Post-cardiac surgery mediastinitis (PSM) is a serious, complex, and multifactorial complication of surgical procedures. Infectious diseases consultation (IDC) has demonstrated improvement in other complex infectious diseases. The objective of the study was to evaluate the impact of IDC in the management and outcome of patients with PSM.
MethodsObservational retrospective study, of adult patients with PSM between January 2010 and June 2021. After January 2016, IDC was performed in all the patients with PSM. The primary endpoint was clinical success, a composite variable of clinical cure, and absence of adverse events, or recurrence. Also, in-hospital stay, and clinical cure was evaluated in patients that received oral sequential therapy (OST).
ResultsA total of 84 patients with PSM were included, 48 pre-IDC and 36 in IDC period. No differences in clinical success were observed between the two periods (pre-IDC 60% vs, IDC 77%, p=0.104). During the IDC period the rate of adequate targeted antibiotic treatment was higher (pre-IDC 71% vs. IDC 94%, p=0.016). Gram-negative bacilli infections (pre-IDC 42% vs. IDC 78%, p=0.002) and polymicrobial infections (pre-IDC 37% vs. IDC 63%, p=0.004) increased in the IDC period. Multivariate analysis did not show any variable associated with clinical success. OST was similar in both periods, and a shorter in-hospital stay was observed in the patients who underwent OST (no-OST, 70 days vs. OST, 44 days, p=0.003).
ConclusionsIDC was related with a higher adequate targeted antimicrobial therapy. We observed that OST offers a promising strategy in the management of this infection.
La mediastinitis poscirugía (MPQ) cardíaca es una complicación grave, compleja y multifactorial de la cirugía. La interconsulta de enfermedades infecciosas (IEI) mejora el manejo y evolución de otras enfermedades infecciosas. Evaluamos el impacto de la IEI en el manejo y evolución de esta dolencia.
MétodosEstudio observacional retrospectivo de pacientes adultos con MPQ entre enero de 2010 y junio de 2021. Después de enero de 2016, se realizó IEI en todos los pacientes. El objetivo principal fue el éxito clínico, variable compuesta de curación clínica y ausencia de eventos adversos o recurrencia. Además, se evaluó la estancia hospitalaria y la curación clínica en los pacientes que recibieron terapia secuencial oral (TSO).
ResultadosSe incluyeron 84 pacientes con MPQ, 48 pre-IEI y 36 en periodo post-IEI. No se observaron diferencias en el éxito clínico entre los 2 periodos (pre-IEI 60% vs. IEI 77%, p=0,104). Durante el periodo IEI, la tasa de tratamiento antibiótico dirigido adecuado fue superior (pre-IEI 71% vs. post-IEI 94%, p=0,016). Las infecciones por bacilos gramnegativos (pre-IEI 42% vs. post-IEI 78%, p=0,002) y las polimicrobianas (pre-IEI 37% vs. post-IEI 63%, p=0,004) aumentaron en el periodo post-IEI. El análisis multivariable no mostró ninguna variable asociada con el éxito clínico. El porcentaje de pacientes con TSO fue similar en ambos periodos, y en estos pacientes la estancia hospitalaria fue menor (sin TSO, 70 días vs. con TSO, 44 días, p=0,003).
ConclusionesLa IEI se relacionó con una mayor adecuación de tratamiento antimicrobiano dirigido. La TSO ofrece una estrategia prometedora en el manejo de esta infección.
Post-cardiac surgery mediastinitis (PSM) is a serious and complex disease,1 with a low frequency (0.25–5%).2 Although there is recent evidence regarding the management of deep sternal wound infections, there are many unsolved questions. Surgical debridement is considered the mainstay of therapy, supported by antibiotic treatment.2,3 In addition, in many centres, negative suction wound therapy (NSWT)/vacuum assisted closure (VAC) is employed as a cornerstone for deep sternal wound infections. PSM is considered a severe complication due to its high mortality, that can reach 14% in the first three months after surgery.1
The risk of suffering a PSM varies depending on the characteristics of the patients. Patients older than 75 years, with obesity or insulin-dependent diabetes mellitus, the use of double mammary arteries, or patients with prolonged ventilatory support have an increased risk of PSM.4–6 Different pathogenic factors have been associated with this severe infection, being wound contamination the most important.4 It is possible that the degree and type of contamination interact with multiple factors for mediastinitis to develop, such as the patient's baseline condition, the need for an urgent surgery, the adequate antimicrobial surgical prophylaxis and postoperative cares.2,4,5,7–9
The impact of infectious diseases consultation (IDC) has been evaluated in different infections, such as in cases of Staphylococcus aureus or enterococcal bacteremia, candidemia and multi-drug resistant infections.10–12 IDC has been related to a reduction in morbidity and mortality rates, in-hospital stay, critical care unit admission, shorter duration of intravenous antibiotic therapy and, adequate source control.13–17 However, there is no data regarding the evaluation of IDC in patients with PSM. The main objective of the study was to analyze the impact of IDC in the management and outcome of patients with PSM.
MethodsObservational-retrospective study, before–after, of adult patients (older than 18 years) with PSM. The study was performed in the Cardiac Surgery Department of the University Hospital Complex of Vigo. Patients with the diagnosis of PSM, identified from the hospital's Coding Service, between January 2010 and June 2021 were included. A single episode of PSM per patient was included. Patients with unconfirmed PSM, patients who died within the first 48h and those with loss of follow-up were excluded.
A shared care programme in patients with PSM was started in January 2016. Previously, the PSM patients were attended by internal medicine physicians without specific formation in infectious diseases. After this date, all patients with PSM were attended by the cardiac surgical team and infectious disease specialists (IDS). The pre-IDC period, from January 2010 to December 2015, was compared with the IDC period, from January 2016 to June 2021. Preoperative screening for nasal carriage of S. aureus was introduced in 2017.
Primary and secondary outcomesThe primary endpoint was a composite variable, clinical success, consisting of clinical cure, and absence of adverse events, or recurrence. Clinical success was compared between both periods.
Secondary objectives were to analyze the in-hospital stay, the percentage of adequate antimicrobial therapy between both periods and to evaluate the evolution of patients who underwent oral sequential therapy (OST).
Variables and definitionsThe diagnosis of PSM was established according to the guidelines of the Centres for Disease Control and Prevention.18 Patients had to meet at least one of the following criteria: (1) Presence of organisms in mediastinal tissue or fluid by culture or a non-culture-based microbiologic method performed for clinical diagnosis or treatment; (2) Evidence of mediastinitis in a gross anatomical or histopathological examination; (3) Signs or symptoms of infection, such as fever (greater than 38°C), chest pain and/or sternal instability, and at least one of the following: purulent drainage from the mediastinal area and/or widening of the mediastinum in an imaging test.
Clinical cure was defined as the resolution of signs and symptoms of infection evaluated 6 months after finishing the antimicrobial therapy. Readmission due to recurrence was defined as the need of a new hospitalization due to the appearance of signs and symptoms of wound infection, after a previous improvement, within 6 months after diagnosis. Side effects were defined as injuries or symptoms resulting from the medication19; specifically: nephrotoxicity, encephalopathy, seizures, haematological events or any other toxicity. We defined the perioperative period as that from surgical planning to full recovery.
Recurrence was defined as the reappearance of signs and symptoms of mediastinitis within 90 days after the start of treatment with the isolation of the same microorganism. Superinfection was considered as the reappearance of signs and symptoms of mediastinitis, with the isolation of a different microorganism, within 90 days after the start or completion of treatment.
The protocol applied by IDS consisted of collecting samples for culture (blood cultures and wound swabs), surgical debridement, and vacuum-assisted closure (VAC) therapy. Empirical antibiotic treatment was intravenous cefepime (2g every 8h) and vancomycin (15–20mg/kg every 12h), adjusting the dosage to obtain 10–20mg/L through level. After receiving the results of microbiological cultures, the treatment was modified to the most appropriate depending on the microorganism, and to avoid adverse effects.
Oral sequential therapy (OST) was carried out after two weeks of intravenous therapy and after observing a satisfactory evolution of the surgical wound. Oral antimicrobial election was based on the antibiogram and bioavailability. The total duration of antimicrobial therapy depended on the evolution of the surgical wound and was, at least, 6 weeks. ECDC criteria of the point prevalence survey database of antibiotic use were employed to establish the adequacy of antibiotic therapy.20 The patients’ clinical and epidemiological characteristics, such as age, sex and comorbidity, determined by Charlson index,21 were collected. The severity of the infection was classified according to SEPSIS-3 criteria.22
MicrobiologyCollection, transport, and processing of samples were carried out following the procedure recommendations of the SEIMC (Spanish Society of Clinical Microbiology and Infectious Diseases). Antimicrobial susceptibility tests were performed using the automated Vitek® system and interpreted following EUCAST breakpoints. The isolation of a coagulase-negative staphylococci was considered a pathogen when the same species was isolated in 3 or more samples. In all other cases the isolation was considered a contaminant.
Statistical analysisThe statistical package SPSS v24.0 (IBM, Chicago, IL) was used for data analysis. A descriptive analysis of all the variables included in the study was performed. Quantitative variables were described as median±interquartile range. Qualitative variables were described by their absolute and relative frequencies. For the analysis of the clinical success, binary logistic regression was performed. The p-value to stablish the statistical significance was<0.05.
Ethical statementWhen conducting the study, we followed the indications of the Declaration of Helsinki. The local Ethical Committee (RESEARCH ETHICS COMMITTEES GALICIA) approved the study protocol (code 2021/389). STROBE recommendations were followed to ensure the reporting quality of the study.
ResultsA total 84 PSM episodes were included: 48 (57%) in the pre-IDC period and 36 (43%) in the IDC period. The PSM incidence rate was similar in both periods (pre-IDC period 14.35 cases per 1000 surgeries and IDC period 13.79 cases per 1000 surgeries).
Baseline characteristicsThe baseline characteristics of the patients are shown in Table 1. Pre-IDC patients had a higher median age (pre-IDC 80 years [69–85] vs. IDC 72 years [67–78], p=0.005) with a lower percentage of males (pre-IDC 23% vs. IDC 53%, p=0.006). No differences were observed in other baseline clinical characteristics, including body mass index, Charlson comorbidity index, or severity of infection at diagnosis.
Characteristics of the patients with post-surgical mediastinitis.
| Total | Pre-IDC (n=48) | IDC (n=36) | p | |
|---|---|---|---|---|
| Median age, years (IQR) | 77 (67–82) | 80 (69–85) | 72 (67–78) | 0.005 |
| Sex, male, n (%) | 30 (36) | 11 (23) | 19 (53) | 0.006 |
| Comorbidities, n (%) | ||||
| Diabetes mellitus | 34 (41) | 17 (35) | 17 (47) | 0.369 |
| Ischaemic heart disease | 35 (42) | 18 (38) | 17 (47) | 0.383 |
| Heart failure | 26 (31) | 15 (31) | 11 (31) | 0.999 |
| Peripheral artery disease | 25 (30) | 13 (37) | 12 (33) | 0.632 |
| COPD | 23 (27) | 13 (27) | 10 (29) | 0.999 |
| Chronic kidney failure | 9 (11) | 3 (6) | 6 (17) | 0.163 |
| Obesity | 38 (45) | 23 (48) | 15 (42) | 0.660 |
| Charlson index≥3, n (%) | 34 (41) | 17 (35) | 17 (47) | 0.369 |
| Surgical procedures, n (%) | ||||
| Urgent cardiac surgery | 11 (13) | 4 (8) | 7 (19) | 0.193 |
| Duration≥5h | 75 (89) | 43 (90) | 32 (89) | 0.999 |
| Double internal mammary artery | 45 (54) | 22 (45) | 23 (64) | 0.124 |
| Perioperative transfusion | 14 (17) | 11 (23) | 3 (8) | 0.137 |
| Reintervention<4 days | 3 (4) | 1 (2) | 2 (6) | 0.574 |
| Perioperative infection | 31 (37) | 16 (33) | 15 (42) | 0.497 |
| Adequate prophylaxis | 57 (68) | 27 (63) | 30 (83) | 0.048 |
| Median time since previous surgery, days (IQR) | 12 (7–21) | 11 (7–24) | 12 (9–20) | 0.477 |
| Clinical manifestations, n, % | ||||
| Pain | 20 (24) | 13 (27) | 7 (19) | 0.469 |
| Fever | 53 (63) | 31 (65) | 22 (61) | 0.821 |
| Purulent drainage | 80 (95) | 47 (98) | 33 (92) | 0.309 |
| Instable sternum | 23 (27) | 10 (21) | 13 (36) | 0.120 |
| Wound dehiscence | 79 (94) | 48 (100) | 31 (86) | 0.012 |
| SOFA≥3, n (%) | 24 (29) | 15 (31) | 9 (25) | 0.628 |
Note: COPD, chronic obstructive pulmonary disease; IDC, infectious disease consultation; IQR, interquartile range; SOFA, Sepsis related Organ Failure Assessment.
Regarding the previous surgery, during the IDC period a greater percentage of adequacy of antibiotic prophylaxis (pre-IDC 63% vs. IDC 83%, p=0.048) was observed. The use of double internal mammary-artery (pre-IDC 45% vs. IDC 64%, p=0.124), urgent cardiac surgery (pre-IDC 8% vs. IDC 19%, p=0.193) or required blood transfusion (pre-IDC 33% vs. IDC 8%, p=0.137) were similar in both periods. The clinical presentation of PSM was similar between both groups; purulent drainage was the most common sign of infection (95%), followed by wound dehiscence (94%) and fever (63%). In addition, median time since previous surgery was similar (pre-IDC 11 days [7–24] vs. IDC 12 days [9–20], p=0.477).
Management of patients with PSMA higher percentage of VAC system (pre-IDC 71% vs. IDC 97%, p=0.001) was performed in the IDC period, and there were no significant changes in surgical debridement (pre-IDC 18% vs. IDC 36%, p=0.082) (Table 2). The vast majority of patients in our series underwent closure of the surgical wound without the need for cover. Only in exceptional cases with great loss of substance they required coverage. In our series, only 3 patients out of the 84 included required it. The choice of need and type of coverage was individualized for each patient. A better adequacy of targeted antibiotic treatment was found in the IDC period (pre-IDC 71% vs. IDC 94%, p=0.016). The duration of antimicrobial therapy (pre-IDC 44 days [28–65] vs. IDC 51 days [33–74], p=0.631) and side-effects (pre-IDC 27% vs. IDC 11%, p=0.100) were similar among both periods. The most common side-effects were acute kidney failure (7 patients, 8%) and cytopenia (5 patients, 6%). There were no differences in the median hospital stays (pre-IDC 37 days [25–52] vs. IDC 39 days [26–76], p=0.426), or in readmission due to poor evolution (pre-IDC 13% vs. IDC 8%, p=0.726).
Management and outcome of patients with post-surgical mediastinitis.
| Global (n=84) | Pre-IDC (n=48) | IDC (n=36) | p | |
|---|---|---|---|---|
| Surgical treatment,n, % | ||||
| Surgical debridement | 22 (26) | 9 (18) | 13 (36) | 0.082 |
| Time until debridement, days (IQR) | 15 (1–48) | 6 (0–54) | 18 (6–49) | 0.458 |
| Use of VAC | 69 (82) | 34 (71) | 35 (97) | 0.001 |
| Delayed closure | 33 (39) | 21 (43) | 13 (36) | 0.509 |
| Time to closure, days (IQR) | 42 (31–72) | 38 (30–66) | 50 (30–77) | 0.417 |
| Antimicrobial treatment,n(%) | ||||
| Appropriate targeted treatment | 62 (74) | 30 (71) | 32 (94) | 0.016 |
| Oral sequential therapy | 66 (79) | 36 (75) | 30 (83) | 0.428 |
| Duration, days (IQR) | ||||
| Oral | 14 (6–21) | 14 (5–21) | 15 (7–21) | 0.236 |
| Total | 46 (31–67) | 44 (28–65) | 51 (33–74) | 0.631 |
| Median hospital stays, days (IQR) | 38 (26–72) | 37 (25–52) | 39 (26–76) | 0.426 |
| Side effects,n, % | 17 (20) | 13 (27) | 4 (1) | 0.100 |
| Cytopenia | 5 | 4 | 1 | |
| Acute kidney failure | 7 | 5 | 2 | |
| Rash | 3 | 2 | 1 | |
| Phlebitis | 3 | 2 | 1 | |
| Clinical outcome,n, % | ||||
| Clinical success | 55 (66) | 28 (60) | 27 (77) | 0.104 |
| Clinical cure | 77 (92) | 44 (92) | 33 (92) | 0.971 |
| Readmission | 9 (11) | 6 (13) | 3 (8) | 0.726 |
| Death≤6 months | 5 (6) | 4 (8) | 1 (3) | 0.386 |
| Superinfection | 20 (24) | 10 (21) | 10 (28) | 0.605 |
Note: IDC, infectious disease consultation; IQR, interquartile range; VAC: vacuum-assisted closure.
The microbiological diagnosis was confirmed by wound swab culture in 90% of the episodes and 25% of the patients had positive blood cultures (Table 3). There was a reduction in PSM without microbiological confirmation in the IDC period (pre-IDC 30% vs. IDC 8%, p=0.015).
Microbiological cultures.
| Global (n=84) | Pre-IDC (n=48) | IDC (n=36) | p | |
|---|---|---|---|---|
| Positive cultures | ||||
| Blood | 22/56 | 13/29 (27) | 9/27 (25) | 0.999 |
| Exudate | 74 (88) | 42 (90) | 32 (89) | 0.999 |
| Negative cultures | 8 (10) | 15 (31) | 3 (8) | 0.015 |
| Isolates | ||||
| Gram-positive cocci | 43 (51) | 23 (48) | 20 (56) | 0.516 |
| S. aureus | 20 (24) | 10 (21) | 10 (28) | 0.605 |
| MRSA | 4/20 | 3/10 | 1/10 | 0.582 |
| CoNS | 11 (13) | 10 (21) | 1 (3) | 0.020 |
| Enterococcus spp. | 13 (16) | 4 (8) | 9 (25) | 0.065 |
| Gram-negative bacilli | 48 (57) | 20 (42) | 28 (78) | 0.002 |
| Enterobacterales | 43 (51) | 17 (35) | 26 (72) | 0.001 |
| ESBL | 6/43 | 2/17 | 4/26 | 0.999 |
| Carbapenemase | 1/43 | 1/17 | 0 | 0.409 |
| Pseudomonas aeruginosa | 11 (13) | 3 (6) | 8 (22) | 0.048 |
| Monomicrobial infection | 33 (39) | 24 (63) | 9 (27) | 0.004 |
Data are shown as n (%).
Note: CoNS, coagulase-negative staphylococci; ESBL, extended spectrum beta-lactamase; MRSA, methicillin resistant S. aureus.
We observed a high increase of Gram-negative bacilli infections in the IDC period (pre-IDC 42% vs. IDC 78%, p=0.002). Not only due to Enterobacterales (pre-IDC 38% vs. IDC 72%, p=0.002), but also due to Pseudomonas aeruginosa (pre-IDC 6% vs. IDC 25%, p=0.025). The percentage of patients with enterococcal infection was similar in both periods (pre-IDC 8% vs. IDC 25%, p=0.065). Finally, a higher percentage of polymicrobial infections was observed in IDC (pre-IDC 37% vs. IDC 63%, p=0.004).
The percentage of multidrug-resistant microorganisms was low in both periods. Methicillin resistant S. aureus (MRSA) was more prevalent in pre-IDC period (pre-IDC 3/35 vs. IDC 1/34) and extended spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae were more common in IDC period (pre-IDC 2/35 vs. IDC 4/34). No isolates with difficult to treat P. aeruginosa were identified.
Clinical successTwo patients were excluded from the clinical success analysis since they died due to a heart failure during follow-up; each patient belonged to each of the study groups. We analyzed the characteristics and evolution of the patients with clinical success, detailed in the Supplementary material (Supplementary material S1). None of the analyzed variables was more frequent in both univariate and multivariate analysis (Table 4).
Oral sequential therapyWe analyzed the characteristics and evolution of the patients who underwent oral sequential therapy (OST), which is detailed in the Supplementary material, S2. Baseline characteristics of patients were similar among both groups. Multidrug-resistant isolates (MRSA and ESBL) were more frequent in the non-OST group.
Duration of antimicrobial therapy was similar among both groups (non-OST 63 days [34–78] vs. OST 44 days [30–60], p=0.074). A shorter in-hospital stay was evidenced in the OST group (non-OST 70 days [33–125] vs. OST 44 days [25–48], p=0.003). We observed no differences in the percentage of readmissions among both groups (non-OST 17% vs. OST 9%, p=0.395). In the non-OST group superinfection was observed more frequently (non-OST 44% vs. OST 18%, p=0.030). Multivariable analysis showed that OST was independently associated with a lower risk of superinfection (OR 0.2, 95%IC [0.07–0.66], p=0.007).
DiscussionThe lack of a clear consensus on PSM management reflects the complex, multifaceted nature of this condition, and underscores the need for further research to establish evidence-based guidelines. The present work is the first that analyses the impact of the IDC in management of patients with PSM. Although we could not demonstrate a greater clinical success after IDC, we observed indirect and beneficious effects of this multidisciplinary approach, with a higher adequate targeted treatment.
Evaluation by IDC has been observed in other studies in patients with complex pathologies, other than PSM.13–17 We observed that IDC did not increase the increase in surgical debridements by surgeons, despite of source control recommendations (pre-IDC 19% vs. IDC 36%, p=0.082). This percentage remains very low, compared to previous publications.23 Among cardiac surgeons, concern about poor evolution of patients after surgical debridement is common. Nowadays, most surgeon teams usually prefer a less aggressive source control, that is, VAC therapy.3 This approach has been related with shorter hospital stays, decreased rates of superinfection, and even with decreased early mortality.3
Another change associated with the IDC was an increase in antimicrobial adequacy, as in other studies.13–17 This also could be related to an improvement in microbiological diagnosis of PSM, since the proportion of patients with negative cultures decreased between both periods (pre-IDC 13% vs. IDC 5%, p=0.007) which probably indicates an improvement in the collection of clinical samples. The duration of antibiotic treatment was shorter than in other studies, where the median treatment duration was 182 days.23,24
We also observed that OST was used in a high proportion of patients (78%) with excellent results (clinical success 71%). This data has not been previously reported in the literature. Probably because most of the publications have been performed by surgeons, and routes of administration of antibiotics are not recorded. In our work, we found a significant reduction in hospital stays in patients with OST (non-OST 70 days vs. OST 35 days, p=0.003). This is in accordance with previous studies related to other infections.25–27 Moreover, we observed a lower probability of superinfection in the follow-up (non-OST 44% vs. OST 18%, p=0.030). This fact could be explained because of an early discharge that decreases the risk of nosocomial infections28 and probably intensive intravenous antibiotic therapy regimens can select antibiotic-resistant organisms and facilitate superinfection.
Over the period of study, we observed a significant increase of Gram-negative bacilli infections (pre-IDC 42% vs. IDC 78%, p=0.002), and a reduction in monomicrobial infections (pre-IDC 63% vs. IDC 27%, p=0.004). The increase in Gram-negative bacilli infections was not only associated with an increase of Enterobacterales, but also of P. aeruginosa. Nonetheless, over the two periods we did not observe an increase of multidrug-resistant strains. Our results contrast with other works where Gram-positive organisms remain the most common aetiology.23,29–31 This could be related to different causes. On the one hand, we used a strict criterion for considering coagulase-negative staphylococci as the causative agents of the infection, since these pathogens are common skin colonizing bacteria. This could explain the 46% of infections due to Staphylococcus epidermidis in a previous cohort.32 On the other hand, the increase of Gram-negative bacilli could also be related with the in introduction of S. aureus screening and decontamination in 2017.
Double internal mammary artery has been classically related with a higher rate of infection.4 In our centre, around 98% of patients with surgical revascularization undergo mammary artery grafting, and surgical site infection remains low (1.4%).
This study has some limitations. In fact, it is a quasi-experimental, before–after study, where data were recorded retrospectively. This could affect the identification of some factors that might have changed over the period of study. Secondly, the study was conducted in a single centre, with a small number of patients, thus the observations could not be applicable to other institutions. Third, though the period of study comprises more than 10 years, we could not include a high number of patients, since mediastinitis is an uncommon complication. This fact has limited the statistical power to find association between IDC and clinical success. Fourth, a high number of infections were caused by Gram-negative bacilli, including P. aeruginosa. No molecular epidemiology studies were performed to rule out an outbreak. Nevertheless, we did not observe a confluence of cases over time due to the same microorganism. Moreover, some of the negative cultures could correspond to anaerobic microorganisms that were not recovered from swabs. Finally, the 6-month follow-up period could have been short for some virulent microorganisms, such as S. epidermidis, Enterococcus spp. and late recurrences could go undetected.
In conclusion, we could not demonstrate that IDC improved the clinical success in patients with PSM. However, indirect results, such as a better microbiological diagnosis, and an increase in adequate targeted treatment and surgical debridement, could lead to better clinical results. During the study period we observed an increase in the number of episodes caused by Gram-negative bacilli and polymicrobial infections. Finally, we observed that OST that has not previously been reported in the management of PSM, offers a promising strategy in the management of this infection.
FundingThis study has no funding.
Conflicts of interestNone to declare.
Data availabilityData archiving is not mandated but will be made available on reasonable request.
The followings are the supplementary data to this article:
