596. The ID Physician Is Out: Are Remote ID E-Consults an Effective Substitute?

Abstract Background Telemedicine (TM) can provide specialty ID care for remote and underserved areas; however, the need for dedicated audio-visual equipment, secure and stable internet connectivity, and local staff to assist with the consultation has limited wider implementation of synchronous TM. ID e-consults (ID electronic consultations or asynchronous™) are an alternative but data are limited on their effectiveness, especially patient outcomes. Methods In the setting of the COVID-19 pandemic and ID physician outage, we were asked to perform ID e-consults at a 380-bed tertiary care hospital located in Blair County, PA. We performed retrospective chart reviews of 121 patients initially evaluated by ID e-consults between April 2020 and July 2020. Follow-up visits were also conducted via e-consults with or without direct phone calls with the patient. Key patient outcomes assessed were length of stay (LOS), disposition after hospitalization, 30-day mortality from initial ID e-consult and 30-day readmission post-discharge. Results The majority of patients were white males and non-ICU (Table 1). The most common ID diagnosis was bacteremia (27.3%, 33/121), followed by skin and soft tissue infections (15.7%, 19/121) and bone/joint infections (14.9%, 18/121) (Figure 1). Table 2 shows patient outcomes. Average total LOS was 11 days and 7 days post-initial ID e-consult. 48.7% (59/121) of patients were discharged home and 37.2% (45/121) to a post-acute rehabilitation facility. 2.5% (3/121) of patients required transfer to a higher level of care facility; none of which were to obtain in-person ID care. The index mortality rate was 3.3% (4/121), which appears to be lower than published data for in-person ID care. The 30-day mortality rate was 4.1% (5/121), which is also comparable to previously reported for ID e-consults. 25.6% (31/121) of patients required readmission within 30 days but only 14.0% (17/121) were related to the initial infection. Table 1. Demographics *Immunosuppressive agents include: Apremilast, Dasatinib, Etanercept, Remicade, Rituximab, and Prednisone >10 mg/day Figure 1. Variety of ID Diagnoses made by e-consults Table 2. Outcomes Conclusion We believe that this is the first report of the implementation of ID e-consults at a tertiary care hospital. Mortality rates appear to be comparable to in-person ID care. In the absence of in-person ID physicians, ID e-consults can be a reasonable substitute. Further study is required to compare performance of ID e-consults to in-person ID consults. Disclosures John Mellors, MD, Abound Bio, Inc. (Shareholder)Accelevir (Consultant)Co-Crystal Pharma, Inc. (Other Financial or Material Support, Share Options)Gilead Sciences, Inc. (Advisor or Review Panel member, Research Grant or Support)Infectious DIseases Connect (Other Financial or Material Support, Share Options)Janssen (Consultant)Merck (Consultant) Rima Abdel-Massih, MD, Infectious Disease Connect (Employee, Director of Clinical Operations) Rima Abdel-Massih, MD, Infectious Disease Connect (Individual(s) Involved: Self): Chief Medical Officer, Other Financial or Material Support, Other Financial or Material Support, Shareholder


Session: P-27. Clinical Practice Issues
Background. Telemedicine (TM) can provide specialty ID care for remote and underserved areas; however, the need for dedicated audio-visual equipment, secure and stable internet connectivity, and local staff to assist with the consultation has limited wider implementation of synchronous TM. ID e-consults (ID electronic consultations or asynchronous™) are an alternative but data are limited on their effectiveness, especially patient outcomes.
Methods. In the setting of the COVID-19 pandemic and ID physician outage, we were asked to perform ID e-consults at a 380-bed tertiary care hospital located in Blair County, PA. We performed retrospective chart reviews of 121 patients initially evaluated by ID e-consults between April 2020 and July 2020. Follow-up visits were also conducted via e-consults with or without direct phone calls with the patient. Key patient outcomes assessed were length of stay (LOS), disposition after hospitalization, 30-day mortality from initial ID e-consult and 30-day readmission post-discharge.
Results. The majority of patients were white males and non-ICU (Table 1). The most common ID diagnosis was bacteremia (27.3%, 33/121), followed by skin and soft tissue infections (15.7%, 19/121) and bone/joint infections (14.9%, 18/121) ( Figure 1). Table 2 shows patient outcomes. Average total LOS was 11 days and 7 days post-initial ID e-consult. 48.7% (59/121) of patients were discharged home and 37.2% (45/121) to a post-acute rehabilitation facility. 2.5% (3/121) of patients required transfer to a higher level of care facility; none of which were to obtain in-person ID care. The index mortality rate was 3.3% (4/121), which appears to be lower than published data for in-person ID care. The 30-day mortality rate was 4.1% (5/121), which is also comparable to previously reported for ID e-consults. 25.6% (31/121) of patients required readmission within 30 days but only 14.0% (17/121) were related to the initial infection. Conclusion. We believe that this is the first report of the implementation of ID e-consults at a tertiary care hospital. Mortality rates appear to be comparable to in-person ID care. In the absence of in-person ID physicians, ID e-consults can be a reasonable substitute. Further study is required to compare performance of ID e-consults to in-person ID consults.
Disclosures Methods. Patient (pt) records were evaluated from Jan 2019 -July 2019 and compared to Jan 2020 -July 2020. Data collected included new OPAT pts, demographics, infection type, location prior to OPAT and therapy characteristics. Statistical analysis was performed using Chi-square test with p< 0.05 considered statistically significant.
Results. Fourteen POICs reported data with a total of 2410 new OPAT pts in 2019 and 1807 in 2020, representing a decrease of 25%. Table 1 shows the comparison of OPAT characteristics between 2019 and 2020. Mean age and gender were similar, but there was a significantly higher percentage of pts ≥65 years treated in 2020 (43% vs. 36%, p< 0.001). Infection type and location prior to OPAT were consistent between 2019 and 2020. Primary antimicrobial use was comparable with the exception of cefepime, which showed a greater use in 2020 (14% vs. 11%, p=0.006). OPAT management differed significantly from 2019 to 2020 with fewer pts completing therapy as prescribed in 2020 (85.9% vs. 88.3%, p=0.021), driven largely by more early discontinuations and switches to oral therapy. Other reasons for those not completing therapy were also significant and due primarily to transfer of care to other settings, most commonly the home (1.9% vs. 2.9%, p=0.029). Overall length of therapy was comparable.

Table 1. Comparison of OPAT in 2019 (Pre-COVID) and 2020 (Post-COVID)
Conclusion. OPAT provided through ID POICs experienced a substantial decrease in pts treated during the first half of 2020 compared to 2019. This was expected with the decline in healthcare services, especially elective procedures. Most pt and treatment characteristics were comparable between years, but interestingly, more elderly received OPAT during the pandemic and fewer completed therapy as planned. Further analysis of these differences can help determine effects of the pandemic on overall health outcomes in the OPAT population. Methods. Records from 18 POICs were queried for GUI pts with ≥1 GNP receiving OPAT from 2018 to 2020. Demographics, pt location prior to OPAT, infection type, year of therapy, and GNP were recorded. Antibiotic resistance patterns were defined as extended-spectrum beta-lactamase (ESBL) or multi-drug resistant (MDR). Chi Square and Fisher's exact test were used to determine if ESBL status was associated with GNP or location prior to OPAT (hospital vs. community). The Cochran-Armitage test was used to analyze temporal trend in ESBL expression. Statistical significance was defined as P< 0.05 for all tests.

A 3-Year Evaluation of Antibiotic Resistance Patterns in Gram-Negative Genitourinary Tract Infections Treated in Outpatient Infusion Centers (POICs)
Results. A total of 634 GNP were identified in 601 pts (mean age: 64±16, 58% female). Infections were 75% complicated urinary tract infection, 20% pyelonephritis, and 5% prostatitis/other. Overall, 56% (n=339) were treated directly from the community and 44% (n=262) following hospital discharge. GNP isolated were 56% E. coli, 19% Pseudomonas spp., 16% Klebsiella spp. and 9% others. Of the 611 GNP with potential to express ESBL, 43% (n=265) were ESBL producers (Table 1). Significantly more ESBLproducing GNP occurred in pts discharged from a hospital prior to OPAT compared to the community (53% vs. 36%, P< 0.001). Overall, the incidence of MDR constituted 36% (n=231) of GNP, which did not differ by location prior to OPAT. Evaluation of ESBL incidence by year showed a significant increase from 2018 to 2020 (P=0.03). Although a slight increase in MDR was noted from 2018 to 2020, this was not significant (Figure 1).