四个环节的数据依存关系Data Dependencies Between the Four Steps
"预测→干预→归因→结算"——这条链路已经成为预防医疗行业的高频词汇。但在大多数语境里，它只是一个漂亮的slogan，背后没有完整的工程实现。"Predict→Intervene→Attribute→Settle" has become a frequent phrase in preventive medicine. But in most contexts, it's just a slogan without complete engineering implementation behind it.
真正理解这条链路，需要理解一件事：这四个环节不是独立的功能模块，而是一条数据因果链——每一个环节的输出，都是下一个环节的必要输入。Understanding this chain requires recognizing one thing: these four steps are not independent feature modules, but a data causal chain — each step's output is the next step's required input.
- 预测→归因：Predict→Attribute:基线风险评分是PSM倾向评分模型的关键协变量。没有可信的风险预测，PSM匹配缺少最重要的混杂控制变量。Baseline risk score is PSM's key covariate. Without credible risk prediction, PSM matching lacks its most important confounder control variable.
- 干预→归因：Intervene→Attribute:PSM需要明确的干预分配记录（T=1/0）。没有结构化的干预轨迹记录，PSM无法执行。"做了干预但没记录"等于没做。PSM needs explicit treatment assignment records (T=1/0). Without structured intervention trajectory records, PSM cannot execute. "Intervened but didn't record" equals not intervening.
- 归因→结算：Attribute→Settle:可结算证据报告的核心内容是PSM输出的ATT及置信区间。没有因果归因，效果数字不具备因果解释力，支付方不接受。The billable evidence report's core content is PSM's ATT and confidence intervals. Without causal attribution, effect numbers lack causal interpretability and payers won't accept them.
ReHealth Core的定位：ReHealth Core's Positioning:市面上不缺预测工具，不缺干预平台，不缺统计分析软件。缺的是把四个环节的数据打通、形成完整因果链路的基础设施。这正是ReHealth Core存在的理由。The market has prediction tools, intervention platforms, and statistical software. What's missing is infrastructure that connects all four steps' data into a complete causal chain. This is exactly why ReHealth Core exists.
"预测→干预→归因→结算"不是四个功能的列表，而是一条数据因果链。ReHealth Core的价值不在于任何单一环节的优秀，而在于四环数据连通这件在行业里独一无二的事。"Predict→Intervene→Attribute→Settle" is not a feature list — it is a data causal chain. ReHealth Core's value is not in any individual step, but in the four-step data connectivity that is unique in the industry.