Points of View Factors Required by Pharmaceutical Companies for Utilization of Medical Information

Printable PDF

The Office of Pharmaceutical Industry Research Natsuko Watanabe, Senior Researcher
The Office of Pharmaceutical Industry Research Norihiro Okada, Senior Researcher

Summary

1. Introduction

The utilization of medical ^{information1)} is an important theme that forms the foundation of modern medicine and is essential for improving medical services for patients. In the pharmaceutical industry, medical information is utilized in various situations throughout the entire value chain, including drug research and development, marketing, and post-marketing efficacy and safety evaluations, but the type and format of information required differs depending on the purpose. In particular, for effective drug development and innovative drug creation, it is important that the quality of medical information is guaranteed from the information construction stage, and a comprehensive infrastructure for utilization of medical information is required.

Looking at recent efforts to utilize medical information, the "Working Group on Secondary Use of Medical and Other Information, ^"2) which has been meeting since November 2023 under the "Study Group on Utilization of Health, Medical and Nursing Care Information," has focused on the development of a legal system and information coordination infrastructure to further promote the secondary use of medical and other information. The main agenda is the development of a legal system and information coordination infrastructure to further promote the secondary use of medical and other information. In addition, a technical working group on secondary use of medical and other information, which deals with technical issues such as how to develop linkage, will be convened under the working group from February 2024, and this is precisely the critical time for the utilization of medical information and the establishment of its infrastructure. However, the working group does not include members related to pharmaceutical companies, and there is a concern that the requests and issues of pharmaceutical companies for the development of medical information infrastructure may be overlooked.

In this paper, we will organize the elements required by pharmaceutical companies for utilizing medical information along the process of medical information formation, and confirm to what extent these elements are fulfilled in the major medical information databases in Japan. The purpose of this report is to clarify the gap between the ideal and reality of medical information utilization from the viewpoint of pharmaceutical companies and to deepen common understanding among various stakeholders in the current urgent need to improve the environment for medical information.

2. elements pharmaceutical companies require when utilizing medical information

The characteristics that pharmaceutical companies should have when utilizing medical information have been mentioned in various regulatory guidelines and publications, including "comprehensiveness, completeness, accuracy, consistency, and reliability" in the IMDRF ^document4) referenced by the FDA in ^{its guidance3} ) and "reliability, completeness, accuracy, consistency, and credibility" in the EMA ^framework5). ⁽ EMA framework5), "reliability, breadth, consistency, timeliness, relevance," and most recently, "completeness, consistency, accuracy, timeliness, stability, contextualization, representativeness, reliability, uniqueness," in a scientific article published in 2024 in "JMIR Medical Informatics " ⁶⁾.

The challenge is how to organically link these various characteristics and organize them into elements that enhance the value of medical information. In this study, we have grouped the above-mentioned characteristics into six elements to be taken into consideration by mutually grouping them with affinity or similarity in light of the following image diagram (Fig. 1) (Fig. 1 and Table 1).

The following is a description of the six factors to be considered in this grouping.

(1) "Traceability of patient information

Continuous collection, updating, and consistent management of medical information for a specific patient over time. This element allows for accurate tracking of a patient's treatment progress and changes in health status. However, it is difficult to keep track of an individual's health status in a single organization, such as a specific medical institution, and creating a system to keep track is costly. In addition, the cooperation of patients and medical institutions is essential to obtain information, and many challenges exist in the stages from generation to collection of information.

(2) "Linkage of various information" (3) "Representativeness of target patients

Integration of different types of medical information to enable a more detailed understanding and analysis of an individual patient's treatment process and health status. The information related to an individual's health condition is diverse, including the patient's medical records, test results, prescription history, medical checkup information, immunization information, and death information, etc. By linking these information, a comprehensive and complete picture of each patient's information can be formed. However, the scattered managers of the information and the lack of identifiers for consolidation are problems, and it is necessary for the database administrator to create a mechanism for efficiently consolidating the information held by patients themselves and by medical institutions.

(3) Representativeness of target patients

The medical information collected should be representative of the patient population for a particular disease or treatment. This element increases the external validity of the information obtained, minimizes bias, and allows generalizable conclusions to be drawn. For the medical information collected to be usable in a variety of settings, it is desirable to obtain information from a diverse patient population. However, as with traceability, the cost of creating a mechanism is enormous, and the need to obtain the approval of patients and medical institutions is problematic, so a mechanism must be created to facilitate the flow between medical institutions and database administrators.

(4) Standardization of information

Medical information collected from different data sources should be organized into a consistent format to enhance the uniformity of information. This element enhances the traceability, diversity, and population richness of medical information described earlier, ensures the accuracy, reliability, and validity of the information, and allows for comparison and analysis of information in various medical settings and studies. However, when each data source uses a different format or code system, the standardization process is complex and requires much coordination and cost to unify, requiring cooperation and agreement among relevant stakeholders from the information generation stage to the database creation stage.

(5) Protection of personal information

Patient privacy must be protected in the utilization of medical information. This element enhances the credibility and social credibility of pharmaceutical companies' activities by utilizing information that is processed in such a way that patients' personal information is anonymized or pseudonymized so that individuals cannot be identified. Date changes and age rounding are implemented through anonymization processing, and some data, such as genomic data and some image data, are deleted in the process of anonymization because they are personal identification codes. Since these processes may affect the accuracy of data analysis, it is necessary to devise ways to maintain the state in which personal identification is not possible without maintaining accuracy and reducing the value of medical information.

(6) "Improvement of accessibility of information

The medical information should remain quickly and easily accessible to users who require it. This element allows information to be provided in real time and enables users to make decisions based on the latest data. However, ensuring the immediacy and timeliness of information requires the development of technical infrastructure and the cost of more frequent updates. In addition, the system should also take into account aspects such as access methods to information, application procedures, access environment, and usage fees, to ensure that users can easily access information.

3. factors to be considered and their importance in pharmaceutical companies

The value chain of a pharmaceutical company can be generally described as a flow from drug discovery research and clinical development to marketing and sales, post-marketing surveillance (PMS), reexamination and reevaluation, etc. Medical information is used to a great extent throughout the value chain. So, in what phases of the pharmaceutical value chain are the six factors mentioned above necessary? We would like to reconfirm the importance of these factors by discussing specific situations in which they are used.

(1) "Traceability of patient information

In pharmaceutical research and development and pharmacovigilance, it is necessary to track the efficacy and safety of patients to whom drugs have been administered. Inevitably, these investigations require information on the patient's history prior to administration, the date of administration, and the capture of endpoints after administration, as well as the flow of time. These patient processes over time are called Patient Journey, and are increasingly expected to capture unmet medical needs and estimate the prognosis of natural treatments, which are important elements in all value chains of pharmaceutical companies.

(2) "Linkage of various information" (3) "Representativeness of target patients

Personal medical information in Japan has not been consolidated, and the information that is currently available is fragmentary in terms of information that represents individual patients. For example, municipalities that manage mortality information do not have information on medical history and treatment details maintained by medical institutions, and thus lack the parameters to build models to evaluate the impact on mortality. As in this case, machine learning models are used in almost all value chains in the pharmaceutical industry, such as AI development of medical images in drug discovery research and market forecasting in marketing, and the variety of parameters is an important factor in improving the accuracy of these forecasts and adjusting confounding factors.

(3) Representativeness of target patients

In order to make decisions using medical information, it is important that the results obtained from the patient population included in the medical information used can be extrapolated to the patient population assumed in the study. At present, most of the medical institutions that have medical information available for use by pharmaceutical companies are large medical institutions, and the proportion of patients with serious conditions tends to be high. If such information is used to survey the actual status of drug treatment, it may not be possible to correctly estimate the overall usage trend. The availability of information with a high level of exclusivity will ensure that the number of patients is sufficient for precise segmentation under a variety of conditions.

(4) Standardization of information

In order to make decisions using medical information, the accuracy of the results estimated from the medical information must be high. In order to increase the representativeness of target patients, it is necessary to analyze information from multiple medical institutions together. However, it can be difficult to obtain correct estimation results due to the fact that information is obtained under different conditions and the definitions of medical terms used in such cases. In particular, it is important to use high-quality information to make decisions regarding pharmaceuticals, because decisions regarding pharmaceuticals involve patients' access to medicines and safety measures.

(5) Protection of personal information

Anonymity processing, which is performed to protect patients' personal information, involves processing information in such a way that specific individuals cannot be identified. The processing of date information used in anonymous processing, as well as the rounding of numerical values for age and clinical test results, leads to a reduction in the accuracy of the estimated results. As with the standardization of information, the use of pseudonymized information, which does not process real data excluding information that identifies specific individuals, contributes to improving the accuracy of the analysis results, since a decrease in the accuracy of estimation is related to patients' access to medicines and safety measures.

(6) "Improvement of accessibility of information

In order to consolidate the aforementioned information and improve the representativeness of target patients and secure the volume of information, pharmaceutical companies will need to promote collaboration with multiple organizations that manage data, but it is difficult for a stand-alone project of a pharmaceutical company to become a hub for information collaboration due to time and cost constraints. Creating an environment where information can be accessed quickly when it is needed will lead to earlier decision-making and provision of information to patients. The overall improvement in accessibility of medical information from pharmaceutical companies is expected to make it possible to use the information throughout the value chain, which was not economically rational and could not be used in a stand-alone project.

Current status of major medical information databases

So, what is the status of development in Japan's major medical information databases? Here, we will focus on four databases: NDB (Database of Receipt Information and Information on Specified Medical Checkups), MID-NET, databases provided by providers certified under the Next Generation Medical Infrastructure Act, and databases provided by private companies. The following matrix (Table 2) shows the degree to which each database satisfies the six factors mentioned earlier.

(1) NDB (Database of Receipt Information and Information on Specific Health Checkups, etc.)

The greatest strength of the NDB is its complete coverage of patients. Under Japan's universal health insurance system, information is collected on a near-total number of patients, regardless of insurance type (employee health insurance, national health insurance, or late-stage medical care system for the elderly). The standardization of information also ensures consistency when information from different regions and medical institutions is integrated for use. 2022 marks the beginning of the assignment of an identifier called ID5, which uniquely traces individuals and enables long-term tracking of patient information ^.8) In terms of "linkage of diverse information," while currently limited to receipts and information on specific medical examinations, efforts are underway to link with public databases such as the Nursing Care DB and DPCDB, as well as with the Next Generation Medical Infrastructure Act, with death information scheduled to be included by ^FY20249).

On the other hand, there are issues with the accessibility of information for pharmaceutical companies. Strict regulations and procedures exist for obtaining and using information, making it difficult to access necessary information quickly, and the information provision format and access procedures are complex, which can be a hurdle for users.

A review of the NDB provision system is already under consideration. For example, the review method is being reviewed to shorten the processing time from application to provision from the current average of 390 days to 7 days in principle, and a mechanism for remote access through the Healthcare and Integrated Care Data Analysis Platform (HIC) is under construction, but the rapid provision of NDB is conditional in that the volume of information is limited. However, the rapid provision of the NDB is subject to conditions such as limited information volume. In addition, with a view to linking with other databases, the application and screening methods differ for each database, and if the various procedures are followed for each database, it is expected to take some time before the use of the database can begin.

(2) MID-NET

The types of information in MID-NET are diverse, including electronic medical records, receipts, and DPC data, and in particular, the availability of clinical laboratory test results, each of which can be consolidated for use. As for other medical information databases, linkage with NCDA operated by the National Hospital Organization began in October 2023, and some data is now ^available10). Another feature of the MRDA is that it provides highly reliable information for which regular standardization has been implemented through a unique mechanism called MRDA.

On the other hand, there are some issues to be addressed. First, regarding linkage with NDB, the PMDA system is not structured to collect the hash-generated information necessary for linkage, so a large-scale modification of the MID-NET cooperating medical institution system is ^required11). Regarding "representativeness of target patients," 10 hospitals of the Tokushukai Group will be added to the current 10 cooperating medical institutions consisting of 7 university hospitals and 3 medical institution groups, but there are limitations in targeting patients in the chronic stage because they are mainly university hospitals and acute stage hospitals, and it is difficult to draw generalizable conclusions on a nationwide scale. It is difficult to draw generalizable conclusions on a national scale. Furthermore, because of the limited number of cooperating medical institutions, there are also issues with traceability to other medical institutions (e.g., before hospital visit, after transfer, etc.) As mentioned above, MID-NET provides highly reliable information, but expanding the size of the patient population will require additional costs and resources, which may spill over into increased user fees, etc. Regarding the accessibility of information, MID-NET users are subject to prior screening regarding the purpose and environment of use, and can only view and analyze the information in limited locations.

In order to solve these issues, standardization at the information generation stage is required, taking secondary use into consideration. Unification of electronic medical record information by medical ^DX12) would help. One of the three pillars of the MID-NET improvement plan is the improvement of convenience, and the revision of guidelines for utilization is under consideration.

(iii) Databases provided by the Next Generation Medical Infrastructure Act certified providers

The databases provided by the Next Generation Medical Infrastructure Act's authorized providers are characterized by their ability to collate and link a variety of information (medical record information, image information, medical checkup information, etc.). In addition, consolidated analysis of public databases such as NDB, Nursing Care DB, and DPCDB is also possible. Furthermore, the ability to create and provide pseudonymized medical information is a major feature not found in other databases.

However, several issues also exist. First, in linking with other databases, it is limited to anonymized processed medical information when linking with NDB, etc., because it also depends on the method of personal information protection in other ^databases13). Regarding "representativeness of target patients," as of April 2024, there are 120 cooperative medical information service providers that provide medical information, distributed in 35 ^{prefectures14)}. As with MID-NET, there are issues in the traceability of other medical institutions, and in terms of accessibility of information, the use of pseudonym-processed medical information requires certification by the government from the perspective of personal information protection. For the accessibility of information, the use of pseudonymized processed medical information requires the acquisition of certification by the government from the perspective of personal information protection, and safety management measures and other obligations are imposed under examination of certification criteria. In terms of "protection of personal information," the system is different from other databases in that medical institutions and other businesses handling medical information notify the person or his/her bereaved family in writing, etc., that medical information will be provided to certified businesses, and if the person does not refuse, the medical information can be provided. This system is different from other databases. There are various opinions on whether this system leads to the protection of the rights of the citizens who provide the information. If more people refuse to provide information, it may affect the representativeness of the target patients by reducing the sample size and biasing the information, and it may also lead to an administrative burden on medical institutions and local governments by accommodating opt-outs.

For the above issues that still remain after the implementation of the revised Next Generation Medical Infrastructure Act, it is expected that discussions will deepen and new measures and systems will be improved in the future. In particular, the introduction of new technologies and methodologies to strike a balance between the protection of personal information and the usefulness of information, the provision of support measures and incentives to promote the participation of medical institutions and local governments, the standardization of electronic medical record information through the promotion of medical DX, and the revision of certification standards and operational flexibility for the use of medical information processed under pseudonyms are also important. It is also important to further promote the utilization of medical information through these efforts.

(4) Private database providers

Private database providers in Japan include JMDC Corporation and Medical Data Vision, Inc. These databases are widely used by pharmaceutical companies because of their standardized processing, high immediacy of information, and high convenience, such as the provision of web tools and the ability to request analysis from the operators.

On the other hand, when the information is derived from insurers, it cannot be traced before or after the transfer of insurance association, and when the information is derived from medical institutions, it cannot be traced before or after the visit to the medical institution in question. In addition, because the database is limited to insurance associations and medical institutions with which the database operator has contracts, there are issues regarding the "representativeness of target patients," and from the perspective of "linking diverse information," although linking with public databases has been discussed, specific measures have not yet been considered.

Looking at the current major medical information databases, while they all have many elements that make them superior, it is difficult to say that there is a data environment that consistently realizes all the elements that should be considered. While it would be ideal to create a unified database that fulfills all the elements, this is currently not easy to do. Therefore, it is expected that the value of medical information will be enhanced by complementing and linking the strengths of the individual databases. More comprehensive and detailed patient information can be obtained by integrating different types of information. For example, causal analysis can be performed by linking past prescription information contained in receipts and DPC data with long-term clinical information after discharge or transfer from a hospital in electronic medical record information. In fact, there have been cases in which electronic medical records, DPC data, and receipts have been used to understand the actual status of treatment and to search for causal factors related to disease progression, ¹³⁾ and we will keep a close eye on the expansion of such cases in the future. The database of electronic medical records provided by MID-NET and other providers certified under the Next Generation Medical Infrastructure Act is at the stage of starting on a small scale and recruiting cooperating medical institutions. It is also hoped that new measures and improvements in the system will lead to further nationwide expansion and increased representativeness. In order to realize the six factors that we have mentioned, we would like to see the development of a legal system and information linkage infrastructure that is aware of database linkage from the stage when medical information is constructed.

5. conclusion

In this paper, we examine the factors that pharmaceutical companies require when utilizing medical information and summarize the current status of major medical information databases in Japan. Medical information plays an important role in pharmaceutical companies throughout the value chain, from drug discovery to post-marketing safety monitoring. If the necessary requirements are not met, problems related to the usefulness and reliability of the information may arise, creating a major bottleneck in value creation for pharmaceutical companies. However, the current medical information databases are scattered and inadequate in various elements, and it is clear that further improvements are needed.

The National Medical Information Platform, which is currently underway, has shown great promise, as evidenced by the fact that an evaluation related to the promotion of medical DX will be newly established in the FY2024 revision of medical ^fees15), but it is desirable to construct a platform that fully communicates the viewpoint of pharmaceutical companies and satisfies their needs as well. It is hoped that through cooperation and discussion among diverse stakeholders to further promote the utilization of medical information, an environment will be created in which pharmaceutical companies can promptly and accurately obtain useful information for drug development and improvement of treatment methods, thereby contributing to the improvement of medical services for patients and the enhancement of the health of the public.

Back to News List

TOP