#7137 Factors associated with response completion of the hospital anxiety and depression scale (HADS) in people with severe acquired brain injury: a logistic regression analysis of routine outcome data

<h3></h3> Anxiety and depression are common symptoms after acquired brain injury that may influence engagement in rehabilitation and subsequent outcomes. The Hospital Anxiety and Depression Scale (HADS) is frequently used to assess these symptoms. However, to date there has been little investigation into the appropriateness of this scale for people with acquired brain injury, especially for those with more severe impairments. Some studies suggest that language or attention problems may present a barrier for completion of the HADS. Using logistic regression, we analysed routine clinical data from <i>N</i> = 916 individuals admitted into specialist neurorehabilitation to investigate the extent to which individual and injury characteristics, as well as clinical presentation on admission (as measured by the Mayo-Portland Adaptability Inventory – MPAI-4), were associated with completion of the HADS over time (i. e. on admission and discharge). Variables that were important factors in independently predicting completion of the HADS included increased verbal communication ability (<i>OR</i> = 0.87, <i>p</i> &lt; 0.01), lower irritability (<i>OR</i> = 0.80, <i>p</i> &lt; 0.01), and better initiation (<i>OR</i> = 0.80, <i>p</i> &lt; 0.001). Older age (<i>OR</i> = 0.99, <i>p</i> &lt; 0.01) and a stroke diagnosis (<i>OR</i> = 0.58, <i>p</i> &lt; 0.01) were associated with lower likelihood of fully completing the HADS. These findings are largely congruent with previous research identifying irritability as a barrier to completion of this measure, in conjunction with impaired verbal communication and shorter length of stay, supporting previous studies and expanding on their findings. We found that this constellation of personal and clinical characteristics collectively had significant predictive power, despite small effects (odds ratios) for some individual predictors and explaining a small proportion of the total variance (<i>Nagelkerke’s R²</i>= 0.15). Nevertheless, these results suggest a patient profile shared by people with mostly moderate to severe brain injury who are unlikely to fully complete the HADS on admission or discharge from post-acute rehabilitation services. Also, our results offer implications for practice when considering the repeated, valid, and reliable assessment of levels of anxiety and depression in this clinical population.