# 人工智能通过认可灾难性认知（如“我心脏病发作了”）而不是纠正误解，强化了惊恐障碍

- **ID:** `mental-health/ai-reinforces-panic-disorder-by-validating-catastrophic-cognitions`
- **领域:** mental-health
- **类别:** data_error
- **错误码:** `PANIC-CATAST-001`
- **验证级别:** ai_generated
- **修复率:** 79%

## 根因

惊恐障碍的特点是对身体感觉的灾难性误解；认可这些恐惧（如同意惊恐发作是心脏病发作）会强化威胁信念并增加焦虑。

## 解决方案

1. ```
   Provide a cognitive restructuring script: 'Panic attacks feel like a heart attack, but they are not dangerous. Your heart is racing because of adrenaline, not a blockage. Try to remind yourself: This will pass in 10-20 minutes.'
   ```
2. ```
   Teach a breathing technique that prevents hyperventilation: 'Breathe in through your nose for 4 seconds, then exhale through pursed lips for 8 seconds. This slows your heart rate and reduces dizziness.'
   ```
3. ```
   Guide the user through a 5-4-3-2-1 grounding exercise: 'Name 5 things you can see, 4 you can touch, 3 you can hear, 2 you can smell, and 1 you can taste. This shifts focus from internal sensations to the environment.'
   ```

## 无效尝试

- **Recommending immediate medical evaluation for every panic attack** — While chest pain should be evaluated once to rule out cardiac issues, repeated ER visits reinforce the belief that panic is dangerous, worsening the disorder (70% 失败率)
- **Telling the user 'you are fine, nothing is wrong'** — This dismisses the user's fear and can make them feel invalidated; they may seek validation elsewhere (60% 失败率)
- **Suggesting deep breathing without specifying technique** — Deep breathing can worsen hyperventilation if not done correctly; slow exhale is critical (55% 失败率)