1
2	Background KEMRI/Wellcome Trust Collaborative Malaria Research Project between Kenyan scientists and counterparts from Oxford University etc since 1989. Kilifi is 60km north of Mombasa, one of poorest districts in Kenya, high malaria prevalence and infant mortality rate. Unit holds large clinical, laboratory, demographic databases ICT has affected the way research is done at the Unit.
3	Previously … Data Management Data recorded on paper q’naires and double entered Standalone Databases for Clinical, Lab, Census data Access only through Data Managers Connectivity Dial-up Access via terrestrial lines through an analogue exchange. Data rate ~ 9,600 bps (approx. 10% achieved). Unreliable, costly, low quality
4	Previously … Information Access (Library) Subscription to paper based journals Medline searches on CD-ROM Circulation of photocopies of papers Physical visit to library Physically meet to discuss papers
5	ICT Infrastructure at the Unit Massive Expansion of PC installations 25 standalone in 1998 to 102 networked in 2002. 3 servers – 2 configured as file, mail, print and web proxy servers and the other as a dedicated database server. Expansion based on demand as a result of increased research activity and need for fast information exchange. Connectivity VSAT link courtesy of the NLM – July 1999 Data rate: 32Kbps outbound, 768Kbps shared (BE ~ 512K)
6
7	Satellite Dish at Kilifi
8	Adding Value to Infrastructure Kilifi Virtual Library Online Admission System Centralised Data Access Laboratory Data Capture System
9	Kilifi Virtual Library In-house centralised system, developed to give researchers easy access to online resources. Librarian updates content Abstracts stored as part of an indexed searchable database Researchers use EndNote to search virtual library Populated using entries from PubMed to standardise entry of references e-TOC groups exchange TOCs from latest journals Journal Web Database (JWD) gives journal info to users
10	Kilifi Virtual Library
11	Kilifi Virtual Library
12	Kilifi Virtual Library
13	Kilifi Virtual Library (JWD)
14	Online Admission System Data entered in real time at point of admission Data entered by FW and MOs Data almost readily available once committed to the DB Single data entry
15	Online Admission System
16	Success Stories Increase in amount of scientific data downloaded (weekly statistics – sawmill logs). Faster, reliable exchange of data, incl. e-groups. Fewer data entry clerks, info available almost immediately, near paper-less environment. Data Dictionaries and Metadata can be shared electronically, making it easy to understand datasets. Data can be accessed from nearly any computer at the unit.
17	Problems Resistance to change Maintenance (down time, unavailable parts, power supply) Weaknesses of single data entry (not yet evaluated) Inappropriate use of systems (sawmill logs)
18	Bandwidth Utilisation Data
19	Bandwidth Utilisation
20	Conclusions and Future Works ICT has changed the way things are done The way data is collected and processed The way data is accessed Access to more online resources Comprehensive study to assess the true impact: Papers published etc Data quality using single entry Reluctance by some scientists to adopt.