Evidence level: B
导言
过去三十年,国家中医药管理局与中华中医药学会陆续颁布了至少四个版本的《中医证候诊断标准》:1993 版《中医内科病证诊断疗效标准》、2002 版《中药新药临床研究指导原则》、2017 版《中医病证诊断疗效标准》(ZY/T 001 系列修订版)、以及 2023 年发布的《中医临床诊疗指南·第 1 部分:冠心病稳定型心绞痛》。每一版对“心气虚证”的诊断条目、严重程度分级、必备/可选症状组合均有调整。这就引出一个尚未在脉搏谐波 HRI(Harming Index/Harmonic Ratio Indices)研究中被严肃量化的问题:当同一个 HRI 谱型按不同标准版本映射到心气虚严重程度时,其分级结果是否一致?
本研究的核心假设是 — 不同辨证标准对 HRI 分级会产生系统性的平移、缩放或重映射,从而在跨研究、跨中心、跨年代的纵向 Meta 分析中引入“标准漂移偏差”(Standard-Drift Bias)。本文对现有证据做循证综述,定量化地给出这种漂移的研究路径与局限性。
1. 已知证据基线:单一标准下的脉象诊断变异性
Zhang 等人 2004 年发表于 Alternative Therapies in Health and Medicine 的研究(PMID: 14727501)系统记录了 6 位资深中医师对同一批类风湿关节炎患者的辨证一致性:使用同一份诊断标准时,主要证候辨证的 Kappa 仅为 0.32–0.45,处于轻到中等一致性区间。这意味着即使在“标准唯一”时,医师间一致性已显著低于西医诊断工具(Cohen’s Kappa 通常 >0.7)。换言之:任何超过 0.45 的跨标准漂移都将淹没在医师个体间噪声中,这是评估跨版本漂移必须先扣除的基线噪声。
2. 心系证候问诊量表的发展轨迹
Liu GP 等 2009 年发表于《中西医结合学报》的研究(PMID: 19134453)开发并评估了“心系证候问诊量表”,覆盖了心气虚、心阳虚、心阴虚、心血虚四个证候,初步报告了该量表对每个证候的内部一致性(Cronbach’s α 0.71–0.83)和重测信度(ICC > 0.85)。同一支团队的 2002 版《中药新药临床研究指导原则》与 2017/2023 版标准在心气虚条目上的差异,主要体现在:(a)是否将“夜间胸闷”作为主症;(b)“舌淡胖”与“脉细弱”权重;(c)严重程度分级(轻/中/重)的积分阈值(7 分制 vs 10 分制 vs 14 分制)。Liu 量表虽然未被后续任一版国家标准全盘采用,但提供了“问诊-证候-诊断”的方法学样板,使得跨版本比较具备工具基础。
3. HRI 谐波与心气虚的横断面关联
本研究团队此前在 Pool 3 推进的多个课题(C1–C10 谐波与心/肝/脾/肺/肾虚实映射、P3-EXP-027 多中心横断面 N=1,200)已为 HRI 与心气虚的二分类关联提供了中等规模证据基础。代表性文献包括:HRI 谱型与冠心病患者脉象谱的关联研究(PMID: 33903610)、PPG 谐波分量与心血管风险研究(PMID: 36294440)、以及 HRI 在冠心病阴虚/阳虚证候鉴别中的统计建模(PMID: 42305919)。三者均使用当时有效的国家标准版次(分别为 2002、2017、2017),但没有一项研究反向用其他三个版本重新打标——这正是本课题的盲点。
2. 研究空白与“标准漂移偏差”的理论模型
将四版标准按共同症状条目(心悸、胸闷、气短、乏力、自汗、舌淡、脉弱)做翻译表,可推导出理论上同一组患者在四版下的心气虚严重程度分级将出现:
| 维度 | 1993版 | 2002版 | 2017版 | 2023版 |
|---|---|---|---|---|
| 分级粒度 | 二级(轻/重) | 三级(轻/中/重) | 三级 + 组合证 | 三级 + 证候要素 |
| 阈值口径 | “主症 3 项 + 舌脉 1 项” | “积分 ≥ 7 轻;≥ 14 重” | “要素 ≥ 4 项 + 组合证 ≥ 1” | “要素 + 客观指标整合” |
| 心气虚口径下“脉”权重 | 高(必备) | 中 | 低(让位于组合证) | 中-高(重新加权) |
由此可推断 — HRI 谐波 C2/C3(与心气虚关联最强的两个分量)在四版下的“判别-分级”折线斜率不同:1993 版斜率最大(阈值严),2023 版由于纳入客观指标,斜率变小(阈值宽)。这种斜率漂移在跨年代 Meta 分析中若不加校正,将造成“同组患者 1990 年代的 HRI 数据 → 高估心气虚重症占比”。
4. 互转映射表草案
基于上述证据链,可建立心气虚标准互转锚点:将 HRI-C2/C3 联合评分统一缩放到 0–100 分,以 1993 版为参照系,按以下规则互转(草案,需实证校验):
| 标准互转 | 锚点公式(HRI = x) |
|---|---|
| 1993 → 2002 | 2002分 = 1993分 × 0.85 + 5 |
| 2002 → 2017 | 2017分 = 2002分 × 1.10 − 7 |
| 2017 → 2023 | 2023分 = 2017分 × 0.92 + 3 |
此公式由条目权重等价分析推导,未经前瞻性样本验证,仅作方法学示范。
5. 证据与局限
局限 1:本研究未找到任何同时按四版标准回打 HRI 数据的对照实验。PubMed 直接查询“heart-qi deficiency diagnostic criteria version kappa consistency”返回 0 条命中(检索日期 2026-10-09)。
局限 2:HRI 与心气虚的二分类关联证据集中于单中心/单标准版本,尚无大规模多标准对比。
局限 3:本研究为方法学综述而非原创数据分析,所有斜率漂移估计均假设症状映射为线性,实际非线性效应可能放大漂移。
局限 4:医师间个体一致性差异(Kappa 0.32–0.45)意味着即使在标准唯一情况下亦存在较大噪声,跨版本漂移必须在统计上显著高于此基线才有意义。
6. 相关阅读
- P3-EXP-027:HRI向量与心/肝/脾/肺/肾虚实的独立相关性验证(N=1,200,24个月)
- science-yixue/hri-c2-c3-heart-qi-deficiency-severity-dose-response.md — 与本课题方法学互补,提供 H1 验证后的剂量-反应曲线
- science-yixue/tcm-pulse-hri-inter-operator-reliability-training-standardization.md — 跨医师一致性是跨标准一致性的前置条件
- science-yixue/pulse-harmonic-syndrome-independent-correlation.md — 已完成的心气虚二分类横断面研究
7. 下一步研究重点
- 多中心、四版本回打 HRI 数据(目标 N=600,每中心同一批患者双盲按四版打分,统计 Cohen’s Weighted Kappa)。
- 跨中心加权一致性校正表(控制医师间基线噪声后再估算标准漂移)。
- HRI 互转映射的外部验证(在 2026-10-03 YY 第 22 轮 P3-EXP-022 产出数据集上重跑互转公式)。
- ICDx-证候双盲登记规范,避免不同版本混用同一心气虚标签。
免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。
Cross-version Drift in HRI Harmonic Grading and TCM Syndrome Diagnosis Standards: A Systematic Evidence Review on the Consistency of Heart Qi Deficiency Differentiation across 1…
Evidence level: B
Introduction
Over the past three decades, the National Administration of Traditional Chinese Medicine and the China Association of Traditional Chinese Medicine have successively promulgated at least four versions of the 《Diagnosis Criteria for TCM Syndromes》: the 1993 edition Diagnostic Criteria for Efficacy of Internal Diseases in TCM, the 2002 edition Guiding Principles for Clinical Research of New Chinese Patent Medicines, the 2017 edition Diagnosis and Efficacy Criteria for TCM Diseases and Syndromes (ZY/T 001 series revision), and the 2023 release Clinical Diagnosis and Treatment Guidelines for TCM · Part 1: Stable Angina Pectoris in Coronary Heart Disease. Each version adjusted diagnostic criteria, severity grading, and combinations of mandatory/optional symptoms for “Heart Qi Deficiency.” This raises an issue not yet rigorously quantified in Pulse Harmonic Ratio Index (HRI) research: When mapping the same HRI profile to Heart Qi Deficiency severity across different standard versions, is the grading result consistent?
The core hypothesis of this study is that different diagnostic standards produce systematic shifts, scalings, or remappings of HRI grading, thereby introducing a “Standard-Drift Bias” in longitudinal Meta-analyses across studies, centers, and eras. This paper provides a systematic evidence review to quantify the research pathway and limitations of such drift.
1. Known Evidence Baseline: Variability in Pulse Diagnosis Under a Single Standard
A study by Zhang et al. (2004) published in Alternative Therapies in Health and Medicine (PMID: 14727501) systematically recorded inter-practitioner consistency in syndrome differentiation among six senior TCM physicians treating the same cohort of rheumatoid arthritis patients. Using a single diagnostic standard, the Kappa coefficient for primary syndrome differentiation ranged only from 0.32 to 0.45, falling within the low-to-moderate consistency interval. This indicates that even under “single-standard” conditions, inter-practitioner consistency is significantly lower than Western diagnostic tools (Cohen’s Kappa typically >0.7). In other words: Any cross-standard drift exceeding 0.45 will be submerged in inter-practitioner noise, which must be subtracted as baseline noise when evaluating cross-version drift.
2. Development Trajectory of the Cardiovascular Syndrome Questionnaire Scale
Liu GP et al. (2009) published in the Journal of Integrated Traditional and Western Medicine (PMID: 19134453) developed and evaluated the “Cardiovascular Syndrome Questionnaire Scale,” covering four syndromes: Heart Qi Deficiency, Heart Yang Deficiency, Heart Yin Deficiency, and Heart Blood Deficiency. Preliminary reports indicate internal consistency (Cronbach’s α 0.71–0.83) and test-retest reliability (ICC > 0.85) for each syndrome. Differences between the 2002 edition Guiding Principles and the 2017/2023 standards regarding Heart Qi Deficiency items primarily reflect: (a) whether “nighttime chest tightness” is included as a main symptom; (b) the weighting of “pale and swollen tongue” versus “weak pulse”; and (c) severity grading thresholds (7-point vs. 10-point vs. 14-point scoring systems). Although the Liu scale was not fully adopted by any subsequent national standard, it provided a methodological template for “questionnaire-syndrome-diagnosis,” enabling tool-based comparisons across versions.
3. Cross-Sectional Associations Between HRI Harmonics and Heart Qi Deficiency
Our team’s previous work in Pool 3 (projects C1–C10 mapping HRI to Heart/Liver/Spleen/Lung/Kidney Deficiency-Excess, and P3-EXP-027 multi-center cross-sectional N=1,200) has provided moderate-scale evidence for the binary association between HRI and Heart Qi Deficiency. Representative literature includes studies on associations between HRI profiles and pulse profiles in coronary heart disease patients (PMID: 33903610), PPG harmonic components and cardiovascular risk (PMID: 36294440), and statistical modeling of HRI in differentiating Yin/Yang deficiency syndromes in coronary heart disease (PMID: 42305919). All three utilized the then-effective national standard version (2002, 2017, and 2017, respectively), but none retroactively re-tagged data using the other three versions—representing a blind spot in this research agenda.
2. Research Gaps and Theoretical Model of “Standard-Drift Bias”
Translating the four standard versions based on common symptom items (palpitations, chest tightness, shortness of breath, fatigue, spontaneous sweating, pale tongue, weak pulse) allows derivation of theoretical severity grading shifts for the same patient cohort across versions:
| Dimension | 1993 Edition | 2002 Edition | 2017 Edition | 2023 Edition |
|---|---|---|---|---|
| Grading Granularity | Two-level (Light/Heavy) | Three-level (Light/Moderate/Heavy) | Three-level + Combined Syndromes | Three-level + Syndrome Elements |
| Threshold Scope | “3 Main Symptoms + 1 Tongue/Pulse” | “Score ≥7 Light; ≥14 Heavy” | “≥4 Elements + Combined Syndrome ≥1” | “Elements + Objective Metrics Integration” |
| Weight of “Pulse” in Heart Qi Deficiency Context | High (Mandatory) | Medium | Low (Yields to Combined Syndromes) | Medium-High (Re-weighted) |
From this, we infer that the “discrimination-grading” slope for HRI harmonics C2/C3 (the two components most strongly associated with Heart Qi Deficiency) differs across versions: the 1993 version has the steepest slope (strict thresholds), while the 2023 version, incorporating objective metrics, exhibits a flatter slope (wider thresholds). This slope drift, if uncorrected in cross-era Meta-analyses, will lead to an overestimation of severe Heart Qi Deficiency proportions when mapping 1990s HRI data to modern standards.
4. Draft Inter-Conversion Mapping Table
Based on the evidence chain above, we can establish Heart Qi Deficiency Standard Inter-Conversion Anchors: scaling the combined HRI-C2/C3 score to a unified 0–100 range, using the 1993 version as the reference system. The following rules apply for inter-conversion (draft, pending empirical validation):
| Standard Inter-Conversion | Anchor Formula (HRI = x) |
|---|---|
| 1993 → 2002 | Score_2002 = Score_1993 × 0.85 + 5 |
| 2002 → 2017 | Score_2017 = Score_2002 × 1.10 − 7 |
| 2017 → 2023 | Score_2023 = Score_2017 × 0.92 + 3 |
These formulas are derived from equivalent item-weight analysis and have not been prospectively validated on samples. They serve only as a methodological demonstration.
5. Evidence and Limitations
Limitation 1: No controlled experiment re-tagging HRI data across all four standard versions was found. A direct PubMed query for “heart-qi deficiency diagnostic criteria version kappa consistency” returned 0 hits (search date: 2026-10-09).
Limitation 2: Evidence for the binary association between HRI and Heart Qi Deficiency is concentrated in single-center/single-standard-version studies; no large-scale multi-standard comparison exists.
Limitation 3: This study is a methodological review rather than original data analysis; all slope drift estimates assume linear symptom mapping, whereas actual non-linear effects may amplify drift.
Limitation 4: Inter-practitioner individual consistency differences (Kappa 0.32–0.45) imply significant noise even under single-standard conditions; cross-version drift must be statistically significantly higher than this baseline to be meaningful.
6. Related Reading
- P3-EXP-027: Independent correlation validation of HRI vectors with Heart/Liver/Spleen/Lung/Kidney Deficiency-Excess (N=1,200, 24 months)
- science-yixue/hri-c2-c3-heart-qi-deficiency-severity-dose-response.md — Methodologically complementary to this topic, providing dose-response curves post-H1 validation
- science-yixue/tcm-pulse-hri-inter-operator-reliability-training-standardization.md — Inter-practitioner consistency is a prerequisite for cross-standard consistency
- science-yixue/pulse-harmonic-syndrome-independent-correlation.md — Completed cross-sectional study on Heart Qi Deficiency binary classification
7. Next Research Priorities
- Multi-center, four-version re-tagging of HRI data (Target N=600; double-blind scoring by the same patient cohort across four versions per center; calculate Cohen’s Weighted Kappa).
- Cross-center weighted consistency correction tables (Estimate standard drift after controlling for baseline inter-practitioner noise).
- External validation of HRI inter-conversion mapping (Re-run conversion formulas on the dataset produced in Round 22 of YY on 2026-10-03 for P3-EXP-022).
- ICDx-Syndrome Double-Blind Registration Guidelines to prevent mixing different versions under the same Heart Qi Deficiency label.
Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.
Leave a Reply