89 5 Discussion This study investigated associations between objective tests of mastication, swallowing, and salivary production and patients reported outcomes. The associations between objective tests and PROs were weak (correlation below 0.40) for all items, except one: a moderate correlation between xerostomia during the day versus total salivary flow at M6 (Spearman’s ρ=-0.441, p=0.001). In addition, none of the items on the EORTC QLQ-H&N35 questionnaire showed a significant correlation to the MAT nor WST. Even when focusing on patients with the highest 10% scores on the MAT or WST, indicating worst masticatory or swallowing performance (see the top part in Figure 2a-c), there was still a large variation of scores on the PROs taking up almost the entire scale. These findings indicate that the objective tests used in this research do not measure the same construct as the used PROs. Comparison with literature Our findings are in line with previous research in patients with HNC that showed that clinical measures and PROs generally correlate poorly.10,12,35 For example, swallowing research showed weak to strong associations between the 100 mL WST and the MD Anderson Dysphagia Inventory (MDADI) questionnaire (a questionnaire similar to the SWAL-QOL).10 Research about salivary measurements found weak associations between salivary flow and xerostomia scores.36 Other research stated that the EORTC quality of life questionnaire provided valuable data on subjective complaints, but that these complaints are not closely correlated with specific objective changes.37 Previous research in patients with Parkinson’s disease and patients with schizophrenia showed that objective and subjective measures were not interchangeable and each has a unique contribution to the problems assessed.38,39 Both objective and subjective measures may predict QoL in these patients.38 Using PROs alone does not seem to measure function the same way functional tests do, and therefore should be combined with other data sources.9 PROs are designed to assess how a patient evaluates his or her functioning rather than actual performance.40 In addition, functional disorders measured via instrumental assessments by clinicians may not have a strong relationship with how patients perceive this disorder. Patients are more likely to rate their symptoms more severely than do clinicians, which can lead to an underestimation of side effects post-treatment.41 Before selecting a measurement method, it is therefore important to identify the purpose of the measurement. For instance, when the effect of swallowing muscle sparing with RT is assessed, it is important to objectively test swallowing function. Whereas when the goal is to evaluate the effect of swallowing muscle sparing RT on perceived swallowing function of patients, it is important to use PROs. When only PROs are measured, it can be difficult to