Statistics
The statistics unit supports the three main programmes, virology, bacteriology and malaria, in collaboration with the TEG (Tropical Epidemiological Group) at LSHTM. The support begins at the project design stage, with experimental design and sample size issues, moving through to the analysis stage at the end of the project. During projects staff of the unit are involved in randomisation tasks and often serve on Data Safety Monitoring Boards to satisfy the GCP requirements. The unit also offers training support to the programs and have run a number of statistics and Stata courses. Future plans include establishing an on-line sample size calculator standardising the many methods and packages that staff use.

In addition to supporting the programs, the unit undertakes applied research relevant to the unit. Current areas of interest are the processing and analysis of laboratory data. Flow cytometry produces huge volumes of data and the unit is developing algorithms for the automatic and objective analysis of FACS data. A software package written in Matlab, which automates data capture from spoligo films has been developed within the unit. Download package and user guide.

The unit also works on a number of coding projects to ensure the integrity of the data, whether it is from laboratory or field data. This involves work in a number of high-level programming languages and complements the work of the Data Management unit.

Data Management
The move to SQL Server based systems is now in full swing and many changes have been implemented. The unit now boasts 3 full-time Database Developers and they are helping to move the technical aspects of our systems considerably. We have also recently started experimenting with OPENCLINICA and open source data management tool, and have two studies piloting this.

Amongst the innovations have been: a double-entry verification system for SQL Server Databases using a web-based tool in ASP.NET; the evolution of a more generic approach to studies (enabling shorter development time and more standardisation); and the (on-going) development of an electronic laboratory sample tracking system.

The Unit decided to use pooled rather than programme-based Data Entry, and we are currently implementing this in a managed fashion. This will allow us to be more flexible with the use of our data entry staff, and free up the Data Managers time to allow them to concentrate on quality issues within their studies. Data Managers will be trained to ensure that issues relating to data quality are paramount for all our studies.

Malaria was the first programme to pilot the use of handheld computers to gather data directly without the use of paper. The project has been running since October 2006, and we are very keen to see success in this field.