---

name: hospital-qcc-pdca
description: |
Knowledge base from (Hospital Quality Control Circle —
Closing Documents and Case Reviews 问题解决流程PSP). Use when running, writing up or reviewing a
QCC / 品管圈 / PDCA improvement project. Triggers: "帮我做品管圈", "how do I calculate
the 目标值", "which QCC type should I use", "review my QCC closing document",
"check my 鱼骨图 / 柏拉图", "what is 圈能力 / 改善重点 / 望差值 / 目标达成率",
"score my QCC report against the competition rubric", "how do I verify 真因",
"write a 标准化作业书", "问题解决型还是课题研究型".
Do NOT use for general Six Sigma / DMAIC / Lean-Kaizen questions, for ISO or
accreditation standards, or for ordinary project planning — this skill is specific to
the hospital 品管圈 method and its Chinese documentation conventions.
cowork:
category: research
icon: BookOpen

医院品管圈结题文档与案例评析

Hospital Quality Control Circle (QCC / 品管圈) — Closing Documents and Case Reviews

A 344-page Chinese practitioner's manual in three parts: the step-by-step procedure for a problem-solving QCC closing document (Ch 1), the task-research variant (Ch 2), the chart toolkit (Ch 3), five complete hospital case reports each followed by a named expert's critique (Ch 4), and the national competition scoring rubrics (Ch 5).

How to Use

• Running a circle → follow Ch 1 (or Ch 2 if the theme is a first-time/breakthrough task); keep cheatsheet.md open for formulas and thresholds.
• Stuck on a specific technique → patterns.md has 13 reusable techniques with trade-offs.
• Drawing a chart → Ch 3 for construction rules and the exact errors reviewers penalise.
• Auditing or scoring a report → Ch 5 rubric first, then the "Tells \& smells" list in cheatsheet.md.
• Term lookup → glossary.md.

Worked example — "our 不合格率 is 23.85%, what target should we set?"

1. Confirm the theme verb is negative (降低) → use the subtraction formula.
2. Take 改善重点 from the Pareto's \~80% band (here 82.52%) — do not invent it.
3. Take 圈能力 from the circle's scored table (here 50.72%).
4. 目标值 = 23.85% − (23.85% × 82.52% × 50.72%) = 13.87%; 改善幅度 = 41.84%.
5. State the basis (可及性 + 先进性) and draw the 改善前/目标值 bar chart.

Missing either multiplier? Say so and go back to the Pareto or the 圈能力 table — never substitute a round number.

When NOT to Use

• Six Sigma / DMAIC, Lean, Kaizen or A3 questions → these use different arithmetic and different artefacts.
• ISO 15189 or accreditation-standard questions → use an accreditation-specific reference.
• Generic project planning, RACI or Gantt questions unrelated to QCC → the Gantt rules here carry QCC-specific constraints.
• Clinical questions arising from the case topics (needlestick prevention, MDRO control, restraint practice) → the cases illustrate method, not clinical guidance; consult current clinical standards.

Core Frameworks \& Mental Models

The 10-step / 13-section chain (问题解决型). 成立品管团队 → 主题选定 → 活动计划拟定 → 现状把握 → 目标设定 → 解析(真因验证) → 对策拟定 → 对策实施与检讨(D) → 效果确认(C) → 标准化(A) → 检讨与改进 → 下期主题 → 附件. Each step's output is the next step's mandatory input; the report fails on broken links, not on weak improvement.

Type selection — do this first. Use the QC STORY 判定表: 5 paired statements scored 强相关2 / 中相关1 / 弱相关0. Higher column wins, and the gap must exceed ⅓ of the higher total, or the judgement is not decisive. Use 课题研究型 when the work is first-time, has no history to survey, aims to break through the current level, or pursues 魅力质量; use 问题解决型 when improving or protecting an existing level. Before either: if a mature standard already exists and simply isn't enforced, enforce it — QCC is the wrong tool.

Theme naming. 动词(正向/负向) + 名词(改善主体) + 衡量指标 — e.g. 降低 + 软式内镜 + 洗消不合格率. The verb's direction selects the target formula.

Three-artefact current-state. 流程图 (scope) → 查检表 (facts, 4–6 items, 5W2H-validated, collected under 三现原则) → 柏拉图. The Pareto's \~80% cumulative band is 改善重点 — a multiplier for the target formula, not just a picture.

Target arithmetic. 负向: 目标值 = 现况值 − (现况值 × 改善重点 × 圈能力). 正向: 目标值 = 现况值 + (标准值 − 现况值) × 改善重点 × 圈能力 (标准值 defaults to 1). Then 改善幅度 = (现况值 − 目标值) ÷ 现况值. At the end: 目标达成率 = (改善后 − 改善前) ÷ (目标值 − 改善前) × 100% and 进步率 = (改善后 − 改善前) ÷ 改善前 × 100%. Attainment is satisfactory at 100%±10% and must be explained outside 80–150%.

圈能力. Either 平均分 ÷ 5.0 × 100% from a 5-3-1 self-score, or the weighted form 年资×40% + 学历×30% + 主题改善能力×30%, with 品管圈经验值 (5 per prior circle, capped at 20) added separately. Label the weights on the table.

The analysis chain. 问题 --[头脑风暴 / 鱼骨图 / 系统图]--> 原因 --[评价法 5/3/1]--> 要因 --[查检 + 柏拉图, or per-cause verification]--> 真因. Use a fishbone when reasoning 结果→原因 and a 系统图 when reasoning 目的→方法. Draw one fishbone per 改善重点, categorise 大骨 as 人机料法环(测), and drill to terminal causes that can be acted on directly. Keep 要因 to ≤10 — needing to merge them afterwards means too many were selected. Only 真因 may receive countermeasures; publishing the 非真因 rejections is evidence of rigour.

Countermeasure design. ≥2 options per 真因, scored 优5/可3/差1 on 重要性·迫切性·圈能力 with the bar at 总分×80%. Merge into 3–5 named 对策群组 with owners and non-overlapping implementation windows, sequenced easy→hard. Record each group on a PDCA card: P (改善前 + 对策内容) / D (owner, specific dates, process) / C (quantified before→after) / A (standardise or re-plan), stamped 对策有效/无效.

Task-research substitutions (课题研究型). 现状把握 → 课题明确化 (望差值 = 期望值 − 现况值; 攻坚点 scored 强5/中3/弱1 on 上级方针·圈的优势·克服能力). 解析+对策拟定 → 方策拟定 (scored on 可行性·经济性·效益性) and 最适方策探究 (实施顺序 → 期待效果预估 → 障碍与副作用判定 + 消除措施 via PDPC → 最适方策). There is no target formula — use 标杆法/同侪调查, 3–5 targets, quantitative and qualitative. Add 模式构建, 课题查新, 成立专家组. Give the D phase ≈50% of the calendar.

Time allocation. Gantt built on 4W1H; 虚线 = plan, 实线 = actual; P≈30% / D>40% / C≈20% / A≈10%; consecutive steps must never overlap; one named owner per row (never "全员"); weeks counted at ≥4 days.

Standardisation is a loop, not a document. 制定 → 贯彻 → 修订 → 再实施, 螺旋式上升. Standards must be quantified, operable and revisable, and free of 适当/加强/注意/随时/小心. Date them after the implementation that justified them. Close with 检讨与改进 across all steps (including weaknesses), the 残留问题 from the Pareto's remaining 20%, the next theme, and an 效果维持 section — the most commonly omitted and most reliably penalised part of the whole report.

Chapter Index

Topic Index

5W2H → Ch 1, 3 · 80/20 法则 → Ch 1, 2 · PDPC → Ch 2, 5 · QC STORY 判定 → Ch 1, 2 · 三现原则 → Ch 1, 5 · 人机料法环 → Ch 1, 3 · 圈能力 → Ch 1 · 对策群组 → Ch 1 · 效果维持 → Ch 4, 5 · 攻坚点 → Ch 2 · 改善重点 → Ch 1 · 无形成果/雷达图 → Ch 1, 3 · 望差值 → Ch 2 · 标准化作业书 → Ch 1 · 柏拉图 → Ch 1, 3 · 查检表 → Ch 1, 3 · 残留问题 → Ch 1 · 特性要因图 → Ch 1, 3 · 现况值 → Ch 1 · 甘特图 → Ch 1, 3 · 目标达成率 → Ch 1, 4 · 真因验证 → Ch 1, 5 · 衡量指标 → Ch 1 · 要因选定 → Ch 1, 4 · 课题明确化 → Ch 2 · 课题查新 → Ch 2, 4 · 评分表/rubric → Ch 5 · 进步率 → Ch 1 · 模式构建 → Ch 2, 4 · 流程图符号 → Ch 3 · 权重评价法 → Ch 1

Supporting Files

• glossary.md — every key Chinese term with an English gloss and chapter reference.
• patterns.md — 13 reusable techniques (theme selection, verification designs, countermeasure grouping, the PDCA card, the effect-confirmation stack) with when-to-use and trade-offs.
• cheatsheet.md — formula card, all numeric thresholds, the six scoring scales, rubric weights, the biggest fixed deductions, and a "tells \& smells" list for spotting a weak report fast.

Guardrails

• Never fabricate a number. 改善重点, 圈能力, 现况值 and every score come from the circle's own tables. If the user has not supplied one, ask for it or leave a marked placeholder — do not assume a typical value.
• Never invent a formula variant. The formulas in cheatsheet.md are the book's; if a situation isn't covered (e.g. a task-research target), say the book gives no formula and name the method it does prescribe (标杆法/同侪调查).
• Attribute rules to their source step or rubric line when advising on a report, so the user can verify against the original document.
• When data is missing or a page is unreadable, state the gap plainly rather than filling it — the source is a scan, and a few tables were not transcribed in full.
• Do not present the case critiques as clinical advice. They audit method; several countermeasures in the cases were criticised as unsafe or outdated.
• Flag, don't overrule. If the user's existing report conflicts with the book (e.g. 要因整合, 目标达成率 140%), point to the rule and the critique that discusses it, and let them decide.

Scope \& Limits

• Source is a scanned Chinese book with no text layer; content was read page by page from images. Chapter and page references are to the printed page numbers.
• Chapter 3 is largely software walkthroughs (Excel/Visio screenshots); the distillation keeps the conventions and error lists, not the click-by-click steps.
• Case coverage includes each report's structure, its headline figures (改善前/目标值/改善后, 目标达成率, 进步率 — all recomputed and verified) and its expert critique. Intermediate per-step tables (individual member scores, full check-sheet tallies, 要因 rankings) were not transcribed — consult the source for those.
• The rubrics are from the national competition; weights may differ in later editions.
• Terminology is preserved in Chinese because the field's vocabulary is Chinese; English glosses are provided throughout.