Business efficiency and quality in the monthly report creation work that involves many people has been improved by using Timer Start and Parallel Processing.
Q. What kind of operations do you use Questetra for?
Social medical corporation Saneikai Tsukazaki Hospital is a general hospital located in Himeji City, Hyogo Prefecture. There are 22 medical departments as well as various surgeries and internal treatments, including ophthalmology, urology, radiology, anaesthesiology, emergency, and rehabilitation, etc.
One of the departments of our hospital that receives particular attention is ophthalmology. About 8,000 surgeries are performed a year and they are engaged in various academic studies.
In the 2020 fiscal year (April 2020 to March 2021) 16 types of surgery were performed, including cataract, vitrectomy, and glaucoma surgery, with a total of 3,674 cataract surgeries performed annually. In terms of academic research, in addition to conducting academic societies and symposiums, we have many academic achievements such as the publication of papers by doctors, orthoptists, nurses, and pharmacists.
We use Questetra to create monthly reports distributed within the ophthalmology department.
The monthly report provides information on the status of surgeries, changes in complication rates, achievement rates, and considerations in each field of ophthalmology such as lacrimal passages, eyelids, and axis misalignment. In addition to doctors in each field, more than a dozen people including designers and proofreaders are involved in creating the report.
Q. What were the challenges you faced in the report creation work?
The problem was that the burden of work for the editing designer who was in charge of conducting the work was heavy.
The designer will ask eight doctors to submit the data that will be included in the report. Then he created content based on the submitted data and asked each doctor to check the completed content. Therefore, communication took a lot of time and effort, such as requesting data submission, urging those who had not submitted data to submit it, and requesting confirmation of created content.
Another issue was that it took a long time to complete the report due to the time and effort required.
Q. How did you solve the problem?
With Questetra, business personnel’s tasks are now automatically assigned in the order they should be processed. As a result, each staff member now only has to handle their assigned tasks. This eliminates the need for designers to request submissions or confirmations, reducing the workload and allowing them to concentrate on their design work.
Also, by using concurrent processing items, tasks related to data collection can now be assigned to all doctors at the same time. As a result, the time to complete the report has been shortened compared to when the designer used to request the submissions individually.
Work handover has also become easier. Questetra itself, in which business progresses according to the workflow diagram, plays the role of a business manual. As a result, it is no longer necessary to explain the work to new staff, and the burden of handing over work has been reduced.
Q. What are you planning to do via Questetra in the future?
We are now considering that Questetra can be used in the process of machine learning in the research and development of artificial intelligence, which we are focusing on.
In our ophthalmology department more than 3 million images and clinical data of more than 70,000 patients have been obtained, and we are developing various solutions by applying this data to machine learning. For example, we have developed an AI that determines whether an eye is left or right by analyzing the image of the eye. Usually, a nurse visually confirms which of the patient’s eyes will be operated on. In addition to this, at our hospital we use a tablet to take pictures of the eyes of patients who have their faces covered with a cloth during surgery and use this AI to check which eye is the target of surgery. This minimizes the risk of misidentifying the eye to be operated on.
Machine learning is the process of finding rules from a large amount of data, and mainly consists of three processes: data preprocessing, model learning, and model verification. Many of the tasks in these processes involve reading images and other tasks that are performed in a set procedure.
In such routine work, we believe that process management is important in order to reduce errors, unevenness, and waste, and to further improve work quality, so we are also considering how we can utilize Questetra in the future.
