Yi-Bo Wu,Jin-Yi Meng,Hong-Xue Chen,Dong-Yun Li,Shu-Xian Xu,Tao Wang

1Department of Pharmacy Administration and Clinical Pharmacy,School of Pharmaceutical Sciences,Peking University,Beijing,China.2College of Humanities and Social Sciences Harbin Medical University,Harbin,China.3School of Pharmacy,Shenyang Pharmaceutical University,Benxi,China.4Department of Emergency,Shanghai East Hospital,Tongji University School of Medicine,Shanghai,China.5College of Art and Media,Tongji University,Shanghai,China.

Abstract Hospital emergency management work is of great importance which not only related to the life and health of the people but also related to social stability.It is necessary to construct and improve the patient satisfaction assessment system so that doctors and patients can interact in a timely and effective manner,and the hospital can alert or respond promptly.This study constructs a closed-loop emergency management system consisting of patient-end,data center and decision-making-end,and apply the feedback information to hospital emergency management.In the preliminary applications,the system has a significant effect on the hospital emergency management mechanism.The researchers will improve the system in the follow-up study.

Keyword:Hospital Emergency Management,Patient Satisfaction,Feedback Mechanism.

Background

In 2006,the State Council issued a series of relevant documents,such as the“Opinions on Comprehensively Strengthening Emergency Management Work”,and provided guidance on the implementation of emergency management work in various places.The working performance of hospital emergency management is not only related to the life and health of people,but also related to social stability [1].In medical institutions,it is important to apply the information provided by the patient satisfaction assessment to emergency management.Chinese patient satisfaction assessment started in the late 1980s [2].Patient satisfaction assessment is a method used by hospitals to evaluate medical services,which can reflect the overall experience of patients in the medical process[3]Because the patient fills out the assessment form as a task;most scales are lack of strict psychometric test;hospital feedback period is long,etc.[4-5],the patient satisfaction assessment system can’t get the patient’s information and solve the problem.And the problem is likely to become an emergency.Therefore,it is necessary to construct and complete the patient satisfaction assessment system so that both doctors and patients can interact in a timely and effective manner,and the hospital can alert or respond promptly.This study initially constructed an intelligent patient satisfaction assessment system to obtain and analyze the patient satisfaction information in real time and applied the feedback to emergency management to achieve the ability of early warning and emergency response.

Methods

System construction process

This study analyzes domestic and foreign patient satisfaction assessment systems and effective feedback mechanisms through former literature [6-13].Base on this,we developed an intelligent patient satisfaction assessment system.Through the consultation with experts (including relevant experts who work in hospitals,medical schools,and the health management system),the assessment system was adjusted.And then the patient satisfaction intelligent assessment system consists of an emergency management system was formed initially.It was used in over 200 hospitals in Shanghai.During the use process,the assessment system is continuously improved through users’feedback.

Evaluation of the result

This study will evaluate the system by the number of how many hospitals put it into use,the number of valid questionnaires,the number of complaints,the time to process complaints,the participation of patients,and the rate of positive feedback.

Results

Intelligent patient satisfaction assessment system

Patient-end.The system consists of patient-end,data center and decision-making-end (Figure 1).The satisfaction assessment is conducted through a web platform.Departments in the hospital post QR code for satisfaction assessment on the wall.Each department has a unique QR code to streamline the patient's answer process.The medical staffs guide the patient or patient’s family to scan and enter the system interface.According to the actual situation of different hospitals,the constructed intelligent patient satisfaction assessment system is slightly different.This study takes the assessment system of a certain grade 2A general hospital in Shanghai as an example to explain the user interface of the Patient-end.

“Do a questionnaire”is linked to the QR code of the WeChat official account of Patient Service Center.After subscribing,users click on the automatically pop-up message to enter the satisfaction assessment questionnaire.The questionnaire is divided into objective and subjective questions.The objective questions are divided into four dimensions include the overall situation,the service,the environment,and the medical staffs,allowing the patient to evaluate the various parts of the medical experience.The questions adopt the Likert five-level rating scale setting option.The options are very satisfied,quite satisfied,so-so,quite unsatisfied,very unsatisfied,and automatically correspond to a score of 5 to 1 in the backend.The subjective question invites the patient to make a holistic comment on the medical service so that the patient can clearly express their thoughts.“Write a comment”links to a more concise assessment interface,and the patient can make a comprehensive assessment and comment on the medical service directly.

The Patient-end also has other functions:①Ward Information Bar.At the top of the user interface is the ward represented by the current QR code,and the patient can confirm the current medical position accordingly;②Scientific Popularization Column.The information bar on the interface has science popularization posters and titles that are displayed cyclically.After clicking,it enters the body of the science popularization article and provides medical information to the patient;③Functional Buttons.There are three functional buttons in the interface.“Hospital Profile”link to the hospital's WeChat official account,users can view the hospital-related information in it.“Question-Answer” link to the Patient Service Center WeChat official account QR code and usage explanation,the users can submit questions about treatment and rehabilitation on it and the hospital will answer timely.The hospital will promptly contact the patient to assist them in solving the problem if it’s necessary.“Condition Tracking”links to the post-hospital condition tracking invitation letter;④Comments Review.The patient clicks on“All Comments” to view all comments and the rate of positive feedback for the hospital or a department.

Data center.The backend can count scores and analysis comments in real time.In the process of the statistical analysis of patient satisfaction,the data will be subjected to overall statistics and departmental statistics.The scores of the objective questions directly produce a visible outcome,and the subjective questions will be evaluated with an emotional sentiment analysis and a word frequency statistical plot(Figure 2).

The Baidu AI platform is used to analyze the emotion tendency,and the model is trained and generated by uploading the artificially labeled positive and negative emotion tendency samples.Use the Java language to call the model through the application programming interface,continuously send the comment data in the form of scripts,and report the returned data to the Excel worksheet,then use MATLAB to read the table and count the proportion of positive and negative emotion data and evaluate confidence.Then,get the final result of the analysis of emotion tendency.

Use Python for word frequency statistics and plotting.After the text is pre-processed,use the Jieba and Collections extensions for word segmentation and word frequency statistics to get the entire comment data and the top ten high-frequency vocabularies in each department's comments.The word frequency histogram of the high-frequency vocabulary is drawn by the Matplotlib extension package,and the word frequency cloud map of the comment data is generated using the Wordcloud extension package.

Decision-making-end.According to the data obtained by the Data Center,it is summarized according to the following 7 types of emergencies (Table 1).From the perspective of emergency management,the data is urgently evaluated for decision-making.If the emergency need response immediately,the staff will inform the medical staffs and the relevant person who is in charge of the logistics department to process the emergency.The result and time of the response will be recorded.

Usage report

Since the establishment of the intelligent assessment system for patient satisfaction,to December 2018,the intelligent patient satisfaction assessment system with emergency management system has covered 14 administrative districts in Shanghai,more than 200 public medical institutions including three grade 3A hospitals,as well as ten private medical institutions;the system receives more than 1.04 million opinions and suggestions from the patients in the pilot hospitals,effectively handling 3,402 complaints,and the average processing time was 12 minutes,alleviating potential doctor-patient conflicts 1941 times.The participation of the patients has increased by 29.3 percentage points.The rate of positive feedback of the pilot hospitals has increased to 94.3%,and the rate of users’ experience has increased by nearly 8%(Table 2).

Discussion

Table 1 Catalog of Incident

The patient satisfaction intelligent assessment system constructed in this study can achieve emergency management in hospitals.This study uses the internet platform to establish a satisfaction assessment system and builds a closed-loop emergency management system consisting of Patient-end,Data Center and Decision-making-end.The Patient-end uses the mobile phone terminal for the questionnaire to improve convenience so that the patient can timely send the satisfaction assessment to the hospital.As soon as the data center receives the patient’s satisfaction objective score and subjective opinions,the analysis will be conducted so that the hospital can know the problem or potential problem.The Decision-making-end guarantees that the hospital will take action to solve the problem or potential problem in good time.

The emergency management mechanism of the patient satisfaction assessment system constructed in this study has a significant effect.As a result,different types of medical institutions covered by the study can be timely and effective in handling complaints,and the participation rate of patients and the rate of positive feedbacks have been significantly improved.Through the patient satisfaction assessment system,patients and their families can talk to the hospitals at any time,so that hospitals can solve potential or existing problems in a timely manner.This timely and effective feedback mechanism can improve the participation of the patient and the rate of positive feedbacks for the hospital,and reduce the risk of emergencies.

However,there are certain limitations in the research.The expert consultation for the system design adopts the form of conference discussion but does not adopt a more rigorous method such as Delphi,and there might be cluster psychology [14] which will lead to certain deviation.For the evaluation of the result,the research only conducted the macro dimension,which does not pay enough attention to the individual.Researchers will try to find solutions to these issues in subsequent research to improve the system.