MII Implementation Guide Microbiology - Local Development build (v2027.0.0-ballot2) built by the FHIR (HL7® FHIR® Standard) Build Tools. See the Directory of published versions

Guidance for Researchers

Microbiological findings are the basis of infection diagnostics and of every statement about antimicrobial resistance. The central document this module represents is the microbiological report a medical laboratory produces: the detection, identification and characterisation of microorganisms and their properties. Because the module is broadly applicable, it does not prescribe consortium-specific scenarios.

Getting Started

If you are new to the MII Core Dataset, we recommend exploring the following resources:

1. Logical Models — Understanding Module Content

The Logical Models provide a high-level, technology-independent description of the data this module covers, split across three models:

• Report — the microbiological report that summarises the investigations it contains
• Investigation — the abstract base carrying what applies to every investigation: identification, status, patient, specimen, timings, and the investigation that triggered it
• Investigation types — the domain variants, grouped into culture, determination, quantitative tests and further properties

They let you understand the scope and structure of the available data without knowing the FHIR details. Their data types and cardinalities are not normative; the profiles settle those.

2. UML Diagrams — Visualizing Data Relationships

The UML Diagrams show how the elements relate to one another. They are particularly useful for:

• seeing how a report, its investigations and their specimens hang together
• identifying the elements relevant to a research question

3. Examples — Seeing FHIR Data in Practice

The Examples show samples of the FHIR resources, among them a three-step diagnostic chain from culture through identification to susceptibility testing, a colony count from a catheter tip, a Gram morphology reported together with its amount, and a positive molecular detection alongside the Ct value it was read from. They demonstrate:

• how microbiological data is represented in FHIR
• what a complete report looks like
• how the resources reference one another

4. Domain Guidance — What to Watch Out for in an Analysis

Four subjects decide whether microbiological data is read correctly:

• Profile Selection and Delimitation — which profile carries which statement. This matters most for negative findings, because "negative" is four different statements here: no growth in culture, a target sought and not found, an assessed pathogen that does not meet a resistance category, and an investigation that yielded no usable result at all. Counting them together produces a wrong denominator.
• Interpretation — how a susceptibility result is assessed, and the norm the assessment rests on. S, I and R are only comparable between sites when the underlying rule set is known, which is why it travels with the result.
• FHIR Profiles – Modelling Notes — how the steps of a diagnostic workflow are linked. Any query that follows an isolate from culture to susceptibility depends on these relations, and the three of them mean different things: triggeredBy a triggered follow-up investigation, derivedFrom a derived result, hasMember a panel.
• Specimen — what was measured from, and which module owns that description.

One pitfall is worth naming here, because it cannot be seen from a resource alone: a coarse microscopic morphology such as Gram-positive coccus is an organism code in SNOMED CT and therefore also a member of the organism value set used for identification. A query asking "which organisms were identified" must filter by profile or test code rather than by value[x] alone, or it will count morphological groups as species. The test codes themselves are disjoint, so no ambiguity remains once the filter is right.