Implications of reducing antibiotic treatment duration for antimicrobial resistance in hospital settings: A modelling study and meta-analysis

Yin Mo; Mathupanee Oonsivilai; Cherry Lim; Rene Niehus; Ben S Cooper

doi:10.1371/journal.pmed.1004013

. 2023 Jun 15;20(6):e1004013. doi: 10.1371/journal.pmed.1004013

Implications of reducing antibiotic treatment duration for antimicrobial resistance in hospital settings: A modelling study and meta-analysis

Yin Mo ^1,^2,^3,^4,^*,^#, Mathupanee Oonsivilai ^1,^2,^#, Cherry Lim ^1,², Rene Niehus ⁵, Ben S Cooper ^1,²

PMCID: PMC10270346 PMID: 37319169

Abstract

Background

Reducing antibiotic treatment duration is a key component of hospital antibiotic stewardship interventions. However, its effectiveness in reducing antimicrobial resistance is uncertain and a clear theoretical rationale for the approach is lacking. In this study, we sought to gain a mechanistic understanding of the relation between antibiotic treatment duration and the prevalence of colonisation with antibiotic-resistant bacteria in hospitalised patients.

Methods and findings

We constructed 3 stochastic mechanistic models that considered both between- and within-host dynamics of susceptible and resistant gram-negative bacteria, to identify circumstances under which shortening antibiotic duration would lead to reduced resistance carriage. In addition, we performed a meta-analysis of antibiotic treatment duration trials, which monitored resistant gram-negative bacteria carriage as an outcome. We searched MEDLINE and EMBASE for randomised controlled trials published from 1 January 2000 to 4 October 2022, which allocated participants to varying durations of systemic antibiotic treatments. Quality assessment was performed using the Cochrane risk-of-bias tool for randomised trials. The meta-analysis was performed using logistic regression. Duration of antibiotic treatment and time from administration of antibiotics to surveillance culture were included as independent variables. Both the mathematical modelling and meta-analysis suggested modest reductions in resistance carriage could be achieved by reducing antibiotic treatment duration. The models showed that shortening duration is most effective at reducing resistance carriage in high compared to low transmission settings. For treated individuals, shortening duration is most effective when resistant bacteria grow rapidly under antibiotic selection pressure and decline rapidly when stopping treatment. Importantly, under circumstances whereby administered antibiotics can suppress colonising bacteria, shortening antibiotic treatment may increase the carriage of a particular resistance phenotype. We identified 206 randomised trials, which investigated antibiotic duration. Of these, 5 reported resistant gram-negative bacteria carriage as an outcome and were included in the meta-analysis. The meta-analysis determined that a single additional antibiotic treatment day is associated with a 7% absolute increase in risk of resistance carriage (80% credible interval 3% to 11%). Interpretation of these estimates is limited by the low number of antibiotic duration trials that monitored carriage of resistant gram-negative bacteria, as an outcome, contributing to a large credible interval.

Conclusions

In this study, we found both theoretical and empirical evidence that reducing antibiotic treatment duration can reduce resistance carriage, though the mechanistic models also highlighted circumstances under which reducing treatment duration can, perversely, increase resistance. Future antibiotic duration trials should monitor antibiotic-resistant bacteria colonisation as an outcome to better inform antibiotic stewardship policies.

Author summary

Why was this study done?

Shortening antibiotic treatment duration is a commonly adopted antibiotic stewardship strategy, with the expectation that it will reduce antimicrobial resistance in treated individuals and in the overall population.
Antibiotic selective pressure acts predominantly on “bystander” colonising bacteria for resistance, and this depends on the spectrum of coverage, pharmacokinetic and pharmacodynamic properties of individual antibiotics.
Empirical evidence and an understanding of the mechanisms by which antibiotic treatment duration effects the emergence and spread of antimicrobial resistance are lacking. Understanding the key factors driving the effect of antibiotic treatment duration on resistance carriage will help to inform future research study designs, antimicrobial stewardship interventions, and resource allocation in multimodal control strategies.

What did the researchers do and find?

We modelled within- and between-host dynamics of colonising “bystander” susceptible and resistant bacteria in response to systemic antibiotic treatment and compared the model findings with a systematic review and meta-analysis.
The meta-analysis found one additional antibiotic treatment day is associated with a 7% absolute increase in risk of resistance carriage when antibiotics administered were not effective against the resistance phenotype in the colonising bacteria.
For treated individuals, the models showed that shortening antibiotic treatment duration is most effective at reducing resistance carriage when resistant bacteria grow rapidly under antibiotic selection pressure and decline rapidly when stopping treatment.
At a population level, shortening antibiotic treatment duration is most effective at reducing resistance carriage in high transmission settings.
Shortening antibiotic treatment duration may increase resistance carriage when the antibiotics administered are effective at eliminating colonising bacteria with a particular resistance phenotype.

What do these findings mean?

Shortening antibiotic treatment duration may increase or decrease colonisation by resistant bacteria, dependent upon individual and combined bacterial and antibiotic characteristics.
The effect of shortening antibiotic treatment duration on colonisation by resistant bacteria colonisation is potentially modest due to short hospitalisation periods and slow decolonisation of resistant bacteria.
These findings can inform antibiotic stewardship programmes to shorten antibiotic treatment and infection prevention and control policies to reduce transmission of resistant bacteria.

Introduction

Reducing treatment duration is a common strategy used to reduce antibiotic consumption and a core feature of antibiotic stewardship programmes (ASPs) [1,2]. This includes optimising definitive antibiotic courses for established bacterial infections while ensuring clinical cure, and rapid discontinuation of empiric prescriptions after bacterial infections are ruled out. Compared with restricting prescriptions, this “back-end” approach of minimising treatment duration is deemed safer for patients and more likely to be accepted by physicians [3]. Numerous randomised trials have concluded that short treatment approaches for common bacterial infections are noninferior to long treatment courses in terms of clinical outcomes [4–6].

The primary motivation behind shortening treatment duration is the expectation that it will reduce antibiotic selective pressure for antimicrobial resistance in treated individuals and, over time, will lead to lower prevalence of resistance at a population level [7]. However, while antibiotic selective pressure as a driver of resistance is not in doubt, there are many gaps in our understanding of the relationship between duration and the prevalence of resistance [8–10]. Specifically, clinical trials of reduced antibiotic treatment duration have reported conflicting effects on resistance carriage at both individual and population levels (Table 1).

Table 1. Details of the antibiotic trials included in the meta-analysis.

No.	Indication for antibiotic treatment	Country	Healthcare setting	Age group	Duration of antibiotics (days)		Antibiotic prescribed	Bacteria detected in surveillance culture	Colonisation site	Duration of follow-up (mean, days)	Total number of participants	Proportion of resistance carriers at the end of treatment or study (resistance carriers/participants who provided samples, %)		Proportion difference between the long and short arms (long—short, 95% confidence intervals)^{^}	Reference
No.	Indication for antibiotic treatment	Country	Healthcare setting	Age group	Short	Long	Antibiotic prescribed	Bacteria detected in surveillance culture	Colonisation site	Duration of follow-up (mean, days)	Total number of participants	Short arm	Long arm		Reference
1	Neonatal late-onset sepsis	European countries	Inpatient	Neonate	7	8	Meropenem vs standard-of-care	Carbapenem-resistant gram-negative bacteria	Gut	28	272	7/94 (7)	19/101 (19)	11 (1, 22)	Lutsar, 2020[37]
2	Otitis media	USA	Outpatient	Children	5	10	Amoxicillin/clavulanic acid	Beta-lactamase producing H. influenzae	Respiratory tract	14	520	28/222 (13)	38/233 (16)	4 (−3, 11)	Hoberman, 2016[38]
3	Uncomplicated urinary tract infection	Turkey	Outpatient	Adults	1	5	Fosfomycin vs ciprofloxacin	Fosfomycin- or ciprofloxacin-resistant Enterobacteriaceae	Urinary tract	7	260	7/77 (9)	12/65 (18)	9 (−3, 22)	Ceran, 2010[39]
4	Uncomplicated urinary tract infection	the Netherlands	Outpatient	Adults	3	5	Trimethroprim	Trimethroprim-resistant E. coli	Urinary tract	3	324	8/66 (12)	12/63 (19)	7 (−7, 21)	Merode, 2005[40]
5	Urinary tract infection in patients with spinal cord injury	Canada	Inpatient	Adults	3	14	Ciprofloxacin	Fluroquinolone-resistant gram-negative bacteria	Urinary tract	41	60	8/30 (27)	5/30 (17)	−10 (−14, 4)	Dow, 2004[41]

Open in a new tab

^{^} Difference in proportions were calculated using two-tailed Z-test for 2 proportions.

Antibiotics promote resistance by killing or inhibiting the growth of susceptible microbes while allowing resistant ones to grow. This selective pressure for resistance most clearly applies to pathogens directly targeted by the antibiotic when used to treat or prevent infections. Examples of pathogens that may become resistant during treatment include Mycobacterium tuberculosis, fluoroquinolone-resistant Enterobacterales, and human immunodeficiency virus, especially with inappropriate dosing and duration [11,12]. More recently, it has been recognised that, for most bacterial pathogens, antibiotic selection pressure largely results from “bystander” selection, which occurs when colonising bacteria are exposed to antibiotics that are not intentionally targeting them [13]. The impact of bystander exposures to antibiotics depends on their spectrum of coverage, pharmacokinetic and pharmacodynamic properties, and bioavailability where the bacteria are located. Antibiotic resistance in colonising bacteria is clinically important because asymptomatic carriage of these bacteria typically precedes invasive infection, and preventing colonisation will reduce associated morbidity and mortality [14–16]. For these reasons, this study focuses on the “bystander” selective pressure resulting from different durations of antibiotic treatment on colonising bacteria.

Emergence of resistance in colonising bacteria may be due to selection of resistant phenotypes that are present at the start of treatment but at low abundance or acquired by transmission during treatment, as a result of horizontal gene transfer, or as a result of de novo mutations. The effect of treatment duration on development of resistance will depend on the parameters governing these processes and their relative importance.

To gain a mechanistic understanding of this relation between antibiotic treatment duration and the prevalence of colonisation with antibiotic-resistant bacteria in hospitalised patients, we modelled within- and between-host dynamics of susceptible and resistant bacteria in response to systemic antibiotic treatment. We then sought to compare the model findings with empirical observations by performing a systematic review and meta-analysis.

Methods and materials

Modelling dynamics of colonising bacteria in response to antibiotic treatment

Three stochastic agent-based models with increasing complexity and biological realism were constructed. From here on, they will be referred to as simple exclusive colonisation model, co-colonisation model, and within-host growth model (Fig 1).

Fig 1 — These flow diagrams describe within-host changes in resistant and sensitive bacteria carried by each individual patient in response to antibiotic consumption, bacterial growth, and transmission events. In the 3 models, all patients were assumed to be always colonised with bacteria. In the exclusive colonisation model, patients can be colonised with resistant bacteria (R), and high or low levels of sensitive bacteria (S or s), but not both resistant and sensitive at the same time due to complete bacterial interference. The co-colonisation model makes a more realistic assumption that patients may be colonised with both resistant and sensitive bacteria at the same time with different levels of abundance. Similarly, in the within-host growth model, patients carry both resistant and sensitive bacteria but their combined populations cannot exceed a maximum carrying capacity. Blue, orange, and pink squares/circles represent the host states when the host is carrying susceptible, a mix of susceptible and resistant, and resistant bacteria, respectively. The red arrows highlight antibiotic killing of resistant bacteria. All transmission and decolonisation events apply to resistant bacteria.

In all models, we simulated individual patients (agents) in a single hospital ward environment. The patients’ colonisation status could change due to (i) differential growth/killing rates of resistant and sensitive bacterial populations within a patient; (ii) transmission events between patients; and (iii) loss of resistance carriage (Fig 1). Because the population size under consideration is small and local fade-out events (when resistant populations reach zero) are potentially important, we considered stochastic implementations of these models (Fig 2 and Section S1.1 in S1 Text), i.e., changes in colonisation status took place with a probability randomly drawn from a distribution.

Fig 2 — This example simulation shows two 10-bedded wards with the same transmission risk of resistant bacteria, patient lengths of stay, proportion of patients who required antibiotic treatment, and proportion of patients who were resistance carriers upon admission. The left column panels represent the ward where antibiotic durations were short (mean of 5 days), while the right column panels represent the ward where antibiotic durations were long (mean of 15 days). Top row panels: Antibiotic treatment types and duration for one single iteration. Each square in the graphs represents one patient day. Light and dark green squares indicate one day of antibiotic effective only against susceptible organisms or both susceptible and resistant organisms. Vertical blue lines represent new admissions to the ward. Second row panels: The carriage status for each patient on a particular day for one single iteration. The increasingly darker shades of orange and green indicate an increasing amount of resistant and susceptible colonising bacteria carried by each patient, respectively. Third row panels: The number of antimicrobial resistance carriers in each ward per day for 40 iterations. Fourth row panel: The cumulative number of resistance carriers in each ward over 60 observation days for 40 iterations. In the third and fourth row panels, each line represents output from one iteration. The thick lines are outputs from the single example iteration illustrated in the top 2 panels.

Within the same ward, antibiotics (administered via both intravenous and oral routes) were prescribed with the same mean duration on or during admission. Exact duration for each antibiotic course was drawn from a uniform distribution (range for short duration 3 to 7 days, range for long duration 14 to 21 days). We considered 2 scenarios. In the first, the resistant phenotype was assumed to be susceptible to one of the administered antibiotics, e.g., carbapenems against third-generation cephalosporin resistance. In the second scenario, there was no effective antibiotic available against the resistant phenotype in the colonising bacteria, e.g., colistin or carbapenem resistance. Subsequently, we refer to these 2 scenarios as “administered antibiotics have activity against resistant and susceptible organisms” and “administered antibiotics have activity only against susceptible organisms.”

The models assumed that antibiotics act within-host by selectively promoting the growth of existing resistant bacteria in the gut [17,18], and between-host by predisposing the treated individual to become more likely to transmit resistant bacteria as a consequence of a higher bacterial load (Fig 1) [19]. Spontaneous decolonisation of resistant bacteria was assumed to only occur when a patient was not receiving antibiotics to which the colonising bacteria were resistant.

In the exclusive colonisation model, patients could be carriers of either susceptible or resistant bacteria but not of both at the same time [20]. In this model, for a carrier of susceptible bacteria to become a resistance carrier required transmission of resistant bacteria from another carrier in the ward, i.e., we did not consider de novo resistance emergence or horizontal transfer from other components of the flora. In the other 2 more complex models, patients could carry both susceptible and resistant bacteria simultaneously. In the co-colonisation model, these abundances were dichotomised into high and low, and only bacteria carried at a high level were able to be transmitted to other patients [21]. In the within-host growth model, bacterial population sizes could vary continuously but had to be above a threshold in order to be capable of transmission to others [22]. We also assumed a total carrying capacity for gram-negative bacteria [23] and note that newly admitted patients may carry fewer bacteria than the carrying capacity as we would expect some patients were prescribed antibiotics prior to admission. The mean total carrying capacity for each iteration was drawn from a lognormal distribution (Table A in S1 Text). In the latter 2 models, an individual could become a resistance carrier from within-host selection of preexisting resistant bacteria through antibiotic consumption.

While the modelling framework we have adopted is quite general, here we focus on parameter space that are appropriate for gram-negative bacteria as they are the commonest cause of urinary tract and bloodstream infections and frequently associated with multidrug-resistant hospital-acquired infections. The predominant mechanisms for resistance dissemination in gram-negative bacteria colonising the gut are horizontal transfer of mobile genetic elements and clonal expansion [24].

We first explored the models by varying pairs of parameters over a grid of values while holding the other parameter values constant. This was followed by global sensitivity analysis using Latin Hypercube sampling and Partial Rank Correlation Coefficient (S1.3 Section in S1 Text). Real-world parameter values were obtained from the literature and used in these explorations (Table A in S1 Text). For those parameter values not found in the literature, we explored the largest reasonable range of parameter values. In all results shown below, each simulation was produced from at least 50,000 iterations (100 repeats of 500 unique sets of parameters selected within the ranges shown in Table A in S1 Text) over 300 days. The computer code is publicly available (https://github.com/moyinNUHS/abxduration_abm).

The models were used to assess how changing duration of antibiotic treatment affected the risk of resistance colonisation at both individual and population levels. Three types of outcomes were considered: resistance carriage among (i) treated patients and, therefore, directly affected by different treatment durations; (ii) overall resistance carriage within a ward population that consisted of treated and untreated patients, i.e., indirectly affected by treatment duration received by those treated patients; and (iii) patients who were not carriers of resistant bacteria when admitted to the ward. To evaluate reducing antibiotic duration as an intervention, all the model outputs were assessed as absolute difference between the short and long wards.

All simulations were performed with R version 4.0.4 (2021-02-15) [25] using packages msm [26], pse [27], and spartan [28]. Code review was done using unit tests for each function’s base scenarios, which can be found under /unit_tests in the source code.

Systematic review

To compare model findings with empirical evidence, we performed a systematic review of antibiotic duration randomised controlled trials, which reported resistance carriage. We searched MEDLINE and EMBASE for randomised controlled trials published from 1 January 2000 up to 4 October 2022, which allocated participants to varying durations of systemic antibiotic treatments. Search strings are provided in Table A in S2 Text. Studies that compared antibiotics to no antibiotic treatment were excluded. Quality assessment of the studies was performed using the Cochrane risk-of-bias tool for randomised trials (RoB 2 tool) [29]. MY and MO reviewed the shortlisted articles, extracted and verified the underlying data for the meta-analysis independently. This study is reported as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline (S1 Prisma Checklist). The funders had no role in the design and conduct of the review. The systematic review reviewers (MY and MO) have no competing interest.

The study methodologies and outcomes were expected to be too diverse to be meaningfully included in a single meta-analysis. Instead, we prespecified a relatively homogeneous group defined by studies that (i) measured asymptomatic carriage of resistant bacteria as an outcome; and (ii) collected cultures at predefined time point(s) for surveillance (not clinically indicated) during follow-up. The specific inclusion criteria and search strategy can be found in Section S2.1 in S2 Text and Table A in S2 Text. The indications for antibiotic treatment in the trials included treatment or prevention of bacterial infections, decolonisation of resistant bacteria, and anti-inflammation for autoimmune diseases. The sites of colonisation included digestive, respiratory, and urinary tracts. We considered resistance carriage when the antibiotic prescribed in the trial was not effective against the specific type of resistance, e.g., ciprofloxacin was prescribed in a trial that monitored carriage of fluoroquinolone-resistant bacteria. No trials monitored resistance phenotypes, which the prescribed antibiotic was effective against.

The specific aims of the systematic review were to (i) evaluate the quality of evidence in the current literature of antibiotic duration effects on resistance carriage; (ii) describe and summarise the findings from these studies; and (iii) compare the findings to the model outputs.

Meta-analysis

From the studies identified in the systemic review, we selected those that monitored resistant gram-negative bacteria carriage for a meta-analysis. The meta-regression analysis was performed using a Bayesian model to estimate the change in the daily risk of acquiring resistant bacteria colonisation per day of antibiotic consumption. The dependent variable in the meta-regression analysis was the number of patients colonised with resistant bacteria in a given arm of a given trial, together with the associated denominator. These data were analysed with a logistic regression using antibiotic duration administered to the patients in each arm and elapsed time from antibiotic administration to surveillance culture as independent variables. The regression model assumed each arm in each trial to be independent (fixed effect). Details of the models can be found in S2.4 Section in S2 Text.

We also considered 2 other models. In the first alternative model, we included the healthcare setting as an additional independent variable. In the second, we allowed slopes and intercepts to vary between trials (random effect). Model comparison was done using the Widely Applicable Information Criterion (WAIC), where lower values indicate improved model fit [30].

We implemented the meta-regression models in JAGS using the R2jags package [31] and performed all analysis in R version 4.0.4 (2021-02-15) [25]. All analysis codes are available at https://github.com/moyinNUHS/abxduration_abm.

Results

Individual and population resistance dynamics with antibiotic treatment

We first consider as an illustrative example of model behaviour a ward where 25% of the patients carried resistant bacteria on admission and half of the patients were given antibiotics upon admission (Fig 3). In the scenario where the patients were given an antibiotic with activity against both susceptible and resistant colonising bacteria (Fig 3, top 2 panels), in all 3 models, carriage of the resistant bacteria declined both in treated individuals and at the population level with longer antibiotic treatment duration.

When antibiotics had activity only against susceptible organisms (Fig 3, bottom 2 panels), the “bystander” selection on resistant bacteria led to an increased prevalence of resistance carriers with longer treatment. More resistance carriers in the ward then acted as a reservoir to further spread resistance to noncarriers.

Key parameters in the relationship between antibiotic duration and resistance carriage

The global sensitivity analysis identified transmission rates and prevalence of resistance on admission as important effect modifiers of the relationship between treatment duration and resistance prevalence (Fig 4, “baseline carriage status” and “resistance transmission” rows). Reducing treatment duration was more effective in reducing resistance carriage in the overall ward population when the transmission rate was high regardless of whether the administered antibiotics were active against resistant organisms.

Fig 4 — Partial rank correlation coefficients for the various parameters are shown in the heatmap. Parameters from the 3 different models are grouped according to the main within- and between-host processes they describe. The 3 panels (separated by black vertical lines) represent the 3 types of outcomes assessed. Within each panel, the left 3 columns refer to the scenario where administered antibiotics have activity only against susceptible organisms; the right 3 columns refer to the scenario where administered antibiotics have activity against resistant and susceptible organisms. Pink-, teal-, and grey-coloured rectangles indicate that as each parameter value increased, the difference in high-abundance resistance carriers between the long- and short-duration wards increased, decreased, and did not change, respectively. AB, antibiotic; GNB, gram-negative bacteria; R, Resistant; S, Susceptible. White rectangles indicate that the parameter was not found in the model or not applicable. *Parameters for the percentage of patients who were administered antibiotics are not relevant for the outcome observed in the treated patients and therefore coloured white.

Conversely, the lower the resistance prevalence on admission, the more effective reducing treatment duration was in reducing resistance carriage at a population level. This is because importation of resistance carriers is not affected by antibiotic treatment. Hence, for the ward population, the lower the prevalence of resistance carriage on admission, the greater effect antibiotic duration has on overall resistance.

However, among patients who were treated and who were noncarriers on admission, reducing treatment duration tended to be more effective when the prevalence of resistance carriage on admission was high. This is because in these subsets of patients, the gain in resistance carriage was mainly through transmission events from existing carriers rather than within-host bystander selection.

Within-host, a rapid growth rate of susceptible bacteria after stopping antibiotics and resistant bacteria while receiving antibiotic treatment resulted in more resistance carriers in the ward administered longer treatment duration compared to the ward administered short duration (Fig 4, “within-host bacterial growth and decolonisation” row).

Within-host dynamics of colonising bacteria with varying antibiotic treatment duration

Bacterial growth and decolonisation rates in the absence of antibiotics were important within-host factors in the relationship between treatment duration and resistance carriage regardless of whether administered antibiotics were active against resistant organisms. When resistance phenotypes are resistant to antibiotic treatment, under both fast and slow growth scenarios, we find that long duration treatment allows the resistant population to become established at a much larger proportion of the total bacterial load at the end of treatment, which in turn results in a longer time for the resistant population to decline to low levels once treatment ceases (Fig 5).

Fig 5 — These graphs illustrate an example where the susceptible and resistant bacteria within the same host were exposed to 15 days (left panel) versus 3 days (right panel) of antibiotic treatment. This antibiotic was capable of killing susceptible bacteria but ineffective against resistant bacteria. All parameter values other than bacterial growth rates were kept the same. Rapid bacterial growth is represented by solid lines. Slow bacterial growth is represented by dashed lines.

During antibiotic treatment, susceptible bacteria declined while resistant bacteria were able to grow. Hence, rapid resistant bacteria growth enabled hosts to transmit resistance onward to others in the ward (Fig 5, solid pink lines). Upon termination of antibiotics, rapid susceptible bacteria recovery suppressed resistant bacteria growth (Fig 5, solid blue lines). However, when susceptible bacteria recovered slowly after termination of antibiotics, patients who were administered antibiotics could remain colonised with resistant bacteria for prolonged periods regardless of the treatment duration (Fig 5, dashed pink lines).

Resistance transmission and importation on the effects of antibiotic duration

In the scenario where administered antibiotics have activity only against susceptible organisms, antibiotic treatment had separate effects on the treated nonresistance carriers and resistance carriers (Fig 6). For the noncarriers, antibiotics increased the risk of acquiring resistant bacteria by reducing the abundance of colonising bacteria to below the carrying capacity. For the existing carriers, antibiotics preferentially promoted the growth of resistant bacteria, increasing these patients’ probability of transmitting resistance to others. Hence, higher transmission rates amplified the effect of treatment duration on increasing resistance carriers by bridging the transfer of resistance from the carriers to the at-risk noncarriers.

Fig 6 — The values shown (represented by the coloured pixels) were derived from the models when they have reached equilibrium. Panel A: Output from co-colonisation model under various parameter values (y-axis) for (i) transmission rate and (ii) prevalence of resistance carriers admitted to the ward with increasing antibiotic duration (x-axis). The coloured pixels indicate the proportion of ward patients with a given carriage status where *S/s* represent high and low proportions of susceptible bacteria and *R/r* represents high and low proportions of resistant bacteria. Increasing antibiotic duration increased the risk of the treated patients acquiring resistance carriage by suppressing the susceptible bacteria population. This resulted in shifts with increasing duration of treatment from S and Sr states to predominantly state s at low transmission rates and to predominantly state sr and sR at higher transmission rates. Panel B: Output from the exclusive colonisation model. (i) Transmission rate (x-axis) and (ii) prevalence of resistance carriers admitted to the ward (y-axis) were varied while other parameters were fixed to explore the difference in proportion of resistance carriers between the wards administered long and short antibiotic duration (coloured pixels). Red, white, and blue pixels indicate that there were more resistance carriers in the long ward (i.e., shortening treatment duration was effective at reducing resistance carriers), no difference between the short and long ward (i.e., shortening treatment duration had no effect on resistance carriers), and more resistance carriers in the short ward (i.e., prolonged treatment duration was effective at reducing resistance carriers), respectively. The top and bottom rows describe the scenarios where administered antibiotics have activity against both susceptible and resistant organisms and only susceptible organisms, respectively. The columns represent the 3 populations in which the outcomes were assessed: (i) treated individuals; (ii) overall ward population; and (iii) noncarriers on admission.

When administered, antibiotics were active against the resistant organisms; there remained areas of parameter space where longer treatment duration led to more resistance (Fig 6, Panel B top row). This was observed at high resistance importation and transmission rates, especially for treated patients and patients who were noncarriers at admission. This is because in these patients, longer antibiotic treatment reduced the abundance of colonising bacteria and increased these patients’ risk of acquiring resistance.

Overall effect of antibiotic duration on resistance carriage

The models showed generally modest and highly heterogenous effects of antibiotic duration on resistance carriage (Fig 7). When antibiotics have activity only against susceptible organisms, longer antibiotic duration resulted in higher prevalence of resistance carriage (Fig 7, right-sided panels). In contrast, when antibiotics have activity against both resistant and susceptible organisms, longer antibiotic duration resulted in either an increase or decrease in the prevalence of resistance carriers (Fig 7, left-sided panels).

Evidence from randomised controlled trials

Initial search of the MEDLINE and EMBASE databases returned 2,649 unique publications. Out of these 206 were randomised trials that compared antibiotic treatment durations. Ten of these trials collected surveillance cultures for colonising bacteria during follow-up visits and were included in the qualitative synthesis. A meta-regression analysis was performed using 5 out of the 10 trials (Table 1). One was excluded because the antibiotic duration in the long arm and the follow-up time were very prolonged compared to the other studies (104 days treatment compared to 5 to 14 days, and 365 days follow-up compared to 3 to 41 days) [32]. One was excluded because the trial reported a mixture of antibiotics and multidrug-resistant bacteria carriage without specifying if the organisms were gram-positive or gram-negative [33]. Three others were excluded because the study only collected gram-positive bacteria in surveillance cultures [34–36]. The PRISMA diagram can be found in the Fig A in S2 Text. Complete data extracted from these randomised trials can be found at https://github.com/moyinNUHS/abxduration_abm.

Among the 10 trials, the indications for antibiotic treatments in 9 were for treatment of infections (urinary tract infections (3), otitis media (1), sepsis (1), acute respiratory illness (4)), and in one was for Pseudomonas spp. in noncystic fibrosis bronchiectasis. Six were performed in outpatient settings with mean follow-up periods ranging from 3 to 365 days; 4 were in hospital settings with mean follow-up periods ranging 28 and 90 days. Five trials enrolled only adults, while the other 5 enrolled only children. All were performed in high- or upper-middle-income countries. Antibiotic durations in the short arms ranged from 3 to 14 days (median 5 days), while in the long arms, the durations ranged from 5 to 104 days (median 10 days).

The meta-analysis found an odds ratio of acquiring gram-negative resistance carriage with one additional day of antibiotics of 1.08 (80% credible interval 1.04% to 1.12%). This translates to an absolute 7% increase in daily probability of acquiring resistance carriage given a baseline daily probability of 0.1 (Fig 8). Sensitivity analyses using data only from studies with surveillance cultures collected up to 30 days after antibiotic treatment and using different priors produced similar results (Table C in S2 Text).

Fig 8 — Daily probability of colonisation by antibiotic-resistant gram-negative bacteria (y-axis) is shown against mean duration of antibiotics (x-axis) reported in each trial. Each colour represents one trial. Each bubble represents a single arm in one trial, where the diameter of the bubble corresponds to the number of participants for the arm in the trial. The analysis allowed the relationship between colonisation and antibiotic duration to vary across each arm in the trials; the blue line corresponds to the mean relationship between the 2 considering all included trials. The light grey shaded areas are the associated 80% and 50% credible intervals.

Three out of 5 studies were deemed to be at risk of bias due to the high proportion of participants who were lost to follow-up and did not provide surveillance samples (Table D in S2 Text). However, all studies were assessed to be at low risk of bias due to deviation from intended interventions.

Discussion

The 3 models reached broadly similar conclusions: Among treated individuals, shortening antibiotic duration is most effective at reducing resistance carriage when the bacterial dynamics respond rapidly to antibiotic treatment. When stopping antibiotics leads to rapid decline in the resistant strain’s abundance, shortening duration is more likely to result in a substantial reduction in the prevalence of resistance [18]. At a population level, high transmission of resistant bacteria between hosts was a key factor in the efficacy of shortening treatment duration at reducing resistance carriage.

Longer antibiotic treatment may decrease or increase the prevalence of resistance carriers, depending (among other factors) on the availability of effective antibiotics against particular resistance phenotypes. The meta-analysis, using data from randomised controlled trials, highlighted the increased incidence of resistance colonisation with treatment duration when there was no effective antibiotic treatment available. The models also showed that even when antibiotics administered are active against a resistance phenotype, longer duration may increase resistance carriers especially among the treated individuals and noncarriers when the prevalence of resistance carriers on admission and transmission rates are high.

We searched MEDLINE and EMBASE for modelling studies that evaluated antibiotic treatment duration on antimicrobial resistance and found 3 related publications. The first study by D’Agata and colleagues is a theoretical agent-based model, which concluded that rapid initiation of treatment and minimising duration were likely to reduce antimicrobial resistance in a hospital epidemic setting [42]. The second study by D’Agata and colleagues focused on within-host dynamics of resistance by effective and ineffective antibiotic killing [43]. It found that shorter and early interruption of antibiotic therapy selected resistant strains when antibiotic killing was effective. The other study is by Blanquart and colleagues, in which antibiotic prescription and antimicrobial resistance data specific to Streptococcus pneumoniae from the Israeli community were used [44]. This study suggested that reducing the number of courses of antibiotics might be a more efficient strategy for reducing antimicrobial resistance than reducing treatment duration. There were no studies that directly addressed the relationship between antibiotic duration and antimicrobial resistance in both individual and population levels specific to healthcare settings.

We did not find any cluster-randomised trials of antibiotic duration that monitored resistance colonisation at the population level. Such trials would have allowed us to compare the individual versus population effects from antibiotic selection. Since most resistance selection in common bacterial pathogens is due to “bystander” selection [45], trials that only look at resistance in bacteria that are causing infections being treated are potentially missing a big part of the picture. Future trials should collect surveillance cultures to evaluate the effect of duration on resistance carriage.

The modest effect of shortening antibiotic duration on reducing resistant bacteria colonisation can be partially explained by the unique patient and antibiotic prescribing characteristics in healthcare settings. These include typically short lengths of stay and a slow rate of decolonisation of resistant bacteria in those previously exposed to antibiotics [17,18]. Admitted patients under antibiotic treatment acquiring resistant bacteria carriage can thus remain as carriers during relatively brief hospitalisation periods even after antibiotic treatment is stopped. This initial rapid increase in risk of acquiring resistance bacteria carriage during the first few days of antibiotic treatment could reduce the effectiveness of shortening treatment duration at reducing resistance.

Our findings highlight an important interaction between shortening antibiotic treatment and reducing transmission of resistant bacteria through infection prevention and control. This is intuitive because in addition to selecting for within-host resistance, antibiotic use also increases the risk of colonisation with resistant bacteria and subsequent prolonged colonisation [17,18]. Shortening treatment duration has the potential to reduce a patient’s risk of acquiring resistant bacteria, while infection prevention and control measures reduce transmission of resistant bacteria to these at-risk patients.

There are important caveats and limitations in our study. Firstly, despite performing a comprehensive systematic review, some parameter values were not available in the literature or were taken from animal or in vitro experiments. In such cases, we explored the largest reasonable range of parameter values. Secondly, emergence and spread of antimicrobial resistance under antibiotic selective pressure are complex and cannot be fully described with any of the models presented. It is reassuring, however, that all 3 versions of the models produced similar conclusions. Thirdly, for the meta-analysis, there were few antibiotic duration trials that monitored resistance carriage as an outcome. This contributed to large credible intervals in the daily resistance colonisation risk with antibiotic treatment. These limitations highlight important gaps in existing literature and the conduct of antibiotic duration trials. Fourthly, our models did not account for de novo mutations during treatment, which are an important source of resistance for certain organisms such as Mycobacterium tuberculosis [46]. Instead, we focused on pathogens in which colonisation with resistant phenotypes is primarily acquired through transmission and horizontal gene transfer. These are highly clinically relevant gram-negative bacteria such as carbapenemase-producing Enterobacterales, extended spectrum beta-lactamase-producing Enterobacterales, and carbapenem-resistant Acinetobacter spp. Lastly, the agent-based models captured both individual and population-level effects of antibiotic selection pressure. However, the trials included in the meta-analysis were individually randomised, and results therefore reflected only the direct individual antibiotic selection effects.

Understanding the key factors driving the effect antibiotic duration on resistance carriage will inform future research study designs, antimicrobial stewardship interventions, and resource allocation in the overall control strategies. The practical implications from our findings are that interventions for shortening antibiotic treatment duration are potentially most effective when antibiotics are stopped as early as feasible without compromising treatment success for the target pathogen, especially in high transmission settings.

Supporting information

S1 Text. Supporting information: Model details and exploration.

Table A. Parameters used in the models and their respective ranges obtained from literature review.

(DOCX)

Click here for additional data file.^{(3.4MB, docx)}

S2 Text. Supporting information 2: Systematic review methodology.

Table A. Search terms used in the systematic review. Table B. Model comparisons. Table C. Sensitivity analysis. Table D. Quality assessment of the randomised controlled trials included in the meta-analysis.

(DOCX)

Click here for additional data file.^{(6.1MB, docx)}

S1 PRISMA Checklist. PRISMA checklist.

(DOCX)

Click here for additional data file.^{(33.1KB, docx)}

Acknowledgments

We would like to thank Anastasia Hernandez-Koutoucheva, Ricardo Sempedro, and Jiraboon Tosanguan for performing code review.

Data Availability

The computer code for all simulations and data for the meta-analysis are publicly available (https://github.com/moyinNUHS/abxduration_abm).

Funding Statement

This work was supported by the Singapore National Medical Research Council Research Fellowship (NMRC/Fellowship/0051/2017 to MY), the UK Medical Research Council (MR/V028456/1 to BSC), the Wellcome Trust (206736/Z/17/Z to CL and 106698/Z/14/Z to Mahidol Oxford Tropical Medicine Research Programme). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Barlam TF, Cosgrove SE, Abbo LM, Macdougall C, Schuetz AN, Septimus EJ, et al. Implementing an antibiotic stewardship program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clinical Infectious Diseases. Oxford University Press; 2016. p. e51–77. doi: 10.1093/cid/ciw118 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.World Health Organization(WHO). Antimicrobial Stewardship Programmes in Healthcare Facilities in Low-and-middle-income Countries, a WHO Practical Toolkit Antimicrobial Stewardship. 2019 [cited 2022 Aug 8]. https://apps.who.int/iris/bitstream/handle/10665/329404/9789241515481-eng.pdf.
3.Chung GW, Wu JE, Yeo CL, Chan D, Hsu LY. Antimicrobial stewardship: A review of prospective audit and feedback systems and an objective evaluation of outcomes. Virulence. 2013;4:151–57. doi: 10.4161/viru.21626 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Sandberg T, Skoog G, Hermansson AB, Kahlmeter G, Kuylenstierna N, Lannergård A, et al. Ciprofloxacin for 7 days versus 14 days in women with acute pyelonephritis: a randomised, open-label and double-blind, placebo-controlled, non-inferiority trial. Lancet. 2012;380:484–90. doi: 10.1016/S0140-6736(12)60608-4 [DOI] [PubMed] [Google Scholar]
5.Chastre J, Wolff M, Fagon J-Y, Chevret S, Thomas F, Wermert D, et al. Comparison of 8 vs 15 Days of Antibiotic Therapy for Ventilator-Associated Pneumonia in Adults. JAMA. 2003;290:2588. doi: 10.1001/jama.290.19.2588 [DOI] [PubMed] [Google Scholar]
6.Uranga A, España PP, Bilbao A, Quintana JM, Arriaga I, Intxausti M, et al. Duration of Antibiotic Treatment in Community-Acquired Pneumonia. JAMA Intern Med. 2016;176:1257. doi: 10.1001/jamainternmed.2016.3633 [DOI] [PubMed] [Google Scholar]
7.Spellberg B, Rice LB. Duration of Antibiotic Therapy: Shorter Is Better. Ann Intern Med. 2019;171:210. doi: 10.7326/M19-1509 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.van Duijn PJ, Verbrugghe W, Jorens PG, Spöhr F, Schedler D, Deja M, et al. The effects of antibiotic cycling and mixing on antibiotic resistance in intensive care units: a cluster-randomised crossover trial. Lancet Infect Dis. 2018;18:401–9. doi: 10.1016/S1473-3099(18)30056-2 [DOI] [PubMed] [Google Scholar]
9.Schuts E, Hulscher M, Mouton J, Verduin C, Stuart J, Overdiek H, et al. Current evidence on hospital antimicrobial stewardship objectives: a systematic review and meta-analysis. Lancet Infect Dis. 2016;16:847–56. doi: 10.1016/S1473-3099(16)00065-7 [DOI] [PubMed] [Google Scholar]
10.Leo S, Lazarevic V, von Dach E, Kaiser L, Prendki V, Schrenzel J, et al. Effects of antibiotic duration on the intestinal microbiota and resistome: The PIRATE RESISTANCE project, a cohort study nested within a randomized trial. EBioMedicine. 2021:71. doi: 10.1016/J.EBIOM.2021.103566 [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Totsika M, Beatson SA, Sarkar S, Phan M-D, Petty NK, Bachmann N, et al. Insights into a multidrug resistant Escherichia coli pathogen of the globally disseminated ST131 lineage: genome analysis and virulence mechanisms. PLoS ONE. 2011;6:e26578. doi: 10.1371/journal.pone.0026578 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Gupta RK, Gregson J, Parkin N, Haile-Selassie H, Tanuri A, Andrade Forero L, et al. HIV-1 drug resistance before initiation or re-initiation of first-line antiretroviral therapy in low-income and middle-income countries: a systematic review and meta-regression analysis. Lancet Infect Dis. 2018;18:346–55. doi: 10.1016/S1473-3099(17)30702-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Sullivan Å. Effect of antimicrobial agents on the ecological balance of human microflora. Lancet Infect Dis. 2001;1:101–14. doi: 10.1016/S1473-3099(01)00066-4 [DOI] [PubMed] [Google Scholar]
14.Troché G, Toly L-M, Guibert M, Zazzo J-F. Detection and Treatment of Antibiotic-Resistant Bacterial Carriage in a Surgical Intensive Care Unit: A 6-Year Prospective Survey. Infect Control Hosp Epidemiol. 2005;26:161–65. doi: 10.1086/502521 [DOI] [PubMed] [Google Scholar]
15.Dubinsky-Pertzov B, Temkin E, Harbarth S, Fankhauser-Rodriguez C, Carevic B, Radovanovic I, et al. Carriage of Extended-spectrum Beta-lactamase-producing Enterobacteriaceae and the Risk of Surgical Site Infection after Colorectal Surgery: A prospective cohort study. Clin Infect Dis. 2019;68:1699–1704. doi: 10.1093/cid/ciy768 [DOI] [PubMed] [Google Scholar]
16.Reddy P, Malczynski M, Obias A, Reiner S, Jin N, Huang J, et al. Screening for extended-spectrum β-lactamase-producing enterobacteriaceae among high-risk patients and rates of subsequent bacteremia. Clin Infect Dis. 2007;45:846–52. doi: 10.1086/521260 [DOI] [PubMed] [Google Scholar]
17.Tacconelli E, De Angelis G, Cataldo MA, Mantengoli E, Spanu T, Pan A, et al. Antibiotic usage and risk of colonization and infection with antibiotic-resistant bacteria: A hospital population-based study. Antimicrob Agents Chemother. 2009;53:4264–69. doi: 10.1128/AAC.00431-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Mo Y, Hernandez-Koutoucheva A, Musicha P, Bertrand D, Lye D, Ng O, et al. Duration of carbapenemase-producing enterobacteriaceae carriage in hospital patients. Emerg Infect Dis. 2020;26:2182–85. doi: 10.3201/eid2609.190592 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Kim S, Covington A, Pamer EG. The intestinal microbiota: Antibiotics, colonization resistance, and enteric pathogens. Immunol Rev. 2017:90–105. doi: 10.1111/imr.12563 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Austin DJ, Kakehashi M, Anderson RM. The transmission dynamics of antibiotic-resistant bacteria: The relationship between resistance in commensal organisms and antibiotic consumption. Proc R Soc B Biol Sci. 1997;264:1629–38. doi: 10.1098/rspb.1997.0227 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Davies NG, Flasche S, Jit M, Atkins KE. Within-host dynamics shape antibiotic resistance in commensal bacteria. Nat Ecol Evol. 2019;3:440–49. doi: 10.1038/s41559-018-0786-x [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Boureau H, Hartmann L, Karjalainen T, Rowland I, Wilkinson MHF. Models to Study Colonisation and Colonisation Resistance. Microb Ecol Health Dis. 2000;2:247–58. doi: 10.1080/08910600050216246 [DOI] [Google Scholar]
23.Niehus R, van Kleef E, Mo Y, Turlej-Rogacka A, Lammens C, Carmeli Y, et al. Quantifying antibiotic impact on within-patient dynamics of extended-spectrum beta-lactamase resistance. Elife. 2020;9:1–20. doi: 10.7554/eLife.49206 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Iredell J, Brown J, Tagg K. Antibiotic resistance in Enterobacteriaceae: Mechanisms and clinical implications. BMJ (Online). BMJ Publishing Group; 2016. doi: 10.1136/bmj.h6420 [DOI] [PubMed] [Google Scholar]
25.R Core Team. R: a language and environment for statistical computing. Vienna, Austria; 2017.
26.Jackson C. Multi-state modelling with R: the msm package 1 Multi-state models 1.1 Introduction. 2019. http://www.jstatsoft.org/v38/i08/.
27.Chalom A, Inácio de Knegt López de Prado P. Package “pse” Type Package Title Parameter Space Exploration with Latin Hypercubes. 2017. https://github.com/andrechalom/pse
28.Alden K, Read M, Timmis J, Andrews PS, Veiga-Fernandes H, Coles M. Spartan: A Comprehensive Tool for Understanding Uncertainty in Simulations of Biological Systems. PLoS Comput Biol. 2013;9:1002916. doi: 10.1371/journal.pcbi.1002916 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Higgins JP, Jelena S, Page MJ, Elbers RG, Sterne JA. Cochrane Handbook for Systematic Reviews of Interventions version 6.2 Chapter 8: Assessing risk of bias in a randomized trial. 2021 [cited 2021 May 6]. https://training.cochrane.org/handbook/current/chapter-08
30.Watanabe S. Asymptotic Equivalence of Bayes Cross Validation and Widely Applicable Information Criterion in Singular Learning Theory. J Mach Learn Res. 2010;11:3571–94. [Google Scholar]
31.Hornik K, Leisch F, Zeileis A, Plummer M. JAGS: A Program for Analysis of Bayesian Graphical Models Using Gibbs Sampling. http://www.ci.tuwien.ac.at/Conferences/DSC-2003/
32.Orriols R, Hernando R, Ferrer A, Terradas S, Montoro B. Eradication therapy against pseudomonas aeruginosa in non-cystic fibrosis bronchiectasis. Respiration. 2015;90:299–305. doi: 10.1159/000438490 [DOI] [PubMed] [Google Scholar]
33.Bouglé A, Tuffet S, Federici L, Leone M, Monsel A, Dessalle T, et al. Comparison of 8 versus 15 days of antibiotic therapy for Pseudomonas aeruginosa ventilator-associated pneumonia in adults: a randomized, controlled, open-label trial. Intensive Care Med. 2022;48:841–49. doi: 10.1007/S00134-022-06690-5 [DOI] [PubMed] [Google Scholar]
34.Barratt S, Bielicki JA, Dunn D, Faust SN, Finn A, Harper L, et al. Amoxicillin duration and dose for community-acquired pneumonia in children: the CAP-IT factorial non-inferiority RCT. Health Technol Assess (Rockv). 2021;25:1–72. doi: 10.3310/hta25600 [DOI] [PubMed] [Google Scholar]
35.Schrag S, Peña C, Fernández J, Sánchez J, Gómez V, Pérez E, et al. Effect of short-course, high-dose amoxicillin therapy on resistant pneumococcal carriage: a randomized trial. JAMA. 2001;286:49–56. doi: 10.1001/jama.286.1.49 [DOI] [PubMed] [Google Scholar]
36.McCallum GB, Fong SM, Grimwood K, Nathan AM, Byrnes CA, Ooi MH, et al. Extended Versus Standard Antibiotic Course Duration in Children <5 Years of Age Hospitalized With Community-acquired Pneumonia in High-risk Settings: Four-week Outcomes of a Multicenter, Double-blind, Parallel, Superiority Randomized Controlled Trial. Pediatr Infect Dis J. 2022;41:549–555. doi: 10.1097/INF.0000000000003558 [DOI] [PubMed] [Google Scholar]
37.Lutsar I, Chazallon C, Trafojer U, De Cabre VM, Auriti C, Bertaina C, et al. Meropenem vs standard of care for treatment of neonatal late onset sepsis (NeoMero1): A randomised controlled trial. PLoS ONE. 2020;15:e0229380. doi: 10.1371/journal.pone.0229380 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Hoberman A, Paradise JL, Rockette HE, Kearney DH, Bhatnagar S, Shope TR, et al. Shortened Antimicrobial Treatment for Acute Otitis Media in Young Children. N Engl J Med. 2016;375:2446–56. doi: 10.1056/NEJMoa1606043 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Ceran N, Mert D, Kocdogan FY, Erdem I, Adalati R, Ozyurek S, et al. A randomized comparative study of single-dose fosfomycin and 5-day ciprofloxacin in female patients with uncomplicated lower urinary tract infections. J Infect Chemother. 2010;16:424–30. doi: 10.1007/s10156-010-0079-z [DOI] [PubMed] [Google Scholar]
40.van Merode T, Nys S, Raets I, Stobberingh EE. Acute uncomplicated lower urinary tract infections in general practice: Clinical and microbiological cute rates after three- versus five-day treatment with trimethoprim. Eur J Gen Pract. 2005;11:55–58. doi: 10.3109/13814780509178238 [DOI] [PubMed] [Google Scholar]
41.Dow G, Rao P, Harding G, Brunka J, Kennedy J, Alfa M, et al. A prospective, randomized trial of 3 or 14 days of ciprofloxacin treatment for acute urinary tract infection in patients with spinal cord injury. Clin Infect Dis. 2004;39:658–64. doi: 10.1086/423000 [DOI] [PubMed] [Google Scholar]
42.D’Agata EMC, Magal P, Olivier D, Ruan S, Webb GF. Modeling antibiotic resistance in hospitals: The impact of minimizing treatment duration. J Theor Biol. 2007;249:487–99. doi: 10.1016/j.jtbi.2007.08.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.D’Agata EMC, Dupont-Rouzeyrol M, Magal P, Olivier D, Ruan S. The Impact of Different Antibiotic Regimens on the Emergence of Antimicrobial-Resistant Bacteria. PLoS ONE. 2008;3:e4036. doi: 10.1371/journal.pone.0004036 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Blanquart F, Lehtinen S, Lipsitch M, Fraser C. The evolution of antibiotic resistance in a structured host population. J R Soc Interface. 2018:15. doi: 10.1098/rsif.2018.0040 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Tedijanto C, Olesen SW, Grad YH, Lipsitch M. Estimating the proportion of bystander selection for antibiotic resistance among potentially pathogenic bacterial flora. Proc Natl Acad Sci. 2018;115:E11988–995. doi: 10.1073/pnas.1810840115 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Palomino JC, Martin A. Drug resistance mechanisms in Mycobacterium tuberculosis. Antibiotics. 2014:317–40. doi: 10.3390/antibiotics3030317 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS Med. doi: 10.1371/journal.pmed.1004013.r001

Decision Letter 0

Louise Gaynor-Brook

4 May 2022

Dear Dr Mo,

Thank you for submitting your manuscript entitled "The effect of reducing antibiotic treatment duration on antimicrobial resistance in the healthcare setting: a modelling study and meta-analysis" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by May 06 2022 11:59PM.

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Louise Gaynor-Brook, MBBS PhD

Senior Editor

PLOS Medicine

PLoS Med. doi: 10.1371/journal.pmed.1004013.r002

Decision Letter 1

Philippa Dodd

11 Jul 2022

Dear Dr. Mo,

Thank you very much for submitting your manuscript "The effect of reducing antibiotic treatment duration on antimicrobial resistance in the healthcare setting: a modelling study and meta-analysis" (PMEDICINE-D-22-01395R1) for consideration at PLOS Medicine.

Your paper was discussed among the editorial team and we invite you to consider publication of your paper in PLOS Medicine's special issue entitled, " Bacterial antimicrobial resistance - surveillance and prevention". Please see the following link for further details: plos.io/AMR.

If you are interested in your paper being published in the special issue please contact me on the email listed at the foot of this letter.

Your paper was sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to invite you to submit a revised version that addresses the reviewers' and editors' comments fully. You will appreciate we cannot make a decision about publication until we have seen the revised manuscript and your response, and we expect to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We hope to receive your revised manuscript by Aug 01 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

Please let me know if you have any questions, and we look forward to receiving your revised manuscript.

Sincerely,

Philippa

Dr Philippa Dodd, MBBS MRCP PhD

Associate editor, PLOS Medicine

pdodd@plos.org

-----------------------------------------------------------

Requests from the editors:

GENERAL

Please report your SR/MA according to the PRISMA 2020 guidelines provided at the EQUATOR site:

http://www.equator-network.org/reporting-guidelines/prisma/

Thank you for including the PRISMA diagram. Please also provide the completed PRISMA checklist. When completing the checklist, please use section and paragraph numbers, rather than page numbers.

We ask you to adapt the title to better match journal style, and suggest: "Implications of reducing antibiotic treatment duration for antimicrobial resistance in hospital settings: A modelling study and meta-analysis".

ABSTRACT

Please structure your abstract using the PLOS Medicine headings (Background, Methods and Findings, Conclusions).

Abstract Background: The final sentence should clearly state the study question.

Please combine the Methods and Findings sections into one section, “Methods and findings”.

Abstract Methods and Findings:

Please ensure that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text (i.e 187 studies Vs 7 included in analyses).

Please include the study design and main outcome measures.

Please include the important dependent variables that are adjusted for in the analyses.

In the last sentence of the Abstract Methods and Findings section, please describe the main limitation(s) of the study's methodology.

Please report your abstract according to PRISMA for abstracts, following the PLOS Medicine abstract structure (Background, Methods and Findings, Conclusions) http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.1001419

Please provide the dates of search, data sources, number of studies included, types of study designs included, eligibility criteria, and synthesis/appraisal methods

Abstract Conclusions:

Please address the study implications without overreaching what can be concluded from the data; the phrase "In this study, we observed ..." may be useful.

Please interpret the study based on the results presented in the abstract, emphasizing what is new without overstating your conclusions.

After the abstract, please add a new and accessible 'Author summary' section in non-identical prose. You may find it helpful to consult one or two recent research papers in PLOS Medicine to get a sense of the preferred style.

MAIN MANUSCRIPT

Please evaluate study quality and risk of bias.

Please evaluate evidence of publication bias.

Please add the following statement, or similar, to the Methods: "This study is reported as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline (S1 Checklist)."

Please clarify the discrepancy between the in the number of studies reported to be included in the analysis (7) and the number of studies referenced in the supp data (9) and provide an updated list as necessary

Please remove COI/funding source/data sharing info from the end of the manuscript and instead include in the manuscript submission form.

Please also remove the 'Role of the funding source' statement from the end of the Methods section.

At line 523, "... systematic review"?

Throughout the text, please style reference call-outs as follows: "... of resistance [5,6]." (noting the absence of spaces within the square brackets).

In the reference list, please convert all italics to plain text.

Where appropriate, please list 6 author names rather than 3, followed by "et al.".

FIGURES

Please provide titles and legends for all figures (including those in Supporting Information files).

Please consider avoiding the use of red and green in order to make your figure more accessible to those with colour blindness

Comments from the reviewers:

*** Reviewer #1:

This interesting manuscript describes a modelling study that examines the impact of duration of antibiotic treatment on antibiotic resistance in Gram negative bacteria using three different agent-based models.

Strengths:

- The manuscript is well written (which is not an easy task given the complexity of the topic) with very thorough methodology

- Understanding the impact of duration of antibiotic therapy on antibiotic resistance is key to better inform antibiotic stewardship decisions

- The manuscript nicely highlights how complex the relationship between treatment duration and antibiotic resistance is and where the knowledge gaps are

Weaknesses:

- As with any model of this complexity a lot depends on the underlying assumptions which given the gaps in the literature remain speculative in many instances. The sensitivity analyses performed and the use of different model which give similar results are reassuring.

- At times this manuscript is difficult to flow which is probably mainly a consequence of the complexity of the subject. At times simple figures may help understanding.

TITLE:

Minor comment: "The effect of reducing antibiotic treatment duration on antimicrobial resistance in the

healthcare setting" Hospital setting rather than healthcare setting is probably preferable?

ABSTRACT:

Major comment: While I understand that space is limited, I feel that the abstract could give more detail regarding the models and underlying assumptions. The way it is currently written it is difficult to eben get a vague idea about what was done.

Minor comment: The hypothetical setting should be described in the abstract.

Minor comment: The term "resistance carriage" should be better defined since it is the main outcome of interest

Minor comment: "7% increase in risk of resistance carriage" Absolute or relative?

INTRODUCTION:

Major comment: I think it would be good to add a paragraph in the introduction about the potential impact on transmission. The association of antibiotic treatment duration and length of stay (and thus probability of transmission / acquisition) would also be worth mentioning.

Major comment: It would also be important to mention that many of the resistant pathogens we care most about (e.g. carbapenemase and extended spectrum beta-lactamase producing Enterobacterales; carbapenem-resistant Acinetobacter) are not selected de novo, but rather acquired through transmission (Unlike fluoroquinolone resistance in Enterobacterales, one will not "create" a carbapenemase producing E. coli even when giving carbapenems for months). I understand that this intended by the "bystander" terminology but I think it would be good to give concrete examples.

Minor comment: "Examples of pathogens which may become resistant with inappropriate treatment include Mycobacterium tuberculosis and human immunodeficiency virus." Rather than using HIV as an example I would mention fluoroquinolone resistance in Enterobacterales.

METHODS AND MATERIALS:

Major comment: It would be good to describe the setup of the modelled hospital in more detail (I assume transfer between units was not simulated?)

Minor comment: "Within the same ward, antibiotics were prescribed with the same mean duration on or during admission." This is obviously quite an oversimplification and should be mentioned as a limitation. Why were the values not sampled from a distribution?

Minor comment "Real-world parameter values were obtained from the literature and used in these explorations (Supplementary material 1 Table 1.2)." These parameters are so essential for the interpretation of the model that they should be part of the main text. Also, some parameters have no references and it should be explained how they were derived.

Minor comment: Was the systematic review registered?

RESULTS:

Minor comment: "Key parameters in the relationship between antibiotic duration and resistance carriage" This section is quite difficult to understand. I acknowledge that is difficult to describe these complex findings simply, maybe some illustrations could help?

DISCUSSION:

Major comment: It should be mentioned as a limitation that both antibiotic use and resistance is much more complex than modelled here (many different antibiotics often in combination for different durations and routes of administrations and dosages over multiple courses with multiple different resistance pathogens with different resistance mechanisms)

Minor comment: "The study by D'Agata et al" This study seems to be missing from the references, idem Blanquart et al. It seems to me that there are two modelling papers by D'Agata looking at treatment duration and resistance (interestingly with somewhat opposing conclusions)

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0004036

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2432019/

FIGURES AND TABLES:

Minor comment: I found figure 1 not simple to understand. Care should be taken to consider readability. The legends of some of the other figures were also difficult to follow (again I acknowledge that this is partly a consequence of the complexity of the topic and models).

REFERENCES:

Minor comments: Some references are not formatted.

Minor comment: There are now also some studies using metagenomics, that show that the impact of treatment duration on intestinal resistance genes for examples is far from obvious https://pubmed.ncbi.nlm.nih.gov/34492446/

*** Reviewer #2 (methodological reviewer):

"The effect of reducing antibiotic treatment duration on antimicrobial resistance in the healthcare setting: a modelling study and meta-analysis" examines said effect in two largely-separate ways: firstly, through mathematical modelling, and secondly through a meta-analysis, which yielded five suitable prior randomised trials. Both methods suggest that reduced antibiotic treatment duration can result in some reduction in resistance carriage. These findings appear plausible, although there are some reservations relating to the representativeness of the mathematical models, and the relatively small sample size of the meta-analysis. Detailed comments follow:

1. It might be clarified as to whether the three models presented (Figure 1) are standard/known from prior work, since there appears extant literature on mathematical/computational modelling for colonisation (resistance), e.g. Boureau, L. Hartmann, T. Karjalainen, I. Rowland, MHF Wilkinson, H. (2000). Models to study colonisation and colonisation resistance. Microbial Ecology in Health and Disease, 12(2), 247-258.

2. In Supplementary S1.2, it is observed that many of the parameters do not appear to have relevant references cited in support of their chosen value ranges. For these parameters, the chosen ranges might be briefly justified.

3. In Supplementary S1.3.1, Figure S2 appears to be labeled as S1. This might be addressed, together with its corresponding reference(s) in the text.

4. In S1.3.2, it is stated that "Non-monotonicities in the correlations between the parameters and the model output were examined visually through scatter plots and calculating the Hoeffding's D measure and Spearman's rank correlation measure"; were any such non-monotonicities found, addressed further?

5. Figure S3 is stated to show scatter plots of the parameters as sampled from hypercubes formed within each parameter space. Then, it is assumed that for each model output (resistance/non-resistance carrier value), each point within the scatterplot for some input variable (e.g. n) represents the estimated relationship between that input variable and the output, with the remaining input variables values being randomly sampled. It is further assumed that each point derives from the "averaged model outputs from [100 iterations] for the same set of parameter combination inputs", as described in S.1.3.1.

If the above understanding is correct, it might be justified as to whether the 500 points (i.e. independent sets of [different] parameter combination inputs, from Line 190) that were used in the scatterplots and associated regression estimates is sufficient. This is because the number of (input) parameter combinations increases exponentially with the number of parameter (i.e. the "curse of dimensionality"). For example, if we consider that a parameter is merely binary (i.e. has a high or low value only), and there are 16 such parameters (as in the smallest exclusive colonisation model), then simply representing all binary input combinations would take 2^16=65,536 points.

An empirical method of testing this might be to re-run the experiments with a different seed/set of 500 unique sets of parameters), and to observe whether similar conclusions are obtained (as presented in Figure 4). This might be warranted as the effects appear minimal for most variables, although there does seem a consistent pattern for p_R on resistant carriers in overall ward population across all three models for example (which however seems somewhat tautological).

6. The caption of Figure S3 further states that the lines in the scatterplots are regression lines obtained from locally estimated scatterplot smoothing. The algorithm/procedure for plotting these lines from the points might be briefly described.

7. Tradeoffs associated with shortening antibiotic treatment duration might also be acknowledged and discussed, as otherwise a reduction to zero duration would appear optimal.

Minor issues:

(Line 78) "This selective pressure for resistance most obviously applies..." might be considered to be rephrased (perhaps as "most directly applies")

(Line 270) "then acted as a reservoirs" might be "reservior"

*** Reviewer #3:

Many thanks for the opportunity to review this interesting and well written manuscript by Mo and colleagues. The manuscript aimed to explore the relationship between antimicrobial duration and colonisation with resistant Gram-negative bacteria in a hospital setting. They aimed to compare the findings from reported models to observations reported in the literature by performing a systematic review and meta-analysis.

This study highlights the challenges of trying to determine the impact of any single intervention (in this case shortening treatment duration) on carriage / propagation of AMR. It approaches this in a logical and open way, acknowledging the limitations of the models used and attempting to triangulate observations where possible. It also reinforces many of the challenges we have in design and assessment of impact of stewardship programmes on AMR. A critical point is who, how often, and where we should be following up patients in trials to try and determine the true impact of reducing antimicrobial consumption on AMR (at the patient, ward, even regional level).

In terms of the models used and moving forwards, it would be interesting to understand whether these could now be applied to a population level, where we often consider antimicrobial consumption and the impact on AMR. For example, modelling a hospital using individual ward data predictions to highlight wards that are safe for stewardship interventions / focus on trying to reduce antimicrobial prescribing. Could it also better inform optimal times for sampling in these studies?

For the systematic review, the methodology is a little light.

Was the manuscript registered on prospero or similar prospective registry prior to starting?

Was any assessment for risk of bias performed? Were the PRISMA reporting guidelines followed for systematic review and meta-analysis?

***

Any attachments provided with reviews can be seen via the following link:

[LINK]

PLoS Med. 2023 Jun 15;20(6):e1004013. doi: 10.1371/journal.pmed.1004013.r003

Author response to Decision Letter 1

15 Aug 2022

Attachment

Submitted filename: abm_abxdur_PlosMed_reply.docx

Click here for additional data file.^{(28.6KB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1004013.r004

Decision Letter 2

Philippa Dodd

2 Sep 2022

Dear Dr. Mo,

Thank you very much for re-submitting your manuscript "Implications of reducing antibiotic treatment duration for antimicrobial resistance in hospital settings: A modelling study and meta-analysis" (PMEDICINE-D-22-01395R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by 4 reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Sep 09 2022 11:59PM.

Sincerely,

Philippa Dodd, MBBS MRCP PhD

Senior Editor

PLOS Medicine

pdodd@plos.org

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Please address all additional reviewer comments below

Throughout, please ensure consistent use of terminology (e.g. antibiotic Vs agent)

Throughout, please ensure consistent use of words (five) or numbers (5) when presenting/discussing numerical values

Abstract

Thank you for adding the following sentence to the abstract “The meta-analysis was performed with a logistic regression ….” Perhaps reword for improved clarity, suggest “The meta-analysis was performed using logistic regression. Duration of antibiotic treatment and time from administration of antibiotics to surveillance culture were included as independent variables” or something similar, as we understand it.

Thank you for adding the following sentence, “Interpretation of these estimates is limited by the low number of antibiotic duration trials which monitored resistance carriage as an outcome, contributing to a large credible interval.” Suggest modifying to “…which monitored carriage of resistant gram negative bacteria, as an outcome…” for clarity and consistency.

Author Summary

Thank you for including an author summary. Please ensure consistent terminology is used throughout – suggest either antibiotic treatment or antibiotic duration or antibiotic treatment duration but not different terms, similarly for antibiotic Vs agent. Some further suggestions are below:

Why was this study done?

* Suggest revising the following sentence to “…this depends on the agent’s spectrum of coverage, pharmacokinetic and pharmacodynamic properties…” – as bioavailability forms part of the aforementioned.

* Consider revising the sentence beginning “Empirical evidence…” to “Empirical evidence and an understanding of the mechanisms by which antibiotic treatment duration effects the emergence and spread of antimicrobial resistance are lacking.” or something similar

* “…factors driving the effect [of] antibiotic…”? consider also adding the words “…[help to] inform…”

What did the researchers do and find?

Some of these statements appear repetitive. Please refine this section to include the studies primary findings.

* Please add [antibiotic treatment] or something similar to the following sentence “The models showed that shortening [ ] duration is most effective at reducing resistance carriage in high transmission settings”

What do these findings mean?

Please revise this section and include two or three main implications of the study findings that you list above.

Some of the points currently in this section might better placed above - some more specific comments are below:

* “Shortening antibiotic duration…” consider revising to “…Shortening the duration of antibiotic treatment may increase or decrease colonization by resistant bacteria depending upon the effectiveness of antibiotics” or something similar.

* Please clarify the following statement “For treated individuals, shortening duration is most effective when resistant bacteria grow rapidly under antibiotic selection pressure and decline rapidly when stopping treatment.” Most effective at reducing resistance?

* The following statement - “The modest effect of shortening treatment duration on resistant bacteria colonisation can be partially explained by typically short hospitalisation stays and a slow rate of decolonisation of resistant bacteria in patients previously exposed to antibiotics” is long and difficult to read/understand. Please either revise to improve accessibility to the reader or remove

* Consider the following sentence, or something similar in place of the current “We found that an important interplay exists between shortening antibiotic treatment and reducing the transmission of resistant bacteria through infection prevention and control” as we understand it

* Please remove the statement on study limitations

Methods & Findings

Thank you for including assessment of the individual risk of bias for each study included in the meta-analysis.

The assessment of publication bias doesn’t appear to be included. While it is good practice to include a minimum of 10 studies when assessing risk of bias using tests for funnel plot asymmetry (if this underpins the omission?), alternative strategies, such as Tang’s regression test (PMID: 29663281) could be an alternative approach in respect of your meta-analysis which includes 5 studies. You may find further guidance here PMID: 10812319 helpful also. An additional comment on the limitations of assessment of publication bias in your study (perhaps in the discussion section of the main manuscript) as appropriate, would also be beneficial.

Figures

Thank you for your attention to our requests regarding the figures.

It appears that throughout, the figures are now rather out of sink with the figure captions. For example line 162, Figure 1…the figure is below the figure caption. The same is seen with all subsequent figures in the main manuscript. Throughout, please place the figures above the figure caption and ensure that captions are clearly distinct from the main text.

Thank you for altering the colour schemes in figure 1. It might be advisable to opt for colours that are very distant from red and green - the pink shading used in figure 1 appears rather red, perhaps varying shades of light and dark blue, for example, or something similar. In addition it is difficult to read the white letters overlying the yellow background so suggest amending for improved readability.

Figures 2 and 3 (heatmaps) are unchanged in colour – perhaps because re-analysis may be required? If they can be changed then please do so otherwise please indicate in your response if and where colour schemes cannot be altered for any reason so that I cease to pester you about it!

Throughout the manuscript, if/when figure colors are altered, please also remember to amend the descriptive text e.g. line 387 in reference to figure 4 “Red, blue and grey coloured rectangles…”

References

Ref #9 please spell out the names of the first 6 authors names

Please ensure all web references have an access date e.g. ref #2

*** If you have any specific questions or require clarification or further assistance please do not hesitate to contact me on the personal email address detailed in this letter ***

Comments from Reviewers:

Reviewer #1: The authors have addressed all my major comments. I congratulate them for this important work.

Reviewer #2 (methodological reviewer): We thank the authors for largely addressing our previous comments. A few minor suggestions might be considered:

1. While 10 repeats for various parameter sets were attempted, the concern about the curse of dimensionality appears to still possibly apply to unexplored combinations of parameters. This might be briefly mentioned as a potential limitation.

2. For the loess, the relevant hyperparameters used for fitting might be included, if relevant.

3. While tradeoffs for optimizing antibiotic treatment duration are now briefly discussed, the condition of "...optimising definitive antibiotic courses for established bacterial infections while ensuring clinical cure, and rapid discontinuation of empiric prescriptions after bacterial infections" appears fairly vague. Have any quantitative guidelines for the tradeoffs been considered?

Reviewer #3: Authors have addressed my previous comments.

No additional comments to add.

Many thanks and congratulations.

Reviewer #4: The authors have addressed the comments I have provided earlier and improved the overall quality of the manuscript significantly. The methodology section has been clarified, especially around the assumptions made to conceptualise the agent based model. The steps taken to simulate the model have now been clearly described to allow audience who were unfamiliar with the method to understand the approach. I believe the revised manuscript fits for publication and have no further comments.

Any attachments provided with reviews can be seen via the following link:

[LINK]

PLoS Med. 2023 Jun 15;20(6):e1004013. doi: 10.1371/journal.pmed.1004013.r005

Author response to Decision Letter 2

5 Sep 2022

Attachment

Submitted filename: abm_abxdur_PlosMed_reply2.docx

Click here for additional data file.^{(22.1KB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1004013.r006

Decision Letter 3

Philippa Dodd

3 Oct 2022

Dear Dr. Mo,

I have discussed the paper with my colleagues and the academic editor and it was also seen again by xxx reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

[LINK]

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Oct 10 2022 11:59PM.

Sincerely,

Philippa Dodd, MBBS MRCP PhD

PLOS Medicine

pdodd@plos.org

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

Please accept my apologies for the delay in responding to you, we are unexpectedly very short staffed currently. Thank you for addressing the previous editorial and reviewer comments. There are some further minor revisions as detailed below that are remain outstanding. If you have any questions related to these please contact me directly via email (pdodd@plos.org) I am happy to discuss via email to expedite the process for you as much as I can.

ABSTRACT

Line 48: “Importantly, shortening antibiotic treatment may increase resistance carriage when antibiotics administered can effectively suppress the colonising bacteria with a particular resistance phenotype….”, is rather confusing to me. I have suggested a revision, below, as per my understanding – please correct this if I am mistaken! “Importantly, under circumstances whereby administered antibiotics can suppress colonizing bacteria, shortening antibiotic treatment may increase the carriage of a particular resistance phenotype.”

Please also amend the abstract in the manuscript submission form when you resubmit

Line 50-53: beginning we “…We identified…” suggest revise to the following as per my understanding:

“We identified 187 randomised trials which investigated antibiotic duration. Of these, 5 reported resistant Gram-negative bacteria carriage as an outcome and were included in the meta-analysis. The meta-analysis determined that a single additional antibiotic treatment day is associated with a 7% absolute increase in risk of resistance carriage (80% credible interval 3 to 11%).”

Notably, the information in the abstract and main manuscript differ:

Line 492: the section titled “evidence from randomized controlled trials” will need revising.

Line 493: “Initial search of the MEDLINE and EMBASE databases returned 2407 unique publications. Out of these 187 were randomised trials which compared antibiotic treatment durations. 7 these trials collected surveillance cultures for colonising bacteria during follow-up visits and were included in the qualitative synthesis.” However in the abstract report the inclusion of 5 studies….

Only later do we find out that a further 2 of these studies were excluded which is a bit confusing. For clarity and transparency the information from line 512 should be reported earlier, at least before we are directed to the table of study characteristics which reports 5 not 7 studies.

Table 1 also only details 5 studies rather than 7…please also update the PRISMA flow charts for study exclusion as necessary

In the abstract, you report the search was performed up to May 2021, can you please update it to within the last 6 months. I suspect there will not be any additional trials to include but please update the search dates in any case and the relevant PRISMA flowcharts etc.

Please include the updated search dates in the main manuscript text where you describe the metanalysis and systematic review (line 277). Please ensure any updates are also made to supplementary files where relevant

AUTHOR SUMMARY

This should be concise, distinct from the abstract and accessible to the lay reader. I have revised, as per my understanding, as below:

Why was this study done?

• Shortening antibiotic treatment duration is a commonly adopted antibiotic stewardship strategy, with the expectation that it will reduce antimicrobial resistance in treated individuals and in the overall population.

• Antibiotic selective pressure acts predominantly on ‘bystander’ colonising bacteria for resistance, and this depends on the spectrum of coverage, pharmacokinetic and pharmacodynamic properties of individual antibiotics.

• Empirical evidence and an understanding of the mechanisms by which antibiotic treatment duration effects the emergence and spread of antimicrobial resistance are lacking. Understanding the key factors driving the effect of antibiotic treatment duration on resistance carriage will help to inform future research study designs, antimicrobial stewardship interventions and resource allocation in the multimodal control strategies.

What did the researchers do and find?

• We modelled within- and between-host dynamics of colonising ‘bystander’ susceptible and resistant bacteria in response to systemic antibiotic treatment and compared the model findings with a systematic review and meta-analysis.

• The meta-analysis found one additional antibiotic treatment day is associated with a 7% absolute increase in risk of resistance carriage when antibiotics administered were not effective against the resistance phenotype in the colonising bacteria.

• For treated individuals, the models showed that shortening antibiotic treatment duration is most effective at reducing resistance carriage when resistant bacteria grow rapidly under antibiotic selection pressure and decline rapidly when stopping treatment.

• At a population level, shortening antibiotic treatment duration is most effective at reducing resistance carriage in high transmission settings.

• Shortening antibiotic treatment duration may increase resistance carriage when the antibiotics administered are effective at eliminating colonising bacteria with a particular resistance phenotype.

What do these findings mean?

• Shortening antibiotic treatment duration may increase or decrease colonisation by resistant bacteria, dependent upon individual and combined bacterial and antibiotic characteristics, and may help to inform antimicrobial stewardship policy making.

Please amend if I am mistaken in my interpretation at any point in editing this summary

GENERIC INFORMATION FOR MODELLING STUDIES

Please check the list below and ensure details are adequately detailed in the manuscript

Please provide a diagram that shows the model structure, including how the disease natural history is represented, the process and determinants of disease acquisition, and how the putative intervention could affect the system.

Please provide a complete list of model parameters, including clear and precise descriptions of [the meaning of each parameter, together with the values or ranges for each, with justification or the primary source cited, and important caveats about the use of these values noted].

Please provide a clear statement about how the model was fitted to the data [including goodness-of-fit measure, the numerical algorithm used, which parameter varied, constraints imposed on parameter values, and starting conditions].

For uncertainty analyses, please state the sources of uncertainties quantified and not quantified [can include parameter, data, and model structure].

Please provide sensitivity analyses to identify which parameter values are most important in the model. Uncertainty estimates seek to derive a range of credible results on the basis of an exploration of the range of reasonable parameter values. The choice of method should be presented and justified.

Please discuss the scientific rationale for this choice of model structure and identify points where this choice could influence conclusions drawn. Please also describe the strength of the scientific basis underlying the key model assumptions.

REFERENCES

Line 569: “…community were used.[39]…” this error is repeated throughout the manuscript, please check and amend throughout as per below:

Please select the PLOS Medicine reference style in your citation manager. Please use square brackets for in-text reference call outs noting the absence of spaces within the square brackets, the space before the first bracket and the punctuation following the second bracket, “…symptomatic [2,8].”

PLOS uses the reference style outlined by the International Committee of Medical Journal Editors (ICMJE), also referred to as the “Vancouver” style. Example formats are listed below. Additional examples are in the ICMJE sample references. Please list up to 6 author names only before et al where more than 8 authors have contributed

Example: Hou WR, Hou YL, Wu GF, Song Y, Su XL, Sun B, et al. cDNA, genomic sequence cloning and overexpression of ribosomal protein gene L9 (rpL9) of the giant panda (Ailuropoda melanoleuca). Genet Mol Res. 2011;10: 1576-1588.

Comments from the Guest Editors:

it's a nice piece of work and I hope you will go ahead and accept it if not already.

Comments from Reviewers:

Reviewer #2: We thank the authors for addressing the previous comments as best as possible, and have no further issues to raise.

Any attachments provided with reviews can be seen via the following link:

[LINK]

PLoS Med. 2023 Jun 15;20(6):e1004013. doi: 10.1371/journal.pmed.1004013.r007

Author response to Decision Letter 3

5 Oct 2022

Attachment

Submitted filename: abm_abxdur_PlosMed_reply3.docx

Click here for additional data file.^{(21.4KB, docx)}

PLoS Med. doi: 10.1371/journal.pmed.1004013.r008

Decision Letter 4

Philippa Dodd

17 Oct 2022

Dear Dr Mo,

On behalf of my colleagues and the Academic Editor, Professor Ramanan Laxminarayan, I am pleased to inform you that we have agreed to publish your manuscript "Implications of reducing antibiotic treatment duration for antimicrobial resistance in hospital settings: A modelling study and meta-analysis" (PMEDICINE-D-22-01395R4) in PLOS Medicine's Antimicrobial Resistance and Surveillance Special Issue.

There is a single, final revision detailed below for you to make prior to publication. It has been a pleasure handling your manuscript throughout the peer review and editorial process, many congratulations on its publication.

Comments from the Editor:

Line 279: "2000 up to 20 May 2021..." you updated your search and amended the search dates in the abstract and supplementary material but not in the main manuscript, please amend accordingly.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Best wishes,

Pippa

Philippa Dodd, MBBS MRCP PhD

Editor

PLOS Medicine

pdodd@plos.org

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Text. Supporting information: Model details and exploration.

Table A. Parameters used in the models and their respective ranges obtained from literature review.

(DOCX)

Click here for additional data file.^{(3.4MB, docx)}

S2 Text. Supporting information 2: Systematic review methodology.

(DOCX)

Click here for additional data file.^{(6.1MB, docx)}

S1 PRISMA Checklist. PRISMA checklist.

(DOCX)

Click here for additional data file.^{(33.1KB, docx)}

Attachment

Submitted filename: abm_abxdur_PlosMed_reply.docx

Click here for additional data file.^{(28.6KB, docx)}

Attachment

Submitted filename: abm_abxdur_PlosMed_reply2.docx

Click here for additional data file.^{(22.1KB, docx)}

Attachment

Submitted filename: abm_abxdur_PlosMed_reply3.docx

Click here for additional data file.^{(21.4KB, docx)}

Data Availability Statement

The computer code for all simulations and data for the meta-analysis are publicly available (https://github.com/moyinNUHS/abxduration_abm).

[pmed.1004013.ref001] 1.Barlam TF, Cosgrove SE, Abbo LM, Macdougall C, Schuetz AN, Septimus EJ, et al. Implementing an antibiotic stewardship program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clinical Infectious Diseases. Oxford University Press; 2016. p. e51–77. doi: 10.1093/cid/ciw118 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref002] 2.World Health Organization(WHO). Antimicrobial Stewardship Programmes in Healthcare Facilities in Low-and-middle-income Countries, a WHO Practical Toolkit Antimicrobial Stewardship. 2019 [cited 2022 Aug 8]. https://apps.who.int/iris/bitstream/handle/10665/329404/9789241515481-eng.pdf.

[pmed.1004013.ref003] 3.Chung GW, Wu JE, Yeo CL, Chan D, Hsu LY. Antimicrobial stewardship: A review of prospective audit and feedback systems and an objective evaluation of outcomes. Virulence. 2013;4:151–57. doi: 10.4161/viru.21626 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref004] 4.Sandberg T, Skoog G, Hermansson AB, Kahlmeter G, Kuylenstierna N, Lannergård A, et al. Ciprofloxacin for 7 days versus 14 days in women with acute pyelonephritis: a randomised, open-label and double-blind, placebo-controlled, non-inferiority trial. Lancet. 2012;380:484–90. doi: 10.1016/S0140-6736(12)60608-4 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref005] 5.Chastre J, Wolff M, Fagon J-Y, Chevret S, Thomas F, Wermert D, et al. Comparison of 8 vs 15 Days of Antibiotic Therapy for Ventilator-Associated Pneumonia in Adults. JAMA. 2003;290:2588. doi: 10.1001/jama.290.19.2588 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref006] 6.Uranga A, España PP, Bilbao A, Quintana JM, Arriaga I, Intxausti M, et al. Duration of Antibiotic Treatment in Community-Acquired Pneumonia. JAMA Intern Med. 2016;176:1257. doi: 10.1001/jamainternmed.2016.3633 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref007] 7.Spellberg B, Rice LB. Duration of Antibiotic Therapy: Shorter Is Better. Ann Intern Med. 2019;171:210. doi: 10.7326/M19-1509 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref008] 8.van Duijn PJ, Verbrugghe W, Jorens PG, Spöhr F, Schedler D, Deja M, et al. The effects of antibiotic cycling and mixing on antibiotic resistance in intensive care units: a cluster-randomised crossover trial. Lancet Infect Dis. 2018;18:401–9. doi: 10.1016/S1473-3099(18)30056-2 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref009] 9.Schuts E, Hulscher M, Mouton J, Verduin C, Stuart J, Overdiek H, et al. Current evidence on hospital antimicrobial stewardship objectives: a systematic review and meta-analysis. Lancet Infect Dis. 2016;16:847–56. doi: 10.1016/S1473-3099(16)00065-7 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref010] 10.Leo S, Lazarevic V, von Dach E, Kaiser L, Prendki V, Schrenzel J, et al. Effects of antibiotic duration on the intestinal microbiota and resistome: The PIRATE RESISTANCE project, a cohort study nested within a randomized trial. EBioMedicine. 2021:71. doi: 10.1016/J.EBIOM.2021.103566 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref011] 11.Totsika M, Beatson SA, Sarkar S, Phan M-D, Petty NK, Bachmann N, et al. Insights into a multidrug resistant Escherichia coli pathogen of the globally disseminated ST131 lineage: genome analysis and virulence mechanisms. PLoS ONE. 2011;6:e26578. doi: 10.1371/journal.pone.0026578 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref012] 12.Gupta RK, Gregson J, Parkin N, Haile-Selassie H, Tanuri A, Andrade Forero L, et al. HIV-1 drug resistance before initiation or re-initiation of first-line antiretroviral therapy in low-income and middle-income countries: a systematic review and meta-regression analysis. Lancet Infect Dis. 2018;18:346–55. doi: 10.1016/S1473-3099(17)30702-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref013] 13.Sullivan Å. Effect of antimicrobial agents on the ecological balance of human microflora. Lancet Infect Dis. 2001;1:101–14. doi: 10.1016/S1473-3099(01)00066-4 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref014] 14.Troché G, Toly L-M, Guibert M, Zazzo J-F. Detection and Treatment of Antibiotic-Resistant Bacterial Carriage in a Surgical Intensive Care Unit: A 6-Year Prospective Survey. Infect Control Hosp Epidemiol. 2005;26:161–65. doi: 10.1086/502521 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref015] 15.Dubinsky-Pertzov B, Temkin E, Harbarth S, Fankhauser-Rodriguez C, Carevic B, Radovanovic I, et al. Carriage of Extended-spectrum Beta-lactamase-producing Enterobacteriaceae and the Risk of Surgical Site Infection after Colorectal Surgery: A prospective cohort study. Clin Infect Dis. 2019;68:1699–1704. doi: 10.1093/cid/ciy768 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref016] 16.Reddy P, Malczynski M, Obias A, Reiner S, Jin N, Huang J, et al. Screening for extended-spectrum β-lactamase-producing enterobacteriaceae among high-risk patients and rates of subsequent bacteremia. Clin Infect Dis. 2007;45:846–52. doi: 10.1086/521260 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref017] 17.Tacconelli E, De Angelis G, Cataldo MA, Mantengoli E, Spanu T, Pan A, et al. Antibiotic usage and risk of colonization and infection with antibiotic-resistant bacteria: A hospital population-based study. Antimicrob Agents Chemother. 2009;53:4264–69. doi: 10.1128/AAC.00431-09 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref018] 18.Mo Y, Hernandez-Koutoucheva A, Musicha P, Bertrand D, Lye D, Ng O, et al. Duration of carbapenemase-producing enterobacteriaceae carriage in hospital patients. Emerg Infect Dis. 2020;26:2182–85. doi: 10.3201/eid2609.190592 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref019] 19.Kim S, Covington A, Pamer EG. The intestinal microbiota: Antibiotics, colonization resistance, and enteric pathogens. Immunol Rev. 2017:90–105. doi: 10.1111/imr.12563 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref020] 20.Austin DJ, Kakehashi M, Anderson RM. The transmission dynamics of antibiotic-resistant bacteria: The relationship between resistance in commensal organisms and antibiotic consumption. Proc R Soc B Biol Sci. 1997;264:1629–38. doi: 10.1098/rspb.1997.0227 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref021] 21.Davies NG, Flasche S, Jit M, Atkins KE. Within-host dynamics shape antibiotic resistance in commensal bacteria. Nat Ecol Evol. 2019;3:440–49. doi: 10.1038/s41559-018-0786-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref022] 22.Boureau H, Hartmann L, Karjalainen T, Rowland I, Wilkinson MHF. Models to Study Colonisation and Colonisation Resistance. Microb Ecol Health Dis. 2000;2:247–58. doi: 10.1080/08910600050216246 [DOI] [Google Scholar]

[pmed.1004013.ref023] 23.Niehus R, van Kleef E, Mo Y, Turlej-Rogacka A, Lammens C, Carmeli Y, et al. Quantifying antibiotic impact on within-patient dynamics of extended-spectrum beta-lactamase resistance. Elife. 2020;9:1–20. doi: 10.7554/eLife.49206 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref024] 24.Iredell J, Brown J, Tagg K. Antibiotic resistance in Enterobacteriaceae: Mechanisms and clinical implications. BMJ (Online). BMJ Publishing Group; 2016. doi: 10.1136/bmj.h6420 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref025] 25.R Core Team. R: a language and environment for statistical computing. Vienna, Austria; 2017.

[pmed.1004013.ref026] 26.Jackson C. Multi-state modelling with R: the msm package 1 Multi-state models 1.1 Introduction. 2019. http://www.jstatsoft.org/v38/i08/.

[pmed.1004013.ref027] 27.Chalom A, Inácio de Knegt López de Prado P. Package “pse” Type Package Title Parameter Space Exploration with Latin Hypercubes. 2017. https://github.com/andrechalom/pse

[pmed.1004013.ref028] 28.Alden K, Read M, Timmis J, Andrews PS, Veiga-Fernandes H, Coles M. Spartan: A Comprehensive Tool for Understanding Uncertainty in Simulations of Biological Systems. PLoS Comput Biol. 2013;9:1002916. doi: 10.1371/journal.pcbi.1002916 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref029] 29.Higgins JP, Jelena S, Page MJ, Elbers RG, Sterne JA. Cochrane Handbook for Systematic Reviews of Interventions version 6.2 Chapter 8: Assessing risk of bias in a randomized trial. 2021 [cited 2021 May 6]. https://training.cochrane.org/handbook/current/chapter-08

[pmed.1004013.ref030] 30.Watanabe S. Asymptotic Equivalence of Bayes Cross Validation and Widely Applicable Information Criterion in Singular Learning Theory. J Mach Learn Res. 2010;11:3571–94. [Google Scholar]

[pmed.1004013.ref031] 31.Hornik K, Leisch F, Zeileis A, Plummer M. JAGS: A Program for Analysis of Bayesian Graphical Models Using Gibbs Sampling. http://www.ci.tuwien.ac.at/Conferences/DSC-2003/

[pmed.1004013.ref032] 32.Orriols R, Hernando R, Ferrer A, Terradas S, Montoro B. Eradication therapy against pseudomonas aeruginosa in non-cystic fibrosis bronchiectasis. Respiration. 2015;90:299–305. doi: 10.1159/000438490 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref033] 33.Bouglé A, Tuffet S, Federici L, Leone M, Monsel A, Dessalle T, et al. Comparison of 8 versus 15 days of antibiotic therapy for Pseudomonas aeruginosa ventilator-associated pneumonia in adults: a randomized, controlled, open-label trial. Intensive Care Med. 2022;48:841–49. doi: 10.1007/S00134-022-06690-5 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref034] 34.Barratt S, Bielicki JA, Dunn D, Faust SN, Finn A, Harper L, et al. Amoxicillin duration and dose for community-acquired pneumonia in children: the CAP-IT factorial non-inferiority RCT. Health Technol Assess (Rockv). 2021;25:1–72. doi: 10.3310/hta25600 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref035] 35.Schrag S, Peña C, Fernández J, Sánchez J, Gómez V, Pérez E, et al. Effect of short-course, high-dose amoxicillin therapy on resistant pneumococcal carriage: a randomized trial. JAMA. 2001;286:49–56. doi: 10.1001/jama.286.1.49 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref036] 36.McCallum GB, Fong SM, Grimwood K, Nathan AM, Byrnes CA, Ooi MH, et al. Extended Versus Standard Antibiotic Course Duration in Children <5 Years of Age Hospitalized With Community-acquired Pneumonia in High-risk Settings: Four-week Outcomes of a Multicenter, Double-blind, Parallel, Superiority Randomized Controlled Trial. Pediatr Infect Dis J. 2022;41:549–555. doi: 10.1097/INF.0000000000003558 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref037] 37.Lutsar I, Chazallon C, Trafojer U, De Cabre VM, Auriti C, Bertaina C, et al. Meropenem vs standard of care for treatment of neonatal late onset sepsis (NeoMero1): A randomised controlled trial. PLoS ONE. 2020;15:e0229380. doi: 10.1371/journal.pone.0229380 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref038] 38.Hoberman A, Paradise JL, Rockette HE, Kearney DH, Bhatnagar S, Shope TR, et al. Shortened Antimicrobial Treatment for Acute Otitis Media in Young Children. N Engl J Med. 2016;375:2446–56. doi: 10.1056/NEJMoa1606043 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref039] 39.Ceran N, Mert D, Kocdogan FY, Erdem I, Adalati R, Ozyurek S, et al. A randomized comparative study of single-dose fosfomycin and 5-day ciprofloxacin in female patients with uncomplicated lower urinary tract infections. J Infect Chemother. 2010;16:424–30. doi: 10.1007/s10156-010-0079-z [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref040] 40.van Merode T, Nys S, Raets I, Stobberingh EE. Acute uncomplicated lower urinary tract infections in general practice: Clinical and microbiological cute rates after three- versus five-day treatment with trimethoprim. Eur J Gen Pract. 2005;11:55–58. doi: 10.3109/13814780509178238 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref041] 41.Dow G, Rao P, Harding G, Brunka J, Kennedy J, Alfa M, et al. A prospective, randomized trial of 3 or 14 days of ciprofloxacin treatment for acute urinary tract infection in patients with spinal cord injury. Clin Infect Dis. 2004;39:658–64. doi: 10.1086/423000 [DOI] [PubMed] [Google Scholar]

[pmed.1004013.ref042] 42.D’Agata EMC, Magal P, Olivier D, Ruan S, Webb GF. Modeling antibiotic resistance in hospitals: The impact of minimizing treatment duration. J Theor Biol. 2007;249:487–99. doi: 10.1016/j.jtbi.2007.08.011 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref043] 43.D’Agata EMC, Dupont-Rouzeyrol M, Magal P, Olivier D, Ruan S. The Impact of Different Antibiotic Regimens on the Emergence of Antimicrobial-Resistant Bacteria. PLoS ONE. 2008;3:e4036. doi: 10.1371/journal.pone.0004036 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref044] 44.Blanquart F, Lehtinen S, Lipsitch M, Fraser C. The evolution of antibiotic resistance in a structured host population. J R Soc Interface. 2018:15. doi: 10.1098/rsif.2018.0040 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref045] 45.Tedijanto C, Olesen SW, Grad YH, Lipsitch M. Estimating the proportion of bystander selection for antibiotic resistance among potentially pathogenic bacterial flora. Proc Natl Acad Sci. 2018;115:E11988–995. doi: 10.1073/pnas.1810840115 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pmed.1004013.ref046] 46.Palomino JC, Martin A. Drug resistance mechanisms in Mycobacterium tuberculosis. Antibiotics. 2014:317–40. doi: 10.3390/antibiotics3030317 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Implications of reducing antibiotic treatment duration for antimicrobial resistance in hospital settings: A modelling study and meta-analysis

Yin Mo

Mathupanee Oonsivilai

Cherry Lim

Rene Niehus

Ben S Cooper

Roles

Abstract

Background

Methods and findings

Conclusions

Author summary

Why was this study done?

What did the researchers do and find?

What do these findings mean?

Introduction

Table 1. Details of the antibiotic trials included in the meta-analysis.

Methods and materials

Modelling dynamics of colonising bacteria in response to antibiotic treatment

Fig 1. Host states and transitions in the exclusive colonisation, co-colonisation, and within-host growth models.

Fig 2. Example output from the co-colonisation model.

Systematic review

Meta-analysis

Results

Individual and population resistance dynamics with antibiotic treatment

Fig 3. Effect of antibiotic duration on antimicrobial resistance at individual and population levels (results shown are from the within-host growth model, but qualitatively similar results are obtained with the other 2 models).

Key parameters in the relationship between antibiotic duration and resistance carriage

Fig 4. Global sensitivity analysis.

Within-host dynamics of colonising bacteria with varying antibiotic treatment duration

Fig 5. Example output from the within-host growth model showing effect of antibiotic duration on competitive growth of susceptible (blue) and resistant (pink) colonising bacteria at different growth rates.

Resistance transmission and importation on the effects of antibiotic duration

Fig 6. Effects of transmission rate and baseline prevalence of resistance carriers at admission on resistant bacteria carriage in the scenario where administered antibiotics have activity only against susceptible organisms.

Overall effect of antibiotic duration on resistance carriage

Fig 7. The difference in proportion of resistance carriers between the wards administered long and short antibiotic duration (x-axis) under different antibiotic treatments.

Evidence from randomised controlled trials

Fig 8. Daily risks of colonisation by antibiotic-resistant gram-negative bacteria given days of antibiotics prescribed reported by the randomised controlled trials included in the meta-analysis.

Discussion

Supporting information

Acknowledgments

Data Availability

Funding Statement

References

Decision Letter 0

Louise Gaynor-Brook

Roles

Decision Letter 1

Philippa Dodd

Roles

Author response to Decision Letter 1

Decision Letter 2

Philippa Dodd

Roles

Author response to Decision Letter 2

Decision Letter 3

Philippa Dodd

Roles

Author response to Decision Letter 3

Decision Letter 4

Philippa Dodd

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases