Conference abstract
Evaluation of the surveillance system in Uganda: a case reliability for the typhoid surveillance data in Koboko - a rural and remote district in Northern Uganda, 2017
Pan African Medical Journal - Conference Proceedings. 2017:3(95).27
Oct 2017.
doi: 10.11604/pamj-cp.2017.3.95.216
Archived on: 27 Oct 2017
Contact the corresponding author
Keywords: Typhoid, integrated disease surveillance and response (IDSR), reliability, surveillance data
Oral presentation
Evaluation of the surveillance system in Uganda: a case reliability for the typhoid surveillance data in Koboko - a rural and remote district in Northern Uganda, 2017
Magdalene Akos Odikro1,&, Doreen Tuhebwe1, Anne Nasubo1, Steven Barnes Ssali1, George Holoya1, Robert Oliga1, Steven Ssendagireza1
1Makerere University, Kampala, Uganda
&Corresponding author
Magdalene Akos Odikro, Makerere University, Kampala, Uganda
Introduction: effective use of data for action largely depends on the quality of surveillance data. Typhoid is an epidemic prone and potentially fatal systemic infection. In Uganda, under integrated disease surveillance and response (IDSR), typhoid data is captured from health facility registers and reported through DHIS-2. From DHIS-2 data, Koboko district has been reporting increasing numbers of suspected typhoid cases since the beginning of 2017. This study ascertained reliability of these reported data and established the factors associated with this variation.
Methods: a cross sectional quantitative and qualitative study of health facilities in Koboko was conducted in June, 2017. Facilities were sampled using stratified random sampling. Quantitative data were collected through review of DHIS-2 records and health facility registers for EPI weeks 16 to 20. Qualitative data was collected using focus group discussions (FGDs) conducted with facility in-charges, records assistants, clinician, nurses and lab technicians. Descriptive analysis were performed and variation between EPI weeks and facilities reported using line graphs. Reliability was determined by assessing variation between DHIS-2 reported and facility level numbers. Concordance was set at < 20% variation. Qualitative data was analyzed using manifest analysis and reasons for variation presented using themes and quotes.
Results: of the 8 facilities visited, 1 was HC IV, 4 were health center IIIs and 3 were health center IIs. Between Epi weeks 16 to 20, all 8 facilities (40 weeks) had variation between facility and DHIS-2 reported numbers in at least one EPI week. Absolute variation ranged from 1 to 18 and from analysis of concordance, variation was present in 13/40 (32.5%) of the weeks. From the FGD, health workers reported lack of training, lack of common understanding of procedures on conducting counts and different naming of diagnosis as reasons for presence for variation. Suggested solutions to prevention of variation included trainings, in-house mentorship, double checking counts and regular support supervision.
Conclusion: there was discordance between health facility typhoid data and data reported in DHIS 2 (variation > 20%) indicating a challenge with data reliability from primary source. IDSR training is recommended for health workers to ensure standardization of operations.
