The project has been developed through two steps and it has been addressed to the four penitentiary institutions in Milan (one remand house, San Vittore, two long-sentenced prisons, Opera and Bollate, and Beccaria, a juvenile prison). A multicenter retrospective epidemiological survey of the four penitentiary institutions has been carried out, to analyze microbiological prevalence and local antibiotic consumption data. Sample included prison population at Milan penitentiary facilities between 2019 (3665 PLP), 2020 (3346 PLP), and 2021 (3316 PLP). Data have been compared over this period, to assess trends in antibiotic consumption and in bacterial resistance. Based on results of the previous analysis, an antimicrobial stewardship program has been implemented. The program involved several health care professionals and provided for the development of operational tools aimed to support antibiotics prescription. Actions undertaken were addressed to a sample consisting of the population detained in Milan prisons since the beginning of the project. In particular, sample included 3540 adults, of which 192 women, 40 young male juveniles and 5 mothers with their 5 children. 40.5% of PLP were foreigners (7). The average age was 41 years, with a wide range of ages, due to the presence of juvenile prison and children up to 6 years hosted at the attenuated custodial institution (ICAM) for mother living in prison. The condition of incarceration at one of the Milan correctional institutions in the years under investigation was the inclusion criterion adopted. All PLPs who were no longer incarcerated by 2019, due to transfers to other prisons or release from prison, have been excluded.

The microbiology unit shared microbiological prevalence data, which have been obtained through the collection and analysis of microbiological sample reports from correctional facilities, referring to the years 2019, 2020, and 2021.

Antibiotic consumption data have been provided by the hospital pharmacy. In particular, the analysis focused on data for the years 2019, 2020, and 2021, which have been extracted from the hospital’s management software. The data have been entered into a database containing the list of antibiotic drugs used in prisons, classified by Anatomical Therapeutic Chemical classification system (ATC) and marketing authorization code (MA code), dosage and pharmaceutical form. In addition, for each drug the database reported the Defined Daily Dose (DDD) value and the annual consumption, defined as the amount of antibiotic, expressed by the number of dosage units, used in penitentiary facilities in the period analyzed. According to the National Plan to Face Antimicrobial Resistance, approved in Italy in 2022, which assesses trends in territorial antibiotic consumption in DDD/1000 population per day (8), the following formula has been applied, to evaluate the consumption trend: DDD value of each antibiotic × the amount of antibiotic used × 1000/number of inmates. As a result, it has been obtained a unique indicator which provided a real and standardized representation of antibiotic consumption, compared with penitentiary population’ size. Values have been summed by grouping the different active agents by ATC, in order to evaluate the overall consumption, according to the different classes of antibiotics.

The project included the development of an Antimicrobial Stewardship Team, including the following professional figures: infectious diseases specialist, pharmacist, nurse, microbiologist and health management physician. Team members attended a series of multidisciplinary meetings to analyze local epidemiological data, to identify the main purposes and working methods, and to periodically assess the advancement of the project.

After the collection and the review of microbiological prevalence and antibiotic consumption data, the Antimicrobial Stewardship Team addressed the following actions:

In 2019, 2020, and 2021 the microbiology unit analyzed a total of 742 microbiological samples coming from Milan prisons: 251 samples have been analyzed in 2019, 232 and 229 in 2020 and 2021, respectively. Most of them were respiratory or urinary samples. The results showed the presence of 7 antibiotic-resistant bacteria. In 2019, only a strain of Proteus mirabilis with a multiresistance profile to penicillins, cephalosporins, ciprofloxacin, colistin, fosfomycin, tigecycline, and to the trimethoprim-sulfamethoxazole association, has been identified. The bacteria infected a prisoner at the San Vittore penitentiary. In 2020, two microorganisms have been isolated in two people living in San Vittore prison: a β-lactamase-producing Enterobacter hormaechei resistant to penicillins and cephalosporins, and an extended-spectrum β-lactamase (ESBL)-producing Escherichia coli strain.

In 2021, the following microorganisms have been isolated in four PLP detained in Bollate and San Vittore prisons:

All microorganisms have been detected in PLP’s urine and processed by urinoculture.

Antibiotic-resistant bacteria accounted for 0.94% of all microbiological isolates coming from Milan prisons, from 2019 to 2021. The analysis showed a progressive increase in antibiotic resistance: in 2019 and 2020 the prevalence was 0.39 and 0.86%, respectively, whereas in 2021 there was an increase of the prevalence up to 1.75% (Table 1).

The analysis conducted by the hospital pharmacy showed a gradual decrease in antibiotic consumption within correctional facilities. In 2019, a total antibiotic consumption of 20600 DDD has been found. In 2020, this figure decreased by 12%, going to 18041 DDD. In 2021, the total consumption of antibiotics decreased further, settling around the value of 15670 DDD, 13% less than the previous year (Figure 1 and Table 2). Consistent with local and national data, this phenomenon was probably related to the lower incidence of infections among inmates detected during the COVID-19 pandemic (9).

Analyzing the consumption trend of the most widely used systemic antibiotics in Milan prisons, in 2021, compared with 2019, a lower use of the associations of penicillins and β-lactamase inhibitors has been observed. In particular, the consumption of amoxicillin-clavulanic acid tablets at the dosage of 875 mg + 125 mg decreased by 21%, going from 9914 DDD in 2019 to 7791 DDD in 2021. However, amoxicillin-clavulanic acid was still the most widely used antibiotic. Macrolides, such as clarithromycin and azithromycin, were also widely prescribed. In this case, the analysis showed a 23% reduction in consumption between 2019 and 2021, although there has been a sudden increase in 2020, especially for azithromycin (from 1246 DDD in 2019 to 2593 DDD in 2020, and then decreased to 1416 DDD in 2021). This significant increase in 2020 was likely due to the fact that azithromycin was initially recommended for the treatment of COVID-19 in non-hospitalized patients (8). Consumption of systemic fluoroquinolones in prisons amounted to 1772 DDD in 2021: it is noteworthy that compared to 2019, in 2020 it dropped by 24%, and in 2021 it decreased further by 11%. The most widely used fluoroquinolone agent was levofloxacin, whose consumption in 2019, 2020, and 2021 was, respectively, of 1853 DDD, 1219 DDD, and 1043 DDD. Among the broad-spectrum penicillins, the most widely prescribed antibiotic was amoxicillin 1000 mg. Its consumption was 2554 DDD in 2019, and it dropped to 1465 DDD in 2021, with an overall reduction of 43%. In contrast, the consumption of tetracyclines increased by 25%. In particular, the consumption of doxycycline increased from 655 DDD in 2019 to 826 DDD in 2021, with a slight decrease of 7% in 2020, compared to the previous year. Table 3 illustrates the consumption trends of all the classes of antibiotics used within the Milan prisons, between 2019 and 2021.

Based on these results, the pharmacy developed both a formulary of antibiotic therapies for correctional facilities and a request form aimed to address highly critical antibiotics’ prescriptions. The antibiotic formulary was aimed at providing a useful tool for clinicians in the empirical choice of antibiotic therapy to deal with more critical infectious diseases which may occur in correctional institutions. All antibiotics used in Milan prisons have been listed and divided by pharmacological class and ATC. As part of the antimicrobial stewardship process, the use of this formulary requires specific training of prison medical staff, which is based on two key points: the need to collect biological samples for etiological identification before starting any antibiotic therapy, and the timely adaptation of the empirical therapy after the microbiological results. The prescription form aimed at ensuring greater control over the prescribing and use of antibiotics, through the direct involvement of the clinician and the pharmacist. It has been designed for prescribing antibiotics which do not belong to the first line of treatment, which have specific prescribing restrictions or that require special precautions in management and administration. Before their use, these agents required approval by an infectious diseases specialist. Some instances are glycopeptides such as vancomycin and teicoplanin, carbapenems, daptomycin, linezolid, colistin, tigecycline, intravenous ampicillin, dalbavancin and latest generation cephalosporins. The form included information about the requesting unit and the prescribing doctor, patient’s data (including age and body weight), the required antibiotic, its dosage, duration of treatment, whether it was prescribed as monotherapy or in combination with other medicines, its clinical indication, specifying whether the treatment was empiric or targeted and the name of the isolated microorganism, any previous antibiotic therapies. Eventually, the last section of the form required specifying whether the prescribed medication had been approved by an infectious diseases specialist. After verifying the request’s appropriateness, according to antimicrobial prescribing criteria, the pharmacist can authorize it. The hospital pharmacy also prepared data sheets for each antibiotic included in the formulary, to provide support to healthcare professionals in prescribing and managing antibiotic therapies. Each data sheet contained information about pharmacokinetics, dosage, interactions, major side effects, methods of preparation and storage. The formulary has been supplemented with specific guidelines for the treatment of the most recurrent infections in prisons: skin and soft tissue infections, STIs, urinary tract infections, otorhinolaryngological infections and pulmonary infections. The guidelines have been developed by the pharmacy, in cooperation with the infectious diseases specialists working in prisons. They referred to both community and nosocomial infections, as PLP could develop infections before and/or during the incarceration. Moreover, they could also contract care-related infections during any hospitalization and, once discharged, they could continue their care within correctional facilities. They focused on antibiotic therapies, distinguishing between empiric therapies and following treatment lines, and indicating which active agents must be chosen once detected antibiotic resistance, specifying any particular contraindication or recommendation for allergic patients or pregnant women who live in prison.