Early Warning Indicators for Virological Failure in the Southern African Context Vincent Marconi, MD Professor of Medicine Emory University School of Medicine.

Early Warning Indicators for Virological Failure in the Southern African Context

Vincent Marconi, MDProfessor of Medicine

Emory University School of MedicineRollins School of Public Health

DisclosureAt the time this presentation was given I had no real or perceived vested interests that related to this presentation nor did I have any relationships with pharmaceutical companies, biomedical device

manufacturers, and/or other corporations whose products or services are related to pertinent therapeutic areas. However, I have

received research funding from ViiV, Gilead and Bayer.

Vincent Marconi

MSFColors

It’s a busy Monday in ART Initiation Clinic…

• Mr. N, 49 M• Recently diagnosed with TB/HIV• Fevers, cough and weight loss• CD4 110• Receiving TB Tx• Starting Regimen 1a

• Ms. S, 34 F• Known HIV for 1 yr• Diarrhea• CD4 210• Received PMTCT, Breastfeeding• Starting Regimen 1b

Since 1995, antiretroviral therapy has averted 7.6 million deaths globally,including 4.8 million deaths in sub-Saharan Africa.

Together, these life-saving medicines have gained approximately 40.2million life-years since the epidemic started.

2013 UNAIDS Gap Report

“Don’t give up the fight!”

1973 Bob Marley

“Perseverance is the foundation of all actions”

6th Century BC Lao Tzu

Do you have any concerns??

1. Which gender is at a higher risk of VF

a) Male

b) Female

Do you have any concerns??

2. Which age group is at higher risk of VF?

a) Young

b) Old

Do you have any concerns??

3. Which clinical factors increase risk of VF?

a) Diarrhea

b) Low CD4

c) Weight loss

d) Breastfeeding

e) PMTCT

f) a and b

g) c and e

Demographic and Clinical Factors

• Men – Drain 2013, Anude 2013• Younger Age – Silverberg 2007,

Weintrob 2008• Low CD4 – Badri 2008• Concurrent OI’s – Marconi 2008• D4t – PHIDISA 2010, McGrath 2012• Diarrhea• PMTCT*

Marconi AIDS Pt Care STD 2014

ART Need and Coverage

• 37.5 million people living with HIV– 25.8 M (69%) in SSA– 6.8 M (18%) in SA

• 15 million people receiving ART (MDGI) in 3/15– 10.7 M in SSA (43%)– 2.9 M in SA (43%)

UNAIDS 2015

Kaiser 2015

Prevalence

2010-2013 New ART

“This early-release guideline makes available two key recommendations that were developed during the revision process in 2015. First, antiretroviral therapy (ART) should be initiated in everyone living with HIV at any CD4 cell count. Second, the use of daily oral pre-exposure prophylaxis (PrEP) is recommended as a prevention choice for people at substantial risk of HIV infection as part of combination prevention approaches”

WHO Guidelines on when to start ART and on PrePSeptember 2015

Public Health Dilemma• Increasing number of individuals

with HIV requiring therapy– Increasing number of infections– Opt-out testing (HCT)– Earlier treatment (per WHO)*– Increasing coverage for treatment

(>40%)– Decreasing mortality among

treated• Limited capacity and resources

to manage existing patients on therapy (funding declining)

*New Guidelines Sept 2015 for all people living with HIV

VF and HIV Drug Resistance: No Small Problem

• Worldwide estimates of 11-47% virologic failure within one year of first ART (1.7 – 7M)*

• 40-95% individuals VF have > 1 major resistance mutation

• 4.4-44.7% of individuals on ART will have drug resistance within one year (0.6 – 6.7M)*

• Over time triple class failure will accumulate

* Calculated for 15M on ART (19% in SA) GAP Report 2014

Upward estimate of 1.3M with at least 1 major resistance mutation in SA

(if everyone was on ART it would be 3M max estimate)

How can we prevent VF?

“One Size Fits All”“Tailored Therapy”

Targeted Approach

Key Clinical/Programmatic Questions

• Can we predict which patients are likely to experience virologic failure?– Before starting– While on treatment

• Can we prevent these patients from experiencing virologic failure?

Population-Level Early Warning Indicators

• Indicators which speak to concerns about HIVDR

– Assess factors at individual clinics which are known to create situations favourable to the emergence of HIVDR

– EWIs provide an alert to clinic and ART programmes -- thus providing an opportunity for corrective action

– Indicators exist for adults and children

High-priority

element

Bennett DE et al., Antivir Ther 2008

HIVDR Early Warning Indicators (EWI)

WHO-recommended HIVDR EWIs

EWI EWI Target

1. Prescribing practices 100%

2. Lost to follow-up at 12 months ≤ 20%

3. Retention on first-line ART at 12 months ≥ 70%

4. On-time drug pick up ≥ 90%

5. On-time appointment keeping ≥ 80%

6. Drug supply continuity 100%

8. Viral load <1000 copies/ml at 12 months ≥ 70%

Bennett DE et al., Antivir Ther 2008

SiteEWI 1: On-time Pill Pick-up

EWI 2: Retention in Care

EWI 3: Pharmacy Stock-outs

EWI 4: Dispensing Practices

EWI 5: Virological Suppression

1 95% 75% 100% 70% -2 70% 50% 100% 15% -3 100% 75% 75% 0% 95%4 85% - 100% 0% 78%5 98% 95% 0% 50%

… … … … … …

100 100% 100% 100% 0% 100%

National Level Reporting

Collated results provide a national levelAt-a-glance assessment of site performance

Countries implementing at least one aspect of the Global HIVDR Strategy and locations of HIVDR testing laboratories

As of 2011, 124 rounds of EWI monitoring in 58 countries in > 2000 clinics

ART Program Use of EWI Results

1. Strengthened record keeping systems• Formation of clinic specific care optimizing committees1

• Validation of existing electronic record keeping systems1, 2,3

• Adjustments in pharmacy record keeping to permit on time pill pick up assessments3

• Pilot of enhanced defaulter tracing to identify patients missing drug pick-ups with the goal of reengaging in care within 48 hours1

• General strengthening of records4,5,6,7,8

2. Seek funding support from partners to scale-up EWI9

3. District teams to support adherence and trace patients LTFU1,10,11

4. Scale-up viral load testing5

5. Regular review of patient pill pick-up and establishment of formal referral system to document transfers-in/out6

1Hong et al. JAIDS 2010; 2 Anna Jonas, MoHSS Namibia, personal communication; 3Dawn Pereko, MSH Namibia, personal communication; 4Jack N et al. CID (in press); 5Ye M et al. CID (in press); 6Daonie e et al. CID (in pres); 7Nhan DT el al. CID (in press); Hedt BL et al., Anti Viral Ther 2008; 9Paula Mundari, Uganda National ART Programme, IAS 2010, Vienna; 10Evelyne B, National ART Program, Burundi, personal communication; 11Anna Jonas, MoHSS Namibia, personal communication.

Adult with Viral load suppressed rate at 6 months

DistrictTarget

FY 2014/15FY 2011/12 FY 2012/13 FY 2013/14 Progress Q3

VLS at 6mFY 2013/14

Amajuba District Municipality 96.5 92.5 94.5 94.1 1,108

eThekwini Metropolitan Municipality 96.5 90.2 92.9 92.8 4,535

Harry Gwala District Municipality 96.5 74.4 78.9 83.5 1,577

iLembe District Municipality 96.5 90.3 91.7 0.0 0

Ugu District Municipality 96.5 92.1 93.2 91.3 3,941

uMgungundlovu District Municipality 96.5 80.2 80.9 84.5 915

Umkhanyakude District Municipality 96.5 92.5 90.5 91.3 1,884

Umzinyathi District Municipality 96.5 82.7 94.5 92.9 369

Uthukela District Municipality 96.5 87.6 89.7 93.1 1,676

Uthungulu District Municipality 96.5 67.5 78.2 83.9 4,250

Zululand District Municipality 96.5 83.3 87.1 92.6 718

KwaZulu-Natal 96.5 84.9 87.7 89.4 20,973

Adult percentage lost to follow up after 6 months ART

DistrictTarget

FY 2014/15FY 2011/12 FY 2012/13 FY 2013/14 Progress Q3

LTF at 6mQ3 FY2013/14

Amajuba District Municipality 10.7 13.6 12.7 18.1 942

eThekwini Metropolitan Municipality 10.7 12.4 14.3 23.2 5,412

Harry Gwala District Municipality 10.7 8.6 10.9 20.4 1,515

iLembe District Municipality 10.7 6.5 11.6 1.9 7

Ugu District Municipality 10.7 8.7 9.7 19.7 2,224

uMgungundlovu District Municipality 10.7 11.4 17.2 18.8 1,068

Umkhanyakude District Municipality 10.7 6.1 8.5 24.0 2,256

Umzinyathi District Municipality 10.7 4.8 8.4 18.4 514

Uthukela District Municipality 10.7 6.9 10.5 15.0 1,482

Uthungulu District Municipality 10.7 9.6 10.8 17.6 2,038

Zululand District Municipality 10.7 8.8 10.9 17.3 665

KwaZulu-Natal 10.7 9.6 11.9 20.0 18,123

Adult with Viral load completion rate at 6 months

DistrictNDoH TargetFY 2014/15

FY 2011/12 FY 2012/13 FY 2013/14 Progress Q3VLD at 6m

FY 2013/14

Amajuba District Municipality 80 54.0 47.9 48.4 11,678

eThekwini Metropolitan Municipality 80 64.6 64.4 67.4 4,872

Harry Gwala District Municipality 80 65.1 55.3 44.1 1,148

iLembe District Municipality 80 50.2 44.0 42.6 23,041

Ugu District Municipality 80 38.6 36.2 32.4 1,178

uMgungundlovu District Municipality 80 26.5 30.6 29.6 4,888

Umkhanyakude District Municipality 80 41.4 39.4 35.4 1,888

Umzinyathi District Municipality 80 33.0 43.8 0.0 0

Uthukela District Municipality 80 37.7 42.9 53.4 4,318

Uthungulu District Municipality 80 38.6 35.2 28.4 1,083

Zululand District Municipality 80 43.4 37.6 32.0 2,064

KwaZulu-Natal 80 17.4 15.4 19.3 397

HIVDR Early Warning Indicators (EWI)

• Programmatic Level*– Prescribing practices– LTFU 12 mos ART– Retention on 1st Line

ART at 12 mos/VL UD– Timely ARV pickup– ARV appointments– ARV shortages– Adherence– Baseline HIVDR

• Individual Level– Pharmacy Refill

Data/Clinic Visits– Pill Counts/Self-

Reported Adherence– Clinical Risk Factors– Baseline Minority

Drug Resistance– Psychosocial Risk

Factors

*WHO recommends (http://www.who.int/hiv/topics/drugresistance/indicators/en/index.html)

Toxicity, AdverseEffects, TolerabilityTreatment Fatigue

Access to Potent cART(Properly prescribed

Combinations)

AcceptanceAdherenceand Uptake

BehavioralSocioeconomic and

Cultural Factors

Pharmacokinetics Absorption Metabolism Drug Interactions

Systemic and Intracellular

Concentration

Increased Immune ActivationImmunologic DeclineDisease ProgressionIncreased TransmissionPoor QOL and High Mortality

Ongoing Viral Replication

Viral ReplicationCapacity, Virulence

and Resistance

Host Immune andIntrinsic Factors

Inhibition of Viral Replication

Decreased Immune ActivationImmune ReconstitutionArrested Disease ProgressionDecreased TransmissionImproved QOL and Survival

Determinants of ART Response

Nachega/Marconi IDDT 2011

Adapted fromMunoz 1996

Socioeconomic, Cultural and Psychological Determinants of Health

Patient

Social Ecological ModelBronfenbrenner 1979

Behavior Paradigm

Ordonez JAR 2012

Barriers to Clinical Care• Poverty/Economic

– Transportation– Food Insecurity– Disability Grants– Poor social support

• Institutional– Long wait times– Negative staff

experiences– Poor health literacy– Limited substance abuse

treatment and mental health facilities

• Sociocultural– Perceived stigmatization– Influence of charismatic

churches– Traditional healers– Gender Inequalities

• Political– Migration– Controversy over

provision of HIV Tx– Unfavorable policies

Kagee J Health Pscyhol, GlobalPublic Health 2010 Western Cape

Barriers to Adherence

• Barriers to Care• Symptoms/QOL• Psychosocial

Peltzer BMC Public Health 2010Bhat Euro J Clin Microb ID 2010Maqutu AIDS Beh 2010Sarna Pub Health Rep 2010Coetzee AIDS Beh 2013

Tired of taking ARVsFear of taking ARVs in front of othersDifficulty swallowingRemembering to take pillsSide effectsCost of meds

MSFColors

What More Do You Want to Know?

• Multiple partners• Lives alone• >60 min from clinic• Taxi driver

• Spouse deceased• 3 children• Lives <30 min from clinic• Domestic worker

Concerns??

4. Which socioeconomic factor increases risk of VF?

a) Partner Status

b) Low Income

c) Long Distance to Clinic

d) Type of Employment

Economic

• > 50% SA live in poverty (HSRC 2004)– 10% living in informal settlements; 40% with extended family; Median household size 4.5 people– Income decline associated with VF in Uganda (Alsan CROI 2011)

• >40% food insecurity (Rose Pub Health Nutr 2002)• Unemployment 25-42% (Kingdon 2004); 80% high school only, 10% middle school• Individuals may trade health for disability grant (Ojikutu JID 2007)• 72% of poor live in rural areas and need to travel long distances to district hospitals (ART rollout sites)

Alsan

Masculinity vs. Dependency

• For men, automobile ownership was a risk factor for VF

• For women, financial insecurity was a risk factor for VF– Unemployment– Non-spouse family paying for

care (employer)– Staying with family other than

spouseHare AIDS Beh 2014

Anna Hare, MD

Claudia Ordonez, MA

Neighborhood Impact

• Neighborhood SES effect independent of individual SES• Implies contextual (not only compositional) effects such as

geography/transportation and culture

Daniella Coker, MPH

Institutional

Vella JAIDS 2010

Braitstein

• Vella JAIDS 2010– Number of new patients per year– Staff training and time commitment– Patient to staff ratio (NS)– Secure area (early on)– Confidentiality

• Available services - especially substance abuse and social (Ncama Int J Nurs Stud 2008)

• Dedicated staff, outreach vehicles, contact <30 d after missed visit (Braitstein JIAS 2012)

• Task shifting, Down referral, Decentralized care*

• Fast Tracking (Geng CROI 2011)• Adherence clubs (Luque CROI 2012)

*Sanne Lancet 2010, Brennan AIDS 2011, Long PLoS Med 2011, Matovu JAIDS 2013, STRETCH NIM-ART BMC 2013Kredo Cochrane 2013

Political• Beliefs about HIV/AIDS• Controversies over provision of HIV treatment• Migration (intra- and inter-national)• “Weak Rights” to the system - access basic services,

housing, health services and employment (Balbo 2005)

• Unequal distribution of healthcare expenditure infrastructural and personnel deficits in public sector– Private (20% of popn) > public (80%) spending by

7x per capita (Goudge 1999)– Lack of comprehensive and integrated care (Jack

JAIDS 2004)

MSFColors

Anything Else?

• Regular Unprotected Intercourse• Frequent alcohol use• Distrustful of healthcare• Feels stigmatized by friends• Sees TH for “low energy”

• Experiences violence at home• No longer attends Church• Depressed• Had negative clinic experiences• Trouble concentrating

Concerns??

5. Which psychosocial factor(s) increases risk of VF?

a) Unsafe sex

b) Intimate Partner Violence

c) Stigma

d) Depression

e) Dementia

f) Poor Clinic Experience

g) All of the above

Sociocultural

• Social marginalization leads to poor retention in care (Goudge SAHARA J 2009)

• Ability to resist stigma and other barriers – impact of Social Capital (Ware PLoS Med 2009, Young HPTN 043 JAIDS 2010, Achieng CROI 2011)– negative attitudes/beliefs about PLHIV (feelings of disgust, blame)– negative perceptions about PLHIV (discrimination)– perceptions of fair treatment for PLHIV (equity)

• Traditional Healers incorporated in ARV programs (Shuster J Comm Health 2009)

Traditional African Medicine• WHO (2008) est 80% Africans use TAM• Babb (2007) 84% TAM use > 1x for HIV;

32% current use• Dahab & Reid (2008) adherence barrier,

under-reporting• Sutherlandia, St. John’s Wort, garlic, and

American ginseng CYP 450 interactions (Mills 2005, Lee 2006, Izzo 2009)

• Potential toxicities (Hsiao 2003)• SARCS and RFVF Study*

– 50-80% have prior to enrollment at SKT– 5-20% have some TAM involvement after ART

initiation– No relationship to drug resistance, virologic failure

or clinical events

Marconi CID 2008Murphy AIDS 2010Sunpath AIDS 2012Marconi AIDS Pt Care STDs 2013 Appelbaum GPH 2014

*

Psychosocial• Number of people in social support network correlated with

adherence (Ncama Int J Nurses Stud 2008)• Relationship factors and treatment supporters enhance

adherence (Nachega JAIDS 2006)• Intimate Partner Violence/Abuse (Dunkle, Jewkes, Pronyk)• Depression (Peltzer BMC Public Health 2010)• Dementia (Joska AIDS Beh 2010) – 42.4% mild neurocognitive

disorder and 25.4% HIV-D in Cape Town starting ART• Stigma/Disclosure (Lyimo BMC Pub Health 2012)• Alcohol Misuse and a partner with HIV (Naidoo BMC

Public Health 2013)

PsychoSocial

Stigma, Faith and Depression

• For Men– Having >1 HIV-positive partner or family

member (OR=2.44, 95% CI 1.01-5.90)– Having >1 family member who died of HIV

(OR=2.98, 95% CI 1.29-6.91)– Disclosing HIV status to friends (OR=3.67,

95% CI 1.46-9.23)• For Women

– Not actively practicing their faith (OR=1.75, 95% CI 1.00-3.06)

– Depression (OR=2.42, 95% CI 1.23-4.77)

Rachel Kearns, MPH

Sally John, PhD

Risk Factors for Virological Failure Study

Henry Sunpath, MD, MPH

Questions

• Who• What• Why• How

− is at risk?− are the barriers?− do these barriers exist?− can we reduce the risk?

Methods• Patients had to be >18 yo and on >5 months of their first

ART regimen (substitutions allowed for toxicity)• Unmatched case-control study

– 158 Cases: VL > 1000 cpm– 300 Controls: (2:1) virologic suppression (VL < 1000 cpm)

• McCord Hospital• Eligible patients were enrolled between October 2010

and June 2012

Marconi AIDS Pt Care STDs 2014

Methods• Data Collection:

– Semi-structured interview in preferred language, coordinator blinded to case/control status• Questionnaire – demographic, socioeconomic (including a wealth

index, employment, education and cohabitants), psychological (including substance abuse, food insecurity, traditional medicine use, safe sex practices, faith, stigma and intimate partner violence), modified ACTG adherence questionnaire, and clinic satisfaction indices

• Kessler 10• Neurocognitive assessment and Pill count

– Study physician history/physical• Symptom screen• Karnofsky score• Clinical information, pharmacy refills and laboratory data from the chart

Marconi AIDS Pt Care STDs 2014

Methods

• Statistical Analysis:– Access was calculated using the medication

possession ratio (MPR)– Adherence was calculated using unannounced pill

counts and expected pill count from the pharm refills– Multivariate model selection was performed by

domain; significant variables were carried over to final models

• Model 1 – Baseline variables• Model 2 – Complete model without Adherence or Access• Model 3 – Complete model with Adherence and Access

Marconi AIDS Pt Care STDs 2014

Participant CharacteristicsCharacteristic Control

(300)Case(158)

P value

Age at enrollment (mean)Gender (%female)

40.971.0

37.152.5

<0.00010.0001

Tuberculosis (%yes)Lipodystrophy (%yes)Recent CD4 count in cells/µL (median)Recent CD4 count (%>350 cells/µL)

54.737.0359.052.0

55.115.2206.022.8

1.00<0.0001<0.0001<0.0001

Mean ART Duration (months)Current ART regimen contains Stavudine (d4T) Zidovudine (ZDV) Other (tenofovir, didanosine, abacavir)Fluconazole use in the past 6 months (%yes)TS use in the past 6 months (%yes)INH or RIF use in the past 6 months (%yes)ETB use in the past 6 months (%yes)

33.0 17.324.758.01.044.79.31.3

24.7 27.815.257.08.963.921.55.7

<0.0001 0.0077  <0.00010.00010.00050.014

Marconi AIDS Pt Care STDs 2014

Employment Status

Employed Full TimeEmployed Part TimeWorking at HomeUnemployed Seeking WorkUnemployed Not Seek-ing Work

80% report having some source of income30% receive some income from family membersMedian number of individuals supported by patient’s income: 3.5

F

atigu

e

H

eada

che

S

wallow

ing d

ifficu

lty

B

ack

joint

pain

S

leep

distu

rban

ces

F

ever

N

ause

a

D

iarrh

ea

A

bdom

inal P

ain

N

one

0%

5%

10%

15%

20%

25%

30%

Symptoms in the past 4 weeks

10% Feel symptoms are ARV related20% Feel symptoms are a barrier to taking ARVs

Probability of VF by Access or AdherenceProbability of Virologic Failure

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

PCAR and MPR

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4

group MPR PCAR

Prob

abili

ty o

f VF

Probability of Virologic Failure

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

PCAR and MPR

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4

group MPR PCAR

Pill Count Adherence (PCAR) or Medication Possession Ratio (MPR)

Wu CHIVR 2014

Peng Wu, MPH

Brent Johnson, PhD

Domain/Risk Factor Model 1 Model 2 Model 3Demographic Age (per 5 year increase) Gender (male)

 0.956**1.995**

 0.837†2.262**

 0.860

2.416**Socioeconomic Education (per 1 year) Transportation (personal) Pay for care (family/spouse)

1.771†1.517

 1.1121.789

 1.1082.034

Psychosocial Faith activity (none) Practice safe sex (<always) Family HIV+ (none) Treatment supporter (yes) Clinic feel pleased (yes) Depression (12+)

 1.634*

---0.620*1.991*

------

 1.722†

5.500***0.593†1.9100.448*

3.136***

 1.802*5.023**0.500*1.7830.509*3.021**

Symptoms and Exam Fatigue Diarrhea Sadness Skin lesions

---  2.532**2.555*

 1.720†

 2.470**2.0791.4091.992*

Medical History Lipodystrophy (yes) Log CD4 (per 1.0 increase)

 ------

 0.428*

0.079***

 0.608

0.078***Medications ARV duration (per 1 month) Recommend HIV clinic Friend vs Family Other vs Family Provider vs Family First Clinic (SKT) ARV training sessions (3+) Adherence counseling 2-4 vs 0-1 5+ vs 0-1 Current Regimen ZDV vs d4T Other vs d4T Recall ARVs (TV/radio) Trimethoprim/Sulfa (yes) Fluconazole (yes) Ethambutol (yes)

 0.995

 0.424*0.446*0.879*0.503†0.350†

---   

0.619*0.489*

---1.625†4.973*2.729

 1.001

 0.311*0.376*0.760*0.440†

0.3700.416

 0.649†0.455†3.519**0.6242.6362.800

 1.008

 0.266*0.397*0.855*

0.3780.419

 0.691†0.435†3.681**

 3.0063.025

Access (0.1) --- ---Adherence (0.1) --- --- 0.763*

ROC Curves for Each MV Model

3

AUC = 0.8881

1

AUC = 0.7824

2

AUC = 0.8867

Marconi AIDS Pt Care STDs 2014

Baseline (While Initiating or Suppressed on ART)

On ART Without Access/Adherence Measures*

On ART With Access/Adherence Measures*

Age

Gender

Faith

Family Member HIV+

Treatment Supporter

Clinic Recommendation

Current Regimen

Fluconazole Use

Depression

Unsafe sex practices

Clinic Experience

Fatigue

Diarrhea

Lipodystrophy

Current CD4 count

ARV Reminders

Depression

Unsafe sex practices

Clinic Experience

Fatigue

Rash

Current CD4 count

ARV Reminders

Adherence

*These factors do not include those that were identified as baseline risk factors.

Proposed Individual-Level EWI

Marconi AIDS Pt Care STDs 2014

Marconi AIDS Pt Care STDs 2014

Institutional, Community and Societal Factors

Access

VL

Adherence

SocioeconomicsComorbid Ill

ness

Psychoso

cial

Medications

“The lines are too long”

“I forget to take my pills”

“I miss appointments because the clinic is too

far to travel”

“I miss appointments because the clinic is

crowded”

“I do not take my pills if I have to take it in front of

others”

“I do not like to take my pills as they make me feel

sick”

“My pastor says I should not take

ARVs”

“I feel too tired to go to the clinic”

Future Directions• Validate these measures in peri-urban and

rural settings• Determine role of minority resistance• Identify impact of drug concentrations• Create a risk calculator

R01 AI098558

KZN HIV Drug Resistance Surveillance Study

Summary• VF and HIVDR are growing global concerns• Population-level EWI are useful for program evaluation but

lack specificity and timeliness for individual patient care• Individual-level EWI at initiation and follow-up assists

patient risk stratification as well as enable targeted and tailored interventions to be employed

• Consider all aspects of the treatment paradigm with a key focus on those impacting adherence

• Pharmacy refills and pill counts are insufficient to predict VF• Important to focus on both structural (institutional and

economic) as well as psychosocial factors when designing interventions for patients

• Need to externally validate model in other settings (rural and peri-urban) and include pharmacokinetics

VF IS ANEMERGENCY

W/ OR W/ORESISTANCE

AcknowledgmentsMcCord Hospital• Sabelo Dladla• Jane Hampton• Helga Holst• Sally John• Roma Maharaj• Phacia Ngubane• Claudia Ordonez• Melisha Pertab• Sifiso Shange• Henry Sunpath

UKZN/DDMRI/RKK/Bethesda• Jaysingh Brijkumar• Kelly Gate• Michelle Gordon• Yunus Moosa• Selvan Pillay

Emory University• Hannah Appelbaum• Daniella Coker• Jonathan Colasanti• Carlos del Rio• Anna Hare• Monique Hennink• Rachel Kearns• Baohua Wu• Peng Wu

Harvard/URMSF/JHU• Brent Johnson• Daniel Kuritzkes• Zhigang Lu• Richard Murphy• Jean Nachega

SupportNIH/NIAID R01 AI098558Emory University CFAR

Harvard University CFARBayer Diagnostics

Gilead Pharmaceuticals

Special Thanks to the staff and patients of Sinikithemba and iThemba Clinics…

…and my family for forbearance.