Simultaneous versus video counting of coughs in hypertonic cough challenges

Abstract

Background: The coughs occurring during cough provocation tests are usually counted at the same time when the test is being conducted, i.e., simultaneously. It is unknown whether cough counting from video recording might increase the accuracy of the cough counting. During recent years, cough challenges with hypertonic aerosols have been introduced. They often provoke very frequent coughing which may complicate the simultaneous cough counting.

Objective: To assess whether cough counting from video recording is superior to simultaneous cough counting in two different hypertonic cough challenges.

Methods: The analysis includes 82 hypertonic saline challenges performed on 66 subjects, providing 1984 observation minutes with both simultaneous and video cough counting. The cough sensitivity was expressed as the osmolality to provoke 15 cumulative coughs (CUM15). The analysis also includes 136 hypertonic histamine challenges performed on 114 subjects providing 5373 observation minutes with both simultaneous and video counting. The cough sensitivity was expressed as the cumulative number of coughs divided by the final histamine concentration administered (CCR). This challenge involved several additional measurements to cough counting.

Results: For the saline challenge, the mean difference between the counting types was 0.0 coughs per minute with 95% limits of agreement of -1.2 to 1.2 coughs per minute. For the hypertonic histamine challenge the respective figures were 0.3 (-1.9 to 2.5) coughs per minute. At high coughing frequency the video counts tended to outnumber the simultaneous counts. The counting type had no effect on the hypertonic saline CUM15 and only a marginal effect on its repeatability. On the contrary, video counting resulted to significantly higher hypertonic histamine CCR values than simultaneous counting (p < 0.001).

Conclusion: The agreement between simultaneous and video counting of coughs is generally good. However, as the coughing frequency increases, simultaneous counting may miss coughs, especially if the nurse has to share his/her attention to several activities simultaneously. Video recording is advisable for the hypertonic histamine challenge but unnecessary for the hypertonic saline challenge. To ensure reliable simultaneous cough counting, cough provocation tests should be performed in a quiet environment, applying as little unnecessary equipment and measurements as possible.

Figure 1

The Bland-Altman plot for cough counts in hypertonic saline challenge. The difference between video vs. simultaneously counted coughs for each observation minute is plotted against the mean of the counted coughs per minute. The solid horizontal line represents the mean difference between the two counting methods and the dashed lines the 95% limits of agreement. The oblique line indicates the regression line.

Figure 2

The Bland-Altman plot for cough counts in hypertonic histamine challenge. The difference between video vs. simultaneously counted coughs for each observation minute is plotted against the mean of the counted coughs per minute. The solid horizontal line represents the mean difference between the two counting methods and the dashed lines the 95% limits of agreement. The oblique line indicates the regression line.

Figure 3

The Bland-Altman plot for CUM15, the index that was used to express cough sensitivity in hypertonic saline challenge. The difference between video vs. simultaneously counted CUM15 is plotted against the mean of the respective index. The solid horizontal line represents the mean difference between the two counting methods and the dashed lines the 95% limits of agreement.

Figure 4

The Bland-Altman plot for CCR, the index that was used to express cough sensitivity in hypertonic histamine challenge. The difference between video vs. simultaneously counted, log-transformed CCR is plotted against the mean of the respective index. The solid horizontal line represents the mean difference between the two counting methods and the dashed lines the 95% limits of agreement.