Discussion

Outpatient antimicrobial parenteral therapy is an increasingly popular therapy modality for patients who need intravenous antimicrobial therapy without acute care, but it is not a benign intervention despite the known benefits. Specific challenges include increased need for monitoring, personal responsibility of the patients to administer antibiotics, and need for institutional support. ^{Reference Chary, Tice and Martinelli1}

In our retrospective study, the most common diagnoses requiring long-term antimicrobials were bacteremia and musculoskeletal infections, and most patients were discharged home with home health services. Although we cannot make inferences on causation, the trend that our patients discharged home were readmitted more frequently than those discharged elsewhere may indicate that patients need more or better home health support. This support could also be provided by telehealth, with an ability to check in on patients more frequently to allow for earlier inteventions before major problems occur. ^{Reference Chary, Tice and Martinelli1}

We noted significant ADEs related to ertapenem and vancomycin, which were 2 of the most-prescribed antibiotics and the agents with the most overall ADEs (Fig. 3). Previous studies have highlighted a favorable safety profile for ertapenem, so additional studies will be needed to further evaluate the ADEs in our cohort of patients. ^{Reference Qureshi, Syed and Doi15,Reference Sanroma, Muñoz, Mirón-Rubio and Aguilera16} Ertapenem is widely used due to its favorable once-daily administration convenient for OPAT; however, this convenience may need to be reassessed in light of ADEs.

Compared to other intravenous antimicrobials, vancomycin is notoriously more difficult to manage in the outpatient setting because of the need to closely monitor drug levels, which affect dosage and frequency of administration. This factor relates to a higher rate of ADEs both in our study and in other studies, often leading to a change in the antimicrobial regimen or to early discontinuation of therapy. ^{Reference Schrank, Wright and Branch-Elliman4} Also, therapeutic implications are associated with obtaining vancomycin trough levels. If the trough is obtained at the incorrect time or there is a delay in sending out the adjusted vancomycin dose, patients may experience days of subtherapeutic treatment. We were not able to find any evidence to support difficulties with obtaining vancomycin levels in the outpatient setting. Among internal data, of 218 vancomycin levels drawn, only 34% fell within the target therapeutic range of 15–20 mcg/L (unpublished data). This finding further supports the need for the use of other anti-MRSA agents that do not require serum-level monitoring (ie, daptomycin), though other factors, such as antimicrobial stewardship and cost or insurance coverage, may need to be considered.

The IDSA guidelines provide recommendations for laboratory monitoring for specific antimicrobials that may be used in OPAT. ^{Reference Norris, Shrestha and Allison8} Our study spanned 2017–2019, with publication of the guidelines in 2018. Thus, for a large portion of our patients, monitoring was conducted at the physician’s discretion. Transaminase elevation was the third most common ADE, occurring in 35% of patients. Transaminase elevation may have been related to the underlying infection or secondary to other medications (ie, acetaminophen), but it is worth considering when developing a laboratory monitoring plan for a patient. However, this factor may be a moot point in future studies because the guidelines are adapted into clinical practice. A balance exists between the guideline recommendations, diagnostic stewardship, and unnecessary laboratory tests, as well as the individual patient. Regardless, Keller et al ^{Reference Keller, Ciuffetelli and Bilker17} reported that following the introduction of an ID-based transition of care for OPAT patients, compared to patients without ID-based care, receipt of laboratory test results was higher (increasing from 37.4% to 94.3%). This finding supports the importance of good teamwork in OPAT.

The rate of adverse effects related to PICC lines in our population was 0.05%. This result is significantly lower than those reported in other studies, though many other studies use days present as evaluation for PICC-related side effects. This low rate may reflect uncaptured events in patients who were rehospitalized. It may reflect robust hospital epidemiology interventions targeted at preventing central-line–associated bloodstream infection, in addition to patient education prior to discharge. In a study of emergency department visits related to OPAT, the most common reason for emergency department visit was vascular-access complications (n = 104), with 53 visits related to occlusion and 19 related to dislodgement. ^{Reference Shrestha, Kim and Rehm18} Few PICC events related to those who went to daily infusion centers or other healthcare centers, and this could be a consideration for some patients who may struggle with PICC care at home though this comes with its own challenges with transportation to infusion centers.

In our OPAT experience, 84 patients were rehospitalized; most of whom were initially discharged home (85.7%). Although this is a reasonable complication rate and we cannot infer causation in this retrospective study, it could be improved. Others note that outpatient follow-up within 2 weeks was associated with lower risk of all-cause 30-day readmission (adjust odds ratio = 0.33; P = .0001), ^{Reference Saini, Ali and Du19} but short-term follow-up is not always feasible for the both the patient and the physician. The role of telehealth visits to “check in” on OPAT patients on a weekly or biweekly bases needs to be explored further. Tan et al ^{Reference Tan, Ingram and Rothnie14} have shown telehealth to be a safe and efficacious way to monitor geographically remote patients in their study of 88 OPAT encounters. We did not evaluate ID follow-up appointments nor the correlation with 30-day readmissions. Studies have demonstrated that improved management of OPAT (ie, led by an ID physician) can reduce the probability of readmission or emergency department visits. Shah et al ^{Reference Shah, Petrak and Fliegelman20} found that in OPAT programs led by ID physicians, patients not only had lower 30-day hospital admission rates but also paid less in total healthcare payments. ^{Reference Shah, Petrak and Fliegelman20} OPAT programs led by ID physicians have also been reported to have more frequent laboratory testing and follow-up as well as fewer errors in prescribing. ^{Reference Keller, Ciuffetelli and Bilker17}

Our study had several limitations. We lacked data from hemodialysis centers because we do not follow patients there after providing initial recommendations. Not all OPAT records contained patient medical record numbers; therefore, some patients were excluded. This study was conducted at a single center, which may limit generalizability. We also lacked data in 3% of patients who were lost to follow-up, and these patients may have had significant adverse effects, such as rehospitalization, ADEs, or death, which are not captured and thus may have skewed the data. Furthermore, the IDSA guidelines for OPAT came out in the midst of our study period, which affected clinical practices regarding OPAT in ways that we cannot account for in this retrospective review.

In conclusion, we examined >600 patients utilizing OPAT in this study, and we observed that most of our rehospitalized patients were initially discharged home. Ertapenem and vancomycin were associated with the most ADEs among our cohort, which was unexpected for ertapenem due to its simple dosing schedule and favorable safety profile. We noted transaminase elevation in 23% of patients, which may suggest that measuring liver function needs to be considered for medication monitoring. Our rate of PICC-line adverse events in our study population was significantly lower than those in other studies, and most occurred in those who were discharged home.

In this dynamic healthcare landscape, we must continue to explore ways to optimize surveillance of our OPAT patients. We must always strive to minimize adverse effects and other barriers that prevent us from achieving our therapeutic goals.