Results:
The initial survey had a 45% resident response rate (22 of 49). Of these residents, five (22.7%) were satisfied with the post-hospitalization primary care follow-up process, 11 (50%) were neither satisfied nor dissatisfied, and six (27.3%) were dissatisfied (Figure 5).
There was a large variation in responses to the free text question asking residents to identify their role in establishing post-hospitalization PCP follow-up. Some residents identified their primary role as communicating with the patients’ PCP, with responses such as “letter to PCP”, “call outside PCP”, “route discharge summary to their (patient’s) PCP’s fax”, or “messaging the PCP to let them know that their patient was hospitalized”. Other residents identified their primary role as ordering, or placing a request for, follow-up appointments with the following responses: “placing an order on discharge”, “usually request follow-up but not involved in scheduling”, “putting in the order”, and “placing an order and hoping it happens”. Lastly, some residents stated that their role in post-hospitalization follow-up was to advise the patient to contact their PCP office to independently make follow-up appointments with responses including: “collect the PCP information and put it in the patient discharge instructions”, “making sure the patient knows to ask outside PCP for appointment”, “advise the patient to see their own PCP within the week”, and “let the patient schedule their own follow-up”. In response to the question “What aspects of the post-hospitalization primary care follow up process work well?” residents primarily identified one component, with nine out of the 21 free text responses stating that establishing patient follow-up was much easier within the Mayo Clinic Health Care Network. Another six out of the 21 residents responded with variations of “I don’t know” or “I am not sure”.
Twenty residents provided free text responses to the question “What aspects of the post-hospitalization primary care follow-up process don’t work well”? Residents identified difficulty in communicating with PCPs (especially when outside the Mayo network), incomplete documentation or difficulty finding documentation of patients’ PCP information in the EMR, and uncertainty about whether follow-up appointments actually occur. Of the 17 residents who answered the question “What do you think would improve how arrangements are made for PCP follow-up?”, six residents suggested increased administrative support, five residents described standardized order sets and protocols, four residents recommended improved means for communication with PCP, one resident asked for clearer documentation of patient’s PCP contact information, and one resident was unsure.
Following our first intervention, 34 out of 49 residents (69.4%) responded to the survey and, although not statistically significant, there was an improvement when compared to the initial survey by 16.7% (p=0.20). Two residents (6.1%) reported being very satisfied with the process, 11 (33.3%) satisfied, 14 (42.4%) neither satisfied nor dissatisfied, and 6 (18.2%) dissatisfied (Figure 5). Following our second intervention, 12 of the 49 residents (24%) responded to the third survey. There was a statistically significant increase in satisfaction rates from baseline by 44% (p= 0.007). Two out of twelve (16.7%) responding residents reported being very satisfied with the post-hospitalization primary care follow-up process, six (50%) satisfied, four (33.3%) neither satisfied nor dissatisfied, and zero residents reported being dissatisfied or very dissatisfied (Figure 5).
Prior to any intervention, the baseline survey showed only one out of 22 (4.5%) responding residents strongly agreed that the time and workload associated with establishing primary care follow-up post hospitalization was not too burdensome. Seven out of 22 (31.8%) agreed, seven (31.8%) neither disagreed or agreed, and seven (31.8%) disagreed (Figure 6). Following the first intervention there was no increase in perceived workload burden (p= 0.58). Three of 34 (8.8%) residents strongly agreed that the workload was not too burdensome, 14 (41.2%) agreed, six (17.6%) neither agreed nor disagreed, 10 (29.4%) disagreed, and one (2.9%) strongly disagreed (Figure 6). After the second intervention, there again was no increase in perceived workload burden from baseline (p= 0.19). Two out of 12 residents (16.7%) strongly agreed, six (50%) agreed, one (8.3%) neither agreed nor disagreed, three (25%) disagreed, and zero residents strongly disagreed (Figure 6).