Schematic of the Monte Carlo method applied.

Background

Puumala hantavirus (PUUV) causes nephropathia epidemica (NE), an endemic form of transient acute renal injury (AKI). Serological testing is the mainstay of diagnosis. It was the aim of the present study to assist decision-making for serological testing by constructing a simple tool that predicts the likelihood of PUUV positivity.

Methods

We conducted a comparative cohort study of all PUUV-tested cases at Aachen University tertiary care center in Germany between mid-2013 and mid-2021. N = 293 qualified for inclusion; N = 30 had a positive test result and clinical NE; N = 263 were negative. Two predictive point scores, the Aachen PUUV Score (APS) 1 and 2, respectively, were derived with the aid of logistic regression and receiver operating characteristic (ROC) analysis by determining the presence of four admission parameters. For internal validation, the internal Monte Carlo method was applied. In addition, partial external validation was performed using an independent historic cohort of N = 41 positive cases of NE.

Results

APS1 is recommended for clinical use as it estimated the probability of PUUV positivity in the entire medical population tested. With a range from 0 to 6 points, it yielded an area under the curve of 0.94 by allotting 2 points each for fever or headache and 1 point each for AKI or LDH>300 U/L. A point sum of 0–2 safely predicted negativity for PUUV, as was confirmed in the NE validation cohort.

Conclusion

Here, we present a novel, easy-to-use tool to guide the diagnostic management of suspected PUUV infection/NE and to safely avoid unnecessary serological testing, as indicated by point sum class 0–2. Since 67% of the cohort fell into this stratum, half of the testing should be avoidable in the future.