Reliability of NIHSS by telemedicine in non-neurologists


  • Conflicts of interests: The authors declare no conflict of interest.

Correspondence: Thierry Moulin, Department of Neurology, CHU Besancon, 2 Boulevard Fleming, 25000, Besancon 25000, France.


Dear Editor, the majority of stroke patients do not receive thrombolysis because they are still managed in general wards in community hospitals rather than in stroke units [1]. This has negative consequences for stroke patients who require rapid assistance and specialist expertise [2-4].

Our prospective study aimed to assess the reproducibility and reliability of National Institutes of Health Stroke Scale (NIHSS) scoring by telemedicine. The tests were performed by two local bedside examiners (neurologists and non-neurologists) and two remote examiners (junior and senior neurologists). All 11 NIHSS items along with their different grades were tested. The mean duration was 15:09 min. The overall video quality was judged as being excellent to fairly good in 96% of cases and none of the 28 patients had to be excluded for technical reasons.

The mean NIHSS scores were 9·32 and 9·14 for the local and remote neurologists, and 9·11 and 8·93 for the local and remote non-neurologists, respectively. Concordance was high between the two neurologists (Fig. 1); for the NIHSS items, there were only two weighted Kappa coefficients <0·75. In nine cases, total NIHSS scores were identical, and a deviation of ≤2 points was observed in only four cases. Fewer concordances were noted between the local and remote non-neurologists (Fig. 1), with two weakly correlated items (facial palsy, ataxia), and six items with a weighted Kappa coefficient <0·75. Total NIHSS scores were identical in five patients and deviation was observed in 11 patients (2 points in six cases and 3 points in five cases). Overall, the maximum difference in NIHSS scoring was always <4 points.

Figure 1.

Global concordances between local and remote neurologists (top panel), and local and remote non-neurologists (bottom panel) (regression line model, validated by the Kendall coefficient).

NIHSS scores can thus be reproduced, obviously, by local and remote neurologists and above all by non-neurologists. It is therefore possible for both neurologists and non-neurologists to remotely and accurately assess the clinical status of a stroke patient and make decisions regarding thrombolysis by telemedicine.