Back

Findings from a Pilot of a Near Real Time Disease Surveillance Tool on the NATO Mission in Kosovo

Lindfield, R.; Biendl, C.; Toth, M.; Ruhl, S.

2023-08-22 public and global health
10.1101/2023.08.21.23293305 medRxiv
Show abstract

NATO has a requirement for a near real time surveillance (NRTS) tool to rapidly identify disease outbreaks but no deployable tool is available. The NATO Centre of Excellence for Military Medicine (MILMED COE) developed a NRTS tool and piloted it on the NATO Mission in Kosovo (KFOR). Ten medical treatment facilities (MTFs) in KFOR piloted the tool which is an app on a smart device. All first presentations were eligible for inclusion into the app. Primary presenting symptoms were mapped to one or more of 40 symptoms. No patient identifiable information was collected. Analysis was available to each MTF and the Medical HQ staff (JMED) which provided a summary of the data collected. Combinations of symptoms resulted in an alert email to the JMED and MTF. An evaluation was conducted at two time points consisting of data analysis and semi-structured interviews with key informants from each participating MTF and JMED team. 1351 patients were entered during the pilot. 851 reports had symptoms recorded. The leading symptoms were pain-in-throat, cough and temperature. Two investigations were launched based on reporting in the app; a cluster of skin lesions and two cases of diarrhoea. No significant outbreaks were reported during the pilot or detected by the app. Interviews revealed that the app was easy to use and the symptom list was comprehensive. The analysis tool was not regularly viewed by MTFs however the JMED team used it frequently. Alerts were considered extremely valuable by both the MTFs and JMED team. The KFOR pilot provided an excellent opportunity to pilot the tool in a NATO Mission and established that the way it collected data and alerted issues worked well. Further validation and analysis of the tool is required before it can be rolled out further.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.