Table 3 Characteristics of Included studies before and during the COVID-19 pandemic

Trivedi (2013) [19]

United States

Cross-sectional study

- AMS core strategies included formulary restriction, antibiotic review, automatic stop orders, preauthorisation, and prospective review with feedback

- AMS supplemental strategies included education, dose optimisation, dose adjustments, guidelines and clinical pathways, parenteral-to-oral switch, streamlining de-escalation, and antimicrobial order forms

- Outcomes measured included antimicrobial resistance patterns (39%), antimicrobial utilisation (36%), antimicrobial costs (35%), Clostridium Difficile infection rates (32%), adverse effects (22%), 17% reported monitoring DDD and 13% reported monitoring DOT

- For a positive trend in outcomes data since the initiation of the ASP, including improved antimicrobial use (74%), decreased antimicrobial costs (63%), improved antimicrobial susceptibility patterns (47%), and 38% used computer software to interface with electronic records facilitated AMS

Kallen (2017) [20]

Netherlands

Randomised clinical trial

- Data extraction and feedback on the overall antibiotic use

- Point Prevalence Study of the European Centre for Disease Prevention and Control (PPS-ECDC) was conducted to provide feedback on validated Quality Indicators (QIs) for appropriate antibiotic use (PPS-QI), such as IV-to-oral switch projects (43%) and projects focusing on appropriate treatment for patients with pneumonia (21%) or the appropriate use of restricted antibiotics (19%)

Primary outcome

The geometric mean LOS was 9.5 days (95% CI 8.9–10.1, N = 4245 patients) at baseline versus 8.7 days (95% CI 8.1–9.2, N = 4195 patients) after intervention while adjusting for dependencies within clusters and potential confounders. After adjusting for the secular trend, the estimated decrease in geometric mean LOS was 0.5 days: 9.5 days (95% CI 8.9–10.1, N = 4245 patients) at baseline versus 9.0 days after intervention (95% CI 8.5–9.6); P < 0.001, N = 4195 patients

Secondary outcomes

DOT per 100 admissions decreased from 1320 (95% CI 1253–1387, N = 4245 patients) at baseline to 1185 (95% CI 1119–1252, N = 4195 patients) after the intervention (P < 0.001). Similar trends were found for days of IV antibiotics. A larger decrease was found for restricted DOT per 100 admissions (P < 0.001). The percentage of patients admitted to the ICU was lower after the intervention (4.8%, N = 201 patients) compared with a baseline (5.9%, N = 251 patients)

Tamma (2021) [21]

United States

Prospective study

Implementation webinars of AMS, antibiotic guidelines, antibiotic time-out, clinical rounds, and antibiotic user guides, identify antibiotic safety and adverse events, antibiotic review, use innovative strategy of the four moments of antibiotic decision-making framework including: make the diagnosis, cultures, and empiric therapy, stop, narrow, change to oral antibiotics and duration

Primary outcome (Unit-Level Antibiotic Use Data):

- Comparing January–February with November–December 2018, antibiotic use decreased from 900.7 to 870.4 DOT per 1000 PD (− 30.3 DOTs; 95%CI, − 52.6 to − 8.0 DOT; P = .008)

- Fluoroquinolone use decreased from 105.0 to 84.6 DOT per 1000 PD across all units between January–February and November–December (− 20.4 DOT; 95%CI, − 25.4 to − 15.5 DOT; P = .009)

Secondary outcome (C difficile identification):

- The number of hospital-onset C difficile Laboratory-identified events per 10 000 PD across the Safety Program cohort was 6.3 for quarter 1, 5.3 for quarter 2, 6.0 for quarter 3, and 5.1 for quarter 4 in the 2018 calendar year in the participating units. The incidence rate of hospital-onset C difficile Laboratory-identified events decreased from quarter 1 to quarter 4 by 19.5% (95%CI, − 33.5%to − 2.4%, P = .03)

Surat (2021) [22]

Germany

Retrospective study

- AMS multidisciplinary committee and regular ward rounds

- Formulary restriction of specific antibiotics (e.g., tigecycline and colistin)

- Creation of selective antibiotic resistogram profiles

- Electronic access to antimicrobial prescribing guidelines, and mobile applications

- Introduction of both surveillance data on AMR and antibiotic consumption rate

- In accordance with the current effective clinical practice guidelines for antimicrobial prophylaxis, the standard prophylactic regime changed from cefuroxime to cefazolin (depending on the procedure, it may differ)

- Further targets involved following antibiotic groups: meropenem, which AMS strived to reduce its usage, and fluoroquinolones, which involved drastic change in hospital’s general antibiotic policy

Primary outcome (The primary endpoint was defined as the total DOT for intraabdominal infections): An overall reduction in the total days on antibiotic therapy (ABT) from a mean of 6.1 days to 4.8 days (p = 0.02) was noted in the antimicrobial stewardship program (ASP) period, decreasing the days of therapy per 100 patient days (DOT/100PD) from 47.0 to 42.2 (p = 0.035)

Secondary outcome (The secondary endpoints included the appropriateness (indication and documentation) of the postoperative antibiotic therapy (PAT), the empiric selection of antibiotics and the frequency of antibiotic changes):

- The rate of patients receiving postoperative antibiotic therapy decreased from 56.8% to 45.2% (p = 0.002) in the ASP period

- A trend of change in the duration of postoperative antibiotic therapy from 8.1 to 7.2 days (p = 0.08) was observed

- The individual assessments of postoperative therapy revealed significantly less inappropriate (no indication) postoperative antibiotic therapy, shortened treatment durations (not significant) and an influence on the choice of antibiotics, with the use of more narrow-spectrum antibiotics

Weston (2012) [23]

United States

Cross-sectional study

- Antibiotic restriction, by using new restriction methods, such as front-end back end, automatic stop orders, ID consult required, verbal approval required

- Antibiotic guidelines and clinical pathways, antimicrobial order forms, streamlining or de-escalation, dose optimisation, parenteral-to-oral switch, and closed formulary

- Combining the results of both surveys, 31 out of 44 (70%) institutions had formal ASP in place. 13 institutions indicated on either survey that they did not have a formal ASP program. 25/38 institutions who responded to the second survey, had had an existing ASP

Mehta (2014) [24]

United States

Quasi-experimental study

Prior authorization and prospective audit with feedback

- The change from prior authorization to prospective audit with feedback was associated with a significant increase both in use of the affected antimicrobials and in overall use of all antimicrobial agents

- Broad-spectrum anti-gram-negative agents that still required prior authorization during both time periods continued to decline in use after the change in ASP

- The overall change in stewardship approach was associated with a significant increase in hospital LOS

- During the pre-intervention period, use of broad-spectrum anti-gram-negative antibiotics was declining at a rate of − 4.00 DOT/1,000-PD per month. However, during the post-intervention period, use increased by 0.80 DOT/1,000-PD per month, indicating that the change in ASP was associated with a slope change of 4.80 DOT/1,000-PD per month (P < .001)

- After decreasing during the 2 years before the ASP change, use of cefepime and piperacillin/tazobactam significantly increased following the transition to prospective audit with feedback by 3.21 DOT/1,000-PD per month (P = .003)

- Overall use of all systemic antimicrobial agents significantly increased after the change in ASP method (P < .001)

- Vancomycin use declined before the intervention but significantly increased after the intervention (P = .005). The use of non-audited antimicrobials significantly increased after the change in ASP methods (P < .001), the slope during the postintervention period continued to decline at − 1.87 DOT/1,000-PD per month

- The LOT of all systemic antimicrobials declined before the intervention by − 2.30 LOT/1,000-PD per month, and, despite a significant increase in slope (P = .029), use continued to decrease after the intervention by − 0.33 LOT/1,000-PD per month

Moriyama, (2021) [25]

Japan

cross-sectional study

- Prospective audit and feedback protocol were observed in 23 (59.0%) hospitals when using broad-spectrum antimicrobials

- Preauthorization was observed in 4 (10.3%) hospitals for using broad-spectrum antimicrobials.—Notification protocols support form was present in 37 (94.9%) for use of broad-spectrum antimicrobials

- The number of hospitals with preauthorization and notification protocols, respectively, using the investigated antibiotics were as follows: broad-spectrum antimicrobials overall 4 (10.3%) and 37 (94.9%); carbapenem 2 (5.1%) and 34 (87.2%); 3rd generation cephalosporin 0 (0%) and 0 (0%); 4th generation cephalosporin 0 (0%) and 10 (25.6%); piperacillin/tazobactam 0 (0%) and 17 (43.6%); and intravenous quinolone 3 (7.7%), and 18 (46.2%)

- Regarding preauthorization and notification protocols, there were no significant differences between small/middle-sized hospitals and large hospitals

- The numbers for hospitals that had intervention procedures within 7 d and 28 d, respectively, for each investigated antibiotic were as follows: broad-spectrum antimicrobials overall 17 (43.6%) and 34 (87.2%); carbapenem 16 (41.0%) and 34 (87.2%); 3rd generation cephalosporin 1 (2.6%) and 11 (28.2%); 4th generation cephalosporin 7n(17.9%) and 20 (51.3%); piperacillin/tazobactam 12 (30.8%) and 23 (59.0%); and intravenous quinolone 13 (30.8%) and 22 (56.4%). Intervention procedures to use broad-spectrum antimicrobials within 7 d were statistically more frequent in small/middle-sized hospitals than in large hospitals with findings as follows: overall, OR = 5.7, 95% CI = 1.4–23.5, p = 0.023; carbapenem, OR = 4.7, 95% CI = 1.1–19.1, p = 0.049; piperacillin/tazobactam, OR = 7.3, 95% CI = 1.3–39.9, p = 0.018; and intravenous quinolone, OR = 8.8, 95% CI = 1.6–48.2, p = 0.008

Thakkar (2021) [26]

India

Prospective study

- The pre-existing components of the hospital antimicrobial stewardship program included generation of antibiogram, formulation/ education and dissemination of antibiotic policies for surgical prophylaxis, community and hospital acquired infections and auditing antibiotics for surgical prophylaxis

- Prospective audit and feedback for the restricted antimicrobials

- Antibiotic restriction using the justification form

- Around 1.4% of admitted patients were put on restricted antimicrobials. The total days of therapy (DOT) were 41.5/1000 inpatient days

- Unjustified use of antimicrobials was reported in 13% and recommendation of the AMS for de-escalation were accepted in 89% by the treatment team

- There was no significant difference between antimicrobial DOT of the restricted antimicrobials between 2018 and 2019

- The colistin susceptibility rates remained stable compared to the previous years

Panditrao (2021) [27]

India

Quasi-experimental study

- Baseline Phase: from April–June 2017

Routine prospective audit and feedback was undertaken

- Intervention Phase: from July–December 2017

The following interventions were added:

Timeout, Correction of doses, continued education for rational use of antimicrobials, Care bundle approach for prevention of hospital-acquired infections (HAIs)

- There was a reduction in the cumulative DDD/1000 PD for all antimicrobials in the intervention phase compared with baseline (baseline phase 1326.3 DDD/1000PD vs. intervention phase 1313.5 DDD/1000PD)

- There was no change in the average number of antimicrobials per individual patient stay in the hospital between the baseline and intervention phases; P = 0.59)

- DOT/1000PD declined from 1112.3 in the baseline phase to 1048.6 days in the intervention phase, while LOT/1000 PD changed from 956.0 in the baseline phase to 936.3 during the intervention phase

- There was a decrease in DDD/1000 PD for antimicrobials such as piperacillin/tazobactam, imipenem, meropenem, clindamycin, levofloxacin, and amikacin, while there was an increase in DDD/1000 PD of vancomycin, colistin, cefoperazone/sulbactam, metronidazole and teicoplanin

- There was a decrease in percentage of carbapenem use in the intervention phase compared with the baseline phase (26.3% vs. 20.9%), whereas there was an increase in the use of polymyxins, particularly colistin (11.1% vs. 6.2%) and glycopeptides (vancomycin and teicoplanin) (12.3% vs 11.0%)

Ababneh (2020) [28]

Jordan

Cross-sectional study

This study quantified antimicrobial use in inpatient settings as part of antimicrobial stewardship program surveillance

- In terms of DDDs, carbapenems (ertapenem, meropenem, imipenem) were the most commonly used agents in a total of 28.0 DDD/100 admissions, followed by glycopeptides (vancomycin, teicoplanin) in a total of 26.8 DDD/100 admissions, piperacillin-tazobactam with 20.5 DDD/100 admissions and ceftriaxone with 14.2 DDD/100 admissions, fluoroquinolones (ciprofloxacin and levofloxacin) in a total of 11.2 DDD/100 admissions

- The highest prescription rate of antibiotics was in the internal medicine wards (49.8 DDD/100 admissions), followed by surgery wards (33.2 DDD/100 admissions), intensive care unit (20.6 DDD/100 admissions), paediatrics (10.5DDD/100 admissions), oncology (10.4DDD/100 admissions)

- Regarding DOTs, piperacillin-tazobactam was the most commonly used agent (27.6 DOT/100 admissions), followed by carbapenems (27.2 DOT/100 admissions), glycopeptides (24.7 DOT/100 admissions), fluoroquinolones (12.4 DOT.100 admissions), and cefazolin (11.4 DOT/100 admissions)

Spernovasilis (2021) [29]

Greece

Cross-sectional study

- Prospective audit and feedback strategy,

along with a case-based education of treating doctors

- Antibiotic review after 24 h, 72 h and 7 days

- Doctors believed that the prospective audit and feedback ASP strategy is more effective and educational than the preauthorization ASP strategy (70.3% and 77.7%, respectively)

- Most respondents (90.6%) agreed that the implementation of an ASP improves the patients’ outcome compared to the absence of such a programme

- Less than 25% of participants agreed that the prospective audit and feedback strategy of the current ASP should change

- More than 80% of respondents agreed that in-person consultation is the preferred practice for the ASP and education

Ashiru-Oredope (2021) [30]

United Kingdom

Cross sectional study

- Audits and Regular surveillance of antimicrobial use/

- Point Prevalence Surveys

- Quality improvement initiatives

- Education, AMS meetings, multidisciplinary team and ward rounds

- Writing non-COVID-19 guidelines

- IV to oral switches

- AMS surveillance activities

- Technology (virtual meetings, virtual platforms, remote working and ward rounds)

- Introduction of novel biomarkers (e.g., Procalcitonin) for differentiating viral and bacterial infections

- The use of hospital electronic prescribing systems facilitated

- AMS activities by antimicrobial pharmacists; allowing them to target their activities, for example, identification of patients receiving excessive durations of antibiotics

- Infection prevention control

- Clinics/out-patient consults and Outpatient parenteral antibiotic therapy (OPAT)

- Changing current inpatient processes such as COVID-19 patients receiving a senior review more quickly

- Prescribing indicators/targets reporting, and Antibiotic Review Kit (ARK)

- AMS committee meeting (formal or informal)

- From qualitative open questions: respondents highlighted core AMS work e.g., reviewing and writing non-COVID-19 guidelines as being the most affected

- Respondents were concerned about increased antibiotic use, delayed IV to oral switches (IVOST), and prolonged antibiotic durations

- The respondents also were concerned that cases of Clostridioides difficile Infection (CDI) were rising in some hospitals

- Stock shortages were also identified as difficult to manage due to overwhelmed supply chains for antibiotics, antivirals and in some cases personal protective equipment (PPE)

- Positive COVID-19 outcomes included: technology being increasingly used as a tool to facilitate stewardship, e.g., virtual meetings and ward rounds

- Another positive outcome was the increased introduction of novel biomarkers (e.g., procalcitonin) for differentiating viral and bacterial infections

- The use of hospital electronic prescribing systems facilitated AMS activities by antimicrobial pharmacists

- There has also been a positive increase in the multidisciplinary team

- Increased awareness of antimicrobial guidelines and improvements seen in infection prevention

Williams (2021) [31]

United Kingdom

Retrospective study

- Seventy-three (33%) patients in the negative PCT group were on antibiotics 48 h following diagnosis of COVID-19 compared with 126 (84%) patients in the positive PCT group (P < 0.001), suggesting good compliance with the guideline

Primary outcome:

- Patients in the negative PCT group received significantly fewer DDDs of antibiotics (both total and per alive day) compared with patients in the positive PCT group (median DDD 3.0 vs 6.8; P < 0.001)

- A significant relationship between PCT and total DDDs remained after accounting for confounders; on average, a patient with PCT > 0.25 ng/mL had almost three-fold more DDDs of antibiotics compared with patients with PCT 0.25 ng/mL [coefficient 2.72, 95% confidence interval (CI) 2.03e3.62; P < 0.001]

Secondary outcomes:

- Sixty-two (28%) patients in the negative PCT group died compared with 54 (36%) patients in the positive PCT group (P.0.021), and 19 (9%) patients in the negative PCT group were admitted to the ICU compared with 28 (19%) patients in the positive PCT group (P.0.007)

- Meropenem was the only carbapenem used in the study population. With specific reference to meropenem consumption, positive PCT was associated with a three-fold increase in the odds of receiving any meropenem during the course of hospital admission (odds ratio 3.16, 95% CI 1.50e6.65; P = 0.002)