pepfar data review...•monthly analysis of dhis data has been on-going ... in interventions are...
TRANSCRIPT
PEPFAR Data Review How PEPFAR is Looking at the Data?
What are the Tools?
Patrick Nadol
CDC/SA
Areas of (re-newed) focus for PEPFAR
Granularity:
Site, age, sex, population, trends
over time
Composite monitoring:
Linkage to Tx, Tx net new
Timeliness
TX_CURR gap: By District and Cumulative gap
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0%
2%
4%
6%
8%
10%
12%
14%
Cu
mu
lati
ve T
X_C
UR
R g
ap
% T
X_C
UR
R g
ap (
dis
tric
t ga
p/t
ota
l gap
)
%_TX_CURR_gap sum_TX_CURR_gap
80% of TX_CURR gap is located in 16/27 districts
50% of TX_CURR gap is located in 6/27 districts
Row Labels
CUMM
TARGETS:
FY2017
RESULTS: FY17-
Q1
RESULTS: FY17-
Q3
CUMM
RESULTS:
FY2017
%
ACHIEVEMENT
STATUS
HTS_TST_Pos+Neg 5,767,766 2,503,472 2,887,251 5,390,723 93%
HTS_TST_Pos 823,111 221,932 236,295 458,227 56%
TX_NEW 965,131 183,957 149,570 333,527 35%
TX_CURR 3,935,927 2,950,894 3,131,935 3,131,935 80%
TX_NET NEW 1,220,164 235,131 181,041 416,172 34%
Clin
ical
cascad
e
PMTCT_STAT 719,446 186,570 170,310 356,880 50%
PMTCT
PMTCT_STAT_Pos 391,670 47,729 50,997 98,726 25%
PMTCT_ART 155,489 45,894 46,120 92,014 59%
PMTCT_EID 154,992 66,172 74,226 140,398 91%
PMTCT_EID_POS_12MO 232 385 580 965 416%
PMTCT_EID_POS_2MO 1,488 317 336 653 44%
PrEP_NEW 1,279 304 698 1,002 78%
VMMC_CIRC 393,465 44,626 145,240 189,866 48%PREVENTION
PMTCT
Q3 Progress: PEPFAR Focus Districts (n=27)
4
Good case-detection with overall high yield
Sub-optimal TX_NEW & TX net new with improving ‘linkage’
post-UTT
Moderate acceleration of
VMMC
Timeliness
Timeliness
• Quarterly analysis of data is insufficient
• Beginning in Q3 (May-June data), PEPFAR is reporting its quarterly data more ‘real-time’ • Moved aware from reporting 1 quarter in arrears
• VMMC is reporting results weekly
• Efforts underway to have TX results reported weekly
• Monthly analysis of DHIS data has been on-going
• Rationale: • Updated NDoH policy supports more timely flow of data from facility to national
level (<30 days) • Allows for more immediate identification of progress & corrective action as
required • Changes in interventions are authorized
FY17 DATIM alignment - progress report
Reporting period Jul-16 Aug-16 Sept-16 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sept-17
Status on
May 15, 2017
Arrears
FY17-Q1
Arrears
FY17-Q2
Reporting period Jul-16 Aug-16 Sept-16 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sept-17
Status on
Aug 15, 2017
Aligned
FY17-Q1
Aligned
FY17-Q3
Reporting period Jul-16 Aug-16 Sept-16 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sept-17
Planned:
Sept 22, 2017
Aligned
FY17-Q1
Aligned
FY17-Q2
Aligned
FY17-Q3
Reporting period Jul-16 Aug-16 Sept-16 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sept-17
Planned:
Nov 15, 2017
Aligned
FY17-Q1
Aligned
FY17-Q2
Aligned
FY17-Q3
Aligned
FY17-Q4
We are
here
Monthly review: TX CURR by district, facility
-
50 000
100 000
150 000
200 000
250 000
300 000
350 000
400 000
gp City of Johannesburg Metropolitan Municipality kz eThekwini Metropolitan Municipality
kz uMgungundlovu District Municipality gp City of Ekurhuleni Metropolitan Municipality
Monthly reporting/review of data
0
2000
4000
6000
8000
10000
12000
14000
Ap
r_1
6
May
_16
Jun
_16
Jul_
16
Au
g_1
6
Sep
_1
6
Oct
_16
No
v_1
6
Dec
_1
6
Jan
_1
7
Feb
_1
7
Mar
_1
7
Ap
r_1
7
May
_17
Jun
_17
kz Cato Manor CHC
TX_CURR
TX_CURR
Granularity
Granularity of reporting and analysis • While district-level results remain critical…reporting and analysis (and
remediation) is down to the level of service delivery (e.g. facility) is essential to reaching goals
• Epi and program data strongly indicate variation in HIV burden, service update, and response by age, sex, population (key-, priority-)….
• …Necessitates the need to understand program coverage by age, sex and to design and implement population appropriate programs to fill these gaps
• International community has recommended reporting by ‘finer’ disaggregations (Ref: WHO. Consolidated Strategic Information Guidelines. 2015)…
• …Beginning in Q1 ‘18 (October 2017), PEPFAR will require reporting on more precise age, sex bands
*Site-Level monitoring: On-line GIS Server (PEPFAR)*
HTC Yield: https://devgeocenter.org/flexviewers/PEPFARSAHTCPositivity/
Linkage to Tx: https://devgeocenter.org/flexviewers/PEPFARSALinkedToTreatment/
Treatment : https://devgeocenter.org/flexviewers/PEPFARSAARTTReatmentGap/
VMMC % of Target Achieved (Q3) – Site Level Analysis
12
13
51% 49% 47% 49% 54%
49% 43%
19% 22% 17% 15%
17% 18%
15%
11% 11%
11% 11%
9% 10%
12%
19% 18%
23% 24%
19% 22%
28%
1% 1% 1% 1% 1% 1% 2%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Jul to Sep 2015 Oct to Dec 2015 Jan to Mar 2016 Apr to Jun 2016 Jul to Sep 2016 Oct to Dec 2016 Jan to Mar 2017
Total
>=50
25:49
20:24
15:19
10:14
Circumcisions by Age Band: 27 Districts Q1 - Q3
Trends in HIV Positivity Rate Dashboard
0%10%20%30%40%50%60%70%80%90%
100%
gp City ofJohannesburgMetropolitanMunicipality
gp EkurhuleniMetropolitanMunicipality
kz eThekwiniMetropolitanMunicipality
kz uMgungundlovuDistrict
Municipality
Male 15-24: TX_CURR (Q2 '17) vs. TX_CURR gap (%)
M_15_24_Q2TXCURR M_15_24_Q2TXCURR_gap
0%
20%
40%
60%
80%
100%
gp City ofJohannesburgMetropolitanMunicipality
gp EkurhuleniMetropolitanMunicipality
kz eThekwiniMetropolitanMunicipality
kzuMgungundlovu
DistrictMunicipality
Female <15: TX_CURR (Q2 '17) vs. TX_CURR gap (%)
F_0015_Q2TXCURR F_0015_Q2TXCURR_gap
0%
20%
40%
60%
80%
100%
gp City ofJohannesburgMetropolitanMunicipality
gp EkurhuleniMetropolitanMunicipality
kz eThekwiniMetropolitanMunicipality
kz uMgungundlovuDistrict
Municipality
Female 15-24: TX_CURR (Q2 '17) vs. TX_CURR gap (%)
F_15_24_Q2TXCURR F_15_24_Q2TXCURR_gap
0%20%40%60%80%
100%
gp City ofJohannesburgMetropolitanMunicipality
gp EkurhuleniMetropolitanMunicipality
kz eThekwiniMetropolitanMunicipality
kzuMgungundlovu
DistrictMunicipality
Female 25+: TX_CURR (Q2 '17) vs. TX_CURR gap (%)
F_25PL_Q2TXCURR F_25PL_Q2TXCURR_gap
Males, Females TX_CURR coverage (Q2 ‘17) and remaining gap to COP17 target by age, sex band
0%
20%
40%
60%
80%
100%
gp City ofJohannesburgMetropolitanMunicipality
gp EkurhuleniMetropolitanMunicipality
kz eThekwiniMetropolitanMunicipality
kzuMgungundlovu
DistrictMunicipality
Male <15: TX_CURR (Q2 '17) vs. TX_CURR gap (%)
M_0015_Q2TXCURR M_0015_Q2TXCURR_gap
0%
20%
40%
60%
80%
100%
gp City ofJohannesburgMetropolitanMunicipality
gp EkurhuleniMetropolitanMunicipality
kz eThekwiniMetropolitanMunicipality
kzuMgungundlovu
DistrictMunicipality
Male 25+: TX_CURR (Q2 '17) vs. TX_CURR gap (%)
M_25PL_Q2TXCURR M_25PL_Q2TXCURR_gap
Viral Suppression (%) by age group: 2015-16
2589 32321 31031 36533
105678 1657327
20434
1883 16853 17992 20693
30698 332623
7326
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
<1 1-9 10-14 15-19 20-24 25+ notstated
% V
LS
AGE CATEGORY
2017
yes No
Current performance line Target line
1109
13059 10197 10412
27625
428394 9198
1660
11278 9808 11467
19071
192959 6116
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
<1 1-9 10-14 15-19 20-24 25+ notstated
% V
LS
AGE CATEGORY
2016
yesNoCurrent performance line
4057
50693 42706 46739
130522 2110487
32143
3668
29978 29415 33090
50777 542708
13830
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
<1 1-9 10-14 15-19 20-24 25+ notstated
% V
LS
AGE CATEGORY
2015
yes No
Current performance line Target line
REF: NHLS/CCMT report July 2017
Composite Measures
Composite indicators e.g. linkages, ‘net_new’
• Individuals move across programs and services which requires monitoring & analysis of ‘composite’ rather than ‘siloed’ indicators • Linkage: HIV testing->HIV Treatment ( ‘direct’ linkage indicator under
consideration)
• Net New on Treatment: change in TROA over time
• TX_NEW-TX net new
• Treatment cascade
• Allows for a holistic picture of program across interventions and for deeper analysis for corrective actions
Time Trend: Linkage to care (Positive test / New ART start), stratified by priority (SOURCE: South Africa DHIS)
19
Impact of UTT implementation:
- Pre-ART Mop Up
- Same Day Initiation
- Linkage Officers
Source: DHIS
0%
20%
40%
60%
80%
100%
120%
140%
160%
Scale-up: Aggressive Scale-up: Saturation
Linkage to TX (TX_New/HTC_Pos): Johannesburg
Oct-Dec 2016 Linkage
Apr-June 2017 Linkage
Linkage (TX_NEW/HTC_POS) by Age/Sex, Q1-Q2: kz eThekwini Metropolitan Municipality
0%
50%
100%
150%
200%
250%
10:14 15:19 20:24 25:49 >=50
Female Male
TX_NEW - TX_NET_NEW, Q1’17
22
-1000
0
1000
2000
3000Ekhuruleni Patient Volume
-3000
-2000
-1000
0
1000
2000
3000eThekwini
Patient Volume
-15000
-10000
-5000
0
5000
City of Johannesburg Patient Volume
-2500
-2000
-1500
-1000
-500
0
500Umgungundlovu Patient Volume
TX_NEW > TX_NET_NEW; potential retention issues
TX_NEW<TX_NET_NEW; potential transfer-in issues
Current/Future Directions
• Accelerated progress towards 90 90 90 is needed
• More timely (e.g. weekly, monthly) review and response to data is essential
• Aggressive monitoring of program, partner performance: • Monthly updating of DHIS Dashboard; quarterly review of cohort
(TX_RET) and viral load data
• Close monitoring of HTC_pos, linkage, TX_NEW, TX_CURR, TX_Retain, TX_net_new, VMMC • Age, sex bands (more precise age bands are forthcoming)
• Site level
• Direct measure of linkage may be forthcoming
• AssessFeedback with remediation plansRepeat
Final thought
• MER represents the minimum and should not be rate-limiting for additional analytics and interpretation to improve the quality of the program.
Acknowledgements
• PEPFAR Strategic Information TWG (CDC, USAID)
• SA NDoH
District TX Net_New Total (secondary x axis)
Site TX Net_New (primary x axis)
Site Q2 ‘17 TX_CURR (circle size)
Partner (circle color)
TX_NET_NEW, Q4’16-Q2’17, by District
-50000
0
50000
100000
150000
200000
Achievement Target
Q3 Progress: PEPFAR ‘Saturation’ Districts (n=4)
Row Labels
CUMM
TARGETS:
FY2017
RESULTS: FY17-
Q1
RESULTS: FY17-
Q3
CUMM
RESULTS:
FY2017
%
ACHIEVEMENT
STATUS
HTS_TST_Pos+Neg 2,065,597 828,458 1,059,260 1,887,718 91%
HTS_TST_Pos 292,716 90,898 93,232 184,130 63%
TX_NEW 336,242 66,716 54,433 121,149 36%
TX_CURR 1,486,960 1,043,202 1,118,564 1,118,564 75%
TX_NET NEW 533,314 89,556 75,362 164,918 31%
Cli
nic
al
cascad
e
PMTCT_STAT 240,901 60,293 57,223 117,516 49%
PMTCT
PMTCT_STAT_Pos 113,316 17,091 17,651 34,742 31%
PMTCT_ART 63,377 16,300 16,994 33,294 53%
PMTCT_EID 63,154 25,517 26,273 51,790 82%
PMTCT_EID_POS_12MO 71 158 203 361 508%
PMTCT_EID_POS_2MO 631 148 153 301 48%
PrEP_NEW 1,279 213 295 508 40%
VMMC_CIRC 173,227 20,890 52,194 73,084 42%PREVENTION
PMTCT
28
-
50 000
100 000
150 000
200 000
250 000
Females Males Females Males Females Males Females Males
Positive
Negative
29
COP 2017 Community Modality Targets
Home Based Index Mobile Other & KP
Entry Streams for ART Enrollment Tested for HIV Identified Positive Newly initiated (90% Linkage)
Community 1,915,892 166,819 150,137
Home Based 440,927 30,674 27,607
Index 482,093 36,770 33,093
Mobile 801,092 76,194 68,575
Other & Key Pop 193,100 23,379 21,041
SAPR 2017 Community Modality Results
8.34%
6.56%
7.92%
7.94%
9.36% 7.06%
9.96% 7.08%