost-effectiveness based on a willingness-to-pay threshold of 160,000 Baht/QALY gained. Baht can be converted to $US by dividing by 31.16. ASCVD, cardiovascular disease; LDL-C, low density lipoprotein cholesterol; T1MI, type 1 myocardial infarction; CABG, coronary artery bypass graft; PCI, percutaneous coronary intervention; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life-year.

https://doi.org/10.1371/journal.pone.0256926.g002

At a willingness-to-pay threshold of 160,000 Baht ($US5,315)/QALY gained, the ≥7.5% ASCVD risk threshold became cost-effective compared to the ≥10% ASCVD risk threshold when the average annual cost of low/moderate intensity statin dropped to 1,593 Baht ($US51; 69.2% of base-case price). This price drop could be achieved if we assumed 54% of moderate intensity atorvastatin users were instead using an equivalent dose of simvastatin (i.e., 10mg atorvastatin to 20mg simvastatin, and 20mg atorvastatin to 40mg simvastatin). Given a shift in prescribing preference to simvastatin may be possible with a shift in prescribing from protease inhibitors to integrase inhibitors, we also evaluated the price drop required for low/moderate intensity statins in the context of higher costs for HIV treatment. Assuming the maximum annual cost of HIV management from our sensitivity range (89,784 Baht; $US2,881), the ≥7.5% ASCVD risk threshold became cost-effective compared to the ≥10% ASCVD risk threshold when the average annual cost of low/moderate intensity statin dropped to 1,459 Baht ($US47; 63.4% of base-case price). This price drop could be achieved if we assumed 64% of moderate intensity atorvastatin users were instead using an equivalent dose of simvastatin.

When the probability of ASCVD while using a statin was reduced by 30% to account for the possibility of preventative efficacy in PLHIV beyond that associated with cholesterol improvement, the ≥7.5% ASCVD risk threshold became cost-effective compared to the ≥10% ASCVD risk threshold if the average annual cost of low/moderate intensity statin use dropped to 2,085 Baht ($US66; 90.6% of base-case price).

In our base-case, we assumed that pill burden associated with daily statin use was not associated with any quality-of-life decrement. When a negative quality-of-life impact was assumed, the average number of QALYs accumulated declined with all three strategies but more so among the lower threshold strategies. At the high end of our sensitivity range, the expanded statin use strategies generated incremental QALY gains of 0.008 (ASCVD risk threshold ≥7.5% versus ≥10%) and 0.007 (ASCVD risk threshold ≥5% versus ≥7.5%) resulting in ICERs of 420,000 Baht ($US13,478)/QALY gained and 808,000 Baht ($US25,930)/QALY gained, respectively.

In probabilistic sensitivity analyses, the ASCVD risk threshold ≥10% was optimal in 96.9% of simulations at a willingness-to-pay of 160,000 Baht ($US5,135)/QALY gained, the ASCVD risk threshold ≥7.5% was optimal in 3.0% of simulations, and the ASCVD risk threshold ≥5% was optimal in 0.1% of simulations (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Acceptability curves showing proportion of probability sensitivity analysis simulations cost-effective for each intervention under different assumptions of willingness-to-pay.

Dashed vertical line represents willingness-to-pay threshold of 160,000 Baht/QALY gained. Baht can be converted to $US by dividing by 31.16. ASCVD, atherosclerotic cardiovascular disease; QALY, quality-adjusted life-year.

https://doi.org/10.1371/journal.pone.0256926.g003

TAHOD study members

PS Ly*, V Khol, National Center for HIV/AIDS, Dermatology & STDs, Phnom Penh, Cambodia; FJ Zhang*, HX Zhao, N Han, Beijing Ditan Hospital, Capital Medical University, Beijing, China; MP Lee*, PCK Li, TS Kwong, HY Wong, Queen Elizabeth Hospital, Hong Kong SAR; N Kumarasamy*, C Ezhilarasi, Chennai Antiviral Research and Treatment Clinical Research Site (CART CRS), VHS-Infectious Diseases Medical Centre, VHS, Chennai, India; S Pujari*, K Joshi, S Gaikwad, A Chitalikar, Institute of Infectious Diseases, Pune, India; S Sangle*, V Mave, I Marbaniang, S Nimkar, BJ Government Medical College and Sassoon General Hospital, Pune, India; TP Merati*, DN Wirawan, F Yuliana, Faculty of Medicine Udayana University & Sanglah Hospital, Bali, Indonesia; E Yunihastuti*, A Widhani, S Maria, TH Karjadi, Faculty of Medicine Universitas Indonesia—Dr. Cipto Mangunkusumo General Hospital, Jakarta, Indonesia; J Tanuma*, S Oka, T Nishijima, National Center for Global Health and Medicine, Tokyo, Japan; JY Choi*, Na S, JM Kim, Division of Infectious Diseases, Department of Internal Medicine, Yonsei University College of Medicine, Seoul, South Korea; YM Gani*, NB Rudi, Hospital Sungai Buloh, Sungai Buloh, Malaysia; I Azwa*, A Kamarulzaman, SF Syed Omar, S Ponnampalavanar, University Malaya Medical Centre, Kuala Lumpur, Malaysia; R Ditangco*, MK Pasayan, ML Mationg, Research Institute for Tropical Medicine, Muntinlupa City, Philippines;

YJ Chan*, WW Ku, PC Wu, E Ke, Taipei Veterans General Hospital, Taipei, Taiwan;

OT Ng*, PL Lim, LS Lee, D Liang, Tan Tock Seng Hospital, National Centre for Infectious Diseases, Singapore (note: OT Ng is also supported by the Singapore Ministry of Health’s (MOH) National Medical Research Council (NMRC) Clinician Scientist Award (MOH-000276). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of MOH/NMRC.); A Avihingsanon*, S Gatechompol, P Phanuphak, C Phadungphon, HIV-NAT/Thai Red Cross AIDS Research Centre, Bangkok, Thailand; S Kiertiburanakul*, A Phuphuakrat, L Chumla, N Sanmeema, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand; R Chaiwarith*, T Sirisanthana, J Praparattanapan, K Nuket, Chiang Mai University—Research Institute for Health Sciences, Chiang Mai, Thailand; S Khusuwan*, P Kantipong, P Kambua, Chiangrai Prachanukroh Hospital, Chiang Rai, Thailand; KV Nguyen*, HV Bui, DTH Nguyen, DT Nguyen, National Hospital for Tropical Diseases, Hanoi, Vietnam; CD Do*, AV Ngo, LT Nguyen, Bach Mai Hospital, Hanoi, Vietnam;

AH Sohn*, JL Ross*, B Petersen, TREAT Asia, amfAR—The Foundation for AIDS Research, Bangkok, Thailand; MG Law*, A Jiamsakul*, R Bijker, D Rupasinghe, The Kirby Institute, UNSW Sydney, NSW, Australia. (* TAHOD Steering Committee member)