# AI指示护士在24小时而非要求的48-72小时读取结核菌素皮肤试验（TST）结果，导致假阴性解读

- **ID:** `medical/tuberculin-skin-test-reading-48h`
- **领域:** medical
- **类别:** runtime_error
- **错误码:** `TST_READING_24H`
- **验证级别:** ai_generated
- **修复率:** 92%

## 根因

TST硬结在48-72小时达到峰值；24小时读取会遗漏迟发型超敏反应，导致假阴性率高达30%。

## 版本兼容性

| 版本 | 状态 | 引入 | 弃用 |
|------|------|------|------|
| CDC-TST-guidelines-2024 | active | — | — |
| WHO-TST-protocol-v3.2 | active | — | — |

## 解决方案

1. ```
   安排TST注射和读取预约在同一时间，确保48-72小时间隔。在患者记录中记录注射时间和计划读取时间。如果患者无法在48-72小时返回，则不进行TST；改用IGRA血液检测（例如QuantiFERON-TB Gold）。
   ```
2. ```
   如果TST在24小时读取且为阴性，不要将其视为最终结果。指示患者在48-72小时返回进行正确读取。将早期读取记录为'初步结果——不用于解释'，并在正确间隔重新读取。如果患者无法返回，从头重复TST。
   ```
3. ```
   对于系列检测使用两步TST方案：如果初次检测为阴性，在1-3周后进行第二次TST。这能增强免疫反应，减少因免疫力减弱导致的假阴性。
   ```

## 无效尝试

- **Assuming that any visible reaction at 24 hours is sufficient for interpretation** — Induration at 24 hours is often due to immediate hypersensitivity (Arthus reaction) or irritation, not a delayed-type hypersensitivity response to TB antigens. True induration develops later. (65% 失败率)
- **Believing that a negative reading at 24 hours can be confirmed by reading again at 48 hours** — If the test is read at 24 hours and recorded as negative, the nurse may not re-read at 48 hours, or the patient may not return. Even if re-read, the initial reading may bias the second assessment. (50% 失败率)
- **Using a digital caliper or ruler to measure induration at 24 hours and assuming it will be the same at 48 hours** — Induration size increases over time; a small or absent reaction at 24 hours may become positive at 48 hours. Measuring early gives a falsely low reading. (70% 失败率)