e-Prevalence Survey

The e-Prevalence survey supports measurement of TB program effectiveness and SDG achievement. First proposed by CheckTB! to the NTP of Ghana in 2012 it allows all participant data - including the dCXR - to be merged in one electronic file that can be sent over any mobile phone network to a central database. This significantly reduces survey cost and time to complete, increases data quality & integrity and therefore set new standards.

A typical survey in a TB high burden setting reaches a sample size of 500-800 clusters with 50,000- 80,000. Normally 12-15 staff members are needed per field team for 6–10 months with 3-4 teams in operation. Average costs of traditional surveys are estimated at US$ 2-3 million including Capital Investment (CXR, Lab, Cars), Human Resource (salaries), Survey Operation (Field and Central), Pre & Post survey events, technical assistance. e-Prevalence allows for significant reductions in operational cost and lead time.

Pre-conditions:

  • Use proven IT technology, mobile phone network and stable PACS software
  • Full electronic (from questionnaire to final report)
  • Use tablets or laptops to capture patient details, perform questionnaires and print barcode labels
  • Full patient data must be retained for patient follow-up should TB be detected
  • Only specific details in prevalence survey data - Age, Sex, Region, etc
  • Digital Chest X-ray (DR) & PACS with automatic image compression
  • Participant data and dCXR to be sent automatically via mobile phone or satellite network
  • Computer Aided Detection - to be used as a classifier normal vs. abnormal, as 2nd opinion on site or as 3rd reading for QC

Process:

Commonly used in high burden countries WHO/STOP TB: Strategy 3

Start
  1. Participant data collected on site electronically with a tablet computer which can be entered automatically into the central database followed by the printing of a unique barcode
  2. Questionnaire (downloading from 1) all applicable c.q. permitted personal data); using the selection menu on the tablet computer for scoring and automated counting of result
  3. On site making of a digital Chest X-ray (downloading from 1) all applicable c.q. permitted personal data, including the CAD abnormality score and related reading results on site)
  4. Automatic back-up on local computer workstation and forwarding of the participant file including questionnaire, CAD score and compressed CXR over 3G or 4G network to central database for data consolidation and in case applicable centralized 2nd reading of the CXR
  5. Fill out fields for smear microscopy and culture results if applicable

Survey Platform

Survey Platform

In above graph the all electronic platform for TB prevalence surveys is illustrated.


Survey differences Analog Digital
Electronic participant data collection no yes
Electronic questionnaire and scoring no yes
Barcoding for unique participant ID no yes
High CXR readability/image quality no yes
CXR image cost (per throughput) € 3.50 - 4.50 € 1.00 - 2.50
Immediate image availability no yes
Picture Archiving & Communication System no yes
Electronic data management & privacy tools no yes
Film & chemical elimination no yes
Remote & Computer Aided CXR reading no yes
Automated data back-up no yes