Management of Clostridioides difficile infection: an Italian Delphi consensus

Abstract Background Clostridioides difficile infection (CDI), a leading cause of nosocomial deaths, is a microbiota-mediated disease. As such, the use of broader spectrum antibiotics, such as vancomycin and metronidazole, can prime the gastrointestinal tract to become more prone to CDI recurrences. Fidaxomicin, a narrow-spectrum antibiotic, has been demonstrated to be superior in preventing recurrence and in preserving the intestinal microbiota; however, widespread employment worldwide has been hindered due to high acquisition costs. Objectives To integrate the currently available guidelines on the management of CDI and to shed light on the timeliest employment of fidaxomicin. Methods An expert panel was gathered to obtain consensus using Delphi methodology on a series of statements regarding the management of CDI and on appropriate antibiotic use. Results Consensus was reached on 21 of the 25 statements addressing the management of CDI. Conclusions Delphi methodology was used to achieve consensus on the management of CDI, on the identification of patients at risk of recurrences or severe infection, and on the most appropriate use of fidaxomicin, with the final aim of fostering clinical practice application of treatment algorithms proposed by previous guidelines, in absolute synergy. It could be an important tool to promote more appropriate and cost-effective CDI treatments in European settings with limited resources, like Italy.


Introduction
Clostridioides difficile infections (CDIs) are a leading cause of nosocomial deaths.This Gram-positive, spore-forming and toxinproducing intestinal bacterium that infects the human gut potentially causing lethal diarrhoea has been designated an 'urgent threat' by the US CDC [1][2][3] and it is under surveillance at European and Italian levels. 4,5The burden of these infections in Italy is underscored by the median hospital incidence density of healthcare-associated CDI (2.9 cases per 10 000 patient-days), which is more pronounced in tertiary care hospitals (5.8 cases/ 10 000 patient-days).This is further emphasized by the recurrence rate at 4 weeks (21% of patients) and the disparity in length of stay [16 (IQR = 13) versus 8 (IQR = 8) days; P < 0.001] between patients with CDI and those without. 6he alarming increase in infections caused by highly pathogenic variants of C. difficile runs parallel to the use of broad-spectrum antibacterial drugs.CDI is, indeed, a microbiota-mediated disease: disruptions in the gut microbiota are critical to CDI development, whereas the restoration of homeostatic bacterial diversity and abundance is essential for recovery. 6Dysbiosis, like the one triggered by broad-spectrum antibiotic uptake, is a precursor to infection, and its persistence often leads to disease recurrence. 7roader spectrum antibiotics such as vancomycin and metronidazole are used to treat CDIs, but these antibacterial agents decimate the normal gut microbiome, paradoxically priming the gastrointestinal tract to become more prone to CDI recurrences [8][9][10] Not surprisingly, treatment of CDI with either metronidazole or vancomycin is associated with recurrence in 20%-30% of patients, which then provides a 50%-60% likelihood of further recurrence. 11n 2011, the macrocyclic antibiotic fidaxomicin became available to treat CDIs. 12Fidaxomicin selectively targets C. difficile without effectively killing crucial gut commensals such as Bacteroidetes, which crowd the human gut microbiome providing protection against C. difficile colonization. 13Fidaxomicin is a narrow-spectrum, macrocyclic antibacterial agent with minimal systemic absorption.It showed higher in vitro activity against C. difficile than vancomycin, with a more prolonged post-antibiotic effect and reduced sporulation and toxin production in vitro and in vivo. 14Most importantly, two prospective randomized controlled trials demonstrated non-inferiority of fidaxomicin versus vancomycin for clinical cure of CDI. 11,12Although fidaxomicin can be associated with treatment failure and relapse after primary infection, it has been demonstrated to be superior in preventing recurrence and in preserving the intestinal microbiota thanks to its narrow-spectrum activity. 15Of note, fidaxomicintreated hospital inpatients proved to be less likely to contaminate their environment (36.8%) than patients treated with metronidazole and/or vancomycin (57.6%). 16espite key advances in therapeutic strategies, CDI remains challenging for clinicians worldwide: 15 beside the management of infrequent cases of fulminant colitis, which carry a high risk of mortality, the most difficult task consists in preventing recurrent infections.Heterogeneity in definitions used for severe and potentially recurrent CDI (rCDI) has been a confounding factor when assessing treatment guideline recommendations and trial outcomes.Consensus between the IDSA, the Society for Healthcare Epidemiology of America (SHEA) and the ESCMID regarding optimal treatment of initial and first recurrence of nonsevere CDI has only recently been reached. 15The recent IDSA/ SHEA update suggests fidaxomicin preferentially over vancomycin for initial CDI, 17 and the latest ESCMID guidelines 18 concur with this recommendation, with vancomycin being acceptable for a first episode, but metronidazole only if the other agents are unavailable.
Although fidaxomicin performed favourably against vancomycin in clinical trials of CDI and has been suggested preferentially over broader spectrum antibiotics by the most recent guidelines, widespread use worldwide has been hindered due to its higher cost.Various recent studies, either industry-supported [19][20][21][22][23] or not, 24,25 showed in different settings and using different health economic models that CDI treatment with fidaxomicin can reduce global healthcare costs.
Extended-pulsed fidaxomicin therapy was more costeffective than vancomycin for first-line treatment of CDI in older patients. 20In fact, higher drug acquisition costs for fidaxomicin were found to be compensated by lower hospitalization costs driven by fewer recurrences, lower costs of complications, and fewer GP visits versus vancomycin.In Italy, for instance, a real-world study showed that the mean cost of a recurrent episode of CDI amounts to €9504.87 ± €8614.11. 26To mitigate the higher acquisition costs of fidaxomicin compared with those of vancomycin, the ESCMID 2021 update on the treatment guidance for CDI suggests applying a risk-stratification strategy in case of economic restraints. 18Identifying patients at risk, however, is challenging.Several prediction models have been developed, yet none has been widely adopted in clinical practice.
This study aimed to integrate current guidelines on the treatment of CDIs focusing on practical issues where clinical evidence is still limited.A panel of Italian experts using the Delphi methods approach, was gathered to obtain consensus on a series of statements addressing (i) the proper use of fidaxomicin; (ii) the impediments to the practical implementation in Italy of the evidence coming from the existing CDI guidelines, and (iii) the strategy of prevention and treatment of recurrent infections.In order to obtain consensus, the Delphi method was employed; this widely accepted technique, built on evidence-based medicine, adopts consecutive iterations using a survey until consensus is reached, 27 allowing clinical recommendations for those areas, like Clostridioides-mediated infections, where clinical-based evidence is still insufficient.

Study design
A modified Delphi process (Figure 1) was organized in the following phases, which were run over a period of 10 months (from June 2023 to March 2024): 1. Desk analysis: a preliminary list of principles of starting points and research questions was drafted by research assistants following analysis of scholarly sources on the topic, local Italian laws addressing the management of CDIs, and the output of a questionnaire covering clinical and organizational subjects submitted to healthcare professionals dealing with CDIs in their daily practice and distributed in the nine main Italian regions: 34 clinicians (including 6 gastroenterologists, 1 geriatrician, 22 infection disease specialists; 5 internists), 4 microbiologists, 13 hospital pharmacists and 1 chief medical officer.2. Identification and selection of a panel of experts: 13 Italian experts (the Scientific Board) were identified by their experience in treatment of CDIs, relevant publications, academic status, clinical practice at recognized centres of excellence, experience in clinical trials, and participation in national and international conferences.The experts were representative of the national territory (seven from Northern Italy, three from Central Italy, and three from Southern Italy); they were specialized in gastroenterology (n = 2), infection diseases (n = 6), internal medicine (n = 1), microbiology (n = 1), health economics and healthcare management (n = 1), geriatrics (n = 1) and hospital pharmacy (n = 1).3. Kick-off meeting and literature review: during the first meeting of the scientific board, the outputs of the desk analysis, including a Review preliminary list of draft statements, were presented, critically discussed and revised.A literature review of articles published in peerreviewed journals allowed identification of red flags for referral on the treatment of CDIs.The search was performed online using the PubMed database from 2015 to 2023.Predefined keywords and inclusion criteria were used; these included: 'fidaxomicin', 'recurrence', 'recurrent clostridium/clostridioides difficile', 'infection', 'vancomycin' and 'guideline'.Only studies published in the English language (unless a specific article in another language was considered relevant by the Board) were included.Letters and abstracts were excluded.
After the kick-off meeting, 23 statements were drafted and submitted for the first round of online voting.4. First, second and third rounds of online Delphi voting: In Delphi Round 1, the 13 experts were asked to express their judgement on the initial 23 statements.Voting was undertaken by email using a 5-point Likert scale to indicate the level of agreement on each statement: 1 = absolutely disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = totally agree.The collected answers were expressed as a percentage response for each item.A total cumulative agreement was defined as the sum of response percentages in items 4 ('agree') and 5 ('absolutely agree').For the purpose of this consensus, a total cumulative agreement ≥75% was considered a priori to represent consensus for each statement.This definition of agreement was based on standards used in previous Delphi studies [28][29][30] During Round 1 voting, experts were also given the opportunity to provide comments and suggest additional items that may not have been included when developing the initial list of statements, with the aim of clarifying any redundancies or issues regarding comprehension or syntax of each statement.
Once Round 1 voting was completed, an analysis of responses and comments was performed.Statements that reached a total cumulative agreement >75% without receiving specific comments or requests of amendment/integration were considered finalized.Statements that reached a total cumulative agreement >75% and received specific comments or requests of rewording/rephrasing were modified accordingly before being submitted to the panel for Round 2 voting.Among statements that did not reach a consensus after Round 1, those that scored ≤75% without specific indications for amendment were excluded.Those that reached a total cumulative agreement >50% and specific indications were modified and/or integrated according to the feedback received before being submitted to the panel for Round 2 voting.
Round 2 voting followed the same process of Round 1. Except for those statements that were already finalized or excluded, each member of the panel expressed their level of agreement for each item and, if deemed necessary, provided comments.
In Round 3, residual statements were discussed in order to nail down a final wording for those statements that were agreed in Round 1 and Round 2 but were suggested to be amended or integrated.Statements that did not reach a consensus in Round 2 but reached a total cumulative agreement >65%, were voted again in Round 3.
An additional statement was added and voted during the revision phase of the manuscript in agreement with the editor's recommendations.

Results
The 23 statements that were drafted after the kick-off meeting (Table 1) spanned the following areas: • diagnosis, including definition of severe infection, frail patient, and patients at risk of recurrences; • management of CDIs in patients at high risk; • benefits of fidaxomicin therapy compared with treatment with broader spectrum antibiotics; • management of CDI and cost monitoring.
At the end of the first round of individual voting, 17 statements of 23 reached a consensus.Among them, statements 1 and 2 were agreed by the panel without suggesting any modifications, whereas the remaining 15 were suggested to be reworded or integrated.Statements indicated in Table 2 as Ex-12, Ex-13 and Ex-23, addressing management of CDI in transplanted patients and patient involvement, failed to reach an agreement and were excluded from the following steps of voting.The residual statements, addressing fidaxomicin supply, cost monitoring and post-hospitalization monitoring strategies, although they were not agreed at first instance, were revised to be submitted again to the panel for the second round of voting.
Following the suggestions received by the panel, in the second round of voting, statement 7 was integrated into statement 6, as both addressing the definition of frailty condition, whereas statement 9 was split into two different ones (8 and 9) to highlight the   The revised 20 statements, net of those that had been already finalized or excluded, were submitted to the panel for the second round of voting.At the end of this further voting step, 12 statements of 18 reached a consensus.Among them, statements 3,  5, 6, 7, 11, 11b, 15, 19, 21 and 22 were agreed by the panel without any further modifications suggested.Statements 4, 9, 10, 17, 18 and 20 reached an agreement as well, but a few rewordings and integrations were suggested.Agreement on the final wording of statement 8 was reached during the revision phase of the manuscript, after the editor's suggestions.Once integrated or amended accordingly, these statements were presented again to the panel and were tacitly approved by each member.Conversely, statement 14, on high-dose vancomycin therapy, and statement 16, addressing availability of fidaxomicin supplies in health clinics, failed to reach a consensus but were revised to be submitted again to the panel for a final round of off-line voting.This latter confirmed disagreement about statement 16, whereas agreement was reached for statement 14.
Statement 12 was added during the revision of the manuscript following the comments of the reviewers.Agreement was reached during the first voting round after dividing the statement into four subsections.

Discussion
Ideally, clinical recommendations should be grounded in evidence obtained from controlled clinical trials, and clinical practice should be guided by both recommendations and clinical trial findings.However, in practical terms, there may be limited research-based evidence available and the implementation of guidelines may face obstacles due to various factors, including cost-saving strategies.The present Delphi study, aligned with the Delphi modified approach, 27 which is based on evidence-based medicine and adopts an anonymous voting process to establish opinion, was employed to achieve consensus on the most appropriate use of fidaxomicin.
This study presents results obtained by an online meeting of 13 Italian professionals with different and complementary expertise in C. difficile and CDI treatment (six infection disease specialists, two gastroenterologists, one internist, one microbiologist, one geriatrician, one health economist and one hospital pharmacist), two rounds of Delphi voting, and a final meeting meant to outline statements addressing the most appropriate treatment of CDI.

Consensus statements on the use of fidaxomicin, in alignment with and supplementing the existing guidelines and on the identification of patients to be treated with fidaxomicin
Consistently, the convenience in adhering to the current guidelines on the diagnostic process 31 and on the management of patients with CDI 32,33 was promptly confirmed by the panel (statements 1 and 2).Moreover, statement 2 stressed the importance of pre-emptive contact isolation of patients suspected to have CDI (i.e. during sample collection).

Review
According to the ESCMID guidelines, 18 severe CDI is characterized by one of the following factors at presentation: fever (i.e.core body temperature >38.5°C), marked leucocytosis (i.e.leucocyte count >15 × 10 9 /L) and rise in serum creatinine (i.e.>50% above the baseline).Additional supporting factors are distension of the large intestine, pericolonic fat stranding or colonic wall thickening (including low-attenuation mural thickening) at imaging.Taking as reference this definition, the panel agreed on the identification of the risk factors for severe infection (statement 3), which take into consideration the chronological age of the patients, their clinical indicators (albuminaemia 1,[34][35][36][37] and Zar score 38,39 ) and comorbidities or other conditions, namely inflammatory bowel disease (IBD), chronic kidney failure, liver failure, diabetes and cardiovascular/pulmonary diseases.The panel agreed on using the definition of severe CDI as reported in the ESCMID guidelines.
Full agreement was reached after two rounds of adjustments on these same parameters (except for the Zar score), which were marked as 'red flags' for risk of recurrence in patients at their first episode of CDI (statement 4).Immunocompromised patients, those recently hospitalized, and the ones under pharmacological treatment for other conditions always need to be identified.Further 'red flags' agreed on by the panel included: the presence or absence of an index episode and the chance to trace back a healthcare-associated origin of the infection.Other indicators with weaker but still substantial evidence for risk of recurrence were agreed to be: severity of the infection, 17 chronic kidney failure, liver failure, diabetes, cardiovascular/pulmonary disease, parenteral nutrition, 34 and use of concomitant antibiotics started during/after CDI diagnosis. 18,34Some members of the panel suggested that patient compliance (and/or caregiver accountability) can play a role in non-hospitalized patients as well.
The crucial role of IBD is highlighted by its identification as a risk factor for both severe and recurrent infections (statements 3 and 4) and by the critical role that CDI can have in these patients when causing flares (statement 5).][55] In terms of predictability of negative outcomes, frailty condition emerged as the most significant risk factor, stronger than the chronological age of the patient; the definition of 'frailty condition' was thoroughly discussed during the kick-off meeting and later finalized in statements 6 and 7. Since frailty level predicts mortality at 90 days more accurately than chronological age and disease severity, 44 the need for a change of mindset in clinicians during daily management of CDIs emerged. 44,56,57The Multidimensional Prognostic Index (MPI) [44][45][46]58 or its screening short version BRIEF-MPI, 47 which requires a very limited amount of time, can be used as effective tools for assessing multidimensional frailty in hospitalized older patients with CDI.
In accordance with available outcomes of clinical trials, 11,12,49,59,60 current guidelines 17,18 and a systematic review, 61 the consensus panel confirmed that fidaxomicin should be recommended as first-line therapy in patients at high risk of recurrence, as defined above, and in patients with rCDI, as reported in statements 8 and 9, since it is associated with a higher sustained response.Alternatively, the combination of vancomycin and FMT was demonstrated to be superior to vancomycin alone in achieving sustained resolution from CDI, as described in a recent clinical trial. 62As reported in statement 11, FMT in combination with standard-of-care (SoC) antibiotics was confirmed to be the preferred treatment option of second or further recurrence of CDI. 18FMT efficacy by itself was confirmed by the panel; this procedure is recommended for rCDI and for patients with severe CDI who have not responded to antibiotic treatment and for whom surgery is not feasible. 18,34Nevertheless, the main drawback of FMT lies in its current availability in a very limited number of hospitals in Italy.
Bezlotoxumab, a fully humanized monoclonal antibody directed against the binding domains of toxin B that is given as a onetime infusion in addition to an SoC antimicrobial, could be an alternative option for breaking the cycle of CDI recurrence and, based on recent evidence, 63 should be considered for: • patients with multiple (≥3) rCDI risk factors, in addition to SoC, regardless of the severity of previous episodes; • patients at first CDI recurrence in addition to vancomycin or fidaxomicin, when fidaxomicin was used to manage the initial CDI episode, independently of rCDI risk factors; • patients with second or multiple CDI recurrences, in centres where FMT is not available. 50is practical suggestion comes from recently published results of a real-world multicentre cohort study, including 442 patients with CDI from 2018 to 2022, collected in 18 Italian centres.This study confirmed the greater efficacy of bezlotoxumab + SoC versus SoC alone for the prevention of rCDI, already seen in previous randomized studies and a similar previous trial emulation performed using observational data 50,64,65 Importantly, in contrast to other studies, this study was conducted in a selected population at high risk of recurrence and included the highest numbers of patients treated with fidaxomicin as SoC, compared with Spanish and US cohorts.
Although not reaching statistical significance, the benefit of bezlotoxumab + SoC on the composite outcome ((30 day recurrence and/or all-cause death) appeared to be attenuated in participants aged <70 years (P = 0.61) and in those who received fidaxomicin as first-line treatment (P = 0.71).For this reason, to balance the risks and benefits of its use, in particular weighing costs in countries with limited resources, the authors suggested limiting bezlotoxumab use in the first CDI episode only to high-risk patients (statements 12, 12b, 12c, 12d).
Agreement was obtained on the superiority of an extendedpulsed fidaxomicin regimen (200 mg oral tablets, twice daily on Days 1-5, then once daily on alternate days on Days 7-25) when compared with standard-dose vancomycin for sustained cure of CDI and to reduce rates of recurrence without additional costs 8,49 (statement 10).Statements on fidaxomicin use for the prophylaxis of Clostridioides difficile-induced diarrhoea (CDAD) in persons undergoing transplantation reached a very low agreement (58% and 9%) after the first round of voting; this is likely explained by the limited literature available-only one randomized controlled trial on this topic was publishedand by the high risk of side effects, including microbiome distortion, linked to its widespread use.Nevertheless, the potential benefit of a fidaxomicin-based prophylaxis in selected patients was acknowledged by some members of the panel and should be investigated further.

Review
The panel recognized that fidaxomicin is rarely associated with resistance events thanks to its limited activity against enteric commensal bacteria.Consistently, employment of high-dose vancomycin was discouraged (statements 13/14) due to selection pressure, which results in resistant strains (e.g.vancomycinresistant enterococci).

Consensus statement on the obstacles involved in managing CDIs in Italy, with a specific focus on prevention and treatment of recurrent infections and the limitations affecting the use of fidaxomicin, in alignment with and supplementing the existing guidelines
Selection, prescription and administration of the most appropriate therapy are often impacted both by clinical considerations and by hospital management obligations.The panel agreed on the importance of antimicrobial stewardship (statement 15) and discussed the convenience of stocking a minimum amount of fidaxomicin in order to provide a timely treatment of CDI; however, this latter did not obtain consensus (Ex 16).Conversely, agreement (77%) was reached after online discussion and two rounds of revision on a statement addressing the importance of cost monitoring of CDI treatment (statement 16); according to the panel, this latter should be performed assessing the whole patient pathway, including testing and other exam costs, hospital readmission rates, and inpatients' and outpatients' costs.0][21][22][23][24] .Despite higher drug acquisition costs for fidaxomicin, these are offset by lower hospitalization expenses resulting from fewer recurrences, reduced complication costs and fewer GP visits compared with vancomycin.For instance, real-world studies revealed that the mean cost of a recurrent episode of CDI is in the range €9504.87 ± €8614.11, 26whereas in a cohort study the total cost attributable to CDI in Rome was €17 714 per patient with recurrence. 66It should be noted that the aforementioned health economic analyses have been supported by grants from industry.Another model set out to analyse the cost-effective sequence of antibiotic sequences as in the studies above and vancomycin/fidaxomicin was found to have a higher probability of being cost effective for an English population with characteristics of the 'average' CDI patient. 25Nevertheless, the ESCMID guideline recommendations for treating an index CDI with fidaxomicin as first-line treatment has been demonstrated by Swart et al. 23 to be cost-effective compared with the NICE treatment strategy (which considers vancomycin as first-line treatment) from the UK National Health Service perspective.
Communication between clinicians and pharmacists is crucial for cost monitoring (statement 17).On a practical level, training in medical wards, and sharing treatment algorithms or integration tools and practical approaches among the hospital's professionals was strongly recommended by the panel.
According to the preparatory analysis of this consensus, one of the issues limiting appropriate management of CDI is the complexity in recurrence identification when the patient is hospitalized in different clinics without a comprehensive medical record.The topic was discussed during the online meeting and agreement was obtained on statements 18 and 19; this highlights how technological supportive systems need to be structured and developed with the users to be effective and readily informative.It emerged also that follow-up at 8 weeks after patient dismissal should be a common practice in order to identify recurrences and implement proper treatment.
Full agreement was also reached on long-term care facilities such as nursing homes when suspecting or after identification of CDI (statement 20), and on the importance of providing patients and caregivers with consistent information about CDI and risk of recurrences when the disease is managed in an outpatient setting (statement 21).Caregivers and family members should also be informed of the risk of transmission as soon as CDI diagnosis is confirmed.

Conclusions
Despite key advances in therapeutic options, CDI remains challenging for clinicians worldwide.The present Delphi study, aligned with the Delphi modified approach, 27 which is centred on evidence-based medicine and adopts an anonymous voting process to establish agreement, was employed to achieve consensus on the management of CDI, on the identification of patients at risk of recurrences or severe infection, and on the most appropriate use of fidaxomicin with the aim of targeting suitability.
The enhanced benefit of this consensus document is that the results were obtained by a multidisciplinary group of 13 Italian professionals with complementary expertise in C. difficile management and treatment (six infection disease specialists, two gastroenterologists, one internist, one microbiologist, one geriatrist, one health economist and one hospital pharmacist).This approach was based on merging clinical experience, open panel online meeting discussions, and literature review of papers, not all of which were included in previous guidelines, due to years of publication, rigorous selection criteria and specific practical issues not addressed before.
Indeed, in practical terms, there may be limited researchbased evidence available on specific issues, and the implementation of guidelines may face obstacles due to various factors, including cost-saving strategies and local or individual behaviours.Correct patient stratification will help mitigate the higher drug acquisition costs for fidaxomicin, which the panel agreed on recommending as first-line treatment for patients at risk thanks to its efficacy and narrow-spectrum antimicrobial activity.
In conclusion, this study aimed to foster clinical practice application of treatment algorithms proposed by previous guidelines, in absolute synergy.It could be an important tool to promote more appropriate and cost-effective CDI treatments in European settings with limited resources, like Italy.

Figure 1 .
Figure 1.Overall flow of the Delphi process that was employed in this study.This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.