Results

Introduction

Background and Objectives

Identifying effective and cost-effective ways to improve adherence to antiretroviral therapy (ART) is critical to maximize the benefits of therapy and use scarce resources most efficiently

Few rigorous ART adherence interventions have been evaluated in low-resource settings; even when effective, few include a cost-effectiveness analysis

Reviews indicate that previous cost-effectiveness analyses of adherence-enhancing interventions are too few in number, use weak methods and poor cost data, and fail to provide complete, clear results1,2

We aimed to analyze the costs and cost-effectiveness of a highly effective ART adherence intervention conducted in a low-resource setting

Methods

We found that an intervention using EDM feedback to inform counseling can improve ART adherence to optimal levels in Chinese patients at a low incremental cost ($247/patient)

Compared to average annual cost of providing ART to a patient in China (estimated at over $2,000), and the social costs of poor adherence, this intervention may be considered a good use of scarce resources

We recommend further analysis of ART adherence interventions and scale-up of those found to be cost-effective in order to treat rising numbers of ART patients most efficiently

Cost-effectiveness of Improving Adherence to Antiretroviral Therapy Using Electronic Drug Monitor

Feedback among HIV-Positive Patients in China: The Adherence for Life (AFL) Study

Lora Sabin,1,2 Mary Bachman DeSilva,1,2 Xu Keyi,3 Davidson H Hamer,1,2,4,5 Kee Chan,6 and Christopher J. Gill1,2

1Center for Global Health and Development, Boston University, Boston, MA, U.S.A.; 2Department of International Health, Boston University School of Public Health, Boston, MA, U.S.A.; 3WHO Collaborating Center for Comprehensive Management of HIV Treatment and Care, Ditan Hospital, Beijing, China; 4Zambia Centre for Applied Health Research and Development, Lusaka, Zambia;

5Infectious Diseases Section, Department of Medicine, Boston University School of Medicine, Boston, MA, U.S.A.; 6Department of Health Sciences, Sargent College of Health and Rehabilitation, Boston University, Boston, MA, U.S.A.

China has one of Asia’s most serious HIV epidemics, with ≈ 740,000 people living with HIV and tens of thousands of new infections occurring annually

Few rigorous ART adherence intervention studies have been assessed in China

In the “Adherence for Life” (AFL) study, we assessed an intervention that involved integrating electronic drug monitor (EDM, via eCAPs) data into HIV clinical care—a process we term ‘EDM feedback’—on ART adherence

The results showed a significant improvement in mean ART adherence in intervention patients3

AFL Study Site Dali 2nd People’s Hospital HIV Clinic, in Dali Old

City, Yunnan Province, China Semi urban area, population mainly from Bai

minority Most HIV infections contracted via injection drug

use (IDU)

AFL Intervention Design 80 ART patients were enrolled, given eCAPs Subjects stratified into high vs. low adherence groups

(based on mean adherence ≥95% in Months 1-5), and randomized in each stratum to intervention or control

Intervention subjects: received EDM adherence data at 6 monthly visits; if adherence was ‘sub-optimal’ (<95%) in previous month, counseled by a clinician using EDM report; otherwise, counseling was optional

Controls: continued standard of care, received counseling if self-reported adherence in previous month was <95%

Adherence measure: incorporated +/- 1 hour dose window, found to be best predictor of CD4 and undetectable VL4

AFL Intervention Effect At month 12, mean adherence was 96.5% in

intervention subjects vs. 84.5% in controls (P-value = 0.003)

Intervention effect largely due to a sharp increase in adherence in previously low adherers (Figure 1)

EDM feedback associated with mean CD4 change (+90 cells/ul in intervention subjects vs. -9 cells/ul in controls (P-value = 0.020))

Analytic Methods Calculated AFL’s financial and economic costs, forecasted

economic cost of a 1-year intervention in 2012 among 500 patients, typical size of an ART clinic in China

Financial analysis: included all project expenditures: 1) equipment & supplies (EDM scanner, eCAPs, user guide); 2) training in eCAP use; 3) shipping; 4) price cuts provided by eCAP supplier Excluded: 1) all research-related costs; 2) control group costs;

3) clinicians’ time (counseling fit into regular workday) China-based costs were converted to US$ (2007 exchange rate),

then added to nominal US$ costsEconomic analyses: to capture societal costs, included:

1) project coordinator/clinicians’ time; 2) opportunity cost of patients’ time for orientation and counseling, valued at mean patient income; 3) full costs (no price cuts) Economic analysis of forecasted 1-year program:

1) no US personnel; 2) addition of local project manager; 3) addition of counselors to orient and counsel patients

All costs adjusted to 2012 pricesCost effectiveness analysis (CEA): Estimated incremental cost per change from sub-optimal to

optimal adherent patient during 6-month program using standard formula: ICEA=(CI–CC)/(AI–AC), (C=total costs; A=change in optimal adherent patients (≥95%); and I/C =intervention/control

Change in adherence based on AFL’s effect size (difference between intervention and control groups in net change in patients with optimal adherence, capturing both increase in intervention arm and avoided loss in control arm

Uncertainty assessed using sensitivity analysis

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

1 2 3 4 5 6 7 8 9 10 11 12

Low adherers, intervention group

Low adherers, control group

High adherers, intervention group

High adherers, control group

Adherence

Study Month

Acknowledgements

Thanks to: Mary Jordan, Billy Pick, David Stanton, Neal Brandes, Connie Osborne, Ray Yip, Ira Wilson, and staff at Med-ic, and our deep appreciation to the medical staff at the Dali Second People’s Hospital and the Dali-based HIV/AIDS patients who participated in the AFL study.

Conclusions

Table 2. Results of one-way sensitivity analysis on the incremental cost of moving an ART patient from sub-optimal adherence to optimal adherence (=95%)

Note: all values expressed in 2012 US$

Variation tested

Estimated incremental cost-effectiveness ratio

Base case $247

Intervention effect varied from 45%-25% (base case = 24%) $188-338

Number of patients participating in intervention varied from 1,000 to 200 (base case = 500)

$126-606

Time spent on counseling varied from 10 minutes to 30 minutes per patient per clinic visit

$246-248

Proportion of patients participating in intervention that require counseling sessions varied from 20% to 80%

$246-248

Cost of EDM data scanner varied from $500 to $2,500 (base case = $1,595)

$241-252

Cost of eCAPs varied from $30 to $80 apiece (base case = $59) $154-314

DaliDali

China

Costs Estimated financial and economic costs of AFL were $7,943 and $9,065 (2007 US$), respectively, or $732 and $836 per patientFixed costs accounted for 75% of total costsFor 2012 scaled-up 1-year program, total and per patient costs were $42,222 and $84 (2012 US$) (Table 1)

Sensitivity AnalysisCost-effectiveness most sensitive to variation in effect size, number of patients participating, and cost of eCAPs

If intervention effect increased to 45%, cost per adherent patient falls by 25% If unit price of eCAPs increased by 1/3, to $80, cost per adherent patient rises 25%.

Other cost variables, such as cost of EDM scanner, and assumptions such as time spent on counseling, do not greatly affect cost-effectiveness

Figure 1: Monthly mean adherence in AFL subjects, stratified by mean adherence in pre-intervention period: high (≥95%) vs. low (<95%)

Table 2: Results of one-way sensitivity analysis on the incremental cost of moving an ART patient from sub-optimal adherence to optimal adherence (≥95%)

Citations1.Elliott RA, Barber N, Horne R. Cost-effectiveness of adherence-enhancing interventions: a quality assessment of the evidence. Ann Pharmacother 2005;39(3):508-15.2.Rosen AB, Spaulding AB, Greenberg D, Palmer JA, Neumann PJ. Patient adherence: a blind spot in cost-effectiveness analyses? Am J Manag Care 2009;15(9):626-32.3.Sabin LL, Desilva MB, Hamer DH, et al. Using Electronic Drug Monitor Feedback to Improve Adherence to Antiretroviral Therapy Among HIV-Positive Patients in China. AIDS Behav 2009.4.Gill CJ, Sabin LL, Hamer DH, et al. Importance of dose timing to achieving undetectable viral loads. AIDS Behav 2009; Published online: April 8, 2009.

Cost-effectiveness (see Table 1)Cost per newly optimally adherent patient in AFL trial: $732 (financial analysis) and $836 (economic analysis)Cost per newly optimally adherent patient in projected 10–year program: $249

Cost item

Financial1 2007

Cost % total

Economic1 2007

Cost % total

Scale-up2 2012

Cost % total

Costs Total Fixed Costs 6060 76.3 6974 76.9 5663 13.4 Program set-up 426 5.4 454 5.0 33 0.1 Training activities 4404 55.4 4705 51.9 3185 7.5 US personnel3 1704 21.5 1704 18.8 - 0 China personnel4 - 301 3.3 485 1.1 Travel5 2200 27.7 2200 24.3 2200 5.2 EDM training fee 500 6.3 500 5.5 500 1.2 Equipment & supplies6 695 8.7 695 7.7 1695 4.0 Technical support/management 500 6.3 920 10.1 250 0.6 Miscellaneous (shipping) 35 0.4 200 2.2 500 1.2 Total variable costs 1710 23.7 2091 23.1 36560 86.6 Intervention delivery personnel time - 0 196 2.2 2939 7.0 Clinic personnel7 160 1.8 2094 5.0 Patient time8 36 0.4 845 2.0 Equipment and supplies9 1710 23.7 1895 20.9 33621 79.6 TOTAL ($2007 US$) 7943 100.0 9065 100.0 TOTAL ($2012 US$) 42222 100.0 Cost per subject2 259 295 84 Variable cost per subject2 61 68 73

Cost-effectiveness Optimal adherent patients (change) (#) 12 12 171 Cost per optimal adherent patient

(2012 US$) 732 836 247

1 Values expressed in 2007 US$, except where otherwise indicated. 2 Values expressed in 2012 US$. 3 Training in EDM use with manufacturer (based in Canada) and training of

clinicians in China. 4 Economic costs include time of clinicians in training; Scale-up model includes

time of training by local trainer. 5 Training activities in Canada and China. 6 Includes EDM scanner and eCAP in Mandarin Chinese. 7 For Economic costs, refers to 3 doctors and 6 nurses; for scale-up model, refers

to 2 nurses who are employed to deliver intervention. 8 Patient time spent on one-time enrollment, and in monthly clinic visits, on

waiting for scanning of EDM and counseling sessions. 9 Includes eCAPs and medication vials (one per patient plus 10% contingency).

Table 1: Costs and cost-effectiveness of AFL intervention