Transcript mass administration

Forecasting trachoma: control,
elimination, or eradication?
Thomas M. Lietman
Kathryn J. Ray
Travis C. Porco
FI Proctor Foundation, UCSF
December 2012
Trachoma
• Leading infectious cause of blindness
(WHO 2002)
• Causative agent Chlamydia trachomatis
• Repeated infection leads to progressive
scarring of the eyelid and mechanical
damage to the cornea
• Infection in children leads to blindness
later in life.
Trachoma (2)
• Progression from follicular and
inflammatory disease
• scarred eyelids
• inturned eyelashes
• secondary bacterial infections lead to
corneal opacity
Healthy eyelid
Severe TF/TI
Scarring
Trichiasis, Corneal opacity
How much less trachoma?
• WHO: annual treatment of all
inhabitants, reduce infection to level
where blindness not a public health
problem.
• Or, should we try to actually reduce the
prevalence of infection to zero?
Important facts
• Ocular infection by C. trachomatis is easily
cured with single-dose azithromycin (95%
efficacy).
• Only humans are infected (there is no animal
reservoir).
• No vaccine is available.
• Clinical signs are unreliable in detecting
infection; laboratory tests are far too
expensive and take far too long.
Trachoma now
• WHO plan to stamp it out as a public
health problem
• Surgery, antibiotics, face-washing,
environment
• The SAFE program
• Mass distribution of azithromycin the
cornerstone
Schachter J, West SK, Mabey D, et al Lancet. 1999 Aug 21;354(9179):630-5
Modeling?
So if we know what causes trachoma…
http://www.pacificu.edu/optometry/ce/courses/13036/antibacterialpg1.cfm
…and we know what stops it…
…?
Mass administration
• Why do we call this program a mass
administration?
• Because no effort is made to try to find
out who actually has the infection and
who does not--everybody gets the
treatment, regardless.
Has modeling helped?
• Lietman et al 1999: Repeat mass
treatment can theoretically eliminate the
infection
• Melese 2004
• Age, immunity,
serotypes,
households,
etc.?
Good data
Practical implications
Current issues
•
•
•
•
•
Assess ongoing efforts
Optimize current efforts
Minimize collateral damage
Consider biological questions
Predict future trends
Assessment
• We want to eliminate infection, but what
should the epidemiology of trachoma
look like during elimination?
• Theoretical models suggest an
approximately exponential
quasistationary distribution (Nåsell;
Lietman 2013 under review)
• Thus, long tails are expected
One of
these
is real data
from a
district
in Nepal,
during
elimination
…the
others are
simulated
geometric
random
variables
Chasing ghosts
• Unpredictability of trachoma at the
village level
• Long tail of the distribution
• Expect transient local hot spots
• The presence of a local hot spot does
NOT imply failure
When can we stop?
• Elementary models suggest as long as
conditions favor transmission, disease
returns unless all cases are eliminated.
• Prevalence
thresholds
possible
Lietman, Epidemics, 2011
Collateral damage
• Mass administration is controversial
because of macrolide resistant
pneumococcus
Maher et al, PLoS1, 2012
Antibiotic minimization?
• Targeting children only leads to
reductions in prevalence, even among
adults
• Can treating children alone actually
eliminate infection?
• Maybe
Other examples
• Does repeat treatment seem to reduce
the efficacy in future years?
• Tanzania data (Kongwa region), PRET
data (S. West)
Liu et al, PLoS NTD, 2013
Predict future trends
A specific model
TANA Trial
• TANA trial: Trachoma Amelioration iN
Amhara
• Lietman Group U10 being conducted in
Ethiopia
• Community randomized trial with four
primary specific aims
Community-specific
prevalence
• Analysis of pooled prevalence from
randomly chosen individuals yields a
prevalence estimate for the
community.
• In the TANA trial, we regard whole
communities as the units of the trial.
Trial
Trial
TANA
•
•
•
•
•
•
May 2006 to March 2007
66,404 people in 48 subkebele
Annual: 50 state teams; 4,437 children
Biannual: 61 ST; 4,462 children
Children: 49 ST; 4,150 children
Delay: 57 ST; 5,166 children
Stochastic epidemic
• Many books now on stochastic models
in epidemiology
• Standard method used here, e.g.
Bailey, Elements of Stochastic
Processes, 1964
State space
• Given a village of size N, let Y be the
number of infected individuals; Y ranges
from 0, 1, … N-1, N.
• Ignore adults for now (low prevalence)
• We examined models with age
structure, partial immunity
State space (2)
0
1
2
…
Infection
Recovery
N-1
N
Model
•
•
•
•
Continuous time
P(Y(t)=i) = pi(t)
Assume population is fixed
Model period between treatments first
Acknowledgments
•
•
•
•
•
Tom Lietman
Teshome Gebre, Berhan Ayele
Jenafir House, Nicole Stoller
Bruce Gaynor, Jeremy Keenan
Zhaoxia Zhou, Vicky Cevallos, Kevin Hong, Kathryn
Ray, Jack Whitcher, Paul Emerson
• Data and Safety Monitoring Committee (W. Barlow,
D. Everett, L. Schwab, A. Reingold, S. Resnikoff)
• Study participants
Acknowledgments, cont’d
Tadege Alemayehu
Tesfaye Belay
Azmeraw Adgo
Melese Temesgen
Gabeyehu Sibhat
Abebe Mekonen
Manalush Berihun
Temesgen Demile
Wosen Abebe
Melkam Andwalem
Mitsalal Aberahraney
Banchu Gedamu
Tessema Eneyew
Muluken Gobezle
Trachoma Projects in Ethiopia
Deb Gill
Melissa Neuwelt
Nandini Gandhi
Cyril Dalmon
Nicolle Benitah
Ying Pan
Lauren Patty
Vivian Schiedler
Ali Zaidi
Dwight Silvera
Isabella Phan
Chihori Wada
David Lee
Harsha Reddy
Kathryn Ray
Rachel May
Alison Skalet
Sara Haug
Andi Hatch
Jesse Biebesheimer
Traci Brown
Laura Cieslik
Anita Gupta
Susie Osaki-Holm
Nazzy Pakpour
Karen Shih
Scott Shimotsu
Kristine Vinup
John Warren
Yinghui Miao
Mariko Bird
Greg Schmidt
Lynn Olinger
Scott Lee
Kevin Hong
Jaya Chidambaram
Allison Loh
Deb Gill
Larry Schwab
Jeremy Keenan
Vicky Cevallos
Lauren Friedly
Bruce Gaynor
Tom Lietman
Kevin Miller
Tisha Prabriputaloong
Michael Saidel
John P. Whitcher
Elizabeth Yi
Michael Yoon
John Warren
Macdara Bodeker
Muthiah Srinivasan
Marilyn Whitcher
Jenafir House
Jon Yang
Nicole Stoller
Charles Lin
Tina Rutar
Colleen Halfpenny
Funding
That Man May See
Bernard Osher Foundation
Bodri Foundation
Harper-Inglis Trust
Peierls Foundation
Jack and DeLoris Lange Foundation
Research to Prevent Blindness
International Trachoma Initiative/Pfizer
NIAID: RO1-AI48789
NIAID: R21-AI55752
NEI: U10-EY016214
Bill and Melinda Gates Foundation
With grateful acknowlegment