Several retrospective cohort studies of TOA management report medical management failure rates, ranging from 18–59% [6, 7]. We report a comparatively low failure rate of 15.6%, which may reflect our institution’s longstanding experience in managing this condition. Alternatively, some studies define primary parenteral antibiotic failure as the need to switch antibiotic regimens during the patient’s admission. Including these cases, our failure rate would have risen to approximately one-fourth of cases. Further, we noted that in ten cases, TOAs were evaluated for image-guided drainage but did not receive drainage due to the location of the TOA, where drainage could not be performed safely. These cases thus received prolonged antibiotic therapy, which ultimately led to symptomatic improvement, but required a prolonged hospital stay. Had these TOAs been successfully drained, our success rate with primary medical management would have been lower.
Abscess size is consistently identified as a factor associated with medical management failure. However, thresholds for this association differ across studies, ranging from TOA size of 4.5 to 8 cm [12, 13]. In our study population, medical management failure was independently associated with the abscess size, with the smallest TOA failing medical management measuring 5.6 cm. Image-guided drainage of TOAs is more effective and associated with fewer complications as compared to operative management of complicated TOAs [13]. If performed for TOAs greater than 5.5 cm in diameter, early drainage could avert primary medical management failure and decrease the length of hospital stays. In our sample, we noted that drainage was not significantly associated with a higher likelihood of resolution of the TOA during follow-up imaging. However, the null finding may have been related to poor follow-up rates whereby those with successful imaging-guided drainage may have been less likely to follow-up. Recommendations for drainage may be strengthened in cases of leukocytosis as our logistic regression modeling noted the independent association of increasing white blood cell count with failure of primary medical management. This finding is supported by three previous studies noting a similar association, citing a threshold of 15x103/uL across all studies [12, 15–17].
Given the need for long-term follow-up and maintenance of records within a single medical system, few studies report on hospital readmission rates for TOA. We noted a readmission rate of 13.6% in our population, which is similar to previous retrospective cohort studies with varying aims and duration of follow-up, reporting rates of 10–24% [18–19]. However, the true readmission rate is unknown given the possibility of readmission to hospitals outside of the electronic medical record of public hospitals in Los Angeles County in emergent settings, as well as out-of-network care in cases where individuals graduate from public insurance. Of note, nearly half of patients with a recorded readmission required multiple admissions for TOA management, suggesting that early procedural intervention may be warranted especially upon hospital readmission.
Nearly three-quarters of TOAs (72%) that were followed in the outpatient setting persisted on initial follow-up imaging. Currently, there are no guidelines on how to evaluate TOAs following initial hospitalization, likely arising from differing timeframes for follow-up amongst existing data, as well as a lack of standard definition of TOA resolution. For example, reported TOA resolution rates in retrospective cohort studies range from 20% at up to two years defined by symptoms or imaging findings [19], to 85% at two weeks defined by improvement in symptoms and exam findings [20]. Beyond ensuring clinical improvement, standardization of follow up should be considered to avoid long term complications of persistent abscess including readmission.
One of the strengths of our study is the fairly large number of TOAs reported and a unique population consisting of mainly reproductive age, Latinx females, largely with public insurance. Our study is limited by its retrospective nature and dependence on non-standardized TOA documentation, limiting our ability to draw conclusions on the full range of factors that might contribute to the failure of medical management. Further, as patients may receive follow-up care from hospitals outside the Los Angeles Department of Health Services system, our estimate of readmission may be an underestimate, especially considering our post-hospitalization attrition rate of 31%. We additionally note that the generalizability may be limited to other groups of population, however, they may still be useful to support practice changes such as the early incorporation of imaging-guided drainage.