REPORT ON

WORKSHOP ON ANALYSIS OF LONGITUDINAL VACCINATION DATA

Nairobi, November 26^th–30^th, 2001

by:

Peter Aaby, Bandim site & Danish Epidemiology Science Center (Scientific Committee)

Philippe Bocquier, French Institute for Research in Africa (Facilitator)

Peter Byass, Dept. of Public Health and Clinical Medicine, Umeå University (Scientific Committee)

Pierre Ngom, African Population and Health Research Center (Scientific Committee)

Joanna Armstrong Schellenberg, Ifakara Health Research Center (Scientific Committee)

Jonathan Sterne, Dept. of Social Medicine, University of Bristol (facilitator)

December 6, 2001

A. Aims

This workshop aimed at building capacity in the INDEPTH sites in the following areas:

·         Design and planning of longitudinal studies, including investigating the effect of vaccines on child survival;

·         Use of Cox regression modeling to analyze longitudinal survival data, taking vaccination as an example;

·         Presentation, discussion and interpretation of findings from the above analyses;

The workshop was also expected to contribute to a minimum standard for collection of vaccination information which would permit INDEPTH sites in the future to carry out multi-site analyses.

B. Software

STATA was the official software of the workshop. During the workshop, each site was provided with a fully licensed copy of STATA and a ‘Getting Started’ manual. These items were retained by each site after the workshop.

C. Data sets

All participants used a dataset provided by the Bandim site. A few sites came with their own data sets and some of the participants managed to carry out some analyses on their own data.

E. Attachments

Attached are the following items:

1.       The revised timetable as used during the workshop

2.       A summary table showing the nature of vaccination data collected by all sites participating to the workshop;

3.       A table summarizing the evaluation of the workshop by participants;

4.       The list of reading materials provided to workshop participants.

5.       The list of workshop participants (this is a PDF file under separate attachment); and

F. Collection of vaccine information in Indepth field sites – status and recommendations

Data collection in relation to time of vaccination: All projects are collecting retrospective information about previously given vaccines through surveys (see Appendix 2). Some projects have given vaccines in connection with trials and Bandim is trying to collect routine information at the time of vaccination from health centres.

Hospital data: At the moment (Nov 2001) only Bandim is collecting data at admission to a pediatric department. Several projects may have the opportunity and interest to do so. A module for data collection at admission is available from Bandim.

Data collection intervals: For most projects the interval is 3-4 months. The length of the interval is clearly associated with the cost of data collection so no general recommendation can be defined but it would be desirable that new projects collect data at 3-4 months intervals to ensure comparability to with existing projects. If data is collected with longer intervals, follow-up time should be censored 3-6 months following the surveys as it cannot be expected that vaccination status at a given moment in infancy will continue to have an influence and other vaccines may be given. 

Source of vaccination information: All projects are using vaccinations cards but there is variability in how the information is used. Main issues are: how to use recalled information if a vaccination card or other documentation is not available. Though recalled and documented vaccination dates can be combined, analyses including only the documented vaccination should be presented.

G. Future suggestions

Technical workshops such as this seem to be an effective strategy for building capacity across the Indepth network. One suggestion for the future could be to stage a more open-ended workshop in which participants could pursue their own analyses and writing with the advice and encouragement of facilitators and colleagues; such an activity would need to be based in an academic institution with good library and internet facilities.

H. Acknowledgements

Funding of the workshop was provided, through the INDEPTH Network, by the Rockefeller Foundation, SIDA/SAREC, and the Mellon Foundation. The workshop was hosted by the African Population and Health Research Center. Although they could not attend the workshop, both the INDEPTH Chair, Dr. Fred N. Binka, and the Director of African Population and Health research Center, Dr. Alex Ezeh, sent messages that were conveyed to all workshop participants.

The organizational aspects of the workshop were perfect thanks to the professionalism of APHRC staff; special thanks to Ms. Anne-Marie Oyuga and Ms. Margaret Shivele of APHRC.

The Scientific Committee (Peter Aaby, Peter Byass, Pierre Ngom and Joanna Schellenberg) and the Facilitators (Philippe Bocquier and Jonathan Sterne) were immense resources to the workshop.

All sites which have responded to the workshop call must be commended for sending in highly qualified participants.

Back to Events