In the era of digital healthcare, efficient collaboration of hospital information systems has become the key to improving the quality and efficiency of medical services. This article delves into how the hospital information SaaS (Software as a Service) system realizes the transformation from scenario adaptation to efficient implementation through multi-terminal collaborative design.
To do a good job in multi-terminal collaboration of hospital information SaaS, it is necessary to have a deep understanding of the medical scenario – know that the most important thing for doctors in the clinic is time, so the PC side should optimize efficiency; Knowing that patients are most afraid of trouble when seeking medical treatment, the mini program should be designed to be lightweight; Knowing that the management needs a global perspective when making decisions, the big screen should do data correlation.
Only by clarifying terminal positioning through user stratification, focusing on core needs through priority division, and opening up data and experience through collaborative mechanisms can hospital information SaaS truly become a link connecting all parties in medical services, and ultimately realize the multiple values of improving medical care efficiency, optimizing patient experience, and improving hospital operation efficiency. In the process of deepening digital medicine, the ability of multi-terminal collaborative product design will also become the core indicator of the competitiveness of hospital information systems.
1. User stratification and terminal adaptation
The user needs of medical scenarios vary significantly, and the choice of terminals must be closely related to their core work scenarios – the key is not what the technology can achieve, but what users need in specific scenarios.
1.1 Medical staff
The work of medical staff is naturally accompanied by high-density information processing: outpatient doctors need to receive 20+ patients in one morning, and each patient’s diagnosis and treatment must retrieve past medical records, issue medical orders, and enter diagnostic results; Inpatient nurses should handle the infusion arrangements, vital sign records, and doctor’s order verification of 10+ beds. These operations involve multi-system data linkage (e.g., electronic medical records, drug databases, testing systems) and require high accuracy (errors can affect treatment safety).
The irreplaceability of the PC side stems from the adaptation to complex scenarios:
- Multi-window collaborative layout: The 15-27 inch screen can display the patient’s basic information column (left), medical record summary (upper right), drug catalog (bottom right), and test report (floating window) at the same time, avoiding frequent page switching. For example, the PC-side HIS system of a tertiary hospital has designed a diagnosis and treatment workbench mode, where doctors can adjust the position of each module by dragging and dropping, and doctors who are accustomed to giving priority to medical records can enlarge the medical record module to 50% of the screen-to-body ratio.
- Efficient input supportThe system also has a built-in doctor’s terminology database – the input sensor can automatically associate acute upper respiratory tract infection, and the input bid is automatically converted to twice a day, shortening the text input time of a single doctor’s order by 40%.
- Complex logic checks: Powerful computing power can support real-time verification, such as drug compatibility contraindication reminders (when prescribing cephalosporin + alcohol preparations, the system will pop up a red pop-up window and mark the relevant provisions of the “Guidelines for the Clinical Application of Antibacterial Drugs”); Dose overrun warning (if the dosage of the child exceeds the standard coefficient of body weight× the second confirmation process will be triggered, and the doctor needs to manually enter the reason).
1.2 Patients
The core demands of patients for medical services are less errands, less waiting, and understanding. In traditional medical treatment, patients often complain of poor experience due to problems such as queuing for 1 hour for registration and 5 minutes for medical treatment to pick up test results on the spot, and the lightweight nature of mini programs/APPs can solve these pain points.
To achieve these three challenges, product managers will only continue to appreciate
Good product managers are very scarce, and product managers who understand users, business, and data are still in demand when they go out of the Internet. On the contrary, if you only do simple communication, inefficient execution, and shallow thinking, I am afraid that you will not be able to go through the torrent of the next 3-5 years.
View details >
The design logic of mini programs/apps revolves around lowering the threshold for operation:
- Middle-aged and elderly friendly designThe homepage adopts large font (default 16 characters, supports enlargement to 20), high-contrast color matching (buttons use white characters on a blue background, avoid light gray background), and hide the entrance to complex functions (such as health files need to click on my second entry).
- Instant information access: The test result push not only contains numerical values, but also comes with a three-stage interpretation – numerical comparison (blood sugar 2mmol/L, normal range 3.9-6.1mmol/L), popular meaning (your blood sugar is at a normal level), and action suggestions (it is recommended to maintain 30 minutes of exercise every day and reduce the intake of high-sugar foods). According to data from a hospital, after incidental interpretation, the patient’s understanding rate of the test results increased from 62% to 91%.
- Scenario-based guidance: On the queuing query page, in addition to displaying the current queue location, it will also prompt dynamically according to the time period (if the current is the peak of treatment, there are 5 patients in front of you, it is estimated that you will wait for 25 minutes, it is recommended to rest in the waiting area first); The navigation function in the hospital adopts a real picture + step guide (such as going down the escalator from the 3rd floor of the outpatient building to the 2nd floor, turn right 50 meters to the laboratory department, and there is a blue sign at the door) to avoid getting lost caused by abstract maps.
1.3 Management
The core needs of hospital management (such as presidents and department directors) are global control and rapid decision-making. They do not need specific diagnosis and treatment details, but need to know whether the outpatient volume is overloaded, which department has the lowest patient satisfaction, and whether the consumables inventory is sufficient to support next week’s surgery.
The design of the visual large screen focuses on data value transmission:
- Hierarchical data layout: The screen is divided into areas by urgent – important – the upper left area displays real-time warnings (such as the operating table utilization rate exceeds 95% and the number of pediatric outpatient patients reaches 80, exceeding the carrying capacity), highlighted by a red flashing border; The upper right area displays the core operating indicators (outpatient/emergency reception volume, number of people in the hospital); The following areas show trend data (outpatient volume curve and patient satisfaction changes in the past 7 days).
- Interaction depth control: Support click-to-drill but limit the level – click on the low satisfaction of internal medicine patients to drill down to the satisfaction of each internal medicine clinic, and then drill down to specific dissatisfaction items (such as waiting time and communication effect), but do not allow viewing the specific evaluation of a single patient, which not only ensures the granularity of decision-making, but also protects patient privacy.
- Data comparison design: The data of the same period last month and the same period last year are marked next to the key indicators (such as the average hospitalization day of 2 days, a decrease of 0.8 days from the previous month and a decrease of 1.5 days from the same period last year), and the trend is intuitively transmitted through the color of the arrow (green up, red down) to reduce the interpretation cost of management.
2. Functional prioritization
The essence of the division of function priorities is to maximize the value of the terminal under limited resources. Not all functions must be implemented on each terminal, but must be accurately matched based on what users need most and what the terminal is best at.
2.1 Priority judgment framework
To determine the priority of functions, it is necessary to combine four dimensions and dynamically adjust according to the scenario:
For example, doctors’ orders are not only high-frequency (dozens of times a day), high-complexity (need to be linked to multiple systems), and high-value (directly affecting treatment safety), so they must be the top priority on the PC side.
2.2 WeChat Mini Program
The core advantage of mini programs is light, so the functional design should be subtracted – only the essential links in the whole process of patient medical treatment should be retained.
2.2.1 Priority 1 (Must be online)
- Quick registration: Support three-level screening of departments, doctors, and time periods, and the doctor’s visit calendar is marked with color (green can be registered, yellow has fewer remaining numbers, and gray is full), and medical insurance payment is free of jumping (directly call the WeChat medical insurance payment interface, no need to switch APP). After the optimization of a hospital, the registration steps were reduced from the original 8 steps to 3 steps (select a department→ select a doctor→ confirm payment).
- Inspection/inspection results push: Push the results to the mini program within 5 minutes after they are generated, with values + range + interpretation (such as blood routine: white blood cells 8×10⁹/L (normal 4-10×10⁹/L), no obvious abnormalities), and support enlarged viewing of the result picture (solve the problem that the details of the report cannot be seen clearly).
- Queue up for medical treatment: Real-time display of the current queue location, the number of people in front, and the estimated waiting time (calculated based on the average consultation time of the doctor in the past 3 days), and provide push reminders when the number is about to arrive (send WeChat service notifications 3 times in advance).
2.2.2 Secondary Priority (Go Live After Core Functions Stabilize)
- Pay online: Support the payment of outpatient expenses (registration fee, examination fee) and hospitalization expenses (deposit, daily list) in stages, and provide detailed splitting of expenses (such as CT examination fee 300 yuan = equipment fee 200 yuan + operation fee 100 yuan) to avoid patients having doubts about the cost.
- Guidelines for medical treatment: The navigation in the hospital adopts AR real scene + voice guidance (the patient turns on the camera, the system superimposes arrow guidance), and virtual road signs are set up at key locations (such as escalators, toilets, and laboratory departments) to solve the problem of incomprehensible text descriptions.
2.2.3 Level 3 Priority (Gradually Launched after Core Functions Stabilize)
- Health science popularization: Push personalized content according to the patient’s disease (such as diabetic patients push blood glucose monitoring time guidelines, hypertensive patients push low-salt diet recipes) to avoid generalizing content and causing user disgust.
- Satisfaction evaluation: After the treatment, star + label evaluation will pop up (such as the doctor’s attitude is good and the waiting time is long), avoid open-ended questions (reduce user input costs), and the evaluation results will be synchronized to the department management background in real time.
2.3 PC HIS system
The PC side is the workbench of medical staff, and the functional design should be added – integrating the scattered diagnosis and treatment links, using integrated thinking to carry complex diagnosis and treatment, and reducing operational breakpoints.
2.3.1 Core Priorities (Must Be Optimized)
- Electronic medical record system: Support structured entry + free text mixed mode, and provide template library for modules such as current medical history (such as fever for 3 days, accompanied by cough, sore throat, and no dyspnea), which doctors can select and modify with one click; Associate past medical records (automatically extract key information from the last 3 visits, such as hospitalization for pneumonia in May 2023) to avoid repeated inquiries.
- Doctor’s order management: Drug dosage support automatic conversion (enter the patient’s weight of 60kg, the system automatically calculates amoxicillin 5g/time, 3 times a day), real-time reminder of contraindications (if the patient has a history of penicillin allergy, the pop-up window is immediately when amoxicillin is prescribed, and cannot be submitted), and can directly trigger the pharmacy dispensing process (after the doctor’s order is submitted, the pharmacy system will automatically receive the list of drugs to be dispensed).
- Panoramic view of patient information: Integrate outpatient, inpatient, examination, imaging and other data, display the basic information card (name, age, allergy history) on the left side, and display the test report image report medication record on the right tab, so that the doctor can obtain complete information without switching the system. A doctor reported that the integrated system saved him 2 hours a day of rummaging through and entering information.
2.3.2 Sub-core priorities (incremental optimization)
- Inspection application: Linked with LIS system, the barcode (including patient ID, project code, and application time) is automatically generated after selecting the project, and the nurse can scan the code to sample, without manual filling; It supports the preservation of commonly used combinations (such as preoperative routines include blood routine, coagulation function, and infectious disease screening), and the doctor can check it with one click.
- Consumables management: Nurses in the operating room can scan the QR code for verification (scan the QR code on the consumables packaging, automatically associate the patient ID and surgery name), the inventory system deducts in real time, and automatically sends a reminder to the consumables administrator when the inventory of a consumables is less than 3 days.
2.3.3 Auxiliary Priority (Add as Needed)
- Notification of internal information in the department: Adopt hierarchical reminders (red pop-up window indicates abnormal blood pressure of 3 bed patients, gray notification indicates 3 p.m. department meeting) to avoid information interference.
- Simple statistical report: Support doctors to view the distribution of diseases of patients today in the past 1 week of prescription volume ranking, assist in self-work review, but do not occupy the core operation interface (hidden in the personal center – data statistics).
2.4 Visualization of large screens: Presenting decision-making data with focused thinking
The value of a large screen is not to display all the data, but only the data that can influence decision-making.
2.4.1 Core Priorities (Must Be Shown)
- Real-time operational metrics: Outpatient/emergency reception volume (updated every 10 minutes), number of people in the hospital, operating table utilization rate (with early warning threshold, if the usage rate exceeds 90% in red), allows management to quickly judge whether the hospital is now operating normally.
- Medical quality indicators: Average hospital stay (compared with industry benchmarks), patient satisfaction (by department), and incidence of adverse events (data in the past 7 days), which intuitively reflect the quality of medical services.
2.4.2 Sub-core priorities (added gradually)
- Resource scheduling data: Bed occupancy rate of each department (red more than 80%, yellow 60%-80%, green < 60%), equipment operating rate (such as the actual use time/planned time of CT machine), and the gap in the scheduling of medical staff (early warning when the number of nurses in a department is lower than 1:0.4 for the bed-to-nurse ratio).
- Financial core data: Outpatient/inpatient income structure (proportion of drugs, examinations, and treatments), and proportion of medical insurance reimbursement to support cost control decisions.
2.4.3 Auxiliary Priority (Later Supplement)
- Historical trend analysis: The change curve of outpatient volume in the past 3 months (indicating influencing factors such as holidays and influenza season) and the quarterly change of patient satisfaction.
- Regional comparison data: Benchmark with key indicators of 3 hospitals of the same level (such as average hospitalization days and average cost per trip) to clarify their own advantages and gaps.
3. Cross-terminal collaboration
The core of multi-terminal collaboration is to let the data run and connect the experience – the patient’s number in the mini program can be immediately seen by the doctor on the PC; The doctor can receive a reminder on the APP for the checkup form opened by the doctor; Real-time data on the large screen, dynamic summary from each terminal.
3.1 Data synchronization
Data desynchronization is the biggest pitfall of multi-terminal collaboration: after the patient’s mini program is registered, the doctor’s PC is not displayed, resulting in a passing number; The nurse updated the patient’s temperature on the PC, and the old data was still on the big screen, which affected the decision-making. Solving this problem requires a two-pronged approach of technical architecture + synchronization mechanism.
3.1.1 Technical architecture
- Data center: Distributed storage + centralized management, the data of all terminals is stored in the patient master data pool, diagnosis and treatment data pool, operation data pool in the middle office, to avoid data silos. For example, only one piece of patient information is stored in the middle office, and each terminal is called through API to ensure that it is updated in one place and everywhere.
- Message queue: RabbitMQ is used as the message middleware, and when the terminal data is changed, a change instruction is sent to the queue (such as patient A completes registration), and the relevant terminal subscribes to the instruction and updates the data. The whole process was completed within 1 second, and a hospital test showed that when 1,000 people registered at the same time, the update delay of the PC reception list was no more than 5 seconds.
3.1.2 Synchronization mechanism
- Format verification: When the terminal submits data, it is verified first (such as ID number must be 18 digits, mobile phone number must be 11 digits), and direct rejection that does not conform to the format to avoid dirty data entering the system.
- Logical checks: The middle office will make a logical judgment after receiving the data (for example, the age of the child patient cannot be filled in at 60 years old, the outpatient registration time cannot be earlier than the current time), and the specific reason will be returned if the verification fails (such as please fill in the correct age of the child).
- Result verification: After data synchronization, the target terminal needs to feedback to the middle office that it has been received and updated, if there is no feedback within 10 seconds, the middle terminal will automatically resend it to ensure that the data is not lost. A hospital used this mechanism to reduce the data synchronization error rate from the original 3% to less than 1%.
3.2 Permission management
The sensitivity of medical data determines that authority must be tightly aligned: patients can only see their own information, doctors can only change the medical records of patients they receive, and management can see hospital-wide data but cannot change specific diagnosis and treatment information.
3.2.1 Role-Terminal-Permission Matrix
Using a matrix model of role + scene, an example is as follows:
3.2.2 Dynamic permission adjustment
- Scenario-based switching: When doctors are in the outpatient clinic, their authority is limited to outpatients; After entering the ward, the system automatically switches to the authority in charge of inpatients through positioning.
- Temporary permission request: When the intern needs to modify the medical record, he needs to submit an application (indicate the reason), and after the teaching doctor approves the PC, the intern will get 2 hours of temporary permission, and the modification content will mark the intern XXX modification, and the teaching doctor will review it.
3.2.3 Authority audit mechanism
The system automatically records all data operations (who, on which terminal, what data is operated, operation time, IP address), and generates a “Permission Usage Report” every week, focusing on checking risks such as unauthorized viewing and abnormal modifications (such as multiple modifications of medical records in the middle of the night) to ensure data security.
3.3 Version iteration
The system iteration of multi-terminal collaboration is most afraid of the different rhythms of each end: the mini program has updated the registration process, and the PC side is not synchronized, resulting in data docking errors; New indicators have been added to the large screen, and the data source terminal has not been transformed, resulting in empty data. Therefore, version iteration must be planned globally and promoted in a coordinated manner.
3.3.1 Before iteration
- Demand research: Interview each end user (doctor, patient, administrator) at the same time to clarify whether a function change affects other terminals. For example, if the mini program plans to add medical insurance electronic voucher registration, it is necessary to confirm whether the PC side supports the verification of the voucher (otherwise the patient will not be able to receive the treatment if the registration is registered).
- Synergy Scheme: Hold a terminal collaboration meeting to clarify the changes of each terminal (such as the mini program is responsible for the development of the registration portal, the PC side is responsible for the development of the certificate verification interface, and the middle office is responsible for data storage adaptation), and formulate the “Cross-terminal Linkage List”.
3.3.2 Iteration
- End-to-end testing: Simulate the complete process (such as the patient registers in the mini program→ the doctor receives the consultation on the PC → issues the examination form→ the patient receives a reminder in the mini program→ the test results are synchronized to the PC and the mini program → large screen data update) to ensure that every link can be connected.
- Stress test: For high-frequency scenarios (such as the peak of registration at 8 a.m.), test the stability of data synchronization when multiple terminals are concurrent (such as 1000 people registering at the same time, whether the PC reception list is updated in real time, and there is no duplicate number source).
3.3.3 After iteration
- Feedback channels: Each terminal sets up a collaborative problem feedback entrance (such as the doctor’s PC can feedback that the mini program registration is not synchronized, and the patient’s mini program can feedback that the checklist has not received a reminder).
- Unified attribution: Classify the feedback and determine whether it is a single terminal problem or a linkage problem (if the checklist is not reminded, it may be that the PC side has not sent a message, or the mini program may not have received it, and it needs to be cross-ended), avoid changing only a single terminal and ignoring the linkage problem).
