在中國的大城市裡，任何去過醫院的人都可能對一個場景記憶猶新：候診大廳裡人山人海，患者們為了見到專家，常常需要經歷漫長的排隊與等待。這幾乎是中國醫療體系長期存在的標誌性難題。面對這一困境，中國正試圖用醫療數位化手段，來重塑其龐大而複雜的醫療體系。從“點評”醫生到AI輔助診斷，變革的浪潮正在逐步到來。然而變革之路並非一帆風順：使用者錯配，線上問診加劇醫生過勞，歷史上的醫療醜聞導致民眾對醫療行業普遍信任度不高，患者資料質量普遍不高導致資料孤島難以訓練出高品質的AI模型等問題都在不斷出現。技術是答案，還是新的問題？醫療數位化，這既是中國醫療改革的新希望，也是最艱巨的挑戰。Dr Chatbot will see you right nowChina wants AI and telemedicine to bolster its health system聊天機器人醫生已上線人工智慧與遠端醫療將如何助力中國醫療體系升級The Economist |《經濟學人》November 29th, 2025China wants AI and telemedicine to bolster its health system HINA'S PUBLIC-HEALTH system is both world-class and woeful. The best hospitals in Beijing and Shanghai have top-notch doctors, the latest drugs and gleaming equipment. But such places make up only around 10% of China's medical institutions. In the country's 33,000-odd township-level health centres,only half of general practitioners (GPs)even have university degrees. In total China spends around 7% of its GDP on health (for comparison Britain shells out around 11%).When ill, Chinese tend to flock to specialists in big cities.That leads to massive queues, perilously overworked doctors and experts wasting their time on common diseases.Vast investment in basic services offered near where people live might help solve the problem, and local governments want to train and hire more doctors. But they are heavily indebted. And recruitmentis tricky because the profession is neither particularly prestigious nor, outside the best hospitals, especially well compensated. One doctor in a hospital in Nantong, in eastern China, complains that overtime is constant and “all essentially unpaid”. Past promises of pay rises have never material-ised,he says.中國的公共醫療體系呈現出一種矛盾景象：既擁有世界頂尖水平，又存在明顯短板。北京、上海的頂尖醫院坐擁一流專家、最新藥物和先進裝置，但這類機構僅佔全國醫療資源的10%左右。在全國約3.3萬個鄉鎮衛生院中，僅半數全科醫生擁有大學學歷。總體來看，中國醫療支出約佔GDP的7%（英國約為11%）。中國人患病時往往湧向大城市專科醫生，導致掛號難、醫生超負荷工作、專家耗費大量時間診治常見病。加大基層醫療投入本可緩解這一問題，地方政府也試圖培養招聘更多醫生，但地方債台高築，且醫生職業吸引力有限——既缺乏崇高社會地位，在非頂尖醫院薪資待遇也普遍偏低。華東地區南通市某醫院醫生抱怨加班成常態，且"幾乎都沒有加班費"，過往漲薪承諾從未兌現。Increasingly China is looking for high-tech solutions to its medical problems. Inrecent years both private companies and government officials have started promoting telemedicine and artificial intelligence(AI).There are still bottlenecks. But China has pressing reasons to make digital healthcare work:its rapidly ageing population will bring with it a wave of chronic diseases that the current system is ill-equipped to handle. By 2050,487m people will be over the age of 60, about 35%of China's population,up from 21% today.Despite earlier government initiatives, telemedicine only took off during the co-vid-19 pandemic. Chinese seized on apps launched by China's tech companies that allowed them to consult doctors either by text or video. One such app, called JDHealth and launched by JD.com, an e-commerce firm, claimed to have 200m active users in the 12 months to June,with an average of over 500,000 online consultations a day.In June Ant Group,an affiliate of Alibaba, a tech giant, launched another appcalled AQ, As of September it had served140m patients and nearly im doctors had offered their services on the platform, according to Ant.中國正日益尋求用高科技手段破解醫療困局。近年來，政府與企業開始大力推廣遠端醫療與人工智慧應用。儘管存在瓶頸，但發展數字醫療對中國具有緊迫意義：快速老齡化將引發慢性病浪潮，現行醫療體系難以應對。預計到2050年，60歲以上人口將達4.87億，佔總人口比例從現在的21%升至35%。儘管政府早有規劃，遠端醫療真正興起於新冠疫情期間。中國科技企業推出的線上問診應用迅速普及，使用者可通過文字或視訊諮詢醫生。電商企業京東旗下的"京東健康"宣稱，截至今年6月的一年內活躍使用者達2億，日均線上問診量超50萬次。科技巨頭阿里巴巴關聯企業螞蟻集團今年6月推出"安鵲健康"應用，截至9月已服務1.4億患者，近百萬醫生在該平台註冊。In theory all this should help bolster local health-care offerings and give rural folkbetter access to faraway experts. "I don't think there's any country in the world that has embraced direct-to-consumer tele-medicine quite like China, "reckons Terence Cheng, of Monash University in Australia. Users are attracted by the "shopping experience" that China's apps offer. Patients can rate doctors and read reviews about their skilfulness.Bottlenecks remain, however. The typical users of telemedicine apps are younger urban residents-in other words, people who already have the best access to China's public-health services. And the doctors who offer their advice on the apps usually use them to earn more money outside their regular hours. So it is hardly helping with overwork. One solution would be for hospitals to get doctors to allocate 5-10%of their daytime hours to telemedicine, ar-gues Tang Shenglan of Duke University in America. That in turn would require tweaks to China's state-backed health insurance to make it easier for doctors to be paid for online work, he says.理論上，這類應用能提升基層醫療服務能力，讓偏遠地區居民接觸頂尖專家。澳大利亞莫納什大學程特倫斯指出："中國對消費者直連式遠端醫療的接納程度全球罕見。"本土應用提供的"購物式體驗"頗具吸引力，患者可對醫生評分並查閱診療評價。但瓶頸依然存在：遠端醫療的主要使用者仍是都市年輕人——本就享有最優質公共醫療資源的群體。平台醫生多利用業餘時間賺取額外收入，無助於緩解本職工作負擔。美國杜克大學湯勝蘭建議，醫院可安排醫生將5%-10%的日間工作時間用於遠端診療，這需要調整醫保支付政策以覆蓋線上服務。Accepting AI health advice is the nextstep. Some in China already get help via chatbots such as those made by DeepSeek, a whizzy startup.China's government wants to push things further. Officials believe AI can significantly improve the capacity and efficiency of primary health-care services. In November it released a plan calling for “full coverage”of AI--powered diagnosis and treatment tools at grassroots health centres by 2030.If local doctors could use AI models to access the latest advice, it could help build patients'trust in the tech,reckons Tien YinWong,who builds such models at Tsing-hua University in Beijing.In turn thatwould result in more local treatment;only complex cases would go to big hospitals. And it could all happen without vast sums being spent on training new doctors. MrWong says officials from several townsand cities have approached his team aboutits models. For now they are tested only at a hospital affiliated with the university.下一步是普及人工智慧醫療諮詢。部分民眾已通過深度求索等明星初創企業的聊天機器人獲取健康建議。中國政府計畫進一步推進，認為人工智慧能顯著提升基層醫療服務能力與效率。去年11月發佈的規劃提出，到2030年基層醫療機構將全面配備AI輔助診療工具。清華大學研發AI醫療模型的黃天蔭教授認為，若基層醫生能借助AI模型獲取最新診療方案，將增強患者對技術的信任，從而讓更多患者留在當地就醫，僅疑難重症轉診大醫院。這種方式無需投入巨資培養新醫生，已有多個市縣醫療機構與其團隊接洽，目前相關模型僅在大學附屬醫院進行測試。China still remains far from AI-super-charged health care.And thanks to a long history of scandals, Chinese tend to be suspicious of the medical profession. A bad AI-powered diagnosis could cause them to lose trust in the new technology too, fears Mr Wong. Another challenge will be getting enough data to feed into AI models. At the moment China's patient data is often poor quality and fragmented between different hospitals and clinics.On the other hand,China has a history of rapid technological adoption.For instance, when Chinese people stopped using cash, they did not switch to plastic but jumped straight to mobile payments, thanks to a combination of nimble techfirms, encouraging officials, a digitally literate population and the lack of established alternatives.The same cocktail offactors may drive technological change inhealth care. In other countries trying tobuild primary health-care systems on a budget,the absence of these ingredients may also make China hard to imitate.中國距離AI賦能醫療的願景仍有差距。由於歷次醫療醜聞的影響，民眾對醫療行業普遍存疑。黃天蔭擔心，一次失敗的AI診斷可能導致公眾對新技術喪失信任。另一挑戰在於資料供給——當前中國患者資料質量參差，且散落於不同醫療機構。但中國具備技術快速落地的土壤：當民眾告別現金支付時，並未轉向信用卡，而是在科技企業創新、政策鼓勵、居民數字素養提升及缺乏成熟替代方案的綜合作用下，直接跨入移動支付時代。這種復合優勢可能同樣推動醫療技術變革。對於其他試圖低成本建設基層醫療體系的國家而言，若缺乏這些要素，中國的經驗或將難以複製。 (菁英學社Elite Aca.)