Labtracker+ , a medical smartphone app for the interpretation of consecutive laboratory results: an external validation study.

OBJECTIVES: When monitoring patients over time, clinicians may struggle to distinguish 'real changes' in consecutive blood parameters from so-called natural fluctuations. In practice, they have to do so by relying on their clinical experience and intuition. We developed Labtracker+ , a medical app that calculates the probability that an increase or decrease over time in a specific blood parameter is real, given the time between measurements.

DESIGN: We presented patient cases to 135 participants to examine whether there is a difference between medical students, residents and experienced clinicians when it comes to interpreting changes between consecutive laboratory results. Participants were asked to interpret if changes in consecutive laboratory values were likely to be 'real' or rather due to natural fluctuations. The answers of the study participants were compared with the calculated probabilities by the app Labtracker+ and the concordance rates were assessed.

SETTING AND PARTICIPANTS: Medical students (n=92), medical residents from the department of internal medicine (n=19) and internists (n=24) at a Dutch University Medical Centre.

PRIMARY AND SECONDARY OUTCOME MEASURES: Concordance rates between the study participants and the calculated probabilities by the app Labtracker+ were compared. Besides, we tested whether physicians with clinical experience scored better concordance rates with the app Labtracker+ than inexperienced clinicians.

RESULTS: Medical residents and internists showed significantly better concordance rates with the calculated probabilities by the app Labtracker+ than medical students, regarding their interpretation of differences between consecutive laboratory results (p=0.009 and p<0.001, respectively).

CONCLUSION: The app Labtracker+ could serve as a clinical decision tool in the interpretation of consecutive laboratory test results and could contribute to rapid recognition of parameter changes by physicians.