The SSCP is not a summary of the CER. It is a publicly accessible regulatory document that will be read by your Notified Body assessor, by healthcare professionals, and by patients. This guide tells you what each section must contain — and what a former NB reviewer will flag if it is missing.
The SSCP is a mandatory public-facing document under EU MDR 2017/745. Understanding its purpose — and who will read it — is the foundation of writing a compliant one.
EU MDR 2017/745 Article 32 requires manufacturers of Class III medical devices and implantable devices to prepare, maintain, and make publicly available a Summary of Safety and Clinical Performance. The SSCP must be validated by the Notified Body before it is published in EUDAMED.
The SSCP applies to:
The SSCP does not apply to Class I, Class IIa, or non-implantable Class IIb devices.
The SSCP is unique among EU MDR technical documents because it must be written for two distinct audiences simultaneously:
Most SSCP failures at Notified Body review are in the lay summary section — manufacturers produce a technically accurate HCP section and then paste an abbreviated version of it as the lay summary, failing the "plain language" requirement of MDCG 2019-9.
The SSCP and CER are separate documents. They cover overlapping information but for different purposes, different audiences, and with different levels of detail. Conflating them — or writing the SSCP as an abbreviated CER — is the most common structural error in SSCP submissions.
| Dimension | CER (Clinical Evaluation Report) | SSCP (Summary of Safety & Clinical Performance) |
|---|---|---|
| Primary audience | Notified Body assessors and internal quality teams. Not publicly accessible. | Healthcare professionals AND lay persons (patients/users). Published in EUDAMED — publicly accessible. |
| Regulatory basis | EU MDR Annex XIV Part A + MDCG 2020-13 | EU MDR Article 32 + MDCG 2019-9 Rev.1 |
| Length | 50 to 300+ pages for Class III devices. Systematic literature review appendices add significant volume. | Typically 10 to 30 pages. Concise by design — must be accessible to lay readers. |
| Clinical evidence depth | Full systematic literature review with quality appraisal, data tables, comparator analysis, and gap analysis. All evidence cited in detail. | Summary of key clinical performance data with essential statistics. Lay summary explains results in plain language without full statistical detail. |
| Benefit-risk analysis | Comprehensive benefit-risk evaluation against state of the art. Detailed residual risk justification referencing risk management documentation. | Brief benefit-risk summary in plain language. States overall conclusion — does not contain the full analytical methodology. |
| PMCF information | Full PMCF Plan with gap-mapping, methodology, timeline, and success criteria. PMCF Report with detailed findings. | Summary of PMCF activities and key findings. States what PMCF is being conducted and why — not the full protocol or data. |
| Update trigger | Annual update expected for Class III. Any significant change or new clinical evidence affecting benefit-risk. | Updated when CER conclusions change, at CE certificate renewal, or following significant device change. NB validates each update before EUDAMED publication. |
| NB review | Full review of clinical evidence methodology, data quality, and conclusions as part of conformity assessment. | NB validates that SSCP content is consistent with approved technical file — it does not re-review the underlying clinical evidence. |
These are the specific checks a Notified Body assessor performs on each SSCP section, based on MDCG 2019-9 Rev.1 requirements and current NB practice. Each section includes the most common sources of NB queries and non-conformities.
The opening section must unambiguously identify the device as described in the technical file. Any discrepancy between SSCP device description, IFU labelling, and UDI registration will generate a query before the NB reads any clinical content.
"The first thing I checked was whether the UDI-DI in the SSCP matched the EUDAMED registration and the technical file exactly. A single character mismatch means the document references the wrong device — everything else becomes irrelevant until it is corrected."
The NB will check that intended purpose and indications in the SSCP are consistent with the IFU, the labelling, and the clinical evaluation scope. Any indication in the SSCP not covered by clinical evidence in the CER is a non-conformity. Contraindications must be complete — a contraindication listed in the IFU but absent from the SSCP is a direct safety issue.
"For implants, I paid close attention to age ranges and anatomical site specifications. An SSCP that stated 'indicated for all patients' when the clinical evidence only covered adults 18-75 years with BMI under 35 was an automatic query — the population gap had to be addressed in both the SSCP and the CER."
This section is the substance of the SSCP. It must present the clinical performance data supporting the device's claims — including key study designs, patient numbers, primary and secondary endpoints, and outcomes with relevant statistics. The NB will cross-reference this section against the CER to verify consistency.
"The most common problem in this section was cherry-picking positive results from studies without acknowledging negative findings or limitations. A clinical section that presented only 'device success rate 98%' without mentioning the 12% reoperation rate in the same study was flagged immediately as inconsistent with the full CER evidence base. The SSCP must be balanced."
The safety section must summarise adverse events, serious adverse device effects (SADEs), and any Field Safety Corrective Actions (FSCAs) from the post-market period. The NB will cross-reference this against EUDAMED vigilance reports and the PSUR. A safety section that states "no adverse events reported" when EUDAMED shows vigilance reports is a critical inconsistency.
"I always checked the SSCP safety section against the PSUR. If the PSUR reported a FSCA in the previous 12 months and the SSCP made no reference to it, that was a direct non-conformity. The SSCP is a public document — it cannot present a more favourable safety picture than the confidential regulatory documentation."
The benefit-risk section must state a clear, evidence-based conclusion on the acceptability of the device's benefit-risk profile. It must reference the state of the art for the indication and explain why residual risks are acceptable relative to clinical benefits. Vague language — "the benefit-risk profile is acceptable" without supporting rationale — will be queried.
The lay summary must present all key information from the SSCP in language accessible to a member of the public with no medical background. This is not a shortened version of the HCP sections — it requires active rewriting in plain language. MDCG 2019-9 recommends readability testing to Grade 8 reading level or equivalent.
"The simplest test I used: if a sentence contained a medical acronym that was not explained in brackets, the lay summary failed. If the sentence 'The device demonstrated 94.3% freedom from MACE at 24 months' appeared without explanation of what MACE is and what this means for a patient, it failed. Every technical term needs a plain-language equivalent in this section."
These are the recurring findings that Dr. Nikhil Khadabadi observed at TUV SUD — patterns that appear across manufacturers of all sizes and device types.
The most prevalent finding. SSCP was produced from the CER at one point in time but the CER was subsequently updated — and the SSCP was not. Indications in the SSCP differ from the IFU. Safety data in the SSCP differs from the PSUR. These are not minor issues — they signal a quality management failure and generate critical non-conformities.
Manufacturers produce a technically accurate HCP section then copy-shorten it for the lay summary — keeping medical terminology, statistical language, and acronyms. MDCG 2019-9 requires plain language accessible to non-medical readers. Submitting a technical document as a lay summary is a routine formal query at all major Notified Bodies.
Many SSCPs contain a single sentence: "PMCF activities are ongoing in accordance with the PMCF Plan." MDCG 2019-9 requires the SSCP to summarise the nature of PMCF activities, what they are designed to investigate, and key findings to date. A generic PMCF sentence will be queried.
SSCPs that present device success rates and primary endpoint results without reporting secondary endpoint failures, reoperation rates, or known adverse event profiles will be flagged as incomplete or misleading. The NB has access to the full CER — discrepancies between SSCP clinical data and CER data tables are always caught.
The SSCP (Summary of Safety and Clinical Performance) is a publicly accessible document required under EU MDR 2017/745 Article 32 for Class III and implantable devices. It summarises the device's safety and clinical performance evidence for both healthcare professionals and lay persons (patients). The SSCP is validated by the Notified Body and published in EUDAMED. It must be updated at each renewal or when new clinical evidence changes the benefit-risk conclusion.
An SSCP is required for: all Class III medical devices, and all implantable devices (including Class IIb implantables). It is not required for Class I, Class IIa, or non-implantable Class IIb devices. The SSCP must be submitted as part of the technical file and published in EUDAMED once the Notified Body has validated it. For legacy MDD-certified devices, an SSCP is required at first EU MDR renewal.
The CER (Clinical Evaluation Report) is the comprehensive internal document — typically 50 to 300+ pages — that systematically evaluates all clinical evidence. It is not publicly accessible. The SSCP is a shorter public-facing summary (typically 10 to 30 pages) drawn from the CER, written for both healthcare professionals and lay persons (patients). They are separate documents with different audiences, different depths of content, and different update requirements. Writing the SSCP as an abbreviated CER is the most common SSCP non-conformity.
The SSCP must be updated when new clinical evidence changes the benefit-risk conclusion (typically triggering an annual CER update for Class III), at each CE certificate renewal, or when a significant device change is made. For Class III devices with annual PSUR requirements, the SSCP should be reviewed annually even if not formally updated. Every update must be validated by the Notified Body before the revised version is published in EUDAMED.
The SSCP draws from the CER and the PMCF programme. These guides cover each in detail.
Dr. Nikhil Khadabadi reviewed SSCP submissions at TUV SUD. Eclevar's medical writing team produces SSCPs that are structured for NB validation — not rewritten after queries. First submission acceptance rate significantly higher than industry average.
