Adapting and pilot testing an evidence-based ARV adherence

intervention for ChinaAnn B. Williams, Honghong Wang, Xianhong Li, Kris Fennie, Jane Burgess

UCLA School of Nursing & Xiangya School of NursingLos Angeles, California, U.S.A. & Changsha, Hunan, China

2

HIV/AIDS Medication Adherence Challenges

• Lifetime duration of treatment

• Frequent (& serious) adverse drug effects

• AIDS stigma

3

Adherence required…

• In order to achieve the optimal virologic, immunologic, and clinical outcomes possible with HAART, the level of adherence required is over 90%. (Bartlett, 2002; Paterson et al., 2000; Singh et al.,

1999)

• This is the equivalent of missing 1 dose per month on a once-a-day regimen.

• Only 6% of patients report full adherence, with a mean level of 56% adherence. (Murphy et al., 2003)

• Adherence may be the variable determining HAART failure or success. (Knobel et al., 1999)

• While patients report a preference for once-a-day dosing, research suggests adherence rates are no better for QD dosing. (Stone et al., 2004)

…and adherence achieved

4

Emerging Resistance

PROBABILITYOF SELECTING FORRESISTANT STRAINS

SUPPRESSION OF VIRAL REPLICATION

0% 100%

MOST DANGEROUS PLACE:PARTIAL SUPPRESSION

5

PATIENTADHERENCE

Systemic drugconcentration

Pharmacokinetics

ViralResistance

Effectivedrug potency

Intracellulardrug

concentration

Rateof viral

replication

PRESERVATIONOF IMMUNEFUNCTION

ANDDELAY INDEISEASE

PROGRESSION

HostFactors

ViralVirulence

6

Background: ATHENA Intervention

• A home-based adherence intervention delivered by a nurse and peer educator team.

• Demonstrated efficacy in the northeastern U.S. in a randomized controlled trial.

7

ClinicianCharacteristics

Adherence

IllnessCharacteristics

PatientCharacteristics

RegimenCharacteristics

8

A focus limited to personal behavior change leads to a programmatic emphasis on individual responsibility for health, at the cost of an examination of individual response-ability, or the capacity of the individual for responding to his or her personal needs or the challenges posed by the environment. Meredith Minkler

Health education, health promotion and the open society: An historical perspective.HEQ, 16: 17-30, 1989

9

ClinicianCharacteristics

Adherence

IllnessCharacteristics

PatientCharacteristics

RegimenCharacteristics

SocialContext

10

Intervention

•Home visit by a peer counsellor and a nurse

• Once a week, first 3 months

• Bi-weekly, months 4-6

• Once a month, months 7-12

• Visits last 15 minutes to one hour

11

Results

TotalN=171

%

Gender

Male 88 51.5

Female 82 48.0

Transgendered 1 0.5

Race/Ethnicity

Native Am 3 1.8

African-Am 59 34.5

Caucasian 72 42.1

Hispanic 32 18.7

Other 4 2.3

Declined 1 0.6

Median (N)

Age 46.8 (171)

Viral Load 400 (171)

CD4 354 (170)

Selected characteristics of ATHENA participants at baseline*

*These characteristics did not differ significantly between the intervention and control groups.

TotalN=171

%

Substance Use1

Current 54 32

Past history 155 911Does not include EtOH; includes marijuana

12

ResultsSubjects with ≥ 90% adherence

0

5

10

15

20

25

30

35

40

45

Baseline 3 6 9 12 15

Months

Pe

rce

nt

Control Intervention

A greater proportion of subjects in the intervention group had adherence greater than 90% at each time point compared to the control group. The difference over time is significant (Extended Mantel-Haenszel Test: 5.80, p=.02)

33 79 42%

33 8240%

12 54 22%

9 37 24%

19 6131%

16 44 36%

19 60 32%

24 66 36%

24 6438%

18 68 27%

14 64 22%

14 60 23%

13

Summary Results

• The intervention group maintained a higher proportion of subjects with adherence greater than 90% over time compared to the control group (p=.02).

• A statistically significant intervention effect on viral load or CD4+ count was not seen.

• There was an statistically signifcant association between >90% adherence and an undetectable viral load over time (p<.03).

14

Conclusions

• Home visits from a nurse and peer counselor significantly improved medication adherence compared to usual care.

• The proportion of individuals with medication adherence >90% was unacceptably low in both control and experimental groups.

15

ATHENA to Ai Sheng NuoReaching around the Globe

Nurses working together to help patients take lifesaving medication

16

Purpose

• To adapt the ATHENA intervention to the social and cultural context of Hunan Province

• To conduct a pilot test of the adapted intervention

17

• HIV/AIDS cases reported through 2010: 10,794• Patients are:

– Rural– Poor– High prevalence of IDU (40% of PLWHA)

Reported HIV infections and AIDS cases in Hunan Province

0

200

400

600

800

1000

1200

1400

1600

1992 1995 1998 2001 2004

Number of HIVInfectionsNumber ofAIDS Cases

HIV/AIDS in Hunan Province

18

HIV/AIDS in Hunan Province

• Free treatment (ARVs) is available

• Medication adherence is a challenge

• Evidence-based interventions to support adherence are limited and were developed for use in different social, cultural, and economic environments.

19

Adaptation Framework

• The ADAPT-ITT Model

– 8 sequential steps– Qualitative and quantitative data

20

Step 1: Assessment

• Cross sectional survey– 7 China CARES sites– 308 respondents

• 20% reported <90% adherence• Associated with current heroin use

21

Step 1: Assessment

• Qualitative data

– Stigma– Family relationships and responsibility– Guilt

22

Step 2: Choosing ATHENA

• Freirian philosophy – Well suited to Chinese

culture– Emphasizes

community context– Known in China

Process

Action

Reflection

Action

23

Step 3 and 4

Administration• Demonstrating the

intervention• Reviewing original

manuals• Consider applicability

to Hunan context

Production• Identify core elements

– Peer educators– Dialogue – Reflection

• Produce plan for adaptation– Emphasis on family– Group activities

24

Step 5: Expert review

• PLWHA, families, and HCWs reviewed proposed intervention

• Concerns: Risk for disclosure & stigma

25

Step 6: IntegrationATHENA to Ai Sheng Nuo

26

Love, Life, Promise

• Ai Sheng Nuo– Family emphasis– Decreased frequency

of home visits– More structured

patient education– Option for group

activities

27

Step 7: Training & developing manual

28

Step 8: Pilot testingMethods

• Randomized controlled pilot– July 2010 – August 2012

• Randomized to intervention or control– Intervention: Monthly visits and interim phone

contact plus standard clinic support– Control: Standard clinic support

29

Study Sites: Hunan Province

• 11th largest province of China, situated in the southeast.

• Commercial sex work and injection drug use are highly prevalent.

• Two clinical sites, in Hengyang City and Changsha.

• Comprehensive evaluation and ARV when indicated.

• However, mental health screening and treatment are not routinely available.

30

Pilot study: SubjectsEligibility

• Living with HIV/AIDS

• Attending one of the two clinical sites

• Self-reporting adherence <90% to prescribed ARVs or to pre-ARV medications (TMP-SMX, multi vitamins)

31

Pilot study: Measures

• A 7-day visual analogue scale

• Social Support Rating Scale

• Center for Epidemiological Studies Depression Scale (Chinese)

• HIV/AIDS Related Stigma Scale.

32

Pilot study: Data collection

• Data were collected in structured face-to-face interviews conducted at the time of a regularly scheduled clinical visit.

• Information regarding ARV regimen, treatment duration, time of diagnosis, CD4 count and HIV-RNA from medical record review.

• Baseline, 6 months, 12 months

33

Results: SubjectsN = 114

ARV status at baseline• 57 reporting <90% adherence to pre-ARV meds• 57 reporting <90% adherence to ARV

Presumed HIV transmission routes– 36% IDU– 40% Heterosexual contact– 11% MTM sexual contact– 2% Transfusion– 11% Unclear

34

Results: Subjects

Male: 82 (72%)

Female: 32 (28%)

Age

< 30 32 (28%)

30 – 45 57 (50%)

> 45 25 (22%)

Married 59 (52%)

High school or college 46 (40%)

Stably Employed 32 (28%)

35

Results: Subjects

• Past or current drug abuse 35 (31%)

• Has disclosed HIV status 84 (75%)

• 2 years or less since diagnosis 90 (82%)

• CD4 <350 cells/mm3 87 (98%)

36

Results: Subjects

• ARV regimens

• AZT + 3TC + NVP or EFV• D4T + 3TC + NVP or EFV• AZT + LPV/r + 3TC• LPV/r + TDF + 3TC

37

Depressive symptoms at baseline

• 66% scored 16 or greater on the CESD-C

• Those in the ARV prep treatment stage were more likely to report significant depressive symptomatology than those for whom ARV had already been prescribed.

(OR = 2.84, 95% CI 1.26, 6.38; p = 0.01)

38

Factors independently associated with depressive symptoms

• History of drug use OR 4.10 (1.11, 15.15) p=.03

• High perception of stigma 1.06 (1.02, 1.09) p=.001

• Lack of stable employment 3.23 (1.01, 10.00) p=.05

• Lack of social support 1.10 (1.03, 1.19) p=.02

39

ResultsSubjects with > 90% adherence

0

10

20

30

4050

60

70

80

90

100

Baseline 6 12

Months

Pe

rce

nt

Control Intervention

A greater proportion of subjects in the intervention group had adherence greater than 90% at both time points compared to the control group. The difference over time is significant (Extended Mantel-Haenszel Test: 8.8, p=.003)

40

Pilot Test

• Biological measures:– No difference between groups:

• Quantitative HIV-RNA• CD4 counts

– Results of ARV resistance studies• No resistance at baseline by standard genotype• Ultra Deep Sequencing ongoing

41

Other findings

• Adherence barriers identified:– Medication side effects– Fear of disclosure– Knowledge deficits– Poor family relationships

42

Conclusions

• Structured approach facilitates adaptation of evidence based interventions.

• In spite of significant cultural differences, adaptation is possible.

• Key barriers to ARV adherence appear to be universal.

• Strategies to improve adherence may differ somewhat, but home based interventions are effective.

43

Future Directions

• Logistics: Mobile communication technology.

• Content: address mental health issues, especially depression.

• Cost of intervention.