Page 81 - Functional impairment and cues for rehabilitation of head and neck cancer patients -

Reliability and validity without spoon
When only including participants, who did not use a spoon during ingestion, results were comparable to results with all participants included. Reliability of duration measurements were good to excellent. The median SPEAD-rate was 2 g/s (range 0-10) for patients and 6 g/s (range 2-11) for healthy participants which corresponds to an effect size of 0.51. Correlation coefficients, only including participants without a spoon, with subjective swallowing outcomes ranged from 0.64 to 0.72 and from 0.44 to 0.64 for objective swallowing outcomes. Correlation coefficients with patient-reported dyspnea, pain and fatigue ranged from 0.20 to 0.29.
Cut-off value
The area under the ROC-curve when the SPEAD-rate was used to discriminate between patients and healthy participants was 0.82 (see Figure 4). The cut-off value with an optimal sensitivity and specificity ratio is 4.2 g/s (sensitivity 80% and specificity 79%). The area under the ROC-curve when the SPEAD-rate was used to discriminate between aspirating and not aspirating patients was 0.79 (see Figure 5). The optimal cut-off value is 1.2 g/s (sensitivity 100% and specificity 57%).
Figure 4 Ability of the SPEAD-rate to discriminate between HNC patients and healthy participants visualized in an ROC-curve. Area under the ROC-curve is 0.82.
3
The SPEAD-test to objectify swallowing capacity
79