What are the main types of clinical notes, and how do you write them well?

Clinical notes are the written or structured records clinicians create to document patient encounters, observations, decisions, and care plans. They serve two purposes at once: a legal record of the care provided, and a communication tool that supports continuity across the wider care team. A note can be as brief as a two-line ward round update or as detailed as a full discharge summary, but each functions as the authoritative account of clinical activity that other clinicians, commissioners, and courts may rely on.

Why do clinical notes matter?

Accurate documentation supports continuity of care, informed decisions, medico-legal protection, and, in many systems, coding and reimbursement accuracy. When notes are incomplete or delayed, clinicians picking up a patient mid-episode lack the context to make safe decisions, and referrals get rejected. NHS England's guidance on documentation identifies documentation burden as a significant driver of clinician burnout, with discharge summaries alone contributing substantially to that load.

What are the main types of clinical notes?

• SOAP notes (Subjective, Objective, Assessment, Plan) are the most widely used structured format across primary and secondary care, valued for a logical structure that's easy to scan and act on.

• POMR notes organise documentation around a patient's active problem list rather than chronological entries, which helps with complex patients on multiple long-term conditions.

• Progress notes form the ongoing record of a patient's status during ward rounds, follow-ups, and admissions.

• Discharge summaries cover the presenting diagnosis, investigations, treatment, and follow-up instructions, and are among the most time-consuming documents clinicians produce. A large-scale study at a Dutch academic hospital, published in eBioMedicine, compared 292 paired physician-written and AI-generated summaries and found no difference in overall quality, though the AI versions scored lower on completeness and higher on conciseness.

• Referral letters summarise history, presentation, and the specific question for the receiving clinician.

• Patient letters are written directly to patients in plain language, distinct from clinical notes written for clinicians.

• Nursing notes capture observations and care delivery from a nursing perspective, often the most frequent record of a patient's condition.

• Operative notes document technique, findings, and complications immediately after a procedure, and carry significant medico-legal weight.

Common formats for clinical documentation

Free-text notes offer flexibility when a situation doesn't map neatly onto a template, but carry well-known limits for data extraction, interoperability, and consistency. A note perfectly legible to its author can be hard for anyone else to search or code.

Structured and templated notes use predefined fields to keep documentation consistent. A multicentre study at Radboud University Medical Center in the Netherlands, published in the Journal of Medical Systems, compared 144 unstructured against 144 structured notes and found structured documentation raised mean quality scores from 64.35 to 77.2 out of 100, alongside better clarity and conciseness.

Structured notes also support clinical coding, which depends on consistent, extractable fields. The trade-off is that rigid templates can feel constraining in complex cases, and there's a documented risk of "note bloat," where templates encourage over-documentation while genuine clinical reasoning gets crowded out.

Voice-dictated notes have featured in documentation for decades, particularly in secondary care, where clinicians dictate for transcription by medical secretaries. This introduces a delay between encounter and finished note, and depends on transcription accuracy. It remains common in radiology and pathology.

AI-generated notes use ambient voice technology: an ai medical assistant listens to a consultation in real time and drafts a structured note from the conversation, rather than requiring the clinician to narrate to a device.

The Dutch eBioMedicine study above found this can match physician-written quality at scale in a fully integrated setting, though the authors note this depended on clinician review and specialty-specific refinement, and that automated quality metrics alone don't reliably capture clinical safety, underscoring the need for a human check before anything reaches the record.

How does clinical documentation differ across care settings?

In primary care, GPs document high volumes of short encounters, often ten to fifteen minutes each, using SOAP notes and structured templates, with documentation burden widely cited as a driver of burnout.

Secondary care spans more note types, including ward round notes, multidisciplinary records, and discharge summaries, often across legacy systems that don't talk to each other, with many clinicians contributing to one record.

Mental health services carry extra considerations: notes may include risk assessments and sensitive disclosures, and some services use BIRP (Behaviour, Intervention, Response, Plan) instead of SOAP.

In private and specialist care, patient letters and referral correspondence tend to carry more weight, and detailed notes also support insurance and billing.

Best practice: what good clinical documentation looks like?

• Write it as soon as you can. A note written hours or days later may not reflect what was actually observed or decided at the time.

• Be accurate, objective, and specific. Avoid vague descriptors like "seems unwell" or "doing better." Standardised terminology, including SNOMED and ICD codes, supports clarity and reuse.

• Use consistent structure and templates, matched to the actual clinical workflow. Structured guidance has been shown to improve both confidence and efficiency in note-writing at every career stage.

• Avoid non-standard abbreviations, which create real risk in handovers, since shorthand obvious in one specialty can be ambiguous in another.

• Write for the whole care team, not just the clinician who wrote it, including GPs and out-of-hours colleagues who'll read it later.

• Amend, don't overwrite. Corrections should be dated additions, preserving the original entry and the timeline's integrity.

These principles apply whether a note is handwritten, typed, dictated, or drafted by an AI assistant. Ambient voice technology and AI-assisted documentation are increasingly part of clinical practice, but the standards that define a good clinical note remain grounded in the same fundamentals: accuracy, clarity, timeliness, and a genuine commitment to the patient's safety and continuity of care.

How AI is changing clinical documentation

Rather than writing notes retrospectively after the patient has left, often under time pressure, clinicians using AI-assisted tools can capture documentation in real time during the consultation, with the AI generating a structured draft from the conversation.

Prospective evaluation of AI-generated hospital course summaries has assessed both safety and burden-reduction potential in real clinical settings, with early findings suggesting comparable quality to physician-written summaries, though the evidence base is still developing. Research has also shown that large language models can give quality improvement feedback on clinical notes, flagging documentation gaps that might otherwise go unaddressed.

Notes drafted in real time are less likely to suffer from recall errors, and a clinician's attention can stay on the patient rather than the screen. These tools produce drafts, not final records, so clinician review and sign-off remain necessary, and output quality depends heavily on transcription accuracy and how well the template fits the encounter.

What should you check for data security and compliance?

Clinical notes are among the most sensitive personal data that exists. In the UK, this falls under UK GDPR, with the Information Commissioner's Office publishing specific guidance on how it applies to AI systems; in the EU, equivalent GDPR obligations apply directly. Before adopting an AI documentation tool, check for ISO 27001 certification, whether it meets UK MHRA or EU Medical Device Regulation requirements where relevant, what access controls limit notes to those with legitimate clinical need, and how data is retained, deleted, and audited on an ongoing basis rather than only when a tool is first adopted.

These principles apply whether a note is handwritten, typed, dictated, or drafted by an AI assistant. Ambient voice technology is increasingly part of clinical practice, but what defines a good note hasn't changed: accuracy, clarity, timeliness, and a genuine commitment to patient safety and continuity of care.

Frequently asked questions

▶ What are clinical notes and what are they used for?

They're the written record of patient encounters, decisions, and care plans, serving both as a legal record and a communication tool for the care team.

▶ What are the main types of clinical notes?

SOAP notes, POMR, progress notes, discharge summaries, referral letters, patient letters, nursing notes, and operative notes, each serving a distinct purpose.

▶ What is the SOAP note format and when is it used?

SOAP stands for Subjective, Objective, Assessment, Plan. It's the most widely used structured note format across primary and secondary care. The Subjective section captures the patient's reported symptoms and concerns, the Objective section records measurable clinical findings, the Assessment section contains the clinician's interpretation and working diagnosis, and the Plan section sets out the intended management. SOAP notes are particularly common in general practice and outpatient settings, though clinicians use them across virtually all clinical disciplines.

▶ What's the difference between free-text notes and structured notes?

Free-text offers flexibility but limits data reuse. Structured notes use predefined fields; a Dutch multicentre study found this raised quality scores from 64.35 to 77.2 out of 100, though rigid templates risk "note bloat" in complex cases.

▶ How does AI-assisted documentation work?

It listens to a consultation and drafts a structured note in real time. A Dutch academic hospital study found AI-generated summaries matched physician quality overall, but still need clinician review before entering the record.

▶ What are the best practices for writing clinical notes?

Good clinical notes should be written as close to the encounter as possible, use precise and factual language grounded in clinical observation, follow consistent structures and templates, avoid non-standard abbreviations, and be written so that colleagues outside the original specialty can understand and act on them. When a note requires correction, the appropriate approach is to add a dated amendment rather than overwrite or delete the original entry.

▶ Why does documentation burden contribute to clinician burnout?

Producing complete records takes substantial time outside patient-facing hours; discharge summaries alone are a significant contributor.

▶ What compliance and data security considerations apply to clinical notes?

Clinical notes contain some of the most sensitive personal data that exists. In the UK and European Union, clinical documentation falls under the General Data Protection Regulation, which requires that patient data is processed lawfully, stored securely, and not transferred outside approved jurisdictions without appropriate safeguards. When evaluating AI documentation tools, organisations should check whether the tool holds relevant certifications such as ISO 27001, whether it meets medical device regulation requirements, what access controls are in place, and how data is retained, deleted, and audited.

▶ How do clinical notes differ across care settings?

In primary care, general practitioners document high volumes of short encounters using SOAP notes and structured templates, with documentation burden widely cited as a driver of burnout. In hospitals, documentation spans a wider range of note types across multiple systems, often contributed to by many clinicians. Mental health services carry specific considerations around risk assessments, sensitive disclosures, and patient access to records. In private and specialist care, detailed patient letters and referral correspondence carry more weight, and comprehensive documentation also supports insurance and billing processes.

▶ Is AI-generated clinical documentation reliable enough to use in practice?

Evidence is growing but tool- and setting-dependent. Look for real-world, EHR-integrated evidence rather than vendor-reported figures alone, and keep clinician sign-off in the workflow.