Source 出处 · Jingui Yaolüe · Doses are not given here: they are set by a licensed practitioner for the individual.
Safety. Two ingredients, and the line that records them contains the most concrete description of what a person brings up that this library has recorded: 久久吐脓如米粥 — in time, pus spat out like rice porridge.
That clause has an address today. Coughing up a large volume of pus is what lung abscess, bronchiectasis with a large daily sputum volume, empyema that has found a bronchus, and necrotising pneumonia look like, and what separates them is imaging — a chest radiograph and a computed tomogram, where a cavity with an air-fluid level is the finding the sentence is describing without knowing it. This library offers no management for acute infection of any kind, and it is worth saying why it insists: the interval between a lung abscess being visible and antibiotics and drainage being started is counted in hours, and the change in outcome brought about by those two things is not something a classical line competes with.
The clause before it, 振寒脉数, is the one that carries the clock. Shivering together with a fast pulse is how sepsis is recognised at the bedside today: temperature, heart rate, respiratory rate, blood pressure, level of consciousness, and a lactate. Rigors are also the specific thing that suggests bacteria in the blood rather than a local collection. The text pairs the two findings that today trigger a screening pathway.
咽干不渴 — dry throat, no thirst — is a small clause with two uses. Classical texts use the presence or absence of thirst to sort conditions from one another, and 不渴 is the branch this line takes. Today, absence of thirst is not a reliable sign that fluid is adequate; the sensation of thirst is documented to weaken with age. A dry throat is also what mouth breathing, medicines and sicca produce, and it is not evidence about the chest.
Three kinds of material brought up from the chest are worth keeping apart, because the line compresses them: pus, which is infection; blood-streaked sputum, which before anything else means tuberculosis and tumour have to be excluded; and saliva, which is neither. A foul smell points to anaerobes. Colour and volume together are more informative than either.
The two materials. 桔梗 is Platycodon grandiflorus. It is the only species in its genus, but the family it belongs to is not made only of it — 党参, 沙参 and 半边莲 are in the same family and are different drugs. Classical materia medica records that 荠苨 was used to pass off as 桔梗, which is the kind of note that matters more now that the trade is global. Platycodon root contains platycosides, saponins; saponins irritate mucosa, and nausea and vomiting after large amounts are reported, while lysis of red cells in vitro is documented for the class without this library claiming it happens by mouth. Platycodon is also a food — the Korean doraji preparation — and this library does not treat the fact that something is eaten as evidence that a formula made from it is safe.
甘草 is licorice root, and the glycyrrhizin in it can produce a state resembling excess mineralocorticoid: salt and water retention, potassium loss and raised blood pressure, described at ordinary amounts over weeks. The same two ingredients appear in the 伤寒论 under the same formula name, for a sore throat that did not settle with licorice alone; the classical corpus gives this pair two indications, and this library records both without deciding between them.
Reliable safety data for pregnancy and breastfeeding are lacking, and this library does not recommend use in either. Chinese-language clinical literature reports observations of this formula in sore throat and in the 肺痈 picture, small in scale and methodologically limited. This library records that such studies exist; it does not record that benefit has been established.
两味药，而记载它们的那句含有本库收录过的**对「吐出什么」最具体的一次描写**：**久久吐脓如米粥**。
这一句今天有它的去处。咳出大量脓，就是**肺脓肿**、**每日痰量很大的支气管扩张**、**破入支气管的脓胸**与**坏死性肺炎**的样子；而把它们分开的是影像 —— 胸片与 CT，其中「空洞伴气液平」正是这句话在不知道的情况下描写的那个发现。本库不对任何急性感染提供处置方案，而这里要说清为什么坚持：**从肺脓肿在片子上可见，到抗生素与引流开始，间隔是以小时计的**，这两件事带来的结局差别不是一条古典条文能与之竞争的东西。
它前面那一句**振寒脉数**才是带时钟的一句。寒战加上脉快，正是今天在床边识别脓毒症的方式：体温、心率、呼吸频率、血压、意识水平与乳酸。寒战还特指**菌血症**的可能，而不只是局部积脓。这条条文把今天会触发一套筛查路径的两个发现并排写在一起。
**咽干不渴**是个短句，却有两处用处。古典用渴与不渴来把病彼此分开，本条走的是「不渴」这一支。而不渴今天不能作为水分充足的可靠证据 —— 口渴感随年龄减弱有文献记载；咽干也见于张口呼吸、药物与干燥综合征，它不是关于胸部的证据。
从胸部上来的三种东西值得分开，因为这一句把它们压在了一起：**脓**（感染）、**痰中带血**（在一切之前要排除结核与肿瘤）、以及**唾液**（两者都不是）。恶臭指向厌氧菌；颜色与量合在一起比单看任一项更有信息量。
两味药。**桔梗**是桔梗（Platycodon grandiflorus），本属仅此一种，但它所属的**科**并不只有它 —— 党参、沙参、半边莲同科而各是各的药。本草记载**荠苨**曾被用来冒充桔梗，在贸易全球化的今天这类按语只会更有用。桔梗根含桔梗皂苷（platycosides），属皂苷类：皂苷刺激黏膜，量大后恶心呕吐有报告；该类成分的体外溶血有记载，本库不主张口服发生。**桔梗也是食物**（朝鲜族「道拉基」），而本库不把「能吃」当作由它组成的方安全的证据。
**甘草**是甘草的根，其中甘草酸可引起**假性醛固酮增多症**：钠水潴留、失钾、血压升高，普通量连服数周亦有报告。同样这两味在《伤寒论》里以同一方名出现，主治甘草汤没能缓解的咽痛 —— 古典给这一对留了两处主治，本库两处都记，不裁定其间关系。
**孕期与哺乳期缺乏可靠的安全性数据**，本库不主张使用。中文临床文献有本方用于咽痛与「肺痈」一类画面的观察报告，规模小、方法学有限。本库记「有这类研究」，不记「疗效已被确立」；本条目也不构成对一次咳脓痰或一次寒战发热的处置。