1
0
mirror of https://github.com/msberends/AMR.git synced 2026-06-24 02:16:23 +02:00
Files
AMR/articles/WISCA.html
2026-06-23 18:00:14 +00:00

612 lines
42 KiB
HTML
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<!DOCTYPE html>
<!-- Generated by pkgdown: do not edit by hand --><html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<title>Estimating Empirical Coverage with WISCA • AMR (for R)</title>
<!-- favicons --><link rel="icon" type="image/png" sizes="96x96" href="../favicon-96x96.png">
<link rel="icon" type="”image/svg+xml”" href="../favicon.svg">
<link rel="apple-touch-icon" sizes="180x180" href="../apple-touch-icon.png">
<link rel="icon" sizes="any" href="../favicon.ico">
<link rel="manifest" href="../site.webmanifest">
<script src="../deps/jquery-3.6.0/jquery-3.6.0.min.js"></script><meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<link href="../deps/bootstrap-5.3.8/bootstrap.min.css" rel="stylesheet">
<script src="../deps/bootstrap-5.3.8/bootstrap.bundle.min.js"></script><link href="../deps/Lato-0.4.10/font.css" rel="stylesheet">
<link href="../deps/Fira_Code-0.4.10/font.css" rel="stylesheet">
<link href="../deps/font-awesome-6.5.2/css/all.min.css" rel="stylesheet">
<link href="../deps/font-awesome-6.5.2/css/v4-shims.min.css" rel="stylesheet">
<script src="../deps/headroom-0.11.0/headroom.min.js"></script><script src="../deps/headroom-0.11.0/jQuery.headroom.min.js"></script><script src="../deps/bootstrap-toc-1.0.1/bootstrap-toc.min.js"></script><script src="../deps/clipboard.js-2.0.11/clipboard.min.js"></script><script src="../deps/search-1.0.0/autocomplete.jquery.min.js"></script><script src="../deps/search-1.0.0/fuse.min.js"></script><script src="../deps/search-1.0.0/mark.min.js"></script><!-- pkgdown --><script src="../pkgdown.js"></script><link href="../extra.css" rel="stylesheet">
<script src="../extra.js"></script><meta property="og:title" content="Estimating Empirical Coverage with WISCA">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.16.11/dist/katex.min.css" integrity="sha384-nB0miv6/jRmo5UMMR1wu3Gz6NLsoTkbqJghGIsx//Rlm+ZU03BU6SQNC66uf4l5+" crossorigin="anonymous">
<script defer src="https://cdn.jsdelivr.net/npm/katex@0.16.11/dist/katex.min.js" integrity="sha384-7zkQWkzuo3B5mTepMUcHkMB5jZaolc2xDwL6VFqjFALcbeS9Ggm/Yr2r3Dy4lfFg" crossorigin="anonymous"></script><script defer src="https://cdn.jsdelivr.net/npm/katex@0.16.11/dist/contrib/auto-render.min.js" integrity="sha384-43gviWU0YVjaDtb/GhzOouOXtZMP/7XUzwPTstBeZFe/+rCMvRwr4yROQP43s0Xk" crossorigin="anonymous" onload="renderMathInElement(document.body);"></script>
</head>
<body>
<a href="#main" class="visually-hidden-focusable">Skip to contents</a>
<nav class="navbar navbar-expand-lg fixed-top bg-primary" data-bs-theme="dark" aria-label="Site navigation"><div class="container">
<a class="navbar-brand me-2" href="../index.html">AMR (for R)</a>
<small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">3.0.1.9061</small>
<button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
<span class="navbar-toggler-icon"></span>
</button>
<div id="navbar" class="collapse navbar-collapse ms-3">
<ul class="navbar-nav me-auto">
<li class="nav-item dropdown">
<button class="nav-link dropdown-toggle" type="button" id="dropdown-how-to" data-bs-toggle="dropdown" aria-expanded="false" aria-haspopup="true"><span class="fa fa-question-circle"></span> How to</button>
<ul class="dropdown-menu" aria-labelledby="dropdown-how-to">
<li><a class="dropdown-item" href="../articles/AMR.html"><span class="fa fa-directions"></span> Conduct AMR Analysis</a></li>
<li><a class="dropdown-item" href="../reference/antibiogram.html"><span class="fa fa-file-prescription"></span> Generate Antibiogram (Trad./Syndromic/WISCA)</a></li>
<li><a class="dropdown-item" href="../articles/AMR_with_tidymodels.html"><span class="fa fa-square-root-variable"></span> Use AMR for Predictive Modelling (tidymodels)</a></li>
<li><a class="dropdown-item" href="../articles/datasets.html"><span class="fa fa-database"></span> Download Data Sets for Own Use</a></li>
<li><a class="dropdown-item" href="../reference/AMR-options.html"><span class="fa fa-gear"></span> Set User- Or Team-specific Package Settings</a></li>
<li><a class="dropdown-item" href="../articles/PCA.html"><span class="fa fa-compress"></span> Conduct Principal Component Analysis for AMR</a></li>
<li><a class="dropdown-item" href="../reference/mdro.html"><span class="fa fa-skull-crossbones"></span> Determine Multi-Drug Resistance (MDR)</a></li>
<li><a class="dropdown-item" href="../articles/WHONET.html"><span class="fa fa-globe-americas"></span> Work with WHONET Data</a></li>
<li><a class="dropdown-item" href="../articles/EUCAST.html"><span class="fa fa-exchange-alt"></span> Apply EUCAST Rules</a></li>
<li><a class="dropdown-item" href="../reference/mo_property.html"><span class="fa fa-bug"></span> Get Taxonomy of a Microorganism</a></li>
<li><a class="dropdown-item" href="../reference/ab_property.html"><span class="fa fa-capsules"></span> Get Properties of an Antibiotic Drug</a></li>
<li><a class="dropdown-item" href="../reference/av_property.html"><span class="fa fa-capsules"></span> Get Properties of an Antiviral Drug</a></li>
</ul>
</li>
<li class="nav-item"><a class="nav-link" href="../articles/AMR_for_Python.html"><span class="fa fab fa-python"></span> AMR for Python</a></li>
<li class="nav-item"><a class="nav-link" href="../reference/index.html"><span class="fa fa-book-open"></span> Manual</a></li>
<li class="nav-item"><a class="nav-link" href="../authors.html"><span class="fa fa-users"></span> Authors</a></li>
</ul>
<ul class="navbar-nav">
<li class="nav-item"><form class="form-inline" role="search">
<input class="form-control" type="search" name="search-input" id="search-input" autocomplete="off" aria-label="Search site" placeholder="Search for" data-search-index="../search.json">
</form></li>
<li class="nav-item"><a class="nav-link" href="../news/index.html"><span class="fa fa-newspaper"></span> Changelog</a></li>
<li class="nav-item"><a class="external-link nav-link" href="https://github.com/msberends/AMR"><span class="fa fa-github"></span> Source Code</a></li>
</ul>
</div>
</div>
</nav><div class="container template-article">
<div class="row">
<main id="main" class="col-md-9"><div class="page-header">
<img src="../logo.svg" class="logo" alt=""><h1>Estimating Empirical Coverage with WISCA</h1>
<small class="dont-index">Source: <a href="https://github.com/msberends/AMR/blob/main/vignettes/WISCA.Rmd" class="external-link"><code>vignettes/WISCA.Rmd</code></a></small>
<div class="d-none name"><code>WISCA.Rmd</code></div>
</div>
<div class="section level2">
<h2 id="why-wisca">Why WISCA?<a class="anchor" aria-label="anchor" href="#why-wisca"></a>
</h2>
<p>When a clinician starts empirical antimicrobial therapy, the
causative pathogen is unknown. The question they need answered is not
<em>“what proportion of</em> E. coli <em>is susceptible to
ciprofloxacin?“</em> but rather <em>“what is the probability that this
regimen will adequately cover whatever pathogen turns out to be causing
my patients infection?”</em></p>
<p>The traditional cumulative antibiogram, as standardised by CLSI M39,
cannot answer that question. It presents susceptibility percentages per
species per antibiotic, but:</p>
<ul>
<li>
<strong>It fragments information by organism.</strong> The clinician
must mentally combine susceptibility rates across multiple species,
weighting by how often each species causes the syndrome, a calculation
nobody does at the bedside.</li>
<li>
<strong>It ignores pathogen incidence.</strong> A species that
causes 2% of infections is given the same visual weight as one that
causes 60%.</li>
<li>
<strong>It does not evaluate combination regimens.</strong> Much
empirical therapy consists of two or more agents, but the traditional
antibiogram only shows monotherapy per organism.</li>
<li>
<strong>It provides no measure of uncertainty.</strong> A reported
“90% susceptible” based on 50 isolates has a 95% confidence interval of
roughly 78-97% (Clopper-Pearson), yet the antibiogram presents it as a
point estimate without context.</li>
</ul>
<p><strong>WISCA</strong> (Weighted-Incidence Syndromic Combination
Antibiogram) resolves all four limitations. It estimates the probability
that a regimen will provide adequate empirical coverage for a given
infection syndrome, weighted by local pathogen incidence, with full
uncertainty quantification via Bayesian inference.</p>
<p>The concept was introduced by Hebert <em>et al.</em> (2012), who
demonstrated that traditional antibiogram susceptibility rates could be
misleading: ciprofloxacin appeared 84% effective against <em>E.
coli</em> in the traditional antibiogram, but WISCA revealed only 62%
coverage for UTI and 37% for abdominal infections, because enterococci
(intrinsically resistant) and other species contribute substantially to
these syndromes. Randhawa <em>et al.</em> (2014) showed that
WISCA-guided regimen selection could improve time-to-adequate-coverage
on the ICU by over 40%. Bielicki <em>et al.</em> (2016) introduced the
Bayesian framework now used in this package, enabling credible intervals
and multi-centre pooling. Cook <em>et al.</em> (2022) applied it
globally across 52 hospitals in 23 countries.</p>
</div>
<div class="section level2">
<h2 id="the-idea">The idea<a class="anchor" aria-label="anchor" href="#the-idea"></a>
</h2>
<p>WISCA asks:</p>
<blockquote>
<p>“What is the <strong>probability</strong> that this regimen
<strong>will cover</strong> the pathogen, given the syndrome?”</p>
</blockquote>
<p>This means combining two quantities:</p>
<ul>
<li>
<strong>Pathogen incidence</strong> in the syndrome (how often each
species causes it),</li>
<li>
<strong>Susceptibility</strong> of each pathogen to the
regimen.</li>
</ul>
<p>We can write this as:</p>
<p><math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext mathvariant="normal">Coverage</mtext><mo>=</mo><munder><mo></mo><mi>i</mi></munder><mo stretchy="false" form="prefix">(</mo><msub><mtext mathvariant="normal">Incidence</mtext><mi>i</mi></msub><mo>×</mo><msub><mtext mathvariant="normal">Susceptibility</mtext><mi>i</mi></msub><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\text{Coverage} = \sum_i (\text{Incidence}_i \times \text{Susceptibility}_i)</annotation></semantics></math></p>
<p>For example, suppose in your hospital:</p>
<ul>
<li>
<em>E. coli</em> causes 60% of UTIs, and 90% of <em>E. coli</em> are
susceptible to a drug.</li>
<li>
<em>Klebsiella</em> causes 40% of UTIs, and 70% of
<em>Klebsiella</em> are susceptible.</li>
</ul>
<p>Then:</p>
<p><math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext mathvariant="normal">Coverage</mtext><mo>=</mo><mo stretchy="false" form="prefix">(</mo><mn>0.6</mn><mo>×</mo><mn>0.9</mn><mo stretchy="false" form="postfix">)</mo><mo>+</mo><mo stretchy="false" form="prefix">(</mo><mn>0.4</mn><mo>×</mo><mn>0.7</mn><mo stretchy="false" form="postfix">)</mo><mo>=</mo><mn>0.82</mn></mrow><annotation encoding="application/x-tex">\text{Coverage} = (0.6 \times 0.9) + (0.4 \times 0.7) = 0.82</annotation></semantics></math></p>
<p>That 82% is a far more clinically meaningful number than the
species-level “90% of <em>E. coli</em>” and “70% of <em>Klebsiella</em>
reported separately in a traditional antibiogram, because it directly
answers the question the clinician actually faces.</p>
<p>But in real data, both incidence and susceptibility are
<strong>estimated from finite samples</strong>, so they carry
uncertainty. A sample of 50 isolates is not a census. WISCA models this
uncertainty <strong>probabilistically</strong>, using conjugate Bayesian
distributions.</p>
</div>
<div class="section level2">
<h2 id="the-bayesian-engine">The Bayesian engine<a class="anchor" aria-label="anchor" href="#the-bayesian-engine"></a>
</h2>
<div class="section level3">
<h3 id="pathogen-incidence">Pathogen incidence<a class="anchor" aria-label="anchor" href="#pathogen-incidence"></a>
</h3>
<p>Let:</p>
<ul>
<li>
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mi>K</mi><annotation encoding="application/x-tex">K</annotation></semantics></math>
be the number of pathogens,</li>
<li>
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>𝛂</mi><mo>=</mo><mo stretchy="false" form="prefix">(</mo><mn>1</mn><mo>,</mo><mn>1</mn><mo>,</mo><mi></mi><mo>,</mo><mn>1</mn><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\boldsymbol{\alpha} = (1, 1, \ldots, 1)</annotation></semantics></math>
be a
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mtext mathvariant="normal">Dirichlet</mtext><annotation encoding="application/x-tex">\text{Dirichlet}</annotation></semantics></math>
prior (uniform, non-informative),</li>
<li>
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>𝐧</mi><mo>=</mo><mo stretchy="false" form="prefix">(</mo><msub><mi>n</mi><mn>1</mn></msub><mo>,</mo><mi></mi><mo>,</mo><msub><mi>n</mi><mi>K</mi></msub><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\boldsymbol{n} = (n_1, \ldots, n_K)</annotation></semantics></math>
be the observed isolate counts per species.</li>
</ul>
<p>Then the posterior incidence is:</p>
<p><math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>𝐩</mi><mo></mo><mtext mathvariant="normal">Dirichlet</mtext><mo stretchy="false" form="prefix">(</mo><msub><mi>α</mi><mn>1</mn></msub><mo>+</mo><msub><mi>n</mi><mn>1</mn></msub><mo>,</mo><mi></mi><mo>,</mo><msub><mi>α</mi><mi>K</mi></msub><mo>+</mo><msub><mi>n</mi><mi>K</mi></msub><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\boldsymbol{p} \sim \text{Dirichlet}(\alpha_1 + n_1, \ldots, \alpha_K + n_K)</annotation></semantics></math></p>
<p>To simulate from this, we use:</p>
<p><math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>x</mi><mi>i</mi></msub><mo></mo><mtext mathvariant="normal">Gamma</mtext><mo stretchy="false" form="prefix">(</mo><msub><mi>α</mi><mi>i</mi></msub><mo>+</mo><msub><mi>n</mi><mi>i</mi></msub><mo>,</mo><mspace width="0.222em"></mspace><mn>1</mn><mo stretchy="false" form="postfix">)</mo><mo>,</mo><mspace width="1.0em"></mspace><msub><mi>p</mi><mi>i</mi></msub><mo>=</mo><mfrac><msub><mi>x</mi><mi>i</mi></msub><mrow><munderover><mo></mo><mrow><mi>j</mi><mo>=</mo><mn>1</mn></mrow><mi>K</mi></munderover><msub><mi>x</mi><mi>j</mi></msub></mrow></mfrac></mrow><annotation encoding="application/x-tex">x_i \sim \text{Gamma}(\alpha_i + n_i,\ 1), \quad p_i = \frac{x_i}{\sum_{j=1}^{K} x_j}</annotation></semantics></math></p>
<p>The Dirichlet is the conjugate prior for multinomial data. With the
non-informative prior
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext mathvariant="normal">Dirichlet</mtext><mo stretchy="false" form="prefix">(</mo><mn>1</mn><mo>,</mo><mn>1</mn><mo>,</mo><mi></mi><mo>,</mo><mn>1</mn><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\text{Dirichlet}(1, 1, \ldots, 1)</annotation></semantics></math>,
the posterior is dominated by the data once sample sizes are reasonable.
With small samples, the posterior is appropriately more diffuse,
reflecting genuine uncertainty, and the resulting credible intervals
will be wider.</p>
</div>
<div class="section level3">
<h3 id="susceptibility">Susceptibility<a class="anchor" aria-label="anchor" href="#susceptibility"></a>
</h3>
<p>Each pathogen-regimen pair has a prior and observed data:</p>
<ul>
<li>Default prior:
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext mathvariant="normal">Beta</mtext><mo stretchy="false" form="prefix">(</mo><mn>0.5</mn><mo>,</mo><mn>0.5</mn><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\text{Beta}(0.5, 0.5)</annotation></semantics></math>
(Jeffreys prior)</li>
<li>Intrinsically resistant pairs:
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext mathvariant="normal">Beta</mtext><mo stretchy="false" form="prefix">(</mo><mn>1</mn><mo>,</mo><mn>9999</mn><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\text{Beta}(1, 9999)</annotation></semantics></math>,
forcing near-zero susceptibility regardless of observed data (based on
EUCAST Expected Resistant Phenotypes)</li>
<li>Data:
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mi>S</mi><annotation encoding="application/x-tex">S</annotation></semantics></math>
susceptible out of
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mi>N</mi><annotation encoding="application/x-tex">N</annotation></semantics></math>
tested</li>
</ul>
<p>The
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mi>S</mi><annotation encoding="application/x-tex">S</annotation></semantics></math>
category could also include values SDD (susceptible, dose-dependent) and
I (intermediate [CLSI], or susceptible, increased exposure
[EUCAST]).</p>
<p>Then the posterior is:</p>
<p><math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>θ</mi><mo></mo><mtext mathvariant="normal">Beta</mtext><mo stretchy="false" form="prefix">(</mo><msub><mi>α</mi><mn>0</mn></msub><mo>+</mo><mi>S</mi><mo>,</mo><mspace width="0.222em"></mspace><msub><mi>β</mi><mn>0</mn></msub><mo>+</mo><mi>N</mi><mo></mo><mi>S</mi><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\theta \sim \text{Beta}(\alpha_0 + S,\ \beta_0 + N - S)</annotation></semantics></math></p>
</div>
<div class="section level3">
<h3 id="final-coverage-estimate">Final coverage estimate<a class="anchor" aria-label="anchor" href="#final-coverage-estimate"></a>
</h3>
<p>Putting it together:</p>
<ol style="list-style-type: decimal">
<li>Simulate pathogen incidence:
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi>𝐩</mi><mo></mo><mtext mathvariant="normal">Dirichlet</mtext></mrow><annotation encoding="application/x-tex">\boldsymbol{p} \sim \text{Dirichlet}</annotation></semantics></math>
</li>
<li>Simulate susceptibility:
<math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><msub><mi>θ</mi><mi>i</mi></msub><mo></mo><mtext mathvariant="normal">Beta</mtext><mo stretchy="false" form="prefix">(</mo><msub><mi>α</mi><mn>0</mn></msub><mo>+</mo><msub><mi>S</mi><mi>i</mi></msub><mo>,</mo><mspace width="0.222em"></mspace><msub><mi>β</mi><mn>0</mn></msub><mo>+</mo><msub><mi>N</mi><mi>i</mi></msub><mo></mo><msub><mi>S</mi><mi>i</mi></msub><mo stretchy="false" form="postfix">)</mo></mrow><annotation encoding="application/x-tex">\theta_i \sim \text{Beta}(\alpha_0 + S_i,\ \beta_0 + N_i - S_i)</annotation></semantics></math>
</li>
<li>Combine:</li>
</ol>
<p><math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mtext mathvariant="normal">Coverage</mtext><mo>=</mo><munderover><mo></mo><mrow><mi>i</mi><mo>=</mo><mn>1</mn></mrow><mi>K</mi></munderover><msub><mi>p</mi><mi>i</mi></msub><mo></mo><msub><mi>θ</mi><mi>i</mi></msub></mrow><annotation encoding="application/x-tex">\text{Coverage} = \sum_{i=1}^{K} p_i \cdot \theta_i</annotation></semantics></math></p>
<p>Repeat this simulation (e.g., 1000 times) and summarise:</p>
<ul>
<li>
<strong>Mean</strong> = expected coverage</li>
<li>
<strong>Quantiles</strong> = credible interval (95% by default)</li>
</ul>
<p>Because each simulation draws from the full posterior, the resulting
distribution of coverage estimates naturally captures the joint
uncertainty in both pathogen incidence and susceptibility. The credible
interval tells you how confident you can be in the coverage estimate,
something a traditional antibiogram never provides.</p>
</div>
</div>
<div class="section level2">
<h2 id="when-to-use-wisca-vs--traditional-antibiograms">When to use WISCA vs. traditional antibiograms<a class="anchor" aria-label="anchor" href="#when-to-use-wisca-vs--traditional-antibiograms"></a>
</h2>
<table class="table">
<thead><tr class="header">
<th>Goal</th>
<th>Recommended approach</th>
</tr></thead>
<tbody>
<tr class="odd">
<td>Guide empirical therapy decisions</td>
<td><strong>WISCA</strong></td>
</tr>
<tr class="even">
<td>Compare regimens for a syndrome</td>
<td><strong>WISCA</strong></td>
</tr>
<tr class="odd">
<td>Evaluate combination regimens</td>
<td><strong>WISCA</strong></td>
</tr>
<tr class="even">
<td>Antimicrobial stewardship (A-team)</td>
<td><strong>WISCA</strong></td>
</tr>
<tr class="odd">
<td>Track resistance trends per species</td>
<td>Traditional / Combination</td>
</tr>
<tr class="even">
<td>AMR surveillance reporting</td>
<td>Traditional / Syndromic</td>
</tr>
<tr class="odd">
<td>Understand species-level epidemiology</td>
<td>Traditional</td>
</tr>
</tbody>
</table>
<p>In short: if the end goal involves a <em>patient</em> who does not
yet have a culture result, WISCA is the appropriate tool. If the end
goal is <em>surveillance</em> of resistance at the species level, the
traditional antibiogram remains fit for purpose.</p>
</div>
<div class="section level2">
<h2 id="practical-use-in-the-amr-package">Practical use in the <code>AMR</code> package<a class="anchor" aria-label="anchor" href="#practical-use-in-the-amr-package"></a>
</h2>
<div class="section level3">
<h3 id="prepare-data">Prepare data<a class="anchor" aria-label="anchor" href="#prepare-data"></a>
</h3>
<div class="sourceCode" id="cb1"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="kw"><a href="https://rdrr.io/r/base/library.html" class="external-link">library</a></span><span class="op">(</span><span class="va"><a href="https://amr-for-r.org">AMR</a></span><span class="op">)</span></span>
<span><span class="va">data</span> <span class="op">&lt;-</span> <span class="va">example_isolates</span></span>
<span></span>
<span><span class="co"># Structure of our data</span></span>
<span><span class="va">data</span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># A tibble: 2,000 × 46</span></span></span>
<span><span class="co">#&gt; date patient age gender ward mo PEN OXA FLC AMX </span></span>
<span><span class="co">#&gt; <span style="color: #949494; font-style: italic;">&lt;date&gt;</span> <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span> <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span> <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span> <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span> <span style="color: #949494; font-style: italic;">&lt;mo&gt;</span> <span style="color: #949494; font-style: italic;">&lt;sir&gt;</span> <span style="color: #949494; font-style: italic;">&lt;sir&gt;</span> <span style="color: #949494; font-style: italic;">&lt;sir&gt;</span> <span style="color: #949494; font-style: italic;">&lt;sir&gt;</span></span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 1</span> 2002-01-02 A77334 65 F Clinical <span style="color: #949494;">B_</span>ESCHR<span style="color: #949494;">_</span>COLI <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #949494;"> NA</span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 2</span> 2002-01-03 A77334 65 F Clinical <span style="color: #949494;">B_</span>ESCHR<span style="color: #949494;">_</span>COLI <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #949494;"> NA</span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 3</span> 2002-01-07 067927 45 F ICU <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>EPDR <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 4</span> 2002-01-07 067927 45 F ICU <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>EPDR <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 5</span> 2002-01-13 067927 45 F ICU <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>EPDR <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 6</span> 2002-01-13 067927 45 F ICU <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>EPDR <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 7</span> 2002-01-14 462729 78 M Clinical <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>AURS <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #5FD7AF;"> S </span> <span style="color: #080808; background-color: #FF5F5F;"> R </span></span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 8</span> 2002-01-14 462729 78 M Clinical <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>AURS <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #5FD7AF;"> S </span> <span style="color: #080808; background-color: #FF5F5F;"> R </span></span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;"> 9</span> 2002-01-16 067927 45 F ICU <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>EPDR <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #BCBCBC;">10</span> 2002-01-17 858515 79 F ICU <span style="color: #949494;">B_</span>STPHY<span style="color: #949494;">_</span>EPDR <span style="color: #080808; background-color: #FF5F5F;"> R </span> <span style="color: #949494;"> NA</span> <span style="color: #080808; background-color: #5FD7AF;"> S </span> <span style="color: #949494;"> NA</span> </span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># 1,990 more rows</span></span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># 36 more variables: AMC &lt;sir&gt;, AMP &lt;sir&gt;, TZP &lt;sir&gt;, CZO &lt;sir&gt;, FEP &lt;sir&gt;,</span></span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># CXM &lt;sir&gt;, FOX &lt;sir&gt;, CTX &lt;sir&gt;, CAZ &lt;sir&gt;, CRO &lt;sir&gt;, GEN &lt;sir&gt;,</span></span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># TOB &lt;sir&gt;, AMK &lt;sir&gt;, KAN &lt;sir&gt;, TMP &lt;sir&gt;, SXT &lt;sir&gt;, NIT &lt;sir&gt;,</span></span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># FOS &lt;sir&gt;, LNZ &lt;sir&gt;, CIP &lt;sir&gt;, MFX &lt;sir&gt;, VAN &lt;sir&gt;, TEC &lt;sir&gt;,</span></span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># TCY &lt;sir&gt;, TGC &lt;sir&gt;, DOX &lt;sir&gt;, ERY &lt;sir&gt;, CLI &lt;sir&gt;, AZM &lt;sir&gt;,</span></span></span>
<span><span class="co">#&gt; <span style="color: #949494;"># IPM &lt;sir&gt;, MEM &lt;sir&gt;, MTR &lt;sir&gt;, CHL &lt;sir&gt;, COL &lt;sir&gt;, MUP &lt;sir&gt;, …</span></span></span>
<span></span>
<span><span class="co"># Add a synthetic syndrome column for demonstration</span></span>
<span><span class="va">data</span><span class="op">$</span><span class="va">syndrome</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/ifelse.html" class="external-link">ifelse</a></span><span class="op">(</span><span class="va">data</span><span class="op">$</span><span class="va">mo</span> <span class="op"><a href="../reference/like.html">%like%</a></span> <span class="st">"coli"</span>, <span class="st">"UTI"</span>, <span class="st">"Non-UTI"</span><span class="op">)</span></span></code></pre></div>
</div>
<div class="section level3">
<h3 id="basic-wisca">Basic WISCA<a class="anchor" aria-label="anchor" href="#basic-wisca"></a>
</h3>
<div class="sourceCode" id="cb2"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="../reference/antibiogram.html">wisca</a></span><span class="op">(</span><span class="va">data</span>,</span>
<span> antimicrobials <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"AMC"</span>, <span class="st">"CIP"</span>, <span class="st">"GEN"</span><span class="op">)</span></span>
<span><span class="op">)</span></span></code></pre></div>
<table class="table">
<thead><tr class="header">
<th align="left">Amoxicillin/clavulanic acid</th>
<th align="left">Ciprofloxacin</th>
<th align="left">Gentamicin</th>
</tr></thead>
<tbody><tr class="odd">
<td align="left">74.2% (72.1-76.1%)</td>
<td align="left">78.4% (75.6-81.1%)</td>
<td align="left">72.5% (70.4-74.6%)</td>
</tr></tbody>
</table>
</div>
<div class="section level3">
<h3 id="use-combination-regimens">Use combination regimens<a class="anchor" aria-label="anchor" href="#use-combination-regimens"></a>
</h3>
<p>Combination regimens are specified with a <code>+</code> separator.
WISCA evaluates whether <em>at least one</em> agent in the combination
covers the pathogen:</p>
<div class="sourceCode" id="cb3"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="../reference/antibiogram.html">wisca</a></span><span class="op">(</span><span class="va">data</span>,</span>
<span> antimicrobials <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"AMC"</span>, <span class="st">"AMC + CIP"</span>, <span class="st">"AMC + GEN"</span><span class="op">)</span></span>
<span><span class="op">)</span></span></code></pre></div>
<table class="table">
<colgroup>
<col width="24%">
<col width="38%">
<col width="36%">
</colgroup>
<thead><tr class="header">
<th align="left">Amoxicillin/clavulanic acid</th>
<th align="left">Amoxicillin/clavulanic acid + Ciprofloxacin</th>
<th align="left">Amoxicillin/clavulanic acid + Gentamicin</th>
</tr></thead>
<tbody><tr class="odd">
<td align="left">74.2% (72.2-76.1%)</td>
<td align="left">88.8% (87.2-90.4%)</td>
<td align="left">90.8% (89.4-92.2%)</td>
</tr></tbody>
</table>
</div>
<div class="section level3">
<h3 id="stratify-by-syndrome">Stratify by syndrome<a class="anchor" aria-label="anchor" href="#stratify-by-syndrome"></a>
</h3>
<p>Use <code>syndromic_group</code> to produce separate WISCA estimates
per clinical stratum. You can pass a column name or any expression:</p>
<div class="sourceCode" id="cb4"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="../reference/antibiogram.html">wisca</a></span><span class="op">(</span><span class="va">data</span>,</span>
<span> antimicrobials <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"AMC"</span>, <span class="st">"AMC + CIP"</span>, <span class="st">"AMC + GEN"</span><span class="op">)</span>,</span>
<span> syndromic_group <span class="op">=</span> <span class="st">"syndrome"</span></span>
<span><span class="op">)</span></span></code></pre></div>
<table class="table">
<colgroup>
<col width="12%">
<col width="21%">
<col width="34%">
<col width="31%">
</colgroup>
<thead><tr class="header">
<th align="left">Syndromic Group</th>
<th align="left">Amoxicillin/clavulanic acid</th>
<th align="left">Amoxicillin/clavulanic acid + Ciprofloxacin</th>
<th align="left">Amoxicillin/clavulanic acid + Gentamicin</th>
</tr></thead>
<tbody>
<tr class="odd">
<td align="left">Non-UTI</td>
<td align="left">70.3% (67.9-72.7%)</td>
<td align="left">86.8% (84.9-88.7%)</td>
<td align="left">88.4% (86.4-90.2%)</td>
</tr>
<tr class="even">
<td align="left">UTI</td>
<td align="left">80.3% (77-83.3%)</td>
<td align="left">88.4% (85.7-90.8%)</td>
<td align="left">91% (88.3-93.3%)</td>
</tr>
</tbody>
</table>
<p>The <code>AMR</code> package is available in 28 languages, which can
all be used for the <code><a href="../reference/antibiogram.html">wisca()</a></code> function too:</p>
<div class="sourceCode" id="cb5"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="../reference/antibiogram.html">wisca</a></span><span class="op">(</span><span class="va">data</span>,</span>
<span> antimicrobials <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="st">"AMC"</span>, <span class="st">"AMC + CIP"</span>, <span class="st">"AMC + GEN"</span><span class="op">)</span>,</span>
<span> syndromic_group <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/grep.html" class="external-link">gsub</a></span><span class="op">(</span><span class="st">"UTI"</span>, <span class="st">"UCI"</span>, <span class="va">data</span><span class="op">$</span><span class="va">syndrome</span><span class="op">)</span>,</span>
<span> language <span class="op">=</span> <span class="st">"Spanish"</span></span>
<span><span class="op">)</span></span></code></pre></div>
<table class="table">
<colgroup>
<col width="12%">
<col width="21%">
<col width="34%">
<col width="31%">
</colgroup>
<thead><tr class="header">
<th align="left">Grupo sindrómico</th>
<th align="left">Amoxicilina/ácido clavulánico</th>
<th align="left">Amoxicilina/ácido clavulánico + Ciprofloxacina</th>
<th align="left">Amoxicilina/ácido clavulánico + Gentamicina</th>
</tr></thead>
<tbody>
<tr class="odd">
<td align="left">Non-UCI</td>
<td align="left">70.4% (68-72.8%)</td>
<td align="left">86.7% (84.6-88.7%)</td>
<td align="left">88.5% (86.5-90.2%)</td>
</tr>
<tr class="even">
<td align="left">UCI</td>
<td align="left">80.3% (77.2-83.5%)</td>
<td align="left">88.4% (85.5-90.8%)</td>
<td align="left">91% (88.4-93.1%)</td>
</tr>
</tbody>
</table>
</div>
<div class="section level3">
<h3 id="interpreting-the-output">Interpreting the output<a class="anchor" aria-label="anchor" href="#interpreting-the-output"></a>
</h3>
<p>Each row shows the estimated empirical coverage for a regimen, with a
95% credible interval. When comparing regimens:</p>
<ul>
<li>
<strong>Overlapping credible intervals</strong> mean there is no
statistically significant difference in coverage. If a narrower-spectrum
regimen overlaps with a broader one, the narrower-spectrum option can be
preferred on stewardship grounds.</li>
<li>
<strong>Non-overlapping credible intervals</strong> indicate a
clinically meaningful difference in coverage.</li>
</ul>
</div>
</div>
<div class="section level2">
<h2 id="sensible-defaults-which-can-be-customised">Sensible defaults, which can be customised<a class="anchor" aria-label="anchor" href="#sensible-defaults-which-can-be-customised"></a>
</h2>
<ul>
<li>
<code>simulations = 1000</code>: number of Monte Carlo draws</li>
<li>
<code>conf_interval = 0.95</code>: coverage interval width</li>
<li>
<code>combine_SI = TRUE</code>: count “I” and “SDD” as
susceptible</li>
</ul>
</div>
<div class="section level2">
<h2 id="practical-considerations">Practical considerations<a class="anchor" aria-label="anchor" href="#practical-considerations"></a>
</h2>
<ul>
<li>
<strong>First isolates only</strong>: always deduplicate using
<code><a href="../reference/first_isolate.html">first_isolate()</a></code> before running WISCA. Repeat isolates
introduce bias.</li>
<li>
<strong>Pathogen selection</strong>: consider filtering with
<code><a href="../reference/top_n_microorganisms.html">top_n_microorganisms()</a></code>. Including rare contaminants
(e.g. CoNS without clinical context) can distort estimates and may
artificially lower coverage (Cook <em>et al.</em>, 2022).</li>
<li>
<strong>Sample size</strong>: coverage estimates become reliable
with approximately 100+ isolates. For smaller datasets, consider pooling
data from multiple sites, but only after verifying that pathogen
distributions are sufficiently similar (Bielicki <em>et al.</em>,
2016).</li>
<li>
<strong>Culture request bias</strong>: WISCA is only as good as the
data it is based on. If cultures are selectively requested (e.g. only
after treatment failure), the dataset will be biased towards resistant
isolates. A robust culture policy is essential for reliable
estimates.</li>
</ul>
</div>
<div class="section level2">
<h2 id="limitations">Limitations<a class="anchor" aria-label="anchor" href="#limitations"></a>
</h2>
<ul>
<li>It assumes your data are representative of the patient population
you are treating</li>
<li>No direct adjustment for patient-level covariates, although these
can be passed onto the <code>syndromic_group</code> argument for
stratification</li>
<li>WISCA does not model resistance trends over time; for that, you
might want to use <code>tidymodels</code>, for which we <a href="https://amr-for-r.org/articles/AMR_with_tidymodels.html">wrote a
basic introduction</a>
</li>
</ul>
</div>
<div class="section level2">
<h2 id="summary">Summary<a class="anchor" aria-label="anchor" href="#summary"></a>
</h2>
<p>WISCA enables:</p>
<ul>
<li>
<strong>Empirical regimen comparison</strong>, answering the
clinicians actual question</li>
<li>
<strong>Syndrome-specific coverage estimation</strong>, stratifiable
by any clinical variable</li>
<li>
<strong>Fully probabilistic interpretation</strong>, with credible
intervals that honestly communicate uncertainty</li>
</ul>
<p>It is available in the <code>AMR</code> package via either:</p>
<div class="sourceCode" id="cb6"><pre class="downlit sourceCode r">
<code class="sourceCode R"><span><span class="fu"><a href="../reference/antibiogram.html">wisca</a></span><span class="op">(</span><span class="va">...</span><span class="op">)</span></span>
<span></span>
<span><span class="fu"><a href="../reference/antibiogram.html">antibiogram</a></span><span class="op">(</span><span class="va">...</span>, wisca <span class="op">=</span> <span class="cn">TRUE</span><span class="op">)</span></span></code></pre></div>
</div>
<div class="section level2">
<h2 id="references">References<a class="anchor" aria-label="anchor" href="#references"></a>
</h2>
<ol style="list-style-type: decimal">
<li>Hebert C, Ridgway J, Vekhter B, Brown EC, Weber SG, Robicsek A.
Demonstration of the weighted-incidence syndromic combination
antibiogram: an empiric prescribing decision aid. <em>Infect Control
Hosp Epidemiol.</em> 2012;33(4):381-388. <a href="https://doi.org/10.1086/664768" class="external-link uri">https://doi.org/10.1086/664768</a>
</li>
<li>Randhawa V, Sarwar S, Walker S, Elligsen M, Palmay L, Daneman N.
Weighted-incidence syndromic combination antibiograms to guide empiric
treatment of critical care infections: a retrospective cohort study.
<em>Crit Care.</em> 2014;18(3):R112. <a href="https://doi.org/10.1186/cc13901" class="external-link uri">https://doi.org/10.1186/cc13901</a>
</li>
<li>Bielicki JA, Sharland M, Johnson AP, Henderson KL, Cromwell DA.
Selecting appropriate empirical antibiotic regimens for paediatric
bloodstream infections: application of a Bayesian decision model to
local and pooled antimicrobial resistance surveillance data. <em>J
Antimicrob Chemother.</em> 2016;71(3):794-802. <a href="https://doi.org/10.1093/jac/dkv397" class="external-link uri">https://doi.org/10.1093/jac/dkv397</a>
</li>
<li>Cook A, Sharland M, Yau Y, Bielicki J. Improving empiric antibiotic
prescribing in pediatric bloodstream infections: a potential application
of weighted-incidence syndromic combination antibiograms (WISCA).
<em>Expert Rev Anti Infect Ther.</em> 2022;20(3):445-456. <a href="https://doi.org/10.1080/14787210.2021.1967145" class="external-link uri">https://doi.org/10.1080/14787210.2021.1967145</a>
</li>
</ol>
</div>
</main><aside class="col-md-3"><nav id="toc" aria-label="Table of contents"><h2>On this page</h2>
</nav></aside>
</div>
<footer><div class="pkgdown-footer-left">
<p><code>AMR</code> (for R). Free and open-source, licenced under the <a target="_blank" href="https://github.com/msberends/AMR/blob/main/LICENSE" class="external-link">GNU General Public License version 2.0 (GPL-2)</a>.<br>Developed at the <a target="_blank" href="https://www.rug.nl" class="external-link">University of Groningen</a> and <a target="_blank" href="https://www.umcg.nl" class="external-link">University Medical Center Groningen</a> in The Netherlands.</p>
</div>
<div class="pkgdown-footer-right">
<p><a target="_blank" href="https://www.rug.nl" class="external-link"><img src="https://amr-for-r.org/logo_rug.svg" style="max-width: 150px;"></a><a target="_blank" href="https://www.umcg.nl" class="external-link"><img src="https://amr-for-r.org/logo_umcg.svg" style="max-width: 150px;"></a></p>
</div>
</footer>
</div>
</body>
</html>