Dr. Kathleen Brady (AACO)'s annual epidemiological update. This presentation was given to the Philadelphia EMA Ryan White Planning Council on Thursday, February 20, 2014.

1. I spy an epidemiologist
Kathleen A Brady, MD
Medical Director/Medical Epidemiologist
AIDS Activities Coordinating Office
Philadelphia Department of Public Health
February 20, 2014

2. The HIV Care Continuum

3. National and Local Engagement in Care
 Data
 National and local HIV Surveillance System
Prevalence (total, diagnosed) – number of persons living with HIV
 Linkage to care


Medical Monitoring Project (MMP)
Retention in care
 Prescribed ART
 Viral suppression


4. Methods
 Prevalence
 HIV diagnosis data
 Data adjustments at the national level
 Back-calculation methods to estimate unaware
 Linkage to Care
 Data reported through December 2012
 Percentage of persons with >1 CD4 or viral load test result
within 3 months of HIV diagnosis

5. Medical Monitoring Project
 MMP is a national probability sample of HIV-infected
persons receiving care in the US in order to:




describe HIV care and support services being received and the
quality of such services
describe the prevalence and occurrence of co-morbidities related to
HIV disease
determine prevalence of ongoing risk behaviors and access to and
use of prevention services among persons living with HIV
identify met and unmet needs for HIV care and prevention services
in order to inform community and care planning groups, health care
providers and other stakeholders
 Philadelphia has participated in MMP since 2005. All
charts of sampled patients are abstracted for clinical
information and patients are offered a voluntary
interview.

6. MMP Population Size Estimates
 States, facilities, and patients sampled with known
probabilities
 Analysis weights include:

Design weights
Inverse of the probability of selection
 Extend inference from sample to reference population

 Non-response adjustment
 Extend inference from respondents to sample
 Sum of weights estimates number of HIV-infected
adults who received at least one medical visit
January-April of a calendar year

7. MMP Definitions
 Retention in care: Number of HIV-infected adults
who received at least one medical care visit between
January and April of the calendar year
 Prescription of antiretroviral therapy (ART):
Documentation in medical record abstraction of any
ART prescription in the past 12 months
 Viral suppression: Documentation in medical
record abstraction of most

8. Philadelphia Engagement in Care, 2009-2010
25000
20000
15000
10000
5000
0
20541
19188
16844
15753
13745
11894
9105
8185
9944
8751
6319
5775
2009
2010

9. Philadelphia Engagement in Care, 2009-2010
120%
100%
80%
60%
40%
20%
0%
100%
82%
100%
76%
63%
54%
56%
49%
38%
37%
2009
2010

10. For every 100 people living with HIV:
Philadelphia
US
Number
Number
100
Diagnosed
100
Diagnosed
80
Are linked to HIV care
82
Are linked to HIV care
45
Stay in HIV care
54
Stay in HIV care
40
Get antiretroviral
therapy
49
Get antiretroviral
therapy
30
Have a very low amount
of virus in their body
38
Have a very low amount
of virus in their body
2010 Data

11. Who is Aware?

14. HIV Prevalence in Philadelphia
(reported thru 6/30/2013)
 19,832 PLWHA (aware)


11,954 AIDS cases
7,878 HIV cases
 Rates (known) vary by
race

 4,353 estimated to be

living with HIV and
unaware
 1.58% Philadelphia
residents estimated to be
HIV+

1.9% of blacks
1.5% of Latinos
0.7% of whites
 Rates vary by sex


2.0% of males
0.7% of females

15. HIV Prevalence in the Philadelphia EMA
(reported thru 6/30/2013)
 27,063 PLWHA (aware)


15,683 AIDS cases
11,380 HIV cases
 Rates (known) vary by
race

 5,941 estimated to be

living with HIV and
unaware
 0.5% Philadelphia EMA
residents estimated to be
HIV+

1.4% of blacks
0.9% of Latinos
0.2% of whites
 Rates vary by sex


0.8% of males
0.3% of females

16. HIV/AIDS Cases by Date of
Diagnosis
AIDS
1308 1302
1178
1177
1200
1200
1001
940 918 928
897
894 895 907 861
821
898
1000
712
800
756
734
652
600
676
750
453
400
226 244
179 198
200
Year
16
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
0
1993
Number of Cases
1400
HIV

17. HIV/AIDS Cases by Sex and Date of
Diagnosis
AIDS Male
HIV Female
HIV Male
Year
17
20
11
20
09
20
07
20
05
20
03
20
01
19
99
19
97
19
95
19
93
1600
1400
1200
1000
800
600
400
200
0
19
91
Number of Cases
AIDS Female

18. HIV Cases by Race/Ethnicity and
Date of Diagnosis
White
AfrAm
Hispanic
700
Number of Cases
600
575
597
536
517
482
500
474
107
106
400
300
200
100
0
2006
171
133
2007
143
127
211
147
2008
125
2009
2010
87
2011
Year
18
103
110
2012

19. HIV Cases by Mode of Transmission
and Date of Diagnosis
Number of Cases
MSM
500
450
400
350
300
250
200
150
100
50
0
2006
457
281
IDU
HetSx
NIR
447
305
334
304
251
325
331
316
314
279
150
127
195
103
75
13
27
28
2007
2008
2009
2010
50
9
2011
Year
19
65
27
2012

20. HIV Cases by Age and Date of
Diagnosis
13-19
20-29
350
40-49
50+
306
300
Number of Cases
30-39
251
250
237
232
266
256
224
150
155
100
128
68
197
194
200
233
212
65
181
134
45
148
155
157
170
50
0
2006
220
158
124
43
161
135
109
39
2007
2008
2009
2010
32
2011
2012
Year
20

23. Summary
 High HIV morbidity in Philadelphia
 Philadelphia epidemic predominantly affects




minority populations
MSM and Heterosexual transmission
predominant modes of transmission
Cases among MSM are increasing
Growing numbers of persons living with HIV and
AIDS
First recent year to see an increase in AIDS cases
23

24. Who is getting infected?

25. Incidence Surveillance
 Collect and STARHS test the diagnostic blood
specimens from all newly diagnosed HIV infections
reported from public and private laboratories and
providers to HIV Surveillance Unit.
 Collect the HIV testing information needed for the
statistical estimates of incidence.
 Calculate population-based estimates of HIV
incidence.
 Use these estimates to identify emerging subepidemics, monitor trends, target prevention
resources and interventions to areas and
populations most heavily affected, and evaluate
programs.

26. Incidence vs. Prevalence
1981
2006
2007
HIV Incidence = the number of individuals newly infected
with HIV within a given period of time (6 - 12
months).
1981
2006
2007
HIV Prevalence = the total number of HIV cases that exist at a
specific time within a specific population.

27. What is STARHS?
Serologic
Testing
Algorithm for
Recent
HIV
Seroconversion

28. Requirements for HIV Incidence
Surveillance
Remnant HIV+
Serum
STARHS
Testing
using BED
Assay
Supplemental Data
Includes:
•Race, sex, mode of
transmission
•Testing history &
reasons for testing
(Calculating weights)
•Any exclusionary
info (AIDS
diagnosis, prior
recent ART)
•Adjust for LFU, QNS
HIV Incidence Estimation

29. CDC STARHS Test Results
 (+) standard test and (+) STARHS test
= long-standing HIV infection
 (+) standard test and (-) STARHS test =
recent HIV infection

30. National Incidence Data
 Estimated 47,500 HIV infections in 2010 in adults




and adolescents (95% CI, 42,000 – 53,000)
Estimated incidence 19.0 infections per 100,000
population
44% among blacks, 21% Latinos
51% among MSM, 38% heterosexual
26% among 13-24 year olds

31. 2011 Local Estimate of
HIV Incidence
 Local estimate of 872 new HIV infections in 2011 in
adults and adolescents (95% CI, 575-1,169)
 Rate is over 3.5 times that of the US estimate

32. HIV Incidence Trends by Demographic Groups
1200
1000
800
Total
Age 13-24
600
Male
Black
400
MSM
200
0
2006
2007
2008
2009
2011

33. Estimated Incidence Rates - 2011
Population Population in
2010 (13 +)
ESTIMATED
Incidence
Estimate,
201
Estimated 95% CI
Case Rate lower
bound
per
100,000
95% CI
upper
bound
MSM
27,841
439
1,476
787
2,162
IDU
37,378
52
139.1
0
332
HET
254,200*
382
129.6
68
191
*Includes persons >13 living in poverty
Data Source: PDPH/AACO HIV Incidence Surveillance Program

34. Incidence Summary
 Includes people unaware of their status.
 Overall, HIV incidence in Philadelphia is stable
 Incidence higher than baseline 2006 data for MSM

35. Who is unaware?

36. Concurrent HIV/AIDS, 2012

37. Concurrent HIV/AIDS, 2012

38. Retention in care

39. Definition: Met Need for Primary Care
 Met Need for Primary Care defined as measurement
of at least one CD4 count and/or one Viral Load
and/or receipt of antiretroviral therapy during a
specified time period

40. Framework
 Input
 Population sizes of those with HIV and AIDS within the service
area
 Care Patterns of those with HIV and AIDS
 Calculated Result
 Number of persons with HIV and AIDS with unmet need

41. Population Sizes
Population Sizes
Value
Data Source(s)
Row A. Number of persons living
with AIDS (PLWA), for
the period of
12/31/2012 in the EMA
15,683
Local eHARS data
Row B. Number of persons living
with HIV (PLWH)/nonAIDS/aware, for the
period of 12/31/2012 in
the EMA
11,380
Local eHARS data
Row C. Total number of
HIV+/aware for the
period of 12/31/2012 in
the EMA
27,063
Local eHARS data

42. Care Patterns
Value
Data Source(s)
Surveillance Data
(Lab Data)
CAREWare
Row D.
Number of PLWA
who received the
specified HIV
primary medical
care during the 12month period of
2012 in the EMA
13,770
Row E.
Number of
PLWH/non-AIDS
who received the
specified HIV
primary medical
care during the 12month period of
2012 in the EMA
8,296
Surveillance Data
(Lab Data)
CAREWare

43. Row F.
Total number of
HIV+/aware who
received the
specified HIV primary
medical care during
the 12-month period
of 2012 in the EMA
22,066

44. Calculated Results
Value
Calculation
1,913
(12.0%)
=A–D
Row H. Number of PLWH/nonAIDS who did not receive
primary medical services
during the 12-month
period of 2011 in the
EMA
3,084
(27%)
=B–E
Row I.
4,997
(18%)
=G+H
Row G. Number of PLWA who
did not receive primary
medical services during
the 12-month period of
2011 in the EMA
Total of HIV+/aware not
receiving specified
primary medical care
services (quantified
estimate of unmet need
in the EMA

45. Met need by demographic groups
100.0%
100.0%
90.0%
80.0%
70.0%
60.0%
76.2%
72.0%
71.7%
86.2%
88.4%85.9%
91.5%
86.4%
90.0%
80.0%
70.0%
77.3%
71.0%
60.0%
50.0%
50.0%
40.0%
40.0%
30.0%
30.0%
20.0%
20.0%
10.0%
10.0%
0.0%
0.0%
HIV
Black
White
AIDS
Hispanic
HIV
Male
AIDS
Female

46. Met need by insurance status
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
77.5%
81.2%
69.3%
93.3%
88.7%
89.9%
67.9%
80.1%
73.7%
60.6%
HIV
Medicaid
Private
AIDS
Other public
Unknown
None

47. Disparities

49. Engagement in Care by Sex, 2010
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
100%
Male
80%
54% 50%
39%
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
100%
Female
86%
53%
45%
33%

51. Philadelphia Engagement in Care, 2010
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
100%
100%
100%
100%
91%
82%
86%
79%
54%
49%
46%
All
Diagnosed
White
Linked
47%
45%
30%
55%
53%
50%
Retained
52%
43%
33%
Black
on ART
Latino
Suppressed

53. Engagement in Care by Mode of
Transmission, 2010
6000
4000
4051
3156
1000
0
4210
3153
3000
2000
4901
4855
4248
5000
1414
1984
1722
1654
1465
1465
1240
935
MSM of Color
Diagnosed
White MSM
Linked
In Care
2586
3039
2170
2365
HET male
On ART
1582
HET female
Suppressed

54. Engagement in Care by Mode of
Transmission, 2010
120%
100%
80%
100%
100%
80%
20%
100%
88%
74%
60%
40%
100%
87%
83%
86%
65%
53%
63%
35%
44%
74%
49%
31%
23%
32%
0%
MSM of Color
Diagnosed
White MSM
Linked
In Care
HET male
On ART
HET female
Suppressed

56. Engagement in Care by Age Group, 2010
120%
100%
80%
60%
100%
100%
100%
78%
82%
81%
50%
44%
82%
56%
53%
43%
51%
46%
40%
20%
4%
0%
18-24
Diagnosed
25-49
Linked
In Care
50+
On ART
Suppressed

57. Engagement in Care Summary
 On ART
 Higher for males than females
 Higher for men who have sex with men (MSM) than for
women who have sex with men (WSM)
 Viral suppression
 Higher for males than females
 Higher for MSM than WSM
 Higher for whites compared to blacks and Hispanics
 Higher for those >50 compared to 18-29 year olds
All P values <0.05

58. Geographic Disparities
BEHIND THE CASCADE: ANALYZING SPATIAL
PATTERNS ALONG THE HIV CARE
CONTINUUM

59. Penn-AACO ECHPP Collaboration
 Complete basic and advanced Geographic Information System
(GIS) Training to staff of the AIDS Activities Coordinating Office
 Developed resources and ongoing collaboration on GIS in HIV
between Penn CFAR and AACO
 Established a permanent Core service to provide GIS support to
HIV investigators

60. Study objectives
Using GIS analytic strategies, we sought to identify areas associated with:
 not linking to care
 not linking to care within 90 days
 not retaining in care
 not achieving viral suppression after HIV diagnosis

61. Methods
 Retrospective cohort
 Data extracted from eHARS
 Inclusion/Exclusion criteria:
 New HIV diagnosis in 2008 and 2009
 Philadelphia address at the time of diagnosis
 Persons with an invalid address or with a prison address at the
time of their diagnosis were excluded

62. Outcomes
 Linkage to Care – Defined as documentation of >1 CD4 or viral load
test results after the diagnosis
 Linkage to Care in 90 days – Defined as documentation of >1 CD4
or viral load test results within 90 days of HIV diagnosis
 Retention in Care – Defined by NQF Medical Visit Frequency
Measure. completing at least 1 medical visit with a provider with
prescribing privileges in each 6-month interval of the 24-month
measurement period, with a minimum of 60 days between medical
visits.


Date of first linkage defined the start of the 24 month measurement period.
We used CD4 and/or viral load as a proxy for HIV medical care visits
 Viral Suppression – Defined as evidence of HIV-1 RNA <200 copies
closest to the end of the 24 month measurement period

63. Variables of Interest
 Age, sex at birth, race/ethnicity, HIV transmission
risk, insurance status at the time of
diagnosis, imprisonment, multiple care
providers, distance to nearest care site
 Spatial Analyses - K function





Analyze a spatial point process
Multiple distance scales
 e.g. clustered at small distances yet dispersed at large
distances
Complete spatial randomness (CSR)
Utilizes all points in a given area
Compare to multiple simulated random processes

64. Results
 1,861 cases, 157 excluded (8%) due to an invalid
address or imprisoned at the time of diagnosis

Excluded persons less likely to be black/Hispanic, more
likely to be >45 years of age, IDU and privately insured
 Among 1,704 person included:
 70% male, 63% black, 30% 45 years or older
 40% heterosexuals, 36% MSM
 82% linked to care
 Among those linked, 75% linked in 90 days and
37% were retained in care
 Among those retained, 72% achieved viral
suppression

65. Multivariate Regression Models for Involvement
in Continuum of Care
Characteristic
Not Linked
to Care
Not Linked
<90 Days
Age at Dx
Not
Retained in
Care
Not Virally
Suppressed
<25
Sex at birth
Male
Race/ ethnicity
Black
Risk Group
IDU
Insurance
Medicare
Uninsured
Uninsured
Yes
Yes
Geographic Area
Black
Hispanic
Yes
Yes
Prison stay
Proximity to care
Multiple care sites
Yes
Yes

66. Geographic Pattern Analysis of HIV Medical Care
Engagement, 2008-2009

67. Summary
 Geographic clustering was independently associated with poor outcomes at
each step along the HIV Care Continuum
 Geographic clusters identified were unique with no geographic overlap between
steps in the Continuum
 Geographic clusters identified have a greater burden of HIV disease compared
to other neighborhoods
 Proximity to HIV medical care was not associated with linkage to care, linkage
in <90 days or retention in care

68. Conclusions
 Community factors related to poverty and community
socioeconomic status may impact HIV treatment outcomes for
individuals in living in geographic clusters
 We hypothesize:



Community norms and social disorder may have a greater effect on
linkage to care;
Access to public transportation and social services may have a greater
effect on retention in care;
And access to pharmacies may have a greater effect on viral
suppression.
 Differences in community factors that influence each step of the
cascade may explain the lack of overlap in hot spots.

69. Next Steps
 Better understand of the characteristics of places that influence access to
HIV medical care and treatment outcomes—mixed methods strategies

Consistent with CDC’s High Impact Prevention program, identification of
geographic clusters could help to specifically target separate linkage,
retention, and adherence interventions in the areas identified with the
greatest need

Philadelphia’s CDC CoRECT application – selected medical providers in the
geographic cluster identified for retention
 Develop new strategies for intervention based upon ecological factors of
the distinct clusters

70. Starting Antiretroviral Therapy in 2012: A Compendium of Interactive Cases
clinicaloptions.com/hiv
What Will It Take to Substantially Reduce
HIV Transmission in an Entire Population?
Undiagnosed HIV
Not linked to care
Not retained in care
ART not required
ART not utilized
Viremic on ART
Undetectable
HIV-1 RNA
•Number of Individuals
•1,200,000
•1,000,000
•800,000
•600,000
•400,000
•200,000
•0
•66%
•19%
•22%
•Current
•DX
90%
•34%
•28%
•Engage
90%
•Treat
90%
•21%
•VL < 50
•Dx,
in 90% Engage, Tx,
and VL < 50
in 90%
•Answer: Treatment AND Prevention
•Gardner EM, et al. Clin Infect Dis. 2011;52:793-800.

71. Quality of Care in Patients Living
with HIV: Performance Measures
in HIV Clinics

72. Background
 As PLWH live longer, ensuring receipt of high
quality care has become increasingly important
 HRSA performance measures include:



HIV specific measures (clinic visits, CD4 and viral load
monitoring, ART, viral suppression, PCP prophylaxis)
Screenings for comorbid conditions( syphilis, GC, Chlamydia,
HBV, HCV, TB, hyperlipidemia, cervical cancer)
Vaccinations (influenza, HBV, pneumococcus
 Little known about the patient and clinic factors that
may contribute to success in meeting performance
measures

73. Methods
 Weighted abstraction data from the 2009 cycle of
MMP in Philadelphia were utilized


Data included for 376 participants (94% of sampled patients)
Facility attributes data included from 24 facilities
 Sociodemographic variables were defined according
to CDC criteria.
 Clinic variables included Patient/FTE ratio, type of
practice, and variables on adherence counseling, case
management were dichotomized into available or not
 Outcome variable – Receipt of HRSA performance
measures was based on HRSA definitions

74. Methods - Continued
 Outcome variable
 described as a continuous variable
 maximum score of 15, where each HRSA performance measure
chosen was given one point
 certain performance measures were excluded:
not well-documented across clinics (ex., oral exam, substance
abuse screening) or
 only have applied to fewer than 50% of patients (ex., cervical
cancer screening, PCP prophylaxis)


A total of 15 of the 27 HRSA performance measures were
included in the outcome variable

75. Results – Sample Characteristics
Characteristic
%
%
Birth outside US
Sex
Characteristic
4.3%
Male
64.6%
HIV risk category
Female
35.8%
Heterosexual
37.2%
MSM
30.9%
Race
White
21.3%
IDU
29.5%
Black
67.3%
Other/Unknown
2.4%
Hispanic
10.9%
Other
0.5%
Insurance status
23.6%
Medicaid
Age
Private
55.6%
18-29
13.6%
Medicare
12.2%
30-39
17.8%
VA
3.7%
40-49
37.8%
Uninsured
8.0%
>50
30.9%
Adherence program
89.4%
60.9%
Case management
59.8%
Ryan White funded
87.8%
AIDS Diagnosis

76. Results – Multivariate Model
Characteristic
Female (male ref.)
IRR (95% CI)
Adjusted IRR (95% CI)
1.01 (0.97-1.04)
0.99 (0.89-1.09)
1.02 (0.93-1.11)
1.04 (0.97-1.10)
0.98 (0.88-1.09)
1.04 (0.96-1.09)
1.00 (0.86-1.16)
1.09 (0.96-1.25)
1.13 (0.77-1.30)
1.19 (1.05-1.35)
1.12 (0.96-1.31)
1.17 (1.01-1.35)
1.02 (0.73-1.45)
1.05 (0.73-1.50)
Race/ethnicity (white ref.)
Black
Hispanic
Age (18-29 ref.)
30-39
40-49
50+
Non-US birth (US birth ref.)

77. Results – Multivariate Model
Characteristic
IRR (95% CI)
Adjusted IRR (95% CI)
1.02 (0.93-1.13)
1.05 (0.93-1.18)
0.99 (0.90-1.10)
1.00 (0.93-1.07)
0.97 (0.93-1.02)
0.98 (0.92-1.06)
0.93 (0.83-1.04)
0.97 (0.88-1.08)
0.98 (0.96-1.01)
1.01 (0.87-1.18)
0.73 (0.64-0.82)
0.91 (0.81-1.03)
HIV Risk Factor (Het ref.)
MSM
IDU
Insurance (Private ref.)
Medicaid
Medicare
VA
Uninsured

78. Results – Multivariate Model
Characteristic
Adherence counseling
Case management
Ryan White funded
IRR (95% CI)
Adjusted IRR (95% CI)
1.21 (1.13-1.28)
1.12 (1.04-1.21)
1.14 (1.03-1.25)
1.12 (1.04-1.14)
1.06 (0.96-1.17)
1.07 (0.92-1.25)
1.08 (1.00-1.17)
1.08 (0.98-1.19)
1.07 (1.02-1.13)
1.14 (1.02-1.28)
1.04 (1.03-1.25)
1.07 (0.92-1.25)
Minimum CD4 Count (<200
ref)
200-349
350-499
>500

79. Results Summary
 The mean number of performance measure met was
8.52 (standard deviation 2.39) of the 15 assessed.
 Older PLWH more likely to meet performance
measures
 HIV-specific measures more likely to be met than for
vaccination related measures and screenings for comorbid conditions
 Patients who attended clinics with adherence
counseling programs and that provided case
management had higher summed performance
scores

80. Limitations
 By using MMP data, we could not address how
clinics tried to improve performance measures
 Focus on one city, results may not be generalizable.
 MMP may underestimate preventative services
received by patients attending other clinics (such as
primary care clinics)
 The performance summary measure


Some measures excluded
All measure given equal weighting in the score but not
necessarily equally important (VL suppression > GC
screening)

81. Conclusions
 Few patients achieved all performance measures
 PLWH more likely to achieve performance measure
related to HIV care (data not shown)
 Future work should focus on how to improve
compliance with performance measures

Investigation of barriers and facilitator to improving care of
PLWH in HIV clinics
 Still need a better understanding of what other clinic
characteristics may be associated with meeting
performance measure metrics

82. The End
QUESTIONS?