A new study suggests the answer may be no. The study by Rosen et alia (http://www.ncbi.nlm.nih.gov/pubmed/26214322) shows that computer use by behavioral clinicians during intake negatively affects the quality of the therapeutic alliance–one of the best predictors of outcome– and continuance in treatment.
It appears that the use of a computer (staring at the screen rather than the client, manipulating a mouse, inputting data) interferes with the development of a therapeutic relationship (e.g. trust, comfort, and understanding) and that in turn impacts whether the client returns for subsequent sessions.
The Rosen study did not indicate type of computer used–desk top, tablet–and the findings are far from established. Another study by Wardia et al(http://www.ncbi.nlm.nih.gov/pubmed/24377404) published last year did not find differences in therapeutic alliance when clinicians used different methods to record data (paper and pencil, ipad, or desktop computer). However, that study did not assess dropout rate and for alliance it used an instrument of very limited psychometric utility: the 4 item session rating scale. In contrast, the Rosen study used a much more robust and established measure, the 12 item (7 likert scale) Working Alliance Inventory.
The takeaway, as I see it, is that clinicians need to be cautious about using a computer, especially during the early sessions of psychotherapy when rapport is being established.