Management of subcutaneous abscesses: prospective cross-sectional study (MAGIC)

Résumé fourni par la source

Subcutaneous abscesses are pus-filled cavities within the dermis and subdermal layers caused by a local bacterial infection, most commonly Staphylococcus aureus1. They are a common presentation, with an incidence of 433 per 100 000 presentations in community practice, leading to 236 per 100 000 admissions to hospital each year2. Surgical drainage is the conventional approach used to treat abscesses in the hospital setting. This aims to achieve complete evacuation of pus, thorough irrigation, and healing by secondary intention. In the UK, no widely established guidance exists for the management of subcutaneous abscesses. The results of a survey of general surgeons in 2018 highlighted stark disparities in the everyday management of abscesses, such as in the choice of anaesthesia, the role of cavity packing, and the role of swabs for culture; in particular, uncertainty surrounding the use of general or local anaesthesia during surgical drainage was shown to exist3. A cohort study at a large teaching hospital later demonstrated the safety of a protocolized care pathway for drainage under local anaesthesia when applied to the majority (84%) of patients with uncomplicated abscesses4. The choice of anaesthesia remains of interest due to its potential impact on health resources, costs, the environment, and emergency care capacity. Another uncertainty is the role of antibiotics after surgical drainage. In the previous survey, the majority of respondents indicated that antibiotics were only indicated sometimes (39.3%) or in rare circumstances (44.3%)3. This was in contrast to a recently published practice recommendation, advocating for the routine use of antimicrobial therapy after surgical drainage, specifically trimethoprim/sulfamethoxazole or clindamycin5. This recommendation was made owing to the results of a systematic review published in 2018, which demonstrated a modest reduction in abscess recurrence with antibiotics across 14 RCTs. In contrast, the review revealed an increase in the risk of gastrointestinal side effects, such as diarrhoea, and also raised concerns about other possible harms, such as from antimicrobial resistance5,6. The aim of this study was to explore the management of subcutaneous abscesses in emergency surgical practice in the UK. In doing so, the study aimed to describe the UK experience of local anaesthesia as the primary anaesthetic modality for surgical drainage and the experience of antibiotic therapy as an adjuvant treatment after surgery. The study protocol was made prospectively available online. The Health Research Authority Decision Tool and a National Health Service (NHS) Research & Innovation Service confirmed that NHS Research Ethics Approval was not required. The study was registered with governance departments at each participating hospital, including confirmation of Caldecott Guardian approval for the safe transfer of non-identifiable data via the Research Electronic Data Capture (REDCap) platform7. The study is reported in line with the STROBE checklist8. A prospective, multicentre, observational study was delivered across NHS hospitals in the UK by the White Rose Surgical Collaborative. No change to routine clinical practice was made. The study opened for data inclusion on 1 September 2022 and closed on 1 December 2022, followed by a 60-day follow-up interval of routinely collected data. The inclusion interval was split into 1-month data collection intervals (Table S1), with participating hospital sites asked to contribute to at least one interval. A hospital-specific, 10-item survey exploring standards of management at each hospital was piloted and disseminated to local lead investigators. Any secondary or tertiary care NHS hospital providing an acute emergency general surgery service in the UK was eligible to contribute. No limits on case volume or local policy for the management of acute subcutaneous abscesses were applied. Included patients were identified from any relevant acute care setting, including accident and emergency departments, acute surgical assessment and ambulatory units, and inpatient wards. Any adult (greater than or equal to 18 years old) presenting with an acute subcutaneous abscess was eligible for inclusion. The abscess location had to be on the posterior neck, anterior or posterior thorax, abdomen, groin, buttock, natal cleft, or proximal limbs. All other abscesses, including peri-anal, breast, distal limb, and facial abscesses, were excluded, as these were considered to demand specialty-specific treatment not provided by an emergency general surgical practice. The remaining exclusion criteria comprised abscesses associated with a previous surgical site, abscesses with known body cavity communication, and lesions or ulcers considered to be of vascular aetiology. The key outcomes of interest were abscess recurrence and return to hospital. Recurrence was defined as any recurrent abscess or surgical-site infection at the same anatomical site. Return to hospital was defined as any re-presentation, irrespective of whether this led to an inpatient admission. Both outcomes were measured using routinely collected data up to 60 days from the index treatment. For patients managed with antibiotics, the index treatment was the time of first antibiotics. For patients managed with surgery, the index treatment was the time of surgery. Explanatory variables were collected to provide a risk-adjusted estimate of outcomes. These included patient-related factors, such as age (years), sex, BMI (kg/m2), smoking status, and relevant co-morbidities (diabetes mellitus, hidradenitis suppurativa, and immunosuppressive therapeutics or conditions), and treatment-related factors, such as pre-hospital and hospital treatment (surgical drainage, needle aspiration, antibiotics, or no treatment). For patients undergoing surgery, anaesthetic and surgical variables were also collected, such as type of anaesthesia (general, local, or other), surgical setting (operating room, emergency department, or ward/treatment area), and operator grade. Patient and treatment characteristics for the entire cohort are presented descriptively as rates, mean(s.d.) values, or median (interquartile range (i.q.r.)) values. Effect estimates are presented as OR (95% c.i.) values. For the subgroup of patients undergoing surgical drainage, multivariable regression models were constructed for each of the clinical outcomes (abscess recurrence and return to hospital) to assess whether either of these was independently associated with: the use of postoperative antibiotics; the use of local anaesthetic; and packing of the abscess cavity. Explanatory variables were entered into the models using a stepwise approach. In addition, to examine whether variation existed between centres, multivariable models containing only patient-level variables along with corresponding multi-level models treating the admitting hospital as a random effect were generated for postoperative antibiotics, anaesthetic type, and abscess packing as outcomes. Residual and standard deviations of the random effects were examined and likelihood ratio tests were performed to identify differences in the models at the centre level. This generated P values to determine whether statistically significant centre-level variation existed for each outcome independent of case mix. Analyses were performed using SPSS® (IBM, Armonk, NY, USA; version 22) and R statistical computing software (R Foundation for Statistical Computing, Vienna, Austria; version 4.3.3), with P < 0.050 indicative of statistical significance. A total of 1637 records were collected from 54 NHS hospitals in the UK. After removal of duplicates, 1631 were considered in the analyses. The median number of records submitted per hospital was 27 (i.q.r. 15–44.5) during a median of 2 (i.q.r. 1–2) data collection intervals. A response to the survey was received from 54 of 54 hospitals (100%). A defined clinical pathway or protocol for the