Table 1. Uses of antibiotic resistance surveillance system data by hospitals, university researchers, pharmaceutical companies and governments. From Critchley *et al.*, 2004 ^[27]^.
```{r tbl1-1} insert_graphic("images/01-t01.svg") ``` -From Critchley *et al.*, 2004 ^[27]^. As an example, ISIS-AR (Infectious disease Surveillance Information System for Antibiotic Resistance) is a Dutch national surveillance program, for which a large number of the Dutch clinical microbiology laboratories provide anonymised data on AMR to the National Institute for Public Health and the Environment (Rijksinstituut voor Volksgezondheid en Milieu, RIVM) ^[40]^. In Germany, ARS (Antibiotic Resistance Surveillance) is a similar laboratory-based national surveillance program, that attempts to enable differential statements according to structural characteristics of health care and regions ^[41,42]^. Both these national surveillance programs provide data for EARS-Net and GLASS of the WHO ^[37,43]^. diff --git a/02-diagnostic-stewardship.Rmd b/02-diagnostic-stewardship.Rmd index 1fb09ae..7a062c5 100644 --- a/02-diagnostic-stewardship.Rmd +++ b/02-diagnostic-stewardship.Rmd @@ -1,4 +1,4 @@ -# Diagnostic Stewardship: Sense or Nonsense?! {#diagnostic-stewardship} +# Diagnostic Stewardship: Sense or Nonsense?! {#ch02-diagnostic-stewardship} Published in Dutch Journal of Clinical Microbiology, 2019 Sep 27, 26:3 (Nederlands Tijdschrift voor Medische Microbiologie; original work in Dutch) @@ -21,12 +21,10 @@ Diagnostic stewardship or diagnostic stewardship programme (DSP), a trending top Although the term DSP was first mentioned in an indexed PubMed article in 2016, articles on antimicrobial stewardship (ASP) have been appearing for 15 years (Figure \@ref(fig:fig2-1)). -```{r fig2-1, fig.cap = "The increase of articles indexed in PubMed. Search strategies: 'antimicrobial stewardship'[Title/Abstract]; 'diagnostic stewardship'[Title/Abstract]; 'antimicrobial resistance'[Title/Abstract]. Source: https://www.ncbi.nlm.nih.gov/pubmed/ (assessed: 2018-05-31).", out.width = "80%"} +```{r fig2-1, fig.cap = "The increase of articles indexed in PubMed. Search strategies: 'antimicrobial stewardship'[Title/Abstract]; 'diagnostic stewardship'[Title/Abstract]; 'antimicrobial resistance'[Title/Abstract]. Source: https://www.ncbi.nlm.nih.gov/pubmed/ (assessed: 2018-05-31). * Extrapolation based on count from 2018-01-01 to 2018-05-31.", out.width = "80%"} insert_graphic("images/02-01.jpg") ``` -\* Extrapolation based on count from 2018-01-01 to 2018-05-31. - Nevertheless, the concept of DSP is neither intended to replace other stewardship concepts (in particular ASP) nor to be an alternative. DSP concerns decision making and goes beyond microbiological diagnostics alone. Kahneman *et al.* ^[2]^ said about decision making: > We think, each of us, that we're much more rational than we are. And we think that we make our decisions because we have good reasons to make them. Even when it's the other way around. We believe in the reasons, because we've already made the decision. ^[2]^ diff --git a/03-introducing-method.Rmd b/03-introducing-method.Rmd index 1bac650..62df80f 100644 --- a/03-introducing-method.Rmd +++ b/03-introducing-method.Rmd @@ -1,6 +1,6 @@ -# Introducing a New, Free, and Independent Method for Standardised, Reproducible and Reliable Analyses of Antimicrobial Resistance Data {#introducing-method} +# Introducing a New, Free, and Independent Method for Standardised, Reproducible and Reliable Analyses of Antimicrobial Resistance Data {#ch03-introducing-new-method} -In preparation +In preparation
Accepted in Journal of Statistical Software (ahead of print)
Table 1. Example of an antimicrobial resistance report.
+ +```{r tbl4-1} +insert_graphic("images/04-t01.jpg") +``` + +Table 2. Example of an antimicrobial resistance report.
+ +```{r tbl4-2} +insert_graphic("images/04-t02.jpg") +``` + +The `AMR` package aims at providing a standardised and automated way of cleaning, transforming, and enhancing these typical data structures (Table 1 and 2), independent of the underlying data source. Processed data would be similar to Table 3 that highlights several package functionalities in the sections below. + +Table 3. Enhanced antimicrobial resistance report example.
+ +```{r tbl4-3} +insert_graphic("images/04-t03.jpg") +``` + +## Antimicrobial resistance data transformation + +### Working with taxonomically valid microorganism names + +Coercing is a computational process of forcing output based on an input. For microorganism names, coercing user input to taxonomically valid microorganism names is crucial to ensure correct interpretation and to enable grouping based on taxonomic properties. To this end, the `AMR` package includes all microbial entries from The Catalogue of Life (http://www.catalogueoflife.org), the most comprehensive and authoritative global index of species currently available ^[16]^. It holds essential information on the names, relationships, and distributions of more than 1.9 million species. The integration of it into the `AMR` package is described in Appendix A. + +The `as.mo()` function makes use of this underlying data to transform a vector of characters to a new class ``‘mo’` of taxonomically valid microorganism name. The resulting values are microbial IDs, which are human-readable for the trained eye and contain information about the taxonomic kingdom, genus, species, and subspecies (Figure 1). + +```{r fig4-1, fig.cap = "The structure of a typical microbial ID as used in the `AMR` package. An ID consists of two to four elements, separated by an underscore. The first element is the abbreviation of the taxonomic kingdom. The remaining elements consist of abbreviations of the lowest taxonomic levels of every microorganism: genus, species (if available) and subspecies (if available). Abbreviations used for the microbial IDs of microorganism names were created using the base` R` function `abbreviate()`.", out.width = "66%"} +insert_graphic("images/04-01.jpg") +``` + +The `as.mo()` function compares the user input with taxonomically valid microorganism names, rates the matching with a score and returns results based on the highest score. This matching score ($m$), ranging from $0$ to $1$, is calculated using the following equation: + +$$m_{(x,n)} = \frac{l_{n} - 0.5 \cdot \min\{ l_n, \operatorname{lev}(x,n) \} }{l_{n} \cdot p_{n} \cdot k_{n}}$$ + +where: + +* $x$ is the user input; +* $n$ is a taxonomic name (genus, species, and subspecies); +* $l_n$ is the length of $n$; +* lev is the Levenshtein distance function ^[17]^, which counts any insertion, deletion and substitution as $1$ that is needed to change $x$ into $n$; +* $p_n$ is the human pathogenic prevalence group of $n$, as described below; +* $k_n$ is the taxonomic kingdom of $n$, set as Bacteria = 1, Fungi = 2, Protozoa = 3, Archaea = 4, others = 5. + +The grouping into human pathogenic prevalence ($p$) is based on experience from several microbiological laboratories in the Netherlands in conjunction with international reports on pathogen prevalence ^[7,18,19]^. **Group 1** (most prevalent microorganisms) consists of all microorganisms where the taxonomic class is Gammaproteobacteria or where the taxonomic genus is *Enterococcus*, *Staphylococcus* or *Streptococcus*. This group consequently contains all common Gram-negative bacteria, such as *Pseudomonas* and *Legionella* and all species within the order Enterobacterales. **Group 2** consists of all microorganisms where the taxonomic phylum is Proteobacteria, Firmicutes, Actinobacteria or Sarcomastigophora, or where the taxonomic genus is *Absidia*, *Acremonium*, *Actinotignum*, *Alternaria*, *Anaerosalibacter*, *Apophysomyces*, *Arachnia*, *Aspergillus*, *Aureobacterium*, *Aureobasidium*, *Bacteroides*, *Basidiobolus*, *Beauveria*, *Blastocystis*, *Branhamella*, *Calymmatobacterium*, *Candida*, *Capnocytophaga*, *Catabacter*, *Chaetomium*, *Chryseobacterium*, *Chryseomonas*, *Chrysonilia*, *Cladophialophora*, *Cladosporium*, *Conidiobolus*, *Cryptococcus*, *Curvularia*, *Exophiala*, *Exserohilum*, *Flavobacterium*, *Fonsecaea*, *Fusarium*, *Fusobacterium*, *Hendersonula*, *Hypomyces*, *Koserella*, *Lelliottia*, *Leptosphaeria*, *Leptotrichia*, *Malassezia*, *Malbranchea*, *Mortierella*, *Mucor*, *Mycocentrospora*, *Mycoplasma*, *Nectria*, *Ochroconis*, *Oidiodendron*, *Phoma*, *Piedraia*, *Pithomyces*, *Pityrosporum*, *Prevotella*, *Pseudallescheria*, *Rhizomucor*, *Rhizopus*, *Rhodotorula*, *Scolecobasidium*, *Scopulariopsis*, *Scytalidium*, *Sporobolomyces*, *Stachybotrys*, *Stomatococcus*, *Treponema*, *Trichoderma*, *Trichophyton*, *Trichosporon*, *Tritirachium* or *Ureaplasma*. **Group 3** consists of all other microorganisms. + +This will lead to the effect that e.g., `"E. coli"` will return the microbial ID of *Escherichia coli* ($m = 0.688$, a highly prevalent microorganism found in humans) and not *Entamoeba coli* ($m = 0.079$, a less prevalent microorganism in humans), although the latter would alphabetically come first. The matching score function is for users available as `mo_matching_score()`. + +If any coercion rules are applied, a warning is printed to the console and scores can be reviewed by calling `mo_uncertainties()`, that prints all other matches with their matching scores. Users can furthermore control the coercion rules by setting the allow_uncertain argument in the `as.mo()` function. The following values or levels can be used: + +* `0`: no additional rules are applied; +* `1`: allow previously accepted (but now invalid) taxonomic names and minor spelling errors; +* `2`: allow all of 1, strip values between brackets, inverse the words of the input, strip off text elements from the end keeping at least two elements; +* `3`: allow all of level 1 and 2, strip off text elements from the end, allow any part of a taxonomic name; +* `TRUE` (default): equivalent to `2`; +* `FALSE`: equivalent to `0`. + +To support organisation specific microbial IDs, users can specify a custom reference `‘data.frame’`, by using `as.mo(..., reference_df = ...)`. This process can also be automated by users with the `set_mo_source()` function. + +#### Properties of microorganisms + +The package contains functions to return a specific (taxonomic) property of a microorganism from the `microorganisms` data set (see Appendix A). Functions that start with `mo_*` can be used to retrieve the most recently defined taxonomic properties of any microorganism quickly and conveniently. These functions rely on the `as.mo()` function internally: `mo_name()`, `mo_fullname()`, `mo_shortname()`, `mo_subspecies()`, `mo_species()`, `mo_genus()`, `mo_family()`, `mo_order()`, `mo_class()`, `mo_phylum()`, `mo_kingdom()`, `mo_type()`, `mo_gramstain()`, `mo_ref()`, `mo_authors()`, `mo_year()`, `mo_rank()`, `mo_taxonomy()`, `mo_synonyms()`, `mo_info()` and `mo_url()`. Determination of the Gram stain, by using `mo_gramstain()`, is based on the taxonomic subkingdom and phylum. According to Cavalier-Smith ^[20]^, who defined the subkingdoms Negibacteria and Posibacteria, only the following phyla are Posibacteria: Actinobacteria, Chloroflexi, Firmicutes and Tenericutes. Bacteria from these phyla are considered Gram-positive - all other bacteria are considered Gram-negative. Gram stains are only relevant for species within the kingdom of Bacteria. For species outside this kingdom, `mo_gramstain()` will return `NA`. + +### Working with antimicrobial names or codes + +The `AMR` package includes the `antibiotics` data set, which comprises common laboratory information system codes, official names, ATC (Anatomical Therapeutic Chemical) codes, defined daily doses (DDD) and more than 5,000 trade names of 456 antimicrobial agents (see Appendix A). The ATC code system and the reference list for DDDs have been developed and made available by the World Health Organization Collaborating Centre for Drug Statistics Methodology (WHOCC) to standardise pharmaceutical classifications ^[21]^. All agents in the `antibiotics` data set have a unique antimicrobial ID, which is based on abbreviations used by the European Antimicrobial Resistance Surveillance Network (EARS-Net), the largest publicly funded system for antimicrobial resistance surveillance in Europe ^[22]^. Furthermore, the `AMR` package includes the antivirals data seta containing antiviral agents, which is also described in Appendix A. + +#### Properties of antimicrobial agents + +It is a common task in microbiological data analyses (and other clinical or epidemiological fields) to work with different antimicrobial agents. The `AMR` package provides several functions to translate inputs such as ATC codes, abbreviations, or names in any direction. Using `as.ab()`, any input will be transformed to an antimicrobial ID of class `‘ab’`. Helper functions are available to get specific properties of antimicrobial IDs, such as `ab_name()` for getting the official name, `ab_atc()` for the ATC code, or `ab_cid()` for the CID (Compound ID) used by PubChem ^[23]^. Trade names can be also used as input. For example, the input values "Amoxil", "dispermox", "amox" and "J01CA04" all return the ID of amoxicillin (AMX): + +```r +as.ab("Amoxicillin") +#> ClassTable 4. Antimicrobial susceptibility test classes.
+ +```{r tbl4-4} +insert_graphic("images/04-t04.jpg") +``` + +The higher the MIC or the smaller the growth inhibition diameter, the more active substance of an antimicrobial agent is needed to inhibit cell growth, i.e. the higher the antimicrobial resistance against the tested antimicrobial agent. At high MICs and small diameters, guidelines interpret the microorganism as “resistant” (R) to the tested antimicrobial agent. At low MICs and wide diameters, guidelines interpret the microorganism as “susceptible” (S) to the tested antimicrobial agent. In between, the microorganism is classified as “susceptible, increased exposure” (I). For these three interpretations the `‘rsi’` class has been developed. When using `as.rsi()` on MIC values (of class `‘mic’`) or disk diffusion diameters (of class `‘disk’`), the values will be interpreted according to the guidelines from the CLSI or EUCAST (all guidelines between 2011 and 2020 are included in the `AMR` package) ^[24,25]^. Guidelines can be changed by setting the guidelines argument. + +```r +# Low MIC value +as.rsi(as.mic(2), "E. coli", "ampicillin", guideline = "EUCAST 2020") +#> ClassTable 5. Example of a multi-drug resistant organism (MDRO) in a data set identified by applying Dutch guidelines.
+ +```{r tbl4-5} +insert_graphic("images/04-t05.jpg") +``` + +#### Multi-drug resistant tuberculosis + +Tuberculosis is a major threat to global health caused by Mycobacterium tuberculosis (MTB) and is one of the top ten causes of death worldwide ^[37]^. Exceptional antimicrobial resistance in MTB is therefore of special interest. To this end, the international WHO guideline for the classification of drug resistance in MTB ^[34]^ is included in the `AMR` package. The `mdr_tb()` function is a convenient wrapper around `mdro(..., guideline = "TB")`, which returns an other ordered `‘factor’` than other `mdro()` functions. The output will contain the `‘factor’` levels `‘Negative’` < `‘Mono-resistant’` < `‘Poly-resistant’` < `‘Multi-drug-resistant’` < `‘Extensively drug-resistant’`, following the WHO guideline. + +## Analysing antimicrobial resistance data + +### Calculation of antimicrobial resistance + +The `AMR` package contains several functions for fast and simple resistance calculations of bacterial or fungal isolates. A minimum number of available isolates is needed for the reliability of the outcome. The CLSI guideline suggests a minimum of 30 available first isolates irrespective of the type of statistical analysis ^[6]^. Therefore, this number is used as the default setting for any function in the package that calculates antimicrobial resistance or susceptibility, which can be changed with the minimum argument in all applicable functions. + +#### Counts + +The `AMR` package relies on the concept of tidy data ^[38]^, although not strictly following its rules (one row per test rather than one row per observation). Function names to calculate the number of available isolates follow these general resistance interpretation standards with `count_S()`, `count_I()`, and `count_R()` respectively. Combinations of antimicrobial resistance interpretations can be counted with `count_SI()` and `count_IR()`. All these functions take a vector of interpretations of the class `‘rsi’` (as discussed above) or are internally transformed with `as.rsi()`. The returned value is the sum of the respective interpretation in the selected test column. All `count_*()` functions support quasi-quotation with pipes, grouped variables, and can be used with `dplyr::summarise()` ^[39]^. + +#### Proportions + +Calculation of antimicrobial resistance is carried out by counting the number of first resistant isolates (interpretation of “R”) and dividing it by the number of all first isolates, see Equation 2. This is implemented in the `proportion_R()` function. To calculate antimicrobial susceptibility, the number of susceptible first isolates (interpretation of “S” and “I”) has to be counted and divided by the number of all first isolates, which is implemented in the `proportion_SI()` function. For convenience, the `resistance()` function is an alias of the `proportion_R()` function, and the `susceptibility()` function is an alias of the `proportion_SI()` function. + +The functions `proportion_R()`, `proportion_IR()`, `proportion_I()`, `proportion_SI()`, and `proportion_S()` follow the same logic as the `count_*()` functions and all return a vector of class `‘double’` with a value between 0 and 1. The argument `minimum` defines the minimal allowed number of available (tested) isolates (default: `minimum` = 30). Any number below the set minimum will return `NA` with a warning. + +For calculating the proportion ($P$) of antimicrobial resistance or susceptibility to one antimicrobial agent, the following equation is used: + +$$P_{(x, o)} = \frac{\sum_{i=1}^k [x_i \in o]}{\sum_{i=1}^k [x_i \in \{R,S,I\}]}$$ + +where $P$ is the proportion of outcome $o$ (that is either “R”, “S”, “I”, or a combination of two of them), where $x$ is a character vector of length $k$ only consisting of values “R”, “S”, or “I” and $[x_i \in o]$ is the indicator function, returning $1$ if the indicator function is true and $0$ otherwise. The denominator must include the collection $\{R,S,I\}$ so that ’`wrong’` elements in $x$ (i.e., elements not being “R”, “S”, or “I”) will not be counted. Thus, the theoretical antimicrobial susceptibility of the vector $x = \{S,S,I,R,R\}$ is: + +$$P_{(x, o = \{S, I\})} = \frac{3}{5} = 0.6$$ + +For the proportion of empiric susceptibility ($s$) for more than one antimicrobial agent, the calculation can be carried out in two ways (Table 6). + +Table 6. Example calculation for determining empiric susceptibility (%SI) for more than one antimicrobial agent.
+ +```{r tbl4-6} +insert_graphic("images/04-t06.jpg") +``` + +The first method is to count the total number of first isolates where at least one agent was tested as “S” or “I” and divide it by the number of first isolates tested where any of the agents was tested (Equation 4). This method will be used when setting `only_all_tested = FALSE` in the `susceptibility()` function: + +$$s_{(x, y)} = \frac{\sum_{i=1}^k [x_i \in \{S,I\} \lor y_i \in \{S,I\}]}{\sum_{i=1}^k [x_i \in \{R,S,I\} \lor y_i \in \{R,S,I\}]}$$ + +where $x$ is a character vector only consisting of values “R”, “S”, or “I” (i.e., ’`agent A’`) and $y$ is another character vector only consisting of values “R”, “S”, or “I” (i.e., ’`agent B’`). + +The second method is to count the total number of first isolates where at least one agent was tested as “S” or “I” and where all agents were tested divided by the number of first isolates tested where all of the agents were tested (Equation 5). This method will be used when setting `only_all_tested = TRUE` in the `susceptibility()` function: + +$$s'_{(x, y)} = \frac{\sum_{i=1}^k [(x_i \in \{S,I\} \lor y_i \in \{S,I\}) \, \land x_i \in \{R,S,I\} \land y_i \in \{R,S,I\}]}{\sum_{i=1}^k [x_i \in \{R,S,I\} \land y_i \in \{R,S,I\}]}$$ + +Based on Equation 2, the overall resistance and susceptibility of antimicrobial agents like gentamicin (GEN) and amoxicillin (AMX) can be calculated using the following syntax. The `example_isolates` data set is an example data set included in the `AMR` package, see Appendix A. The `n_rsi()` function is analogous to the `n()` function of the `dplyr` package. It counts the number of available isolates, but only includes observations with valid antimicrobial results (i.e., “R”, “S”, or “I”): + +```r +library("dplyr") +example_isolates %>% + summarise(r_gen = proportion_R(GEN), + r_amx = proportion_R(AMX), + n_gen = n_rsi(GEN), + n_amx = n_rsi(AMX), + n_total = n()) +#> r_gen r_amx n_gen n_amx n_total +#> [1] 0.2458221 0.5955556 1855 1350 2000 +``` + +This output reads: the resistance to gentamicin of all isolates in the `example_isolates` data set is $P{(x = GEN, o = \{R\})} = 24.6\%$, based on $1855$ out of $2000$ available isolates. This means that the susceptibility is $P{(x = GEN, o = \{S,I\})} = 75.4\%$. The susceptibility to amoxicillin is $P{(x = AMX, o = \{S,I\})} = 40.4\%$ based on $1350$ isolates. + +To calculate the effect of combination therapy, i.e., treating patients with multiple agents at the same time, all `proportion_*()` functions can handle multiple variables as arguments as defined in Equation 4 and 5. For example, to calculate the empiric susceptibility of a combination therapy comprising gentamicin (GEN) and amoxicillin (AMX): + +```r +example_isolates %>% + summarise(si_gen_amx = proportion_SI(GEN, AMX), + n_gen_amx = n_rsi(GEN, AMX), + n_total = n()) +#> si_gen_amx n_gen_amx n_total +#> [1] 0.931843 1921 2000 +``` + +This leads to the conclusion that combining gentamicin with amoxicillin would cover $s{(x = GEN, y = AMX)} = 93.2\%$ based on $1921$ out of $2000$ available isolates, which is $17.8\%$ more than when treating with gentamicin alone ($P{(x = GEN, o = \{S,I\})} = 75.4\%$). With these functions, exact calculations can be done to evaluate the empiric success of treating infections with one or more antimicrobial agents. + +## Design decisions +The `AMR` package follows the rationale of tidyverse packages as authored by Wickham *et al.* ^[40]^. Most functions take a `‘data.frame’` or `‘tibble’` as input, support piping (%>%) operations, can work with quasi-quotations, and can be integrated into `dplyr` workflows, such as `mutate()` to create new variables and `group_by()` to group by variables. Although the `AMR` package integrates well into tidyverse workflows, it can also be used with base `R `only. To this extent, the `AMR` package was developed to be independent of any other `R `package to ensure and maintain sustainability. + +The `AMR` package supports multiple languages. Currently supported languages are English, Dutch, French, German, Italian, Portuguese, and Spanish. The system language will be used if the language is supported but can be overwritten with `options(AMR_locale = ...)`. Multi-language support affects language-dependent output of functions such as `mo_name()`, `mo_gramstain()`, `mo_type()`, and `ab_name()`. + +The `AMR` package uses S3 classes, object oriented (OO) systems available in R. They allow different types of output based on the user input. The `AMR` package introduces 5 S3 classes (`‘mo’`, `‘ab’`, `‘rsi’`, `‘mic’`, and `‘disk’`) to increase the convenience when working with antimicrobial susceptibility data. + +## Reproducible example + +We consider the problem of working with antimicrobial resistance data from three different hospitals between 2011-01-01 and 2020-01-01. After loading the `AMR` package and additional tidyverse packages to allow transformation and plotting, we load the `example_isolates`_unclean example data from the `AMR` package into the global environment and assign it a new name. + +```r +library("dplyr") +library("tidyr") +library("AMR") + +options(AMR_locale = "en") + +data <- example_isolates_unclean + +glimpse(data) +#> Rows: 3,000 +#> Columns: 8 +#> $ patient_idTable 1. Required R packages for RadaR.
+ +```{r tbl5-1} +# insert_graphic("images/05-t01.jpg") +``` + +The input data should be structured in a dataset format, where each variable is one column and each observation is one row. This follows the concept of “tidy data,” as defined by Hadley Wickham [26]. Table 2 displays the set of variables underlying RadaR’s functionality. In our setting for the development of RadaR, these variables originated from three different data sources: administrative data from the hospital data warehouse, microbiological data from the laboratory information system, and antimicrobial prescription data from the computerised prescriber order entry system. The data preparation and cleaning process are very specific for each data source, dependent on local data standards, and difficult to generalise. Therefore, Table 2 represents the final variables and formats for the analysis and use with RadaR, referring to the “tidy data” concept above and to the R package collection tidyverse for the preparation process [26,27]. Additional variables are calculated and transformed using the packages lubridate and zoo for time points and intervals, and AMR for antimicrobial (group) names, microbial isolate names, first isolate identification, and resistance analysis [28-30]. Microbiological resistance is calculated per antimicrobial substance or as co-resistance if more than one substance is selected. + +Table 2. Input variables for RadaR.
+ +```{r tbl5-2} +# insert_graphic("images/05-t02.jpg") +``` + +The input data should be structured in a dataset format, where each variable is one column and each observation is one row. This follows the concept of “tidy data,” as defined by Hadley Wickham [26]. Table 2 displays the set of variables underlying RadaR’s functionality. In our setting for the development of RadaR, these variables originated from three different data sources: administrative data from the hospital data warehouse, microbiological data from the laboratory information system, and antimicrobial prescription data from the computerised prescriber order entry system. The data preparation and cleaning process are very specific for each data source, dependent on local data standards, and difficult to generalise. Therefore, Table 2 represents the final variables and formats for the analysis and use with RadaR, referring to the “tidy data” concept above and to the R package collection tidyverse for the preparation process [26,27]. Additional variables are calculated and transformed using the packages lubridate and zoo for time points and intervals, and AMR for antimicrobial (group) names, microbial isolate names, first isolate identification, and resistance analysis [28-30]. Microbiological resistance is calculated per antimicrobial substance or as co-resistance if more than one substance is selected. + +RadaR can be used for graphical exploratory data analysis. Differences in LOS are displayed by a Kaplan-Meier curve in conjunction with a log-rank test, using the survminer package [32]. Time trends for number of admissions, antimicrobial consumption, and resistance counts per year, quarter, or month, are visualised in run charts using the qicharts2 package [33]. Nonrandom variation in these run charts is tested using Anhøj’s rules [34]. + +RadaR has been developed in macOS High Sierra (1.4 GHz, 4 GB RAM), and it was successfully tested in Windows 7 (3.2 GHz, 8 GB RAM) and Linux (Ubuntu 16.04.4 LTS, 3.4 GHz, 12 GB RAM). A running example version has been deployed to shinyapps.io, a publicly available web hosting service for R Shiny apps [35]. The entire source code of RadaR is freely accessible on GitHub [36]. We intend to integrate suggestions and feedback coming from its users and the R community. RadaR was developed using data of patients admitted to the University Medical Center Groningen, Groningen, the Netherlands. Data were collected retrospectively, and permission was granted by the ethical committee (METc 2014/530). RadaR can be used locally in protected environments or hosted on the web, provided appropriate measures have been taken to guarantee data protection, depending on national regulations. + +## Results + +### Overview + +We have developed RadaR, a web-based software app providing an intuitive platform for rapid analysis of large datasets containing information about patients’ admission, antimicrobial use, and results of microbiological diagnostic tests. This software app can help users (i.e., AMS team members) find answers to questions, such as “What are the most commonly used antimicrobials at an institution/specialty/department and have they changed over time?”, “Were adequate microbiological diagnostics performed at the start of antimicrobial treatments?”, “What are the most frequent microorganisms found and their resistance patterns in different departments?”, and “Can we identify priority areas within a hospital where antimicrobial or microbiological diagnostic use has the largest room for improvement?”. + +### Application Design + +RadaR is designed in the form of a web browser–based dashboard that most users are familiar with from typical websites and web-based tools (see Figure 1). The basis of RadaR’s functionality is filtering datasets and producing analytical graphs according to selection criteria defined by the user. Any calculations and data aggregation are based on single observations of individual patients. To identify and analyse groups of patients, 17 different selection criteria can be found in the sidebar (Table 3). The output of RadaR is grouped into four panels (patient, antimicrobials, diagnostics, and outcome) that each comprise three to four output boxes displaying the results. + +Table 3. Selection criteria in sidebar.
+ +```{r tbl5-3} +# insert_graphic("images/05-t03.jpg") +``` + +All output is based on the selection criteria defined by the user in the sidebar. Each new selection and any change need to be confirmed by clicking the confirm selection button (see Figure 1). Users can navigate among the different analysis panels by clicking the respective button. + +```{r fig5-1, fig.cap = "Application design."} +# insert_graphic("images/05-01.jpg") +``` + +Results are shown in bar charts, density plots, run charts, a bubble plot, and a Kaplan-Meier curve for LOS in hospital. Each panel further displays a table summarizing the respective data analyses. All output boxes and their content are described in Table 4. Most output boxes include modification options that can be identified by small gear icons (see Figure 1). These clickable icons allow for further specification of the generated plots and tables. Users can compare different groups (e.g., antimicrobial use by antimicrobial agent, resistance patterns per isolate, or LOS by specialty) or modify the plots (e.g., switch from count to proportion, change the chart type, or show or hide the legend). Plots and tables can be downloaded through download buttons as PNG files for plots and CSV, Excel, or PDF files for tables. + +Table 4. Output boxes for analysis results.
+ +```{r tbl5-4} +# insert_graphic("images/05-t04.jpg") +``` + +Finally, two datasets (antimicrobial/admission data and microbiological data) of the user-defined selection can be downloaded from the sidebar menu in a CSV-file format for further analysis (e.g., retrieving a list of patient numbers of the selected patient group). + +### Development Process + +RadaR has been developed in close contact with the AMS team and senior consulting specialists at the University Medical Center Groningen, Groningen, the Netherlands, to meet the needs and requirements of this user group. Subsequently, all members of the European Society of Clinical Microbiology and Infectious Diseases Study Group for Antimicrobial Stewardship (ESGAP) were asked to evaluate and test the software app through a running web-based example of RadaR and by filling out a web-based survey. The ESGAP comprises around 200 members from more than 30 countries worldwide. A total of 12 members from 9 different countries took part in the evaluation. This yielded important information on user experiences with the software app, which in turn led to further improvements that are reflected in the version we presented in this report. In a next phase, RadaR will be tested in different settings of ESGAP members and other interested partners using locally available data (e.g., an 837-bed tertiary care hospital in the Netherlands and a 750-bed tertiary care hospital in Greece). + +### Workflow + +RadaR was developed and tested with a dataset of all patients admitted to our institution, a 1339-bed academic tertiary referral hospital, within the years of 2009 to 2016, comprising over 180,000 admissions. For simulation purposes and web-based user testing, we have created a test dataset of 60,000 simulated patients. This sample dataset allows testing of RadaR’s functionality, but it does not produce meaningful results. + +A typical example workflow with RadaR comprises 6 steps (with examples from the test dataset). They are listed below: + +1. Define the selection: For example, patients receiving intravenous second- or third-generation cephalosporins as first treatment for at least two days, starting within the first two days of hospital admission from any specialty in all years in the dataset. +2. Patients’ panel: Identify the total number of patients and the subspecialties with the highest number of included patients (e.g., 537 patients selected in total, with 97 patients from internal medicine). Investigate patients’ gender and age distribution. +3. Antimicrobials panel: Identify the total use of the initial cefuroxime treatment in DDD and DOT per 100 bed days (e.g., 4.51 and 1.5, respectively). Stratify the results by subspecialty and identify the highest number of DDD and DOT per 100 bed days (e.g., highest use by DDD and DOT in internal medicine). +4. Diagnostics panel: Check if the selected microbiological diagnostic test (e.g., blood culture test) has been performed on the same day as the start of the treatment (defined in the sidebar). Investigate the proportion of tests performed over the years and investigate which subspecialty performs best compared with others (e.g., paediatrics). Check which microorganisms (as first isolates) were found in the selected diagnostic specimens (the most common isolate: Escherichia coli). Investigate the proportion of isolates resistant to cefuroxime (8.9%) and analyse the trend over time. +5. Outcome panel: Check for patterns of differences in LOS in the defined patient group by subspecialties or performed diagnostics (e.g., highest mean LOS of 7.8 days in Surgery). +6. Refine the selection: Investigate a subgroup of the original selection. For example, select only the top three subspecialties by number of patients and repeat step 2 to 5. + +### Customisation + +For setting up RadaR in a new environment after data preparation, users only need to perform the following four steps: + +1. Downloading R and RStudio [21,22], which are free to use and open-source software +2. Download or copy and paste RadaR’s source code [36] into three files in RStudio – global.R, server.R, and ui.R +3. In global.R, manually edit the paths for the prepared datasets to be imported into RadaR +4. Run the app in RStudio with the calling the function runApp() in the console or by clicking the green run app button. This will download and install the required R packages needed for the app if they have not been installed previously, and this will create the final dataset for analysis. The RadaR interface will open in the RStudio viewer pane or in a new window of the standard browser of the user’s operating system. + +RadaR’s appearance has been customised using a cascading style sheets (CSS) script [37] that is loaded into the app upon its start. This script needs to be saved into a subdirectory of the directory of the three main files (global.R, server.R, and ui.R) called www. We recommend RStudio’s project function to create a single project for RadaR and to store all information in this project directory. Users with experience in using CSS can fully alter RadaR’s design by changing the underlying CSS script. + +## Discussion + +### Principal Findings + +We have developed a web-based software app for rapid analysis of diagnostic and antimicrobial patterns that can support AMS teams to tailor their interventions. It has been designed to enhance communication of relevant findings while being easy to use. This also applies to users without extensive prior software skills, as it follows usual web browser user experiences. Moreover, it has been developed using open-source software. It is therefore free to use and accessible for download. In our experience, this system can be adapted to new settings within one day, when the required data (Table 2) are available. + +Commercial software for infection management is available (e.g., Epic Antimicrobial Stewardship Module, TREAT Steward). These offer extensive options for filtering, analysing, and visualizing EHRs with real-time connections to hospital data infrastructures and have been shown to be useful in clinical practice [38]. However, it is difficult to compare functionalities of these tools because of their non–open-source nature. This fact, along with the required budget to purchase the software, drastically limits their use. We are convinced that transparent software development can support the adoption of data-driven developments while enhancing optimal quality of care and patient safety, which is crucial in the light of new data-driven developments of using EHRs [39,40]. + +The global nature of infections further calls to develop software tools applicable in resource-limited settings [41]. Open-source approaches for data analysis, such as RadaR, have advantages over traditional methods, such as Excel or SPSS. Hughes et al described those in their report of a software app for RNA-sequencing data analysis [15]. They highlight aspects that were also fundamental for the development of RadaR. First, R allows transparent, reproducible, and sustainable data analysis through scripts that can easily be shared and changed. This can build the basis for collaboration, and this enforces the spirit of open science (also through the strong collaborative R community on the web). Second, R is open source and free to use; therefore, it also enables use in resource-limited settings. Finally, Shiny empowers users to interact with the data, making even very large datasets quickly interpretable. + +Innovative approaches used in supporting infection management by leveraging EHRs are being investigated [17-19]. Reporting on AMR, antimicrobial use, and hospital infections (e.g., for quality assurance) is well established, but it is important to integrate these data sources in an approach that allows detailed filtering options on all input. Merely looking at antimicrobial use alone or comparing aggregated results (e.g., total amount of a specific antimicrobial substance per hospital correlated with the total count of a resistant isolate) will result in loss of information or even misleading interpretation. Detailed data and calculations on the basis of each individual patient are crucial to draw informed conclusions. Unfortunately, the abovementioned infection management approaches [17-19] either depend on additional commercial software for data visualisation or the source code is not openly available. We want to encourage others to turn toward available open-source software solutions, such as R, for an increased potential of collaboration and transparency. However, their strength is the connection to real-time data flows. This enables the prospective use and increases their usability for daily clinical practice. RadaR is currently still limited to retrospective data analysis because of a changing hospital data infrastructure in our setting. Technically, it is feasible to connect R-based software apps such as RadaR to real-time hospital data infrastructures running with clinical data standards [42]. For a start, access to static data extraction is often easier and faster to achieve. RadaR can be used to advocate the use of data visualisation tools and improved accessibility of hospital data sources. Until connection to real-time hospital data is established, RadaR can support users as a stand-alone option for retrospective data analysis in infection management. Next steps will involve testing in multiple settings and forming a user and research group to continue and expand the use of open-source technology and open science principles in infection management. + +### Conclusions + +RadaR demonstrates the feasibility of developing software tools for infection management and AMS teams in an open-source approach, making it free to use, share, or modify according to various needs in different settings. RadaR has the potential to be a highly useful tool for infection management and AMS in daily practice. + +## Acknowledgements {-} + +The authors would like to thank the ESGAP executive committee for supporting the evaluation of RadaR in the ESGAP study group and all its members for their valuable input, suggestions, and comments. Furthermore, the authors wish to thank Igor van der Weide, Jan Arends, and Prashant Nannan Panday for their great support in obtaining required data at our institution that built the basis for the development of RadaR. The authors also thank the online R community as well as the valuable comments, suggestions, and input from reviewers that they have received to improve RadaR. + +RadaR was developed as part of a project funded by the European Commission Horizon 2020 Framework Marie Skłodowska-Curie Actions (grant agreement number: 713660-PRONKJEWAIL-H2020-MSCA-COFUND-2015). + +## Conflicts of interests {-} + +None declared. + +## References {-} + +1. Dyar OJ, Huttner B, Schouten J, Pulcini C, ESGAP (ESCMID Study Group for Antimicrobial stewardshiP). What is antimicrobial stewardship? Clin Microbiol Infect 2017 Nov;23(11):793–798. PMID:28882725 +2. Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, Srinivasan A, Dellit TH, Falck-Ytter YT, Fishman NO, Hamilton CW, Jenkins TC, Lipsett PA, Malani PN, May LS, Moran GJ, Neuhauser MM, Newland JG, Ohl CA, Samore MH, Seo SK, Trivedi KK. Implementing an antibiotic stewardship program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis 2016 May 15;62(10):e51–77. PMID:27080992 +3. Davey P, Marwick CA, Scott CL, Charani E, McNeil K, Brown E, Gould IM, Ramsay CR, Michie S. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev 2017 Feb 9;2:CD003543. PMID:28178770 +4. National Institute for Health and Care Excellence. Antimicrobial stewardship: systems and processes for effective antimicrobial medicine use. https://www.nice.org.uk/Guidance/NG15. Archived at: http://www.webcitation.org/73ybLqgCI +5. Schuts EC, Hulscher MEJL, Mouton JW, Verduin CM, Stuart JWTC, Overdiek HWPM, van der Linden PD, Natsch S, Hertogh CMPM, Wolfs TFW, Schouten JA, Kullberg BJ, Prins JM. Current evidence on hospital antimicrobial stewardship objectives: a systematic review and meta-analysis. Lancet Infect Dis 2016 Jul;16(7):847–856. PMID:26947617 +6. Stichting Werkgroep AntibioticaBeleid. SWAB Guidelines for Antimicrobial Stewardship 2016. http://www.swab.nl/swab/cms3.nsf/uploads/FAF701FBF4FCBDD6C12580EC0037D292/$FILE/20161228%20SWAB%20Guidelines%20for%20Antimicrobial%20Stewardship%20-definitief.pdf . Archived at: http://www.webcitation.org/73ycxKUfO +7. Pulcini C, Binda F, Lamkang AS, Trett A, Charani E, Goff DA, Harbarth S, Hinrichsen SL, Levy-Hara G, Mendelson M, Nathwani D, Gunturu R, Singh S, Srinivasan A, Thamlikitkul V, Thursky K, Vlieghe E, Wertheim H, Zeng M, Gandra S, Laxminarayan R. Developing core elements and checklist items for global hospital antimicrobial stewardship programmes: a consensus approach. Clin Microbiol Infect. PMID:29625170 +8. Sarkies MN, Bowles K-A, Skinner EH, Mitchell D, Haas R, Ho M, Salter K, May K, Markham D, O’Brien L, Plumb S, Haines TP. Data collection methods in health services research: hospital length of stay and discharge destination. Appl Clin Inform 2015 Feb 18;6(1):96–109. PMID:25848416 +9. British Society for Antimicrobial Chemotherapy. Antimicrobial stewardship: From Principal to Practice. Birmingham, United Kingdom: British Society for Antimicrobial Chemotherapy; 2018. http://bsac.org.uk/antimicrobial-stewardship-from-principles-to-practice-e-book/. Archived at: http://www.webcitation.org/73ydEbLNu +10. Pulcini C, Morel CM, Tacconelli E, Beovic B, de With K, Goossens H, Harbarth S, Holmes A, Howard P, Morris AM, Nathwani D, Sharland M, Schouten J, Thursky K, Laxminarayan R, Mendelson M. Human resources estimates and funding for antibiotic stewardship teams are urgently needed. Clin Microbiol Infect 2017 Nov;23(11):785–787. PMID:28778544 +11. Howard P, Pulcini C, Levy Hara G, West RM, Gould IM, Harbarth S, Nathwani D, ESCMID Study Group for Antimicrobial Policies (ESGAP), ISC Group on Antimicrobial Stewardship. An international cross-sectional survey of antimicrobial stewardship programmes in hospitals. J Antimicrob Chemother 2015 Apr;70(4):1245–1255. PMID:25527272 +12. Dik J-WH, Poelman R, Friedrich AW, Panday PN, Lo-Ten-Foe JR, van Assen S, van Gemert-Pijnen JEWC, Niesters HGM, Hendrix R, Sinha B. An integrated stewardship model: antimicrobial, infection prevention and diagnostic (AID). Future Microbiol 2016;11(1):93–102. PMID:26323589 +13. Staib A, Sullivan C, Jones M, Griffin B, Bell A, Scott I. The ED-inpatient dashboard: Uniting emergency and inpatient clinicians to improve the efficiency and quality of care for patients requiring emergency admission to hospital. Emerg Med Australas 2017 Jun;29(3):363–366. PMID:27592365 +14. Smith CM, Hayward AC. DotMapper: an open source tool for creating interactive disease point maps. BMC Infect Dis 2016 Apr 12;16:145. PMID:27066780 +15. Hughes LD, Lewis SA, Hughes ME. ExpressionDB: An open source platform for distributing genome-scale datasets. PLoS One 2017 Nov 2;12(11):e0187457. PMID:29095940 +16. GitHub. https://github.com . Archived at: http://www.webcitation.org/73yb4rqQT +17. Simpao AF, Ahumada LM, Larru Martinez B, Cardenas AM, Metjian TA, Sullivan KV, Gálvez JA, Desai BR, Rehman MA, Gerber JS. Design and Implementation of a Visual Analytics Electronic Antibiogram within an Electronic Health Record System at a Tertiary Pediatric Hospital. Appl Clin Inform 2018 Jan;9(1):37–45. PMID:29342478 +18. Lesho EP, Waterman PE, Chukwuma U, McAuliffe K, Neumann C, Julius MD, Crouch H, Chandrasekera R, English JF, Clifford RJ, Kester KE. The antimicrobial resistance monitoring and research (ARMoR) program: the US Department of Defense response to escalating antimicrobial resistance. Clin Infect Dis 2014 Aug 1;59(3):390–397. PMID:24795331 +19. Simões AS, Maia MR, Gregório J, Couto I, Asfeldt AM, Simonsen GS, Póvoa P, Viveiros M, Lapão LV. Participatory-Implementation of an antibiotic stewardship programme supported by an innovative surveillance and clinical decision-support system. J Hosp Infect 2018 Jul 30; PMID:30071264 +20. Molloy JC. The open knowledge foundation: open data means better science. PLoS Biol 2011 Dec;9(12):e1001195. PMID:22162946 +21. RStudio. https://www.rstudio.com/ . Archived at: http://www.webcitation.org/73ybbedpd +22. R Core Team. R: A Language and Environment for Statistical Computing. https://www.R-project.org/. Archived at: http://www.webcitation.org/73ydeuUVX +23. Chang W, Cheng J, Allaire JJ, Xie Y, McPherson J. shiny: web Application Framework for R. https://CRAN.R-project.org/package=shiny. Archived at: http://www.webcitation.org/ 73ydh83pW +24. shinyapps.io. https://www.shinyapps.io/. Archived at: http://www.webcitation.org/ 73ybe2prd +25. Stallman R. GNU General Public License, version 2. 1991. https://www.gnu.org/licenses/gpl-2.0.html. Archived at: http://www.webcitation.org/73ydum67R +26. Wickham H. Tidy Data. Journal of Statistical Software. Foundation for Open Access Statistics; 2014;59(10):1–23. [doi: 10.18637/jss.v059.i10] +27. Tidyverse. https://www.tidyverse.org/. Archived at: http://www.webcitation.org/ 76WntlTmp +28. Grolemund G, Wickham H. Dates and Times Made Easy with lubridate. Journal of Statistical Software 2011;40(3):1–25. [doi: 10.18637/jss.v040.i03] +29. Zeileis A, Grothendieck G. zoo: S3 Infrastructure for Regular and Irregular Time Series. Journal of Statistical Software. 2005. p. 1–27. [doi: 10.18637/jss.v014.i06] +30. Berends MS, Luz CF. AMR: An R package to simplify the analysis and prediction of Antimicrobial Resistance (AMR) and work with antibiotic properties by using evidence-based methods. [doi: 10.5281/zenodo.1749847] +31. WHO Collaborating Centre for Drug Statistics Methodology. Guidelines for ATC classification and DDD assignment. Oslo, Norway; 2018. https://www.whocc.no/filearchive/publications/ guidelines.pdf. Archived at: http://www.webcitation.org/73yeFdXyM +32. Kassambara A, Kosinski M. survminer: Drawing Survival Curves using “ggplot2”. https://CRAN.R-project.org/package=survminer. Archived at: http://www.webcitation.org/ 73yeLlWLx +33. Anhoej J. qicharts2: Quality Improvement Charts. https://CRAN.R-project.org/ package=qicharts2 . Archived at: http://www.webcitation.org/76nVgPWG5 +34. Anhøj J. Diagnostic value of run chart analysis: using likelihood ratios to compare run chart rules on simulated data series. PLoS One 2015 Mar 23;10(3):e0121349. PMID:25799549 +35. RadaR shinyapps.io. https://ceefluz.shinyapps.io/radar/. Archived at: http://www. webcitation.org/73ybyEBHu +36. RadaR GitHub. https://github.com/ceefluz/radar. Archived at: http://www.webcitation.org/ 73yc2PgeW +37. RadaR CSS. https://github.com/ceefluz/radar/blob/master/www/radar_style.css. Archived at: http://www.webcitation.org/73ycOiZA5 +38. Pettit NN, Han Z, Choksi A, Bhagat PH, Pisano J. Using the Epic® Antimicrobial Stewardship (ASP) Module to Optimize Antimicrobial Stewardship Interventions. Open Forum Infectious Diseases 2017 Oct 1;4(suppl_1):S263–S264. [doi: 10.1093/ofid/ofx163.580] +39. Xiao C, Choi E, Sun J. Opportunities and challenges in developing deep learning models using electronic health records data: a systematic review. J Am Med Inform Assoc 2018 Jun 8; PMID:29893864 +40. Pirracchio R, Cohen MJ, Malenica I, Cohen J, Chambaz A, Cannesson M, Lee C, Resche-Rigon M, Hubbard A, ACTERREA Research Group. Big data and targeted machine learning in action to assist medical decision in the ICU. Anaesth Crit Care Pain Med 2018 Oct 16; PMID:30339893 +41. Hahn E, Blazes D, Lewis S. Understanding How the “Open” of Open Source Software (OSS) Will Improve Global Health Security. Health Secur 016 Jan;14(1):13–18. PMID:26889576 +42. Hong N, Prodduturi N, Wang C, Jiang G. Shiny FHIR: An Integrated Framework Leveraging Shiny R and HL7 FHIR to Empower Standards-Based Clinical Data Applications. Stud Health Technol Inform 2017;245:868–872. PMID:29295223 diff --git a/docs/404.html b/docs/404.html deleted file mode 100644 index 1ecd007..0000000 --- a/docs/404.html +++ /dev/null @@ -1,237 +0,0 @@ - - - - - - -The page you requested cannot be found (perhaps it was moved or renamed).
-You may want to try searching to find the page's new location, or use -the table of contents to find the page you are looking for.
-As the burden of antimicrobial resistance (AMR) is continuously increasing, reliable and reproducible data and data analysis are of utmost importance. Conducting AMR data analysis is challenging since it requires (1) a thorough understanding of (clinical) epidemiology; (2) expertise in (clinical) microbiology and infectious diseases; (3) experience in microbiological data analysis; (4) availability of reference data, such as the biological taxonomy of microorganisms and defined daily doses (DDD) for antimicrobials; and (5) availability of (inter-)national guidelines and software methods to apply them. Furthermore, data stored in laboratory information systems lack the right structure, (inter-) national guidelines for interpreting raw laboratory test results cannot be easily applied, and scientifically reliable reference data about microorganisms and antimicrobial agents are not readily available. To fill this gap, we developed a free, independent, and open-source software solution to cover all those aspects of working with AMR data. The AMR package for R enables AMR data analysis for research and clinical workflows alike. Through an online survey package users reported more reproducibility of analysis results (83%), more reliable outcomes of AMR analyses (72%), and new or improved insight into AMR patterns (61%). The AMR package was also used to support clinical decision-making (44%) and for clinical research (28%). Our first insights into the usage and the usability of the AMR package confirm that this package is fulfilling its intended aim, as regional, national, and international organisations already use the package to support clinical decision-making in infection management. The flexible open-source design also enables rapid integration of updated guidelines (e.g., new EUCAST breakpoints) and setting-specific adaptations are encouraged. Together, the AMR package for R can thus empower any specialist in the field working with AMR data by providing a comprehensive toolbox of solutions for AMR data analyses.
-As the burden of antimicrobial resistance (AMR) is continuously increasing, surveillance programs with reliable and reproducible data and data analysis methods are of utmost importance for controlling and streamlining efforts to curb AMR [1,2]. To guide these efforts and to support clinical decision-making and infection-control interventions, AMR data analysis has to be conducted in a clinically and epidemiologically sensible way [3]. Conducting AMR data analysis is challenging since it requires (1) a thorough understanding of (clinical) epidemiology; (2) expertise in (clinical) microbiology and infectious diseases; (3) experience in microbiological data analysis; (4) availability of reference data, such as the biological taxonomy of microorganisms and defined daily doses (DDD) for antimicrobials; and (5) availability of (inter-)national guidelines and software methods to apply them.
-Moreover, AMR data analysis is often also hindered by three key aspects. Firstly, data stored in microbiological laboratory information systems (LIS) are typically not readily suitable for (epidemiological) data analyses. LIS were initially designed to fit result registration and billing purposes rather than AMR data analysis. Consequently, fundamental requirements for (epidemiological) data analyses are often lacking, such as isolate selection criteria, phenotypic determination of (multi-)drug resistance, and the ability to extract data for analysis in an automated, structured, fast, and reliable way. Moreover, data analyses that require data from multiple LIS sources (e.g., in multi-centre studies) face major barriers in data aggregation which, to the best of our knowledge, cannot be solved by currently available commercial software solutions. Besides, as applications of artificial intelligence are expected of being increasingly developed in the coming years, also in clinical microbiology, microbiological data technologies and structures need to become compatible for these future applications.
-Secondly, AMR data analysis depends on (inter-)national standards and guidelines for the interpretation of raw laboratory measurements and the reporting of AMR results. In Europe, guidelines from the European Committee on Antimicrobial Susceptibility Testing (EUCAST) are the predominantly implemented set of rules in clinical microbiological laboratories [4,5]. LIS need to be well-maintained to be able to integrate continuous guideline updates. In our experience, this maintenance can often not be guaranteed and depends on the availability of local or external software support services. This is further hindered by the current distribution of manually formatted guidelines in Microsoft Excel and Portable Document Format (PDF) formats that are not often readily machine-readable. LIS maintainers, in collaboration with clinical staff, are therefore forced to manually implement updated guidelines which can be time-consuming and error-prone
-Thirdly, reliable AMR data analysis depends on taxonomic reference data to interpret raw LIS data using AMR interpretation guidelines, such as EUCAST Expert Rules and EUCAST Clinical Breakpoints [5,6]. Unfortunately, typical LIS contain local, static taxonomic data. We found that these data are often poorly maintained. We collected the taxonomic names of bacteria used in clinical reports from seven different public health institutions in the Netherlands which cover microbiological diagnostics in hospitals and primary care for 15% of the total Dutch population. The taxonomic names were compared to publicly available and authoritative reference databases; the Catalogue of Life and the List of Prokaryotic names with Standing in Nomenclature (LPSN, previously known as the Deutsche Sammlung von Mikroorganismen und Zellkulturen, DSMZ) [7,8]. We found that all participating institutions reported taxonomic names in clinical reports that did not match current taxonomic standards according to reference databases. For example, Enterobacter aerogenes and Enterobacter massiliensis were renamed Klebsiella aerogenes and Metakosakonia massiliensis respectively in 2017 [9,10]. LIS that are not kept up to date are consequently not entirely compatible with recent interpretation guidelines. Given that AMR guidelines are strongly based on the microbial taxonomy (some rules only apply to a specific genus, other rules apply to a specific family) it is crucial that this information is correct and kept up to date. In the studied institutions, the lag between the reported taxonomic names and the taxonomic standard was up to 41 years as of March 2021.
-Previously, no dedicated software solution was available to address all aforementioned aspects. To fill this gap, we developed a free, independent, and open-source software solution to cover all those aspects of working with AMR data. The AMR package for R [11] provides functionalities that enable standardised and reproducible workflows from any raw LIS data to results ready to publish, for research and clinical workflows alike. The AMR package for R was developed with a team of contributors from 12 public health organisations in seven countries aiming to be used in any research or clinical setting where (epidemiological) data analysis of microorganisms, AMR, or antimicrobial agents is required. It is independent of any other software solution and was designed to work in any setting, including those with limited computational and financial resources.
-With this AMR package, we aimed at providing: (1) tools to simplify AMR data cleaning, transformation, and analysis; (2) methods to easily incorporate (inter)national guidelines; and (3) scientifically reliable reference data, including the aforementioned aspects. The AMR package enables standardised and reproducible AMR data analysis with the application of evidence-based rules (e.g., EUCAST expert rules for intrinsic resistance), the selection of first isolates, the translation of various codes for microorganisms and antimicrobial agents, determination of (multi-)drug-resistant microorganisms, and the calculation of antimicrobial resistance rates, prevalence, and future trends. The AMR package supports all EUCAST MIC/disk diffusion interpretation guidelines from 2011 until 2021 and EUCAST Expert rules versions 3.1 (2016) and 3.2 (2020) [12,13] In addition, the AMR package supports all CLSI MIC/disk diffusion interpretation guidelines from 2011 until 2019 (non-veterinary only). For all mentioned guidelines, files readable for LIS are provided for easy implementation.
-As of 30 April 2021, the AMR package for R has been downloaded from 162 countries since its first release in early 2018 (Figure 1), according to data from a popular public repository where users can download R packages. After 19 releases, the median number of downloads per release is 2,548 (range: 269-5,050).
-
--Figure 3.1: Countries (grey, n = 162) with registered downloads of the AMR package for R between March 2018 and April 2021. Sources: cran.rstudio.org and cloud.r-project.org. -
-A technical validation of the AMR package has been accepted for publication [11]. Additionally, it has been clinically and epidemiologically validated in a tertiary care hospital and across seven clinical microbiology laboratories in the Netherlands [Berends et al., unpublished, see chapter 6 and 7 of this thesis]. Moreover, the AMR package has already been used in several scientific publications that focused on different aspects in the field of AMR [14–17].
-Popular statistical software such as SPSS, Stata and SAS, focus on a broad implementation of statistical functions but are proprietary software, disallowing users to freely use, modify, or share the software. This also prohibits extending the software by unaffiliated developers. Since R is free, open software and extendible, users and developers can contribute to the software, to which end the AMR package is a practical example.
-Other free software alternatives for AMR data analysis exist, for example WHONET, a free microbiology laboratory database software supported by the WHO [18]. WHONET allows manual data entry from LIS reports and provides AMR interpretation using recent CLSI and EUCAST guidelines with a particular focus on AMR surveillance. Results from WHONET can also be shared to surveillance programs such as the European Antimicrobial Resistance Surveillance Network (EARS-Net) and the WHO Global Antimicrobial Resistance Surveillance System (GLASS). Yet, the latest release, WHONET 2020, does not provide tools for cleaning and transforming data and relies on outdated EUCAST guidelines. Furthermore, we found a lag between the included taxonomic database and the current taxonomic standard of up to 59 years (median 7 years). Another alternative of a free software program is Epi Info which is provided by the United States Centers for Disease Control and Prevention (CDC) and aims at public health practitioners and researchers [19]. While Epi Info provides statistical and epidemiological methods for analysing data, it does not offer tools nor reference data for working with AMR test results or antimicrobial drugs, thus, ruling out the option for dedicated AMR data analysis. With the AMR package for R, an open and dedicated software solution is available that covers all aspects of working with AMR data.
-In July 2020, we published a survey on the website created for this package (https://msberends.github.io/AMR) to seek voluntary feedback from package users about user backgrounds and usage of the AMR package. Until December 2020, 18 participants completed the survey. Participants have used the AMR package in Australia, Colombia, Egypt, France, Germany, Haiti, India, Mali, Mexico, the Netherlands, Nigeria, Philippines, Spain, Sweden, and the United Kingdom.
-Participants were asked to rate their experience in the statistical programming language R and in using the AMR package on a scale from 1 (not experienced/useful) to 10 (very experienced/useful). The overall experience in R was reported with a median of 7 (range: 4-9)., whereas Ssuit ability for AMR analyses using the AMR package was rated with a median of 9 (range: 6-9). The participants rated the usefulness of the AMR package for their work with a median of 9 (range: 5-9). The convenience of the included software functions was rated with a median of 8 (range: 6-9) and the documentation of the AMR package was rated with a median of 8.5 (range: 7-10). Of all participants, 83% reported more reproducibility of analysis results and, 72% reported more reliable outcomes of AMR analyses (Figure 2). Notably, 61% reported new or improved insight into AMR for their institution or region. The AMR package was also used to support clinical decision-making (44%) and for clinical research (28%). Furthermore, 66% reported a faster and streamlined analysis workflow and 39% reported improved communicating analysis results. In 33%, participants started using R more often because of the capabilities that the AMR package provides.
-
--Figure 3.2: The outcome of the survey amongst 18 participants. MIC: minimal inhibitory concentration, MDRO: multidrug-resistant organism, SNOMED: Systematised Nomenclature of Medicine. -
-Aside from AMR data analysis, most participants (78%) used the AMR package as a reference for the taxonomy of microorganisms. It was also regularly used for interpreting raw MIC and disk diffusion values (56%) and applying EUCAST expert rules (67%). This is in line with the original aims of the AMR package development.
-AMR data analysis is dependent on (inter-)national guidelines and reliable (reference) data on the one hand but constrained by diverse and often inadequate data analysis tools and poor data quality on the other. We aimed to address these dependencies and constraints by introducing the AMR package for R for standardised and reproducible AMR data analyses. Our first insights into the usage and the usability of the AMR package confirm that this package is fulfilling its intended aim. Regional, national, and international organisations already use the AMR package to support clinical decision-making in infection management by gaining new or improved insights into resistance levels. We invite others to make use of our open-source approach and adapt it to their needs. The advantages of sharing open-source software such as the AMR package allow for a collaborative, transparent use and further development that can lead to more standardised analysis processes for AMR data. The flexible open-source design also enables rapid integration of updated guidelines (e.g., new EUCAST breakpoints), and setting-specific adaptations are encouraged. Together, the AMR package for R can thus empower any specialist in the field working with AMR data by providing a comprehensive toolbox of solutions for AMR data analysis.
-AMR - An R Package for Working with Antimicrobial Resistance Data
+-There is no doubt that the mortality was much diminished, as I said before, by the flight of the population, which commenced soon after the outbreak; but the attacks had so far diminished before the use of the water was stopped, that it is impossible to decide whether the well still contained the cholera poison in an active state, or whether, from some cause, the water had become free from it.
For this reason, Snow hypothesised that the local outbreak was caused by poisoned water coming from a water pump. To investigate the number of cases, he drew one of the most well-known data visualisations in epidemiology, Figure 1.1 (top). In this then-novel form of data visualisation, he counted the cases per household and denoted them as stacked rectangles. This resulted in his conclusion that there had been no particular outbreak or prevalence of cholera in that part of London except among the persons who were in the habit of drinking the water of one specific water pump: the one on Broad Street. The handle of the pump was removed on the day following his briefing to the local government, leading to an end of the outbreak.
-With the advancements in information technology, heatmaps would nowadays be a more effective way to visualise geographic trends, Figure 1.1 (bottom). Using modern map data as illustrated, the incredible accuracy of Snow’s drawing of London from 167 years ago is also highlighted. The type of investigating geographic trends in health and disease is nowadays known as spatial epidemiology.
+For this reason, Snow hypothesised that the local outbreak was caused by poisoned water coming from a water pump. To investigate the number of cases, he drew one of the most well-known data visualisations in epidemiology, Figure 1.1 (top). In this then-novel form of data visualisation, he counted the cases per household and denoted them as stacked rectangles. This resulted in his conclusion that there had been no particular outbreak or prevalence of cholera in that part of London except among the persons who were in the habit of drinking the water of one specific water pump: the one on Broad Street. The handle of the pump was removed on the day following his briefing to the local government, leading to an end of the outbreak.
+With the advancements in information technology, heatmaps would nowadays be a more effective way to visualise geographic trends, Figure 1.1 (bottom). Using modern map data as illustrated, the incredible accuracy of Snow’s drawing of London from 167 years ago is also highlighted. The type of investigating geographic trends in health and disease is nowadays known as spatial epidemiology.
@@ -162,7 +282,7 @@
Figure 1.1: Visualisations of the ‘Broad Street cholera outbreak’ in London in 1854. Top: original map as drawn by John Snow. Bottom: Snow’s original map with a self-made heatmap visualisation overlay, based on the geographic position of the cases. The blue circles (n = 13) indicate the location of the water pumps.
Spatial epidemiology is one example of the many different specialities in the field of epidemiology. Another example is the direct consequence of Snow’s work: infectious disease epidemiology, which has developed widely since the nineteenth century and has become the de facto standard for researching diseases and their health effects caused by pathogens (i.e., bacteria, viruses and fungi). Since this speciality concerns pathogens, it is a domain shared by the fields of epidemiology and clinical microbiology (Figure 1.2). Moreover, infectious disease epidemiology can be split into two subspecialties: clinical (infectious disease) epidemiology and microbial epidemiology. The former focuses on the properties of the disease (such as the burden of disease caused by infection, or the disease-related mental and financial costs), while the latter focuses on the properties of the pathogen (such as the credibility of its source, antimicrobial resistance and pathogenicity).
+Spatial epidemiology is one example of the many different specialities in the field of epidemiology. Another example is the direct consequence of Snow’s work: infectious disease epidemiology, which has developed widely since the nineteenth century and has become the de facto standard for researching diseases and their health effects caused by pathogens (i.e., bacteria, viruses and fungi). Since this speciality concerns pathogens, it is a domain shared by the fields of epidemiology and clinical microbiology (Figure 1.2). Moreover, infectious disease epidemiology can be split into two subspecialties: clinical (infectious disease) epidemiology and microbial epidemiology. The former focuses on the properties of the disease (such as the burden of disease caused by infection, or the disease-related mental and financial costs), while the latter focuses on the properties of the pathogen (such as the credibility of its source, antimicrobial resistance and pathogenicity).
Applying microbial epidemiology was barely possible in the days of John Snow, for the lack of scientific knowledge about pathogens and the lack of advancement in information technology. Antibiotics were not discovered yet, the cause of cholera was undetermined, and scientists had no clue about the infectivity and pathogenicity of different bacteria. However, what John Snow did in 1854 ‘clinical epidemiologically,’ is in essence quite equal to what we currently do on a large scale during the COVID-19 pandemic. Information technology required to attain this large scale has brought us not only the possibilities to look beyond regional, national and international borders but to observe, analyse and understand pandemics in real-time. Methods we develop and use today can be implemented on the other side of the world tomorrow. This is an important advantage in modern infectious disease epidemiology, as is also illustrated in this thesis.
Microbial epidemiology has an important focus on observing and analysing (1) the microorganisms that cause infections and the human site of origin, (2) the intrinsic or acquired antimicrobial resistance they manifest, and (3) their infectivity and pathogenicity. As any type of microorganism – bacteria, viruses and fungi (including yeasts) – can cause infections in humans, microbial epidemiology is not limited to a certain type of microorganism. Nonetheless, there tends to be a stronger focus on bacteria and fungi, which are more easily isolated at a clinical microbiology laboratory than viruses and can be tested for phenotypical antimicrobial resistance in a routine diagnostic setting. Based on these diagnostic findings, treatment guidelines are developed and evaluated. This in itself urges microbial epidemiology to be employed in a routine setting as well, to make sure that treatment guideline development continually has a solid epidemiological basis.
For this interpretation of raw AMR test results, international guidelines exist. The most often applied guidelines are supplied by the Clinical and Laboratory Standards Institute (CLSI) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST) [18,19]. In Europe, an increasing number of clinical laboratories apply EUCAST guidelines, as it was shown that the coverage of EUCAST guidelines among these laboratories was 73.2% in 2013, and only a few European countries did not use the EUCAST methodology in 2019 [20,21]. According to the World Health Organisation (WHO), guidelines from CLSI and EUCAST are adopted by 94% of all countries reporting AMR to the Global Antimicrobial Resistance Surveillance System (GLASS) of the WHO [22].
-Generally, AMR is defined as the proportion of resistant microorganisms (R) among all tested microorganisms of the same species (R + S + I). The CLSI and EUCAST guidelines define the interpretations for the most common combinations of pathogenic microorganisms and antimicrobial agents. For example, the EUCAST 2021 guideline considers ciprofloxacin against Escherichia coli to be susceptible when either the MIC is at most 0.25 mg/L or when a diffusion disk with 5 µg has a growth inhibition zone of at least 25 millimetres (Figure 1.3).
+Generally, AMR is defined as the proportion of resistant microorganisms (R) among all tested microorganisms of the same species (R + S + I). The CLSI and EUCAST guidelines define the interpretations for the most common combinations of pathogenic microorganisms and antimicrobial agents. For example, the EUCAST 2021 guideline considers ciprofloxacin against Escherichia coli to be susceptible when either the MIC is at most 0.25 mg/L or when a diffusion disk with 5 µg has a growth inhibition zone of at least 25 millimetres (Figure 1.3).
In 2017, EUCAST implemented the area of technical uncertainty (ATU) for certain microbial species/antibiotic combinations, to warn laboratory staff that the interpretation of routine susceptibility testing is uncertain [23]. For example, disk diffusion results from the combination of any species in the order of Enterobacterales with amoxicillin/clavulanic acid are considered unreliable for a zone diameter of 19-20 mm in the latest EUCAST interpretation guideline [24]. EUCAST advises to rerun the test, perform an additional test, or to report this uncertainty with a clear warning [23].
@@ -215,9 +335,10 @@ Figure 1.3: Interpretation of 100 random minimum inhibitory concentrations (top)
With the current WHO surveillance program GLASS, the overall coverage of AMR is continuously being monitored for most countries of the world [37]. For Europe, the prevalence of AMR on the country level is monitored by national surveillance programs that share their data with the European Centre for Disease Prevention and Control (ECDC), an agency of the European Union [38]. Their surveillance program European Antimicrobial Resistance Surveillance Network (EARS-Net) is the largest publicly funded system for AMR surveillance in Europe. Public access to descriptive data (maps, graphs and tables) are available through the ECDC Surveillance Atlas of Infectious Diseases [38], which was also consulted for multiple studies in this thesis. While the ECDC estimated in 2009 that bacterial infections caused by MDROs were responsible for 25,000 extra deaths per year [39], others found that there is a large discrepancy between the real count of deaths attributable to MDROs and the subsequent alarmist predictions, based on data from over 500 studies [35].
Although surveillance programs allow for signalling significant differences and shifts in AMR rates, additional AMR data analyses and AMR surveillance studies are strict requirements to fully understand the continuous development in AMR rates as there is no “ideal” surveillance system covering all aspects [28]. Nonetheless, the desire to continuously monitor, analyse, model and predict AMR, has led to the increased development and use of local, regional, national and international surveillance systems [27]. Critchley et al. have inventoried the requirement set by different types of users (Table 1).
On the local level, clinical microbiology laboratories should communicate AMR surveillance data to healthcare providers in an understandable manner. Since MDROs can migrate between healthcare institutions, countries and continents by migrating people, local healthcare providers should be aware of local, regional, national and international surveillance program implementations and their ensuing results on AMR. On the other hand, such surveillance program implementations should be well-designed, well-maintained, longitudinal, and involve an appropriate collaboration with local laboratories over time [27].
-
-From Critchley et al., 2004 [27].
+Table 1. Uses of antibiotic resistance surveillance system data by hospitals, university researchers, pharmaceutical companies and governments. From Critchley et al., 2004 [27]. +
+
As an example, ISIS-AR (Infectious disease Surveillance Information System for Antibiotic Resistance) is a Dutch national surveillance program, for which a large number of the Dutch clinical microbiology laboratories provide anonymised data on AMR to the National Institute for Public Health and the Environment (Rijksinstituut voor Volksgezondheid en Milieu, RIVM) [40]. In Germany, ARS (Antibiotic Resistance Surveillance) is a similar laboratory-based national surveillance program, that attempts to enable differential statements according to structural characteristics of health care and regions [41,42]. Both these national surveillance programs provide data for EARS-Net and GLASS of the WHO [37,43].
In academia, the free and open-source statistical language R is an increasingly popular tool for analysing study results and developing new scientific methods, especially in medical fields such as human genetics, health decision sciences, and proteomics [44–47]. Even more so, a new type of study seems to currently arise where researchers from different medical fields publish tutorials on how to acquire new insights using R as a programming language [48–50]. In 2020, R ranked 8th in the TIOBE index, a global initiative to measure the popularity of programming languages, while it ranked 73rd in 2008 [51].
R was developed for statistical computing and graphics supported by the R Foundation for Statistical Computing [52,53]. It is freely available under the GNU General Public License v2, meaning that it may be used for both private and commercial purposes in any way, but not for patent purposes. As a statistical package, it is comparable to the proprietary software programs Stata, SAS and SPSS [54]. However, as opposed to these proprietary software programs, R has an open file format and can read data from any source, including files from other software programs, and websites. Moreover, the ‘base’ functions of R are extendible by users who develop so-called packages for R. The Comprehensive R Archive Network (CRAN) that hosts and maintains R through the R Foundation for Statistical Computing, accepts package submissions from users and subjects users to a peer-review submission process and a strict repository policy [53,55]. As of May 2021, the CRAN package repository features 17,671 available packages.
-Not only the popularity of using R has increased over the last decade. The number of developed packages has also increased strongly over the last years, especially since 2016 (Figure 1.4). This is probably attributed to a rather new integrated desktop environment (IDE) to use R, called RStudio [56]. RStudio is also the name of the corporation that developed the RStudio IDE and authored the so-called tidyverse, a collection of R packages (such as dplyr and ggplot2) that are specifically designed to ease data importing, tidying, manipulating, visualising, and programming, as well as to improve code reading [57–59]. The tidyverse can be used for most data analytical tasks and has been the method of choice for numerous (clinical) studies, including those presented in this thesis.
+Not only the popularity of using R has increased over the last decade. The number of developed packages has also increased strongly over the last years, especially since 2016 (Figure 1.4). This is probably attributed to a rather new integrated desktop environment (IDE) to use R, called RStudio [56]. RStudio is also the name of the corporation that developed the RStudio IDE and authored the so-called tidyverse, a collection of R packages (such as dplyr and ggplot2) that are specifically designed to ease data importing, tidying, manipulating, visualising, and programming, as well as to improve code reading [57–59]. The tidyverse can be used for most data analytical tasks and has been the method of choice for numerous (clinical) studies, including those presented in this thesis.
@@ -236,7 +357,7 @@ Figure 1.4: The number of R packages by date of the last update over the last te
Studies within this thesis were geographically organised or initiated in the Northern cross-border region of the Netherlands and Germany, Figure 1.5. According to the German philosopher Liessmann, there are only national borders defined by humans, but no natural borders [67]. He explained that borders as man-made conventions are never absolute, but that it is always possible to cross them. Despite the existing territorial border, there are many similarities in the Netherlands and Germany today, but just as many and clear differences, especially concerning the healthcare sector. A German patient can become a patient in the Netherlands just as quickly as a Dutch patient can in Germany. Since pathogens know no borders, patient protection and infection prevention must not stop at borders [68]. The Netherlands and Germany have, among many other matters, apparent differences within the healthcare system in general and in terms of AMR, especially concerning MDRO definitions and infection prevention guidelines. To study these differences, INTERREG programs enable cross-border, transnational and interregional cooperation. INTERREG is one of the central instruments in European cohesion and regional policy, with which the development differences between the European countries in the border regions should be reduced and economic cohesion strengthened. It aims to ensure that national borders are not an obstacle to the balanced development and integration of the European territory [69]. One of its programs, EurHealth-1Health, was a large research project that aimed to facilitate working together in battling AMR and MDROs and to empower sustainable collaborations across the border.
+Studies within this thesis were geographically organised or initiated in the Northern cross-border region of the Netherlands and Germany, Figure 1.5. According to the German philosopher Liessmann, there are only national borders defined by humans, but no natural borders [67]. He explained that borders as man-made conventions are never absolute, but that it is always possible to cross them. Despite the existing territorial border, there are many similarities in the Netherlands and Germany today, but just as many and clear differences, especially concerning the healthcare sector. A German patient can become a patient in the Netherlands just as quickly as a Dutch patient can in Germany. Since pathogens know no borders, patient protection and infection prevention must not stop at borders [68]. The Netherlands and Germany have, among many other matters, apparent differences within the healthcare system in general and in terms of AMR, especially concerning MDRO definitions and infection prevention guidelines. To study these differences, INTERREG programs enable cross-border, transnational and interregional cooperation. INTERREG is one of the central instruments in European cohesion and regional policy, with which the development differences between the European countries in the border regions should be reduced and economic cohesion strengthened. It aims to ensure that national borders are not an obstacle to the balanced development and integration of the European territory [69]. One of its programs, EurHealth-1Health, was a large research project that aimed to facilitate working together in battling AMR and MDROs and to empower sustainable collaborations across the border.
@@ -341,7 +462,7 @@ Figure 1.5: Geographic overview of three Euregio’s that make up most of the Du