There were also cases where a category only partially corresponded to one of the predicted sources of paper generation (e.g., sensorimotor preferences) or did not correspond at all (e.g., security). These are represented in the last row of Table 3 (“Other”). For example, ‘sensorimoter preferences’ is represented by (e), “not well integrated into workflow” , in cases where participants used paper because they preferred having paper notes with which to walk into and use in the patient’s room. However, the case where a nurse preferred to “hear” paper being dropped in her basket (indicating a new unscheduled patient had checked in) is not represented by the predicted sources of paper generation in fig. 1. Since frameworks and models are place-holders in time to describe current knowledge, the 11 categories outlined in Table 2 give a more comprehensive and descriptive representation of factors that lead to the persistent use of paper with the EHR at the human-technology integration level. These factors help us understand workflow and cognitive reasons individuals use paper rather than the EHR; this knowledge is important for informing future EHR; this knowledge is important for informing future EHR design.