Surgery for Trichiasis, Antibiotics to prevent Recurrence (STAR) Clinical Trial methodology.

Abstract

Trachoma is the leading infectious cause of blindness worldwide. Surgery is available to correct trichiasis, which results from repeated episodes of infection with C. trachomatis. However, trichiasis recurrence rates post-surgery are very high. Methods for reducing post-surgical trichiasis recurrence need to be explored. This paper outlines the design of the Surgery for Trichiasis, Antibiotics to prevent Recurrence (STAR) Clinical Trial currently underway in Ethiopia. The STAR trial, funded by The National Eye Institute, is a randomized, controlled clinical trial of antibiotic use at time of trichiasis surgery, comparing topical tetracycline to single-dose azithromycin for the surgical patient and single-dose azithromycin for the surgical patient and all household members. The primary outcome is trichiasis recurrence at one-year. Data from this trial will be critical in helping to determine future policy on antibiotic treatment for C. trachomatis following surgery.

Full text

Page 1

Ophthalmic Epidemiology, 12:279–286, 2005
Copyright? c Taylor & Francis Inc.
ISSN: 0928-6586
DOI: 10.1080/09286580591005769
ORIGINAL ARTICLE
Surgery for Trichiasis, Antibiotics to Prevent
Recurrence (STAR) Clinical Trial
Methodology
Emily S. West, PhD
Dana Center for Preventive
Ophthalmology, Johns Hopkins
University, Baltimore, MD, USA
Wondu Alemayehu,
MD, MPH
ORBIS International, Addis
Ababa, Ethiopia
Beatriz Munoz, MS
Dana Center for Preventive
Ophthalmology, Johns Hopkins
University, Baltimore, MD, USA
Muluken Melese, MD, MPH
ORBIS International, Addis
Ababa, Ethiopia
Alemush Imeru, MPH
and Sheila K. West, PhD
Dana Center for Preventive
Ophthalmology, Johns Hopkins
University, Baltimore, MD, USA
ABSTRACT Trachoma is the leading infectious cause of blindness worldwide.
Surgery is available to correct trichiasis, which results from repeated episodes of
infection with C. trachomatis. However, trichiasis recurrence rates post-surgery
areveryhigh.Methodsforreducingpost-surgicaltrichiasisrecurrenceneedtobe
explored.ThispaperoutlinesthedesignoftheSurgeryforTrichiasis,Antibiotics
topreventRecurrence(STAR)ClinicalTrialcurrentlyunderwayinEthiopia.The
STAR trial, funded by The National Eye Institute, is a randomized, controlled
clinical trial of antibiotic use at time of trichiasis surgery, comparing topical
tetracycline to single-dose azithromycin for the surgical patient and single-dose
azithromycin for the surgical patient and all household members. The primary
outcome is trichiasis recurrence at one-year. Data from this trial will be critical
in helping to determine future policy on antibiotic treatment for C. trachomatis
following surgery.
KEYWORDS
azithromycin; Chlamydia trachomatis; STAR Clinical Trial; trial methodology; Ethiopia
Trachoma surgery; trachoma recurrence; antibiotic therapy; tetracycline;
INTRODUCTION
Trachoma, an ocular disease caused by infection with Chlamydia trachomatis,
istheleadinginfectiouscauseofblindnessworldwide.Thediseaseisseenalmost
exclusivelyinpoorerareasofdevelopingcountries.Theactivestageofthedisease
presents in early childhood as chronic conjunctivitis, with hyperemic lids and
presenceoffolliclesontheuppertarsalplate.Inhyper-endemicareas,morethan
50% of children are typically infected or show signs of disease.1−3Repeated
infections cause trachomatis scarring which may result in entropion (inward
deviation of the eyelid) and trichiasis (in-turned lashes). Trichiasis primarily
affects adults, although it can be seen in young children in hyper-endemic
communities. In areas where trachoma is endemic, the prevalence of trichiasis
is often more than 5% among adults.1,3,4
Data suggest that persons at greatest risk of developing trichiasis are indi-
viduals with scarring and active or persistent infection.4−6Women are two to
Received 10 December 2004
Accepted 24 April 2005
Address correspondence to Emily S.
West, PhD, Wilmer Room 116, Wilmer
Eye Institute, Johns Hopkins School of
Medicine, 600 N Wolfe St., Baltimore,
MD, 21205, USA. E-mail:
ewest@jhsph.edu
279
Ophthalmic Epidemiol Downloaded from informahealthcare.com by JHU John Hopkins University on 06/03/10
For personal use only.

Page 2

four times more likely develop trichiasis than men.1−3,7
Recent studies have documented C. trachomatis infec-
tion in adults, including those who do not have clinical
signs of disease.6,8−10Data on increasing risk of scar-
ring and trichiasis in those with infection support the
importanceofthepresenceofC.trachomatis,eveninthe
absence of clinically active disease, in the progression
of this process. One of the mainstays of trachoma con-
trol programs is surgery to correct trichiasis.11However,
under the best of circumstances, the recurrence rate of
trichiasis following surgery is disappointingly high, es-
timated at 17% or more.12−17
Recently, studies have begun to investigate the rea-
sons for these high recurrence rates. One hypothe-
sis is that ongoing infection with C. trachomatis leads
to recurrence. One study in Tanzania following surgi-
cal cases two or more years post-surgery demonstrated
that individuals with trichiasis recurrence were more
likely to have trachomatis inflammation than individ-
uals without recurrence. Participants living with two or
morehouseholdmemberswithtrachomawere3.7times
more likely to have recurrence than individuals liv-
ing alone or with no other household members with
trachoma.16A case-control study in Nepal also demon-
strated an association between infection and trichiasis
recurrence: 82% of individuals with recurrence were in-
fected at 12 months post-surgery, while none of the
controls, individuals with trachomatis scarring, were
infected.17
The standard treatment for active trachoma is either
a single dose of oral azithromycin or a one- to two-
month course of topical tetracycline twice daily. Clin-
ical trials have shown equivalent efficacy of the two
treatments;18−21however, given the significant differ-
ence in treatment length and the problem of poor com-
pliance, azithromycin is likely to be more effective.18
Treatmentfortrachomaatthetimeofsurgerycanridthe
patient of C. trachomatis infection; nonetheless, trichi-
asis surgery patients who return to trachoma-endemic
families are at risk of re-infection, with the possibil-
ity of ongoing scarring and development of recurrent
trichiasis. Thus, treating all household members of sur-
gical cases may be of significant importance. This clin-
ical trial allows us the opportunity to investigate the
value of treating the patient with or without treating the
household with long-acting systemic antibiotics com-
pared to use of topical antibiotics in reducing trichiasis
recurrence.
METHODS
Study Design
The STAR trial is a randomized, single-masked trial
ofantibioticusefollowingtrichiasissurgeryinapopula-
tion of individuals living in the Wolayta Soddo zone of
the Southern Nations Nationalities and Peoples Region
(SNNPR) of Ethiopia.
Aims
This study has two specific primary aims:
(1) To determine if there is a difference in trichiasis
recurrence rates at one-year post-surgery between
trichiasis surgery cases treated with topical tetracy-
cline (usual care) and surgical cases treated with oral
azithromycin.
(2) To determine if there is a difference in one-
year recurrence rates comparing surgical patients
treated with azithromycin versus surgical patients
whose household members were also treated with
azithromycin.
Secondary aims of the trial focus on evaluation of
other risk factors for recurrence, such as severity of
trichiasisatbaselineandsurgicalfactorssuchassurgeon
and surgical complications (e.g., excess bleeding).
IRB Approval
This project was approved by the Johns Hopkins
MedicalInstitutionsInstitutionalReviewBoard(JHMI-
IRB) and the Ethiopian Science and Technology Com-
mission (ESTC). JHMI IRB approval was granted con-
tingent on approval by the ESTC. The study meets the
tenets of the Declaration of Helsinki.
Ethiopia: Background
Ethiopia is among the countries with the high-
est prevalence of trachoma worldwide. Some stud-
ies of active disease report prevalence of trachoma as
high as 90% among children aged less than 10.22−24
Until recently, surgical services to correct trichiasis
were largely unavailable and estimates suggest that
952,500 individuals in Ethiopia are in need of surgical
services.25
Villages in Ethiopia are clustered into groups. A clus-
ter of villages within a defined proximity is called a
“Woreda” and is similar to a “county” according to US
E. S. West et al. 280
Ophthalmic Epidemiol Downloaded from informahealthcare.com by JHU John Hopkins University on 06/03/10
For personal use only.

Page 3

definition. Each Woreda has an average population of
150,000. Primary healthcare in the region is provided
by Community Health Assistants (CHAs). Typically,
one CHA is assigned to provide healthcare to approxi-
mately four villages. Each CHA works at a health post,
where healthcare is administered. Integrated Eye Care
Workers (IECWs) are nurses or health assistants who
are trained for one month in primary eye care, includ-
ing trichiasis surgery. Within this group, nurses are high
school graduates trained for at least two years in general
nursing. Health assistants are high school graduates and
have received 18 months’ training in basic healthcare.
Regional health centers provide additional healthcare
resources at the Woreda level, and often are equipped
with electricity. For this trial, trichiasis surgery was per-
formed at the health centers by certified IECWs and
follow-up visits are conducted at the health posts.
Wolayta Soddo is located in a hilly region with clay-
like soil. The rainy season extends from June until
September. During this time period, some roads are
impassable, and few individuals are willing to travel
to the health center. Moreover, planting occurs, and
households–largely subsistence farmers—are very busy.
These issues imposed limitations on the times for
follow-up.
Interventions
Treatment
Participants were randomly assigned to one of three
treatment arms using a specially designed SAS program
that incorporated permuted block sizes. Each surgical
participant received one of the following treatments
post-surgery:
Two tubes of topical tetracycline 1%, with instruction
for twice daily administration for six weeks (the cur-
rent standard of care in Ethiopia).
A single one-gram dose of azithromycin
A single one-gram dose of azithromycin for the surgical
case plus a weight-based (20 mg/kg up to 1 gram)
dose for everyone living within the same house as
the surgical case.
All consenting household members were requested
to participate in an exam for trachoma and were of-
fered azithromycin treatment for trachoma control at
one year post-surgery, or when the surgical case reached
endpoint; whichever occurred first.
Surgery
All
Rotation
guidelines.11
workers (IECWs) performed trichiasis surgery. Prior
data from other countries suggest that the technical
skill of individuals performing surgery is an important
factor for a good surgical outcome.26Therefore,
although the surgery itself is not part of this trial, we
required that each IECW undergo certification by
an outside examiner prior to trial commencement to
ensure that surgical failure was not an overwhelming
contributor to recurrence. Eight IECWs received
certification from the examiner using a certification
program developed for this trial.27The three IECWs
with the highest recommendation were selected to
complete all surgeries for the trial. Two IECWs were
responsible for all operations during each enrollment
period. Participants were randomized to operating
IECW on a 1:1 basis, using permuted blocks of two
and four in order to control for any IECW-related
factors.
For bilateral cases, IECWs completed surgery on the
right eye first, followed by surgery on the left eye. The
study eye was also selected randomly. For unilateral
cases, clearly only one eye was eligible, and that eye
was assigned as the study eye. For bilateral cases with
both eyes eligible, the study eye was based on the pa-
tient’s id-number: the right eye was assigned when the
last digit was an even number and the left eye was as-
signed when the last digit was odd.
participants
(BTR)
underwent
surgery
Local, trained
Bilamellar Tarsal
WHO
eye
according
integrated
to
care
Eligibility and Ineligibility Criteria
The eligibility criteria (Table 1) were designed to
maximize the number of participants eligible for the
study while minimizing the number of individuals
lost to follow-up and conforming to Ethiopian health
TABLE 1
Eligibility and Ineligibility Criteria
Eligibility
criteria
Presence of upper lid trichiasis
age 18 or older
Ineligibility
criteria
Plans to move out of the district within one
year
Allergy to tetracycline
Pregnancy
Other household members participating in
the trial
Previous report of trichiasis surgery in the
only eligible eye at time of enrollment
281STAR Methodology
Ophthalmic Epidemiol Downloaded from informahealthcare.com by JHU John Hopkins University on 06/03/10
For personal use only.

Page 4

regulations. The primary eligibility criteria were upper-
lid trichiasis in at least one eye with no previous surgery
and age 18 years or more in order to provide consent as
an adult. As more than 90% of trichiasis occurs in indi-
vidualsaged18andolder,1thisagecriterionwasnotfelt
to interfere with generalization of findings. Although
many obstetricians in the US prescribe azithromycin as
the drug of choice for treatment of pregnant women
with genital Chlamydia,28and several studies have
shown that azithromycin is well-tolerated during preg-
nancyanddoesnotincreasesideeffects,29−31thedrugis
classifiedasacategoryBagent.Thiscategorizationisas-
signed to drugs that have not been specifically studied
in pregnant women, but have low likelihood of birth
defects. Indeed, in Europe, the drug is currently rec-
ommended for the treatment of genital chlamydia in
pregnant women,32and a clinical trial of azithromycin
use in pregnant women was recently conducted in the
UnitedStates.33However,theEthiopianhealthauthori-
tieshavenotapprovedazithromycinforuseinpregnant
women in Ethiopia, and women who self-report preg-
nancy are excluded from the trial. Household members
whoreportedpregnancyweregiventopicaltetracycline.
Sample Size
Sample size calculations were based on our primary
outcome variable: trichiasis recurrence at one year post-
surgery. We assumed a two-sided alpha of 0.05 and 80%
power for the comparison of each azithromycin group
tothetetracyclinegroup.Recurrenceinthecontrolarm
was estimated conservatively to be as low as 17%.12−15
We assumed that a 40% reduction in recurrence, or a
10% recurrence rate in the patient-based azithromycin
arm, would provide a meaningful difference between
the control group and the patient-based azithromycin
group. We further decided that a 40% reduction be-
tweenthepatient-basedandhousehold-basedtreatment
arms was meaningful, for an expected recurrence rate of
6% in the household-based treatment arm. Assuming a
20% loss to follow-up, we estimated that 475 subjects
were needed for enrollment in each treatment arm.
Recruitment and Consent
All local community health assistants (CHAs) had
been trained in trichiasis identification as part of the
Ethiopian national program for referral of cases for
surgery. For this trial, each CHA was assigned a specific
week to bring all individuals with trichiasis in his/her
jurisdiction to the health center. To accommodate the
TABLE 2
Enrollment Periods
Time frame Woredas participating
November 2002–December 2002
March 2003–April 2003
October 2003–November 2003
Soddo Zuria
Soddo Zuria, Gesupa
Humbo, Gununo
rainy season, three surgical periods were necessary, as
outlined in Table 2. STAR team members trained in
trichiasis identification screened presenting individuals
for severity of trichiasis and previous eyelid surgery. All
individuals meeting initial screening criteria were fur-
therevaluatedforfinaleligibilityandwillingnesstopar-
ticipate. Interested individuals were read an informed
consent form in Amharic or Wolaytinga and were asked
to provide a fingerprint to confirm their willingness to
participate. The CHA for the village witnessed con-
sent for all participants from his area. Ineligible per-
sons with trichiasis were still offered trichiasis surgery at
no cost.
Ocular Assessments
Visual Acuity
Distance visual acuity for each eye was determined
using a tumbling ‘E’ chart patterned on a standard
ETDRS chart. As electricity could not be guaranteed,
acuity measurements were taken with the chart fully il-
luminated by the sun. Visual acuity was tested at three
meters with the participants presenting refractive cor-
rection. If the participant could not correctly identify
three of the five letters on the first line, testing was con-
ducted at one meter. If participants could not read the
first line of the chart at this distance, hand motion and
light perception were evaluated.
Trichiasis
Trichiasis was defined according to WHO criteria:
one or more lashes touching the globe and/or evidence
of epilation. Epilation is a common practice performed
in this area. Trichiasis patients typically carry around a
tweezers-like instrument and will have one of their fam-
ilymembersplucktheirin-turnedeyelashesonaregular
basis. Evidence of epilation was defined as the visible
presence of eyelash stubs. Patient history alone was al-
lowed for purposes of obtaining surgery; however, his-
tory of epilation alone was not considered sufficient ev-
idence of epilation for purposes of outcome assessment
as there would be uncertainty if the epilation occurred
prior to the surgery or afterwards. Severity of trichiasis
E. S. West et al. 282
Ophthalmic Epidemiol Downloaded from informahealthcare.com by JHU John Hopkins University on 06/03/10
For personal use only.

Page 5

was graded based on an expanded grading scheme from
the WHO simplified assessment.34Evidence of epila-
tionwasrecorded.Iflashesweretouchingtheglobe,the
number of lashes touching any part of the globe and
the number touching only the cornea were recorded. If
more than ten lashes were touching, 10+ was recorded.
Theupperlidwasdividedintothreeequalsections.The
location of the lashes touching the globe was recorded
based on the origin of the deviated lashes: nasal, cen-
tral, and/or temporal. Entropion was graded as present
if there was evidence of inward deviation of the upper
eyelid margin with at least one eyelash base not visi-
ble at primary gaze. Participants were asked to report
the duration of trichiasis for each eye. Active trachoma
status was assessed using the WHO simplified grading
scheme.35
Ocular Chlamydial Infection
A swab was taken of the enrolled eye in order to test
for presence of Chlamydia trachomatis infection. One
individual everted the lid and another individual, who
never touched the patient, held a Dacron swab one-
third of the way down the shaft and rubbed the tip
across the upper tarsal plate three times, turning the
swab each time she crossed the tarsal plate. Next, the
person who everted the upper lid exposed the lower
fornix and the tip of the same swab was rubbed three
times across the lower conjunctival surface. The person
collecting the specimen was instructed not to touch
the participant at any time during the process and to
make sure to avoid touching any portion of the swab
stick other than the top edge, which was broken off and
discarded when the swab was inserted into the storage
tube. Swabs were placed in individually labeled tubes
in M-4 transport medium and stored on ice until trans-
ported to the study freezer (−3◦C) at the end of the
day. The person who everted the lid washed her hands
as if preparing to perform surgery between patients to
maintain sterility. After every 20 swabs, the person re-
sponsible for handling the swab stick had her hands
swabbed in order to check for potential contamination
of the specimens.
Quality of Life
The final 400 enrolled participants were asked to
complete a quality of life questionnaire that included
questions on ability to participate in visually intensive
daily activity and how their ocular condition limited
these activities. The questionnaire was based on a pre-
vious quality of life instrument used in rural Africa36
with questions adapted to relate specifically to Ethiopi-
ans,basedonfieldtesting.Thisquestionnairewasasked
again at six months post-surgery.
Primary Outcome
The primary outcome for this trial is trichiasis recur-
rence at one year. Recurrence is defined as a minimum
of one lash touching the globe of the eye or evidence of
epilation, as defined above. Recurrence was assessed at
two and six weeks post-surgery (considered surgical fail-
ure), and six and twelve months post-surgery. The pri-
maryoutcomevariablewasevaluatedbyateammember
who did not participate in baseline exam and treatment
in order to ensure that masking was maintained. Addi-
tionally, the examiner and accompanying team mem-
bers were instructed not to discuss prior treatment with
the participants. Of note, the examiner did not speak
thelocallanguageandcouldnothaveunderstoodtreat-
ment discussions.
Study Visits
Surgical Participants
After surgery, participants were scheduled for follow-
up visits by STAR team members at their local health
post up to four times, including suture removal. Study
visits were scheduled at two weeks (suture removal), six
weeks, six months and twelve months following surgery
(Table 3). The study visit window for the two-week
visit opened seven days after surgery and closed one
week after the ideal visit date. For the six-week visit,
the window opened on the six-week anniversary of the
surgery and closed six weeks later. The windows for the
six and twelve month visits were one month on either
TABLE 3
Data Collection Schedule
266 12
Data items Baseline Weeks Weeks Months Months
Surgical cases
Visual acuity
Trachoma status
Trichiasis status
Ocular swab
Quality of life
Household
Members
Trachoma status
Ocular swab
X
X
X
X
X2
X
X
X
X
X
X
X
X
X
X1
X2
X3
X3
X1
X1
1: Unless trichiasis recurs sooner, then at time of the scheduled exam
when recurrence was found.
2: Final 400 participants enrolled only.
3: For household members in the household treatment arm only.
283STAR Methodology
Ophthalmic Epidemiol Downloaded from informahealthcare.com by JHU John Hopkins University on 06/03/10
For personal use only.