Cascade Dissemination v7BATYAslides

Two Complementary Approaches to
the HIV Care Cascade
ICAP Data Dissemination Meeting
Matthew Lamb & Molly McNairy
March 13, 2014
NUMBER OF SELECTED ACTIVITIES AT SUPPORTED FACILITIES,
JULY-SEPTEMBER, 2013
3500
Number of facilities
3000
3062
2500
2516
2000
1500
1000
1150
1279
1153
895
500
0
1495
519
Total
distinct
facilities
ART
Care
Source: ICAP Site Census, November 2013
Note: Some facilities offer more than one activity.
PMTCT
HIV testing
TB
for TB
screening
patients
for HIV
patients
Lab
662
48
Early Infant HIV Testing
Diagnosis
and
Counseling
291
PwP
HRH
CUMULATIVE ENROLLMENT IN HIV CARE AND
TREATMENT, BY QUARTER
2,000,000
As of September 2013:
1,800,000
1,787,682 enrolled in
care
971,893 initiated ART
1400
1200
1,400,000
1,474 clinics
1,200,000
1000
1,000,000
800
800,000
600
600,000
400
400,000
200
200,000
0
Sep-04
0
Jun-05
Mar-06 Dec-06 Sep-07
HIV care
Jun-08
Mar-09 Dec-09 Sep-10
ART
Source: ICAP URS November 2013.
Note: Includes adults and children. Data are from all clinics ever supported.
Jun-11
Clinics ever reporting
Mar-12 Dec-12 Sep-13
Number of clinics
Number of patients
1,600,000
1600
CUMULATIVE PEDIATRIC ENROLLMENT
IN HIV CARE AND TREATMENT, BY QUARTER
180,000
160,000
159,734 enrolled in
care
86,010 initiated ART
120,000
1400
1200
1,474 clinics
1000
100,000
800
80,000
600
60,000
400
40,000
200
20,000
0
Sep-04
0
Jun-05
Mar-06 Dec-06 Sep-07
HIV care
Source: ICAP URS November 2013.
Note: Data are from clinics ever supported.
Jun-08
Mar-09 Dec-09 Sep-10
ART
Jun-11
Clinics ever reporting
Mar-12 Dec-12 Sep-13
Number of clinics
Number of patients
140,000
1600
As of September 2013:
MEDIAN CD4 COUNT AT BASELINE, 6 AND 12 MONTHS AFTER
ART INITIATION, BY COUNTRY, JULY-SEPTEMBER, 2013
400
331
326
300
343
287
266
249
CD4 Cell Count (cells/ul)
241
228
185
200
209
185
180
148
138
116
100
0
170 166
Overall
n=4902
Kenya
n=1423
Mozambique
n=359
Baseline
6-month
Ethiopia
n=1801
Tanzania
n=1168
12-month
Source: ICAP URS November 2013
Notes: This reflects cumulative cohort data for patients followed 12 months. Baseline CD4 count is an average of 6 and 12
month cohort baseline data. Data are only from clinics currently supported and reporting.
98
Cote d'Ivoire
n=151
Number of pregnant women attending ANC and receiving HIV
testing, April 2007 – September 2013
5,000,000
4,500,000
4,000,000
As of September 2013:
4,660,865 attending
ANC
4,067,373 HIV tested
3000
2500
3,282 clinics
3,000,000
2000
2,500,000
1500
2,000,000
1,500,000
1000
1,000,000
500
500,000
0
Jun-07
0
Dec-07
Jun-08
Dec-08
Jun-09
ANC visit
Source: ICAP URS November 2013.
Notes: Data from clinics ever supported.
Dec-09
Jun-10
HIV tested
Dec-10
Jun-11
Dec-11
Clinics ever reporting
Jun-12
Dec-12
Jun-13
Number of clinics
Number of women
3,500,000
3500
TYPE OF ART REGIMEN FOR PMTCT, BY QUARTER
Percent of women
100%
4% 2% 4% 3% 6% 5%
12% 10% 11% 13% 8%
18%
12% 5% 6% 6%
22%
27%
21% 20% 22%
31% 34%
25%
6% 16% 26%
10% 9%
7%
10%
8%
6%
11%
13%
9%
6%
11%
20%
7%
6%
41%
50%
62% 59%
67%
75%
81% 83%
75% 74% 71%
67%
61% 64% 60% 63%
54%
32%
0%
Dec-09 Mar-10 Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Sep-13
Multi-drug regimens
Newly initiating ART
Already on ART
Source: ICAP URS November 2013
Notes: Multi-drug regimens include AZT/sd-NVP, AZT/3TC tail postpartum, HAART at 34 weeks.
Data are only from clinics currently supported and reporting.
SD-NVP
Number of new HIV patients screened for TB,
April 2007-September 2013
1,000,000
As of September 2013:
900,000
1600
949,693 enrolled in
care
568,917 TB screened
800,000
700,000
1400
1200
1,566 clinics
600,000
1000
500,000
800
400,000
600
300,000
400
200,000
200
100,000
0
Mar-07
Mar-08
New care enrollment
Source: ICAP URS November 2013.
Notes: Data are from clinics that ever reported.
Mar-09
Mar-10
Mar-11
TB screening at enrollment
Mar-12
Mar-13
Clinics ever reporting
0
Number of clinics
Number of patients
1800
Number of people tested for HIV,
as of September 2013
11,000,000
10,000,000
As of September 2013
9,000,000
9,536,893 HIV tested
8,000,000
476,472 HIV+
1,500
1,936 facilities
6,000,000
1,000
5,000,000
4,000,000
3,000,000
500
2,000,000
1,000,000
0
-
Mar-07
Dec-07
Sep-08
Jun-09
Mar-10
Tested and received results
Source: ICAP URS November 2013.
Notes: Data are only from clinics currently supported and reporting.
Dec-10
HIV+
Sep-11
Jun-12
Mar-13
Facilities ever reporting
Number of clinics
Number of people
7,000,000
2,000
PROPORTION OF PATIENTS TESTED WHO WERE HIV+ AND
ENROLLED IN CARE, BY POINT OF SERVICE,
JULY-SEPTEMBER 2013
60%
51%
Proportion of patients
50%
40%
28%
28%
30%
24%
20%
20%
13%
10%
10%
3%
2%
0.8%
0%
VCT*
n=52611
Other**
n=65108
Inpatient
n=35119
HIV+
Outpatient
n=454103
Outreach
n=2219
Enrolled in HIV care
Source: ICAP URS November 2013
Notes: *Most data from VCT clinics comes from Ethiopia, which may decrease the overall prevalence found in VCT clinics. **Other includes: laboratory; family
planning, well-baby, STI, and TB clinics; casualty/trauma; palliative care/hospice; customer care, cervical cancer screening; HIV care / ART (family testing); blood
donation center; and other points of service.
2,500,000
Number of tests conducted, by type of test,
as of September 30, 2013
As of September 2013: 1,334,098 CD4 count
606 laboratories
454,868 CD4 %
2,011,452 HIV rapid
tests
494,055 AFB
800
700
600
Number of laboratories
Number of tests
2,000,000
900
1,500,000
500
400
1,000,000
300
200
500,000
100
0
Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Sep-11 Dec-11 Mar-12 Jun-12 Sep-12 Dec-12 Mar-13 Jun-13 Sep-13
HIV rapid tests
Source: ICAP URS November 2013.
Notes: Data are from labs ever supported.
CD4 count
CD4%
AFB
Clinical labs ever reporting
0
Upcoming SI Webinars
March 27:
Care and treatment and lab PFacTS
Other:
URS, logic models, DHIS and surveillance
Overview
• Background
– “HIV continuum of care”
– How the continuum of care is typically measured
• Measure the continuum using ICAP data
1. Traditional HIV Care Cascade
2. Alternative HIV Care Cascade
• Use of both cascade approaches to inform a
comprehensive assessment of patient
outcomes and program performance
HIV Continuum of Care
Link
ART
Eligible
• A theoretical framework outlining essentials components of
HIV care and treatment
• Failure in any one step results in overall system failure:
morbidity, mortality and HIV transmission
McNairy, El-Sadr AIDS 2012
Why do we care about the continuum?
Patients in MILLIONS
2
1
1 million
0.5
HIV Care
ART
1.8 million
1.5
0
2005 2006
2007 2008
2009 2010
2011 2012
2013
ICAP Care and Treatment Programs (2005-2013)
Background: Traditional Cascade
The traditional HIV care cascade is a tool to map the HIV continuum of
care using a series of steps
– Each step is conditional on the previous step
16
– Focus on treatment (ART) with common endpoint of viral load suppression
Literature: Cascades, Cascades, Cascades
Gardner et al, CID 2011, CDC MMWR 2012, Kaygamba et al, Plos One 2012, Mugglin et al, Trop Med 2012
Background-2
• The traditional cascade provides incredibly useful information.
• However, incomplete insight as it does not account for:
– Outcomes for patients not on ART (Pre-ART)
– The multiplicity of reasons for losses between steps
– Timeframes for achievement of each step of the cascade
• HIV programs expanding to include Pre-ART patients and
there is a need to evolve the traditional cascade to include
outcomes of these patients
18
Purpose of Study
• To measure the traditional HIV care cascade among
adult HIV-infected patients enrolled in ICAP-supported
HIV care and treatment facilities in sub-Saharan Africa
• To develop an alternative cascade approach to:
– Account for outcomes of all patients (Pre-ART+ART)
– Identify reasons for losses across each step of cascade
– Measure outcomes over time (i.e., 3, 6, 12 months)
19
Study Population and Setting
• Routinely collected data on
390,603 adults (>15 years)
enrolled in HIV care at 217
ICAP-supported care and
treatment sites in 4 countries
from 2005-2011
• Participants in The Identifying
Optimal Models for HIV Care in
Africa Study
Kenya:
85 clinics
101,938 patients
Rwanda:
41 clinics
32,105 patients
Tanzania:
57 clinics
55,057 patients
Mozambique:
34 clinics
201,503 patients
20
Methods
Steps in this study’s cascade were adapted for available data
Traditional Cascade
Available Study Data
HIV infected
Not available
HIV diagnosed
Not available
Linked to care
• Enrolled in care
Retained in care
• Assessed for ART eligibility (CD4/WHO Stage)
On ART
• ART eligible using national guidelines 2005-2011
• Initiated ART
Suppressed viral load • ART retention at 12 months
21
Methods-2
Population
•
Pre-ART: any patient not on ART (ART-ineligible, unknown eligibility,
ART-eligible not initiated)
•
ART: any patient with documented initiation of ART
Outcomes
•
Loss to follow-up (LTF)
 ART: not attending a care visit > 6 months
 Pre-ART: not attending care visit > 12 months
•
•
Death and transfers based on data in patient records
Retention: Percent known to be alive, not LTF, and attending clinic
Methods
•
Kaplan-Meier to estimate cumulative incidence of retention after
ART initiation
22
Results
Traditional Cascade
400,000
Number of adult HIV patients
390,603
89%
23%
345,839
300,000
200,000
48%
167,523
100,000
70%
70%
117,525
78%
91,211
Assessment ART eligible Initiated ART 12m
ART Retention
Retention
Enrolling in HIV CD4/WHO
ART 12
atafter
12 months
Care
stagefor
within
ART
eligibility
months,
by
initiation
eligibility
N = 390,603 patients at 217
clinics in Kenya, Mozambique, Tanzania, and Rwanda
24
determination
Traditional Cascade Findings
400,000
Number of adult HIV patients
390,603
345,839
300,000
Outcomes of these patients
not traditionally included
200,000
167,523
100,000
70%
117,525
91,211
-
ART Retention
Enrolling in HIV Assessment
CD4/WHO
ART eligible Initiated ART 12m
Retention
atafter
12 months
forwithin
ART 12
Care
stage
ART
eligibility
months,
by
initiation
eligibility
N = 390,603 patients at 217
clinics in Kenya, Mozambique, Tanzania, and Rwanda
25
determination
Who is lost: after enrollment?
Number of adult HIV patients
400,000
390,603
345,839
(89%)
300,000
89% LTF
3% Death (recorded)
4% Transfer
3% Pre-ART retention
200,000
167,523
(43%)
100,000
117,525
(30%)
91,211
(23%)
-
Enrolling in
HIV Care
Assessment
CD4/WHO
ART eligible
CD4/WHO
for within
ART
stage
within
stage
eligibility
1212
months,
monthsby
eligibility
determination
ARTRetention
Retention
Initiated ART 12m
atafter
12 months
ART
initiation
Who is lost: after assessment for ART eligibility?
Number of adult HIV patients
400,000
390,603
345,839
(89%)
300,000
200,000
48% LTF
2% Death
26% ART initiation
8% Transfer
16% Pre-ART retention
167,523
(43%)
100,000
117,525
(30%)
91,211
(23%)
-
Enrolling in
HIV Care
Assessment
CD4/WHO
ART eligible
CD4/WHO
for within
ART
stage
within
stage
eligibility
1212
months,
monthsby
eligibility
determination
ARTRetention
Retention
Initiated ART 12m
atafter
12 months
ART
initiation
Traditional Cascade Findings
400,000
Number of adult HIV patients
390,603
89%
345,839
300,000
200,000
48%
167,523
100,000
70%
70%
117,525
78%
91,211
-
ART Retention
Enrolling in HIV Assessment
CD4/WHO
ART eligible Initiated ART 12m
Retention
atafter
12 months
forwithin
ART 12
Care
stage
ART
eligibility
months,
by
initiation
eligibility
N = 390,603 patients at 217
clinics in Kenya, Mozambique, Tanzania, and Rwanda
28
determination
Traditional Cascade By Country
100%
All countries end
up at similar
endpoint
(22%-26%)
80%
60%
40%
20%
0%
Enrolling in HIV Assessment for
Care
ART eligibility
Overall
Rwanda
ART eligible
Tanzania
ARTretention
retentionat
Initiated ART Art
at12
12months
months
Kenya
Mozambique
Traditional Cascade
What it can tell us?
• % of patients achieving each
step
• % of patients lost across
each step
• Which step is the “weakest”
link in the cascade
• Use of common desired
cascade endpoint (VLS, ART
retention) which is essential
for optimized patient
outcomes
What is missing?
• Outcomes of all patients
(Pre-ART, transfers)
• Reasons for losses across
each step
• Timeline
Alternative Cascade Approach
• All patients followed across time
– Each step begins not from previous step but from enrollment
– Outcomes of the entire cohort are described
• Outcomes are categorized
Optimal
Suboptimal
Poor
Retained, Transferred
Retained but did not receive optimal
care, missing data
LTF, Death
Alternative Cascade
100%
390,603
% of adult HIV patients
Optimal Outcomes
80%
Retained ART-ineligible
58%
60%
Initiated ART
51%
49%
Retained on ART
40%
Transfer out
20%
0%
Enrollment
3 months
6 months
12 months
Time (months) since HIV care enrollment
Alternative Cascade
100%
390,603
% of adult HIV patients
Suboptimal Outcomes
80%
Retained ART-eligible but not
initiated
Retained indeterminate
eligibility
60%
40%
20%
0%
Enrollment
3 months
6 months
12 months
Time (months) since HIV care enrollment
Alternative Cascade
% of adult HIV patients
100%
390,603
80%
Poor Outcomes
LTF (Pre-ART)
Death (Pre-ART)
60%
LTF (ART)
40%
Death (ART)
20%
0%
0 months
3 months
6 months
12 months
Enrollment
Time (months) since HIV care enrollment
Alternative Cascade Summary
100%
Optimal
% of adult HIV patients
20%
80%
60%
Suboptimal
37%
100%
44%
Poor
22%
12%
7%
40%
58%
51%
49%
20%
0%
0 months
3 months
6 months
12 months
Enrollment
Time (months) since HIV care enrollment
56% of
patients
retained
Alternative Cascade: Mozambique
100%
Optimal
% of adult HIV patients
23%
80%
Suboptimal
44%
52%
Poor
25%
60%
12%
100%
8%
40%
53%
20%
44%
41%
6 months
12 months
0%
Enrollment
3 months
Time (months) since HIV care enrollment
201,503 patients
(52% of all patients)
34 Clinics
(16% of all clinics)
100%
Alternative Cascade: Tanzania
Optimal
% of adult HIV patients
19%
35%
80%
Suboptimal
41%
20%
60%
12%
Poor
8%
100%
40%
61%
54%
55,057 patients
(14% of all patients)
51%
20%
57 Clinics
(26% of all clinics)
0%
Enrollment
3 months
6 months
12 months
Time (months) since HIV care enrollment
Alternative Cascade: Kenya
Optimal
100%
% of adult HIV patients
21%
Suboptimal
33%
80%
40%
Poor
21%
13%
60%
8%
100%
40%
101,938 patients
(26% of all patients)
58%
54%
52%
6 months
12 months
20%
0%
Enrollment
3 months
Time (months) since HIV care enrollment
85 Clinics
(39% of all clinics)
Alternative Cascade: Rwanda
% of adult HIV patients
100%
5%
8%
10%
5%
80%
Optimal
14%
3%
Suboptimal
Poor
60%
100%
87%
40%
84%
83%
32,105 patients
(8% of all patients)
41 Clinics
(19% of all clinics)
20%
0%
Enrollment
3 months
6 months
12 months
Time (months) since HIV care enrollment
Summary
Cascade Approach
Cascade Endpoint
Strengths
Traditional
• 23% (91,211)
achieve ART
retention at 12
months
• Measures drop-off
between steps
• Common desired
endpoint for ART
patients
Alternative
• 49% (190,122)
have optimal
outcomes at 12
months
• 56% (218,738)
retained in care
• Follows all patients
• Stratifies poor
outcomes into preART and ART
• Includes time
40
Summary by Country
Cascade Approach
Traditional
Alternative
Cascade Endpoint
Conclusion
•
•
•
•
Kenya: 24%
• Similar outcomes
Mozambique: 22%
across countries
Rwanda: 26%
Tanzania: 26%
•
•
•
•
Kenya: 52%
• Very different
Mozambique: 41%
outcomes across
Rwanda: 83%
countries
Tanzania: 51%
Conclusions
Complementary approaches
• Traditional cascade
– focuses on treatment with overarching programmatic
goal of continued successful retention and adherence
– Identifies modifiable bottlenecks in achieving that goal
• ART eligibility  ART initiation
• Alternative cascade
– Focuses on outcomes of all patients (Pre-ART and ART)
– Pre-ART losses are the majority of poor outcomes
– Shows large heterogeneity between country programs
Conclusions-2
• Neither study cascade includes the desired start point(s)
– HIV infection
– HIV diagnosis
– Linkage to HIV care
• The ultimate value of HIV care cascades is to inform the
design and implementation of multicomponent
interventions that achieve desired patient outcomes and
population impact
Applying to ICAP…next steps
• Utility for
– routine program monitoring and evaluation?
– assessing impact of interventions targeting key
steps in the continuum of HIV care?
– Other applications?
Questions?