相关数据的评估与分析

在临床评价计划（CEP）中，“相关数据的评估与分析”的内容是阐述方法学的核心部分。它旨在清晰、系统地规划在临床评价过程中，将如何识别所有与待评价医疗器械相关的关键数据，如何严格评估这些数据的适用性、质量和科学有效性，以及最终如何综合分析这些经过评估的数据，以得出关于器械安全性、性能和受益风险比的结论。本章节所描述的计划和方法，是确保整个临床评价过程科学、客观、可重复并符合法规要求的关键。

撰写本章节时，应重点突出评估与分析过程的系统性、客观性和预设标准。

明确“相关数据”的范畴及识别策略

首先，CEP需要清晰界定对于本次临床评价而言，哪些数据被视为“相关数据 (pertinent data)”。这一定义应紧密围绕先前章节已明确的临床评价范围、待评价器械的特性、预期用途以及需要通过临床证据支持的具体宣称。

在此基础上，应简要重申计划从哪些主要途径获取这些相关数据。正如在CEP其他部分（如“制造商持有数据”、“文献检索策略”等）已有所提及，这些数据来源通常包括（但不限于）：

• 由制造商产生和持有的临床前数据（如台架测试、动物实验、生物相容性数据等）；
• 针对待评价器械或等同/类似器械的已发表科学文献；
• 制造商申办的临床研究数据（如有）；
• 上市后监测（PMS）和上市后临床跟踪（PMCF）数据；
• 来自等同器械的临床数据（如适用）；
• 风险管理过程的输出。

本章节应概述识别所有这些相关数据的具体策略和方法，特别是对于文献数据，应提及将采用系统性的文献检索和筛选方法。明确的纳入和排除标准对于确保数据筛选的客观性和一致性至关重要，应在此处概述或引用。

规划数据的评估过程 (Data Appraisal)

数据评估是临床评价的核心步骤之一，旨在判断所识别数据的质量及其对评价结论的贡献程度。CEP必须详细规划这一评估过程。

预设的评估标准： 强调将采用一套预先定义的、客观的评估标准来系统地评价所有收集到的数据。这些标准应确保评估过程的透明性和一致性，并尽可能减少偏倚。评估标准通常会覆盖以下几个维度：

• 相关性 (Relevance)： 评估数据是否直接关联到待评价器械（或其等同/类似器械）、其预期用途、目标人群、以及需要评价的特定安全性和性能终点。
• 质量与方法学可靠性 (Quality and Methodological Soundness)： 深入考察数据的产生方法，包括研究设计（如随机对照试验、观察性研究、病例系列等）、实施过程、数据收集的完整性、统计分析的适当性，以及潜在的偏倚来源（如选择偏倚、信息偏倚、混杂因素等）、随机误差的影响、信息披露的充分性等。
• 对符合性论证的贡献度 (Contribution to Demonstration of Conformity)： 根据数据的相关性和质量，判断其在证明器械安全性和性能方面的权重和价值。

评估标准的统一应用： 强调这些预设标准将统一应用于所有来源的数据（包括制造商持有的数据、文献数据、临床研究数据等），并且对正面（支持性）和负面（不支持或提示风险）的数据都将给予同等权重的客观评估。

数据的差异化使用： 认识到并非所有数据都能同等适用于所有评价目的。例如，某些研究设计上存在缺陷、不足以证明器械性能的数据，可能仍包含对安全性分析有价值的信息，反之亦然。CEP应说明在评估时会考虑这种差异化的适用性。

规划数据的分析策略 (Data Analysis)

在数据经过严格评估并被认为适用后，CEP必须规划如何对这些数据进行综合分析，以形成关于器械安全性、性能和受益风险的整体结论。

• 证据的综合与整合： 说明计划如何将来自不同来源（如临床前、临床研究、文献、PMS等）的、经过评估的证据进行系统性地整合与综合，以构建一个全面、连贯的证据体系。
• 临床证据缺口的识别与处理： 阐述在数据分析过程中，将如何识别临床证据中可能存在的缺口（即某些方面的安全或性能未能得到充分数据支持）或不同数据来源之间的不一致性。同时，应说明计划如何应对这些情况，例如，通过调整PMCF计划来进一步收集数据、在风险管理中重新评估相关风险，或在结论中明确承认存在的局限性。
• 受益风险分析： 这是数据分析的最终目标之一。CEP应明确指出数据分析将导向对待评价器械受益风险特征的全面评估，并结合当前技术发展水平（SOTA）和可替代治疗方案，判断其受益风险比是否可接受。
• 与通用安全和性能要求（GSPR）的符合性论证： 数据分析的计划应直接指向证明器械符合MDR中相关GSPR（特别是GSPR 1和GSPR 8）的要求。

临床证据水平的合理性论证

根据MDR法规第61条的要求，制造商必须明确并论证为证明其器械符合相关GSPR所必需的临床证据水平。CEP的本部分应前瞻性地阐述这一点。

声明证据水平： 清晰说明根据待评价器械的特性（如创新性、风险等级、技术成熟度等）及其预期用途，制造商认为何种水平和类型的临床证据是充分且适当的。

论证理由： 提供支持这一证据水平判断的详细理由。例如，对于成熟技术、低风险器械，可能更多依赖文献数据和临床前数据；而对于高风险或创新器械，则可能需要专门的上市前临床研究数据。

评价目标概述： 可以概述本次临床评价旨在通过所规划的证据水平来确认哪些具体方面，例如：

• 器械是否按制造商的预期和规定发挥性能。
• 器械的风险特征是否与可比器械或当前SOTA相当。
• 器械对于患者和使用者是否安全。
• 器械的整体受益风险比是否可接受。

过程的迭代性与最终报告

应明确数据评估和分析过程可能具有迭代性。在评估过程中发现的新信息或提出的新问题，可能需要调整临床评价的范围、优化计划，甚至补充额外的数据收集和分析工作。

最后，CEP应简要说明，所有相关数据的识别、筛选、评估、分析过程及其最终结论都将被完整、透明地记录在临床评价报告（CER）中。CER将基于这些分析，全面阐述器械的安全性、临床性能、临床受益，并就是否已获得充分的临床证据来支持器械的预期用途和符合性做出最终判断。

---

Within the Clinical Evaluation Plan (CEP), the “Appraisal and Analysis of Pertinent Data” contents are core methodological information. It aims to clearly and systematically plan how all pertinent clinical data related to the device under evaluation will be identified during the clinical evaluation process, how the suitability, quality, and scientific validity of this data will be rigorously assessed, and finally, how this appraised data will be collectively analyzed to draw conclusions regarding the device’s safety, performance, and benefit-risk profile. The plans and methods described in this chapter are key to ensuring that the entire clinical evaluation process is scientific, objective, repeatable, and compliant with regulatory requirements.

When drafting this chapter, the emphasis should be on the systematic nature, objectivity, and pre-defined criteria of the appraisal and analysis process.

Defining the Scope of “Pertinent Data” and Identification Strategy

Firstly, the CEP needs to clearly define what constitutes “pertinent data” for this specific clinical evaluation, closely aligning with the previously established clinical evaluation scope, the characteristics of the device under evaluation, its intended use, and the specific claims that need to be supported by clinical evidence.

Building on this, a brief reiteration of the main pathways planned for acquiring this pertinent data should be provided. As mentioned in other parts of the CEP (such as “Data Generated and Held by the Manufacturer,” “Literature Search Strategy,” etc.), these data sources typically include (but are not limited to):

• Preclinical data generated and held by the manufacturer (e.g., bench testing, animal studies, biocompatibility data).
• Published scientific literature pertaining to the device under evaluation or equivalent/similar devices.
• Data from manufacturer-sponsored clinical investigations (if any).
• Post-Market Surveillance (PMS) and Post-Market Clinical Follow-up (PMCF) data.
• Clinical data from equivalent devices (if applicable).
• Outputs from the risk management process.

This part should outline the specific strategy and methods for identifying all such pertinent data. Particularly for literature data, it should mention that systematic literature search and screening methodologies will be employed. Clear inclusion and exclusion criteria are crucial for ensuring the objectivity and consistency of data screening and should be outlined or referenced here.

Planning the Data Appraisal Process

Data appraisal is one of the core steps of clinical evaluation, aimed at judging the quality of the identified data and its contribution to the evaluation conclusions. The CEP must detail this appraisal process.

Pre-defined Appraisal Criteria: Emphasize that a pre-defined, objective set of appraisal criteria will be used to systematically evaluate all collected data. These criteria should ensure the transparency and consistency of the appraisal process and minimize bias as much as possible. Appraisal criteria typically cover the following dimensions:

• Relevance: Assessing whether the data directly pertains to the device under evaluation (or its equivalent/similar devices), its intended use, target population, and the specific safety and performance endpoints under evaluation.
• Quality and Methodological Soundness: Thoroughly examining the methods used to generate the data, including study design (e.g., randomized controlled trials, observational studies, case series), conduct, completeness of data collection, appropriateness of statistical analysis, and potential sources of bias (e.g., selection bias, information bias, confounding factors), the impact of random error, adequacy of information disclosure, etc.
• Contribution to Demonstration of Conformity: Based on the relevance and quality of the data, judging its weight and value in demonstrating the safety and performance of the device.

Uniform Application of Appraisal Criteria: Stress that these pre-defined criteria will be uniformly applied to all data sources (including manufacturer-held data, literature data, clinical investigation data, etc.), and that both favorable (supportive) and unfavorable (non-supportive or indicative of risk) data will be objectively appraised with equal weight.

Differential Use of Data: Recognize that not all data may be equally suitable for all evaluation purposes. For instance, data from studies with design flaws that are insufficient to demonstrate device performance might still contain information valuable for safety analysis, and vice versa. The CEP should state that this differential applicability will be considered during appraisal.

Planning the Data Analysis Strategy

Once data has been rigorously appraised and deemed suitable, the CEP must plan how this data will be collectively analyzed to form overall conclusions about the device’s safety, performance, and benefit-risk profile.

• Evidence Synthesis and Integration: Explain how appraised evidence from various sources (e.g., preclinical, clinical investigations, literature, PMS) will be systematically integrated and synthesized to build a comprehensive and coherent body of evidence.
• Identification and Handling of Clinical Evidence Gaps: Describe how, during data analysis, any potential gaps in clinical evidence (i.e., certain aspects of safety or performance not adequately supported by data) or inconsistencies between different data sources will be identified. Also, outline the planned approach to address these situations, for example, by adjusting PMCF plans to gather further data, reassessing relevant risks in risk management, or clearly acknowledging limitations in the conclusions.
• Benefit-Risk Analysis: This is one of the ultimate goals of data analysis. The CEP should clearly state that the data analysis will lead to a comprehensive assessment of the benefit-risk profile of the device under evaluation, and whether this profile is acceptable when considered against the current state-of-the-art (SOTA) and available alternative treatments.
• Demonstration of Conformity with General Safety and Performance Requirements (GSPRs): The data analysis plan should be directly aimed at demonstrating the device’s conformity with the relevant GSPRs of the MDR (particularly GSPR 1 and GSPR 8).

Justification for the Level of Clinical Evidence

As per MDR Article 61, the manufacturer must specify and justify the level of clinical evidence necessary to demonstrate conformity with the relevant GSPRs for their device. This part of the CEP should prospectively address this.

Declaration of Evidence Level: Clearly state what level and type of clinical evidence is considered sufficient and appropriate by the manufacturer, based on the characteristics of the device under evaluation (e.g., novelty, risk class, technological maturity) and its intended use.

Rationale for Justification: Provide a detailed rationale supporting this judgment of evidence level. For example, for mature technology, low-risk devices, greater reliance might be placed on literature data and preclinical data; whereas for high-risk or innovative devices, dedicated pre-market clinical investigation data might be necessary.

Overview of Evaluation Objectives: An outline can be provided of what specific aspects this clinical evaluation aims to confirm through the planned level of evidence, for example:

• Whether the device performs as intended and specified by the manufacturer.
• Whether the risk profile of the device is comparable to that of similar devices or the current SOTA.
• Whether the device is safe for both patients and users.
• Whether the overall benefit-risk profile of the device is acceptable.

Iterative Nature of the Process and Final Reporting

It should be clarify that the data appraisal and analysis process can be iterative. New information discovered or new questions raised during the appraisal may necessitate adjustments to the clinical evaluation scope, refinements to the plan, or even supplementary data collection and analysis.

Finally, the CEP should briefly state that the complete process of identifying, screening, appraising, and analyzing all pertinent data, along with the final conclusions, will be fully and transparently documented in the Clinical Evaluation Report (CER). The CER will, based on these analyses, comprehensively describe the device’s safety, clinical performance, clinical benefits, and make a final judgment on whether sufficient clinical evidence has been obtained to support the device’s intended use and conformity.