Page 1 of 46
SNP WORKING PAPERS
Social Networks Project Working Papers
Social
.N.e.t.w.o.r.k.s.
project
No. 6, June 2004
Protocol for biomarker testing in
the 2004 Malawi Diffusion and
Ideational Change Project
Simona Bignami-Van Assche, Kirsten Smith,
Georges Reniers, Philip Anglewicz, Rebecca
Thornton, Li-Wei Chao, Alex Weinreb, Susan
Watkins, Irving Hoffman, and the MDICP STI
Team
Page 2 of 46
Social Networks Project Working Papers is an
unreviewed and unedited prepublication series
reporting on research in progress.
Social Networks Project Working Papers are
circulated for comment and to inform interested
colleagues about the work in progress on the
survey data collected by the Project in Kenya
and Malawi. The Social Networks project is
funded by National Institute of Health Grants
HD/MH41713 on “Gender, Conversational
Networks and Dealing with AIDS” and
HD044228 on “AIDS/HIV risk, marriage and
sexual relations in Malawi." The Project has
also received support from NIH/ORA; NIH/NIA;
The Rockefeller Foundation; The Transnational
Cooperation Foundation, Government of
Germany; The Research Foundation, University
of Pennsylvania; and the Carolina Population
Center/USAID Evaluation Project.
The views expressed are those of the authors and
not necessarily those of the Project, the National
Institutes of Health or any other of the Project’s
funding agencies.
For more information about the Social Networks
Project and to submit manuscripts for
publication, contact:
Social Networks Project
239 McNeil Building
University of Pennsylvania
3718 Locust Walk
Philadelphia, Pennsylvania 19104
Email: social_networks@pop.upenn.edu
Website: www.pop.upenn.edu/Social_Networks
Page 3 of 46
2
TABLE OF CONTENTS
1. Introduction............................................................................................. 3
2. Summary of testing procedures..................................................................... 4
2.1 Target sample....................................................................................... 4
2.2 Tests.................................................................................................. 4
2.3 Collection of biomarkers.......................................................................... 4
2.4 Analysis of biomarkers............................................................................ 5
2.5 Results dissemination.............................................................................. 5
2.6 Treatment............................................................................................ 5
2.7 Incentives.......................................................................................................................... 6
3. Pre-survey operations................................................................................. 7
3.1 Informing the communities........................................................................ 7
3.2 Recruitment of nurses.............................................................................. 7
3.3 Training of nurses................................................................................... 8
3.4 Pilot study............................................................................................ 8
3.5 Creation of identifiers for STI/HIV testing...................................................... 9
4. Collection of biomarkers.............................................................................. 9
4.1 STI team.............................................................................................. 9
4.2 Data collection design.............................................................................. 10
4.3 Labeling of STI questionnaire and specimens................................................................... 11
4.4 STI questionnaire............................................................................................................... 11
4.5 Informed consent for STI and HIV testing...................................................... 11
4.6 Specimens’ collection.............................................................................. 13
4.7 Recording specimens taken.............................................................................................. 15
4.8 Incentives experiment....................................................................................................... 16
4.9 Delivery of specimens to the nurses’ supervisor................................................ 17
4.10 Delivery of Testing Sheets to the survey staff................................................. 18
4.11 Logging............................................................................................. 18
4.12 Call-backs.......................................................................................... 19
5. Analysis of biomarkers................................................................................ 20
5.1 Delivery of specimens to Lilongwe Central Hospital........................................................ 20
5.2 Delivery of results to survey staff................................................................. 21
6. Dissemination of results to respondents............................................................ 21
6.1 Voluntary Counseling and Testing (VCT) centers............................................. 21
6.2 Identification of respondents...................................................................... 21
6.3 Treatment for STIs other than HIV............................................................... 22
6.4 Counseling for HIV................................................................................. 23
Appendices.................................................................................................... 24
A. Personnel list........................................................................................................ 25
B. Supervisors’ form to evaluate contact with communities....................................... 27
C. STI questionnaire................................................................................................................. 28
D. Informed consent forms.............................................................................. 37
E. STI/HIV Testing Sheet............................................................................... 43
F. VCT Sheet............................................................................................................................ 44
G. STI/HIV Transfer Sheet............................................................................. 45
Page 4 of 46
3
1. INTRODUCTION
The 2004 Malawi Diffusion and Ideational Change Project (MDICP-3) will collect and
analyze biomarkers for sexually transmitted infections (STIs) and HIV. The testing will
be conducted for all respondents in the original MDICP sample, plus an additional
sample of adolescents that will be added in 2004. This report presents the protocol that
will be implemented to collect the STI/HIV specimens.
The rationale for the collection of behavioral information and STI/HIV
biomarkers for the same individual in the MDICP-3 is threefold:
1) monitoring disease burden and behavior change over time,
2) assessing and calibrating self-reported data,
3) explicating pathways and elaborating linkages between social environment and
health.
The study will provide the first population-based information on trends in
prevalence of STIs. Although little is known about the distribution of non-HIV STIs in
rural Malawi, the most common are said to be genital ulcer disease, gonorrhea,
chlamydia, trichomonas, syphilis, and herpes, in that order (Agabu, personal
communication). The focus of the MDICP-3 is on curable1 STIs. For females the
biomarkers are for chlamydia, gonorrhea and trichomonas; for males, chlamydia and
gonorrhea.2
HIV data will also provide an opportunity to validate estimates of HIV
prevalence in rural Malawi routinely produced on the basis of data collected among
pregnant women in antenatal surveillance clinics.
The testing method that will be used in the MDICP-3 is not anonymous. The
respondents’ unique Biomarker ID will be kept on every specimen that is collected
from them, and it will be linked to the person’s computerized data. Identifying
individuals who were tested is necessary to inform them of their test results. However,
we will ensure confidentiality of the respondents who agree to be tested. No personal
identifier will be kept on the specimens (like the name of the respondent or the village
where he or she lives). In addition, at every stage of data collection (including results’
1
STIs for which curative treatment exists. Non-curable STIs are, for example, herpes or HIV. 2
The diagnosis of genital ulcer disease requires a clinical exam, which is not feasible for a large and
relatively dispersed rural sample.
Page 5 of 46
4
dissemination), only the MDICP biomarker coordinators and selected members of the
STI team will have access to the respondent’s identifying information.
2. SUMMARY OF TESTING PROCEDURES
2.1 Target sample
The target sample for the STI/HIV testing includes all respondents who will be
interviewed by the MDICP-3, that is:
- the respondents included in the original MDICP-1 sample (about 1500 married
women and their husbands) and the respondents added in the MDICP-2 (i.e. all
co-wives of polygamous men, as well as new spouses of respondents who had
remarried between MDICP-1 and MDICP-2);
- a sample of adolescents aged 15-24 (approximately 600 males and 600
females);
- a sample of approximately 500 adolescents aged 15-24 in Balaka District that
will be included in the ACASI study, carried out by Barbara Mensch and Paul
Hewitt from the Population Council.
2.2 Tests
For women, vaginal swabs will be used to test for chlamydia, gonorrhea, and
trichomonas. For men, urine will be used to test for chlamydia and gonorrhea. The HIV
tests will be done with oral swabs.3
The selection of these tests has been done according to guidelines of the
Malawian Ministry of Health and in conjunction with specialists at from the UNC
Project based at Lilongwe Central Hospital. The MDICP STI/HIV testing protocol has
been approved by the IRB in Malawi as well as in the United States.
2.3 Collection of biomarkers
The team involved in the MDICP-3 data collection will include a group of trained
nurses who will be responsible for the collection of biomarkers.
3
Roche PCR will be used for gonorrhea, chlamydia, and trichomonas. ORASURE saliva tests will be
used for HIV, and positive results from ORASURE will be confirmed through Western Blot on the same
specimen.
Page 6 of 46
5
Once the biomarkers have been collected, the nurse will take a Polaroid picture
of the respondent, write the respondent’s ID number on the back of it, and leave it with
the respondent, explaining that he/she has to show this picture in order to obtain the
results. The nurse will also provide information about when the respondent will be able
to receive the results.
In the evening all specimens will be logged and refrigerated. Three times a
week the specimens will be transported to the UNC laboratory in Lilongwe for analysis.
2.4 Analysis of biomarkers
The biomarkers will be analyzed by the laboratory run by the University of North
Carolina at Chapel Hill (UNC) in Lilongwe Central Hospital (LGH), under the
direction of Irving Hoffman and Francis Martinson.
2.5 Results dissemination
Results will be communicated to respondents approximately five to seven weeks after
collection of the specimens by the nurses. The approach chosen for results
dissemination uses temporary Voluntary Counseling and Testing (VCT) centers set up
in each village, staffed by nurses trained in HIV testing and counseling.
Results will be made available throughout each study area in small portable
tents, which provide privacy. There will be several VCT (tent) sites in each surveyed
area, so that all respondents’ homes will be within approximately five kilometers from
at least one site. A few days before the results are to be available, the VCT team will
visit all respondents in their homes to communicate the location of the tent and the
specific week the team will be in their village.
The VCT team will have a list of Biomarker ID numbers with corresponding
test results, but without respondents’ names or any other identifying information. In
order to get their results, respondents will have to present to the nurse the Polaroid
picture they had taken when the specimens where collected. The nurse will use the
picture to make sure that nobody else other than the respondent is given the test results.
These procedures should preserve confidentiality, as the results will only be known to
the respondent.
2.6 Treatment
One single-dose treatment for each STI will be provided to all respondents who are
Page 7 of 46
6
infected.4
Infected respondents will also receive treatment for up to two of their sexual
partners. All respondents who come to get their test results will receive vitamins.
Respondents who test positive for HIV will be offered counseling.
By providing all with some treatment, we hope to preserve confidentiality and
prevent speculation about who is infected and who is not. All respondents, whether
they are infected or not, will be counseled by the nurses about the symptoms of STIs
and encouraged to go to the local clinic should they experience these symptoms.
2.7 Incentives
The MDICP-3 biomarker data collection is linked to a randomized experiment focused
on the incentives for VCT uptake.
The MDICP incentives experiment will be carried out at the moment of the
collection of biomarkers from the respondents. After the nurse has completed collecting
the specimens, the respondents will be asked to choose an amount of monetary
compensation randomly from two plastic bags, one for STIs and one for HIV.5
Each
plastic bag will contain bottle tops where a monetary amount is written on the back,
with a total value ranging from 0 to 300 Kwacha (approximately $3.00). Each patient
will choose a random amount that he/she will be able to collect conditional on receiving
STI and/or HIV test results. The monetary amount will be copied on a voucher, and
will be distributed when the respondent comes to get his/her test results at the VCT
center. The fact that individuals select the amount of their voucher provides a
transparent allocation of incentives given the randomness of the experiment.
4
Azithromycin for chlamydia, metronidazole for trichomonas, Ciprofloxacin for gonorrhea. Women
who are pregnant and who have gonorrhea will receive azithromycin instead of Ciprofloxacin (see
Section 6.4).
5
To determine whether having an additional incentive specifically linked to the HIV test had a different
effect from a having single incentive linked to receiving either test result, in the third site, Rumphi, only
one incentive was given. In that site, respondents received the amount of the incentive if they consented
to and received results for either STI or HIV testing.
Page 8 of 46
7
3. PRE-SURVEY OPERATIONS
3.1 Informing the communities
The collection of biomarkers will be preceded by visits to each of the village chiefs in
the sample to inform the community about the coming survey. These visits will be
carried out by an experienced supervisor trained specifically for the purpose.
Since the communities are quite familiar with the MDICP, emphasis will be
placed upon explaining the reasons for testing for STIs and HIV, the procedures for
specimen collection and results dissemination, and the confidentiality of the results.
Village heads will also be briefed about the timetable of the survey and STI/HIV
testing.
At the end of each visit, the supervisor will fill in a form to keep track of the
issues raised, to improve communication and explanations for the succeeding villages
and field sites, and ultimately to give back the results of the testing. This short form
(presented in Appendix B) has been designed on the basis of those used by the Mali
Demographic and Health Survey to evaluate informed consent (Yoder and Konaté
2002).
3.2 Recruitment of nurses
The selection of nurses is important to the project for a number of reasons. First,
because the demand for nurses is high in Malawi and health centers are often
understaffed, it is important that nurses are not hired away from their professional jobs.
In addition, because of the sensitive nature of HIV testing, it is important that nurses
are able to speak the local language in each of the study sites but, to preserve
confidentiality, who are also outsiders to the community to the extent that it is possible.
Nurses will be recruited before fieldwork begins by Rebecca Thornton with the
assistance of personnel from Kamuzu College of Nursing. Two types of nurses will be
hired for the project: younger nurses who recently graduated from one of the health
science schools with a degree in nursing (from Zomba or Blantyre), and more veteran
health professionals who took time off from their job. Because nurses tend to have the
same practical training, we will hire on the basis of past research experience,
experience in VCT and HIV counseling, as well as personal qualities such as
willingness to work in more difficult and rural conditions. In total, approximately 60
nurses will be recruited, including supervisors for the three sites.
Page 9 of 46
8
3.3 Training of nurses
Nurses’ training for the MDICP-3 survey will consist of a combination of classroom
training and practical experience that will provide the nurses with the background
necessary to perform these tasks during a field survey. The nurses’ training will be
carried out by the staff of the UNC Project at Lilongwe Central Hospital and by the STI
team of the MDICP.
The nurses will have five days of training before the beginning of the fieldwork
operations in Balaka and Mchinji,6
during which the various sections of the Nurse’s
Training Material and of this Protocol will be reviewed. Two days will be carried out
by the staff of the UNC lab and will focus on management of STIs, lab procedures and
pre- and post-test counseling. The remaining days of training will be carried out by the
MDICP staff and will focus on the survey procedures associated with the collection of
biomarkers, with particular focus on the informed consent for STI testing. Nurses will
learn how to collect and label the specimens, how to complete the necessary forms, and
how to tell the respondents when and where they can get their results.
3.4 Pilot study
A small pilot study of the collection of biomarkers will follow the nurses’ training. In
order to avoid sample contamination, the respondents included in the pilot study will
belong to a village neighboring, but not included in, the MDICP sample. The main aim
of the pilot study is to assess:
1) the impact of the study on respondents (especially in terms of response rates);
2) the acceptability of vaginal swabs among women;
3) the consent forms and the overall efficacy of the STI/HIV collection,
4) the efficiency of testing done by the UNC laboratory,
5) the effectiveness of VCT pre- and post-counseling for HIV.
This assessment will serve as the basis to modify, if necessary, our protocol, in order to
achieve higher response rates and smoother procedures.
The pilot study will be carried out in the first site to be surveyed (Mchinji
District), with a target sample of 30 respondents (15 men and 15 women). Additional
pilots will be conducted prior to the start of fieldwork in the other two sites, in which
6
Nurses hired to work in the third site, Rumphi, will attend the Balaka training.
Page 10 of 46
9
nurses will be grouped together and work closely with nurse supervisors to practice the
procedure of biomarker collection. Questions and concerns regarding procedure will be
addressed in meetings following these pilots and before additional fieldwork.
3.5 Creation of identifiers for STI/HIV testing
In order to ensure confidentiality, a set of identifiers specific to biomarker data
collection (Biomarker ID numbers) will be created for each respondent prior to the
beginning of fieldwork. This set of Biomarker IDs will be different from the Survey
IDs used for the survey data collection. The Biomarker IDs will be used to label all
specimens, so that at no point will it be possible for the survey personnel to link the
survey information with the biomarker data. A master file that links this set of
Biomarker IDs with the original Survey IDs of respondents in the sample will be kept
only by the MDICP biomarker coordinators. Only after the end of data collection in
each site will the Survey IDs be linked with the Biomarker IDs.
4. COLLECTION OF BIOMARKERS
4.1 STI team
In each survey site, there will be 4 teams (5 in Balaka, because of its larger size) to
collect biomarker data. Each team will be composed of 5-6 nurses and 1 nurse
supervisor.
In addition to the nurses’ team, in each village there will be typically 1-2
‘scouts’, local villagers hired because of their extensive familiarity with village
residents, who will help to identify and locate respondents.7
In each village, the scout
will work first with the MDICP survey team, and subsequently with the STI team.
Depending on the logistics of the site, in some locations scouts will be village-specific,
whereas in others they will cover multiple villages. In addition to identifying and
locating respondents, scouts will make appointments with respondents for future nurse
visits.
Bike messengers will aid the STI team by providing transport when households
are located far apart.
7
Village headmen will assist in identifying potential scouts in each village.
Page 11 of 46
10
4.2 Data collection design
To reduce confusion amongst team members and respondents, an effort will be made to
avoid having both MDICP survey interviewers and nurses working simultaneously in
the same village. Respondents will be interviewed first by members of the MDICP
survey team. The MDICP interviewers will make the initial contact with respondents
and tell them to expect a visit from a nurse in the coming days.8
Once all the respondents in a village have been interviewed by the MDICP
survey team, the nurses will begin the STI data collection. First, the STI team will
meet with the village scout. The MDICP field logs, listing each person in the village
who have been interviewed by the MDICP and should thus be approached by the STI
team, will be given to nurse supervisors. The experience of the scout and the
documentation kept by the MDICP survey team will guide the STI team in finding
respondents, to ensure that each team approaches the same people. After arriving in the
village and meeting the scout, the nurse supervisor will assign nurses to respondents,
and provide each nurse with a set of STI labels for each respondent.
Upon meeting a respondent, the nurse will fill the STI questionnaire, discuss the
tests, ask for their informed consent and, if consent is given, collect the STI/HIV
specimens. If the respondent has accepted to be tested for HIV, the nurse will also
provide pre-counseling. In sum, the primary responsibilities of the nurse include:
- filling the STI questionnaire (see Section 4.4);
- informing eligible respondents about the test and asking for consent (see Section
4.5);
- collecting the specimens—urine samples and oral swabs for men; vaginal and oral
swabs for women (see Section 4.6);
- completing the necessary forms (see Section 4.7);
- telling respondents where and when they can go if they want to be given the
results.
Following a completed attempt with a respondent, the nurse will go back to the
nurse supervisor and be given a new set of ID’s and the location of a new respondent
from the scout.
8
Because the interviewer will obtain the initial consent and conduct the initial interview, he/she will also
be responsible for giving respondents the gift (a bag of sugar for each respondent interviewed) that is
provided to thank them for their cooperation.