System and method for predicting supply and demand in clinical trials

Abstract

To provide systems and methods for demand and supply forecasting for clinical trials.SOLUTION: In a clinical trial supply forecasting system, a method carried out by a supply forecasting system implemented in the form of one or more servers, databases, or the like that are electronically accessible to users may include receiving, from a user, one or more electronic files that contain parameters of a clinical trial, the parameters comprising a quantity of patients, a plurality of sites, and one or more confidence values, calculating according to the quantity of patients, a respective demand profile for each of the plurality of sites, calculating according to the one or more confidence values, a buffer quantity for each of the plurality of sites, calculating a supply plan for the clinical trial according to the demand profiles and the buffer quantities, and transmitting the supply plan to the user.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims the benefit of priority of U.S. Provisional Patent Application No. 62 / 563,283, filed on September 26, 2017, and U.S. Utility Patent Application No. 16 / 143,049, filed on September 26, 2018, and all of its disclosure content is incorporated herein by reference.

[0002] Each embodiment described herein relates to the supply and management of clinical trials, and more particularly, to systems and methods for predicting the supply and demand of clinical trials.

Background Art

[0003] During the past decade, a plurality of systems for managing the pharmaceutical supply of clinical trials have been developed. Currently, many pharmaceutical companies rely on such systems for formulating production plans, optimizing costs, managing subject recruitment activities, etc. Such systems include, for example, independently created Excel spreadsheets or Excel spreadsheet templates, prediction systems based on average values, systems using Monte Carlo simulations, etc. Generally, systems using Monte Carlo simulations are considered to be the most accurate.

Summary of the Invention

Means for Solving the Problems

[0004] Each embodiment described in this specification relates to a system and method for predicting the supply and demand of clinical trials. One embodiment of a computer-implemented method includes: receiving, by a server, from a user, one or more electronic files including parameters of a clinical trial, including the number of patients, a plurality of implementation facilities, and one or more reliability values; calculating, by the server, a demand profile for each of the plurality of implementation facilities according to the number of patients; calculating, by the server, a buffer amount for each of the plurality of implementation facilities according to the one or more reliability values; calculating, by the server, a supply plan for the clinical trial according to the demand profile and the buffer amount; and transmitting, by the server, the supply plan to the user.

[0005] One embodiment of a computer-implemented method includes: receiving, by a server, from a user, one or more electronic files including parameters of a clinical trial, including the number of patients, a plurality of implementation facilities, and one or more reliability values; calculating, by the server, a supply plan for the clinical trial according to the parameters; transmitting, by the server, the supply plan to the user; after transmitting the supply plan to the user, establishing electronic communication with a computer operating the clinical trial and a randomization and trial supply management (RTSM) system; outputting a command to the RTSM system via an application programming interface (API) of the RTSM system; receiving, in response to the command, data from the RTSM system; identifying, according to the data received from the RTSM system, the situation of the clinical trial; comparing the situation of the clinical trial with the supply plan; determining that the clinical trial is not in line with the supply plan; and automatically transmitting to the user an electronic notification including something indicating that the clinical trial is not in line with the supply plan.

[0006] One embodiment of a method implemented on a computer can include the server receiving from a user one or more electronic files including an electronic spreadsheet file and one or more text files, the electronic spreadsheet file including parameters of a clinical trial including the number of patients, a plurality of implementation facilities, and one or more reliability values; converting data in the electronic spreadsheet file into one or more tables in an electronic text format; inserting the one or more tables in the electronic text format into the one or more text files to create one or more completed text files; performing natural language processing on the one or more text files to identify parameters of the clinical trial; calculating, by the server, a supply plan for the clinical trial according to the parameters; and transmitting, by the server, the supply plan to the user.

[0007] One embodiment of a non-transitory computer-readable medium can store instructions that, when executed by a processor, cause the processor to receive from a user one or more electronic files including parameters of a clinical trial including the number of patients, a plurality of implementation facilities, and one or more reliability values; calculate a demand profile for each of the plurality of implementation facilities according to the number of patients; calculate a buffer quantity for each of the plurality of implementation facilities according to the one or more reliability values; calculate, by the server, a supply plan for the clinical trial according to the demand profile and the buffer quantity; and transmit the supply plan to the user.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3A

Figure 3B

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

DETAILED DESCRIPTION OF THE INVENTION

[0009] The RTSM system is a system that can predict the demand for clinical material supply at a clinical trial facility, which is a facility where patients participate in a clinical trial, pre-distribute appropriate drugs to the clinical trial facility in preparation for administration, and additionally send drugs to the clinical trial facility as needed. Usually, this function is called the "resupply algorithm". The control of the resupply algorithm can be performed by parameters such as the minimum level (safety stock), the maximum level (resupply level), the short-window, and the look-ahead window. However, for a pharmaceutical company, the management of these parameters - that is, calculating the predicted parameter values around the world for a specific clinical trial and further tracking and updating those predicted parameter values - is a heavy burden.

[0010] Over the past decade, predictive tools have been developed for a certain clinical trial to help pre-plan supply needs and continuously adjust resupply parameters (and project future prospects). However, existing prediction systems have many drawbacks. For example, some of the existing predictive tools use simulation technology as the basis of their predictive power. Such predictive tools usually need to incorporate the same or similar parameters for a variety of RTSM products because they need to approximate the behavior of the RTSM resupply algorithm to be predicted. However, it is almost impossible for such predictive tools to perfectly match the RTSM resupply algorithm to be predicted. As a result, when using known prediction systems, important supply decisions may be made based on inappropriate information (for example, predictions that do not match the actual usage in clinical trials due to algorithm differences between the prediction system and the management system), which may lead to wasteful over-supplies and stockouts. Furthermore, usually, the RTSM system and the clinical trial supply management system are completely separate systems. Therefore, during the implementation of a clinical trial, predictions at the late stage of the trial are not performed very frequently (about once every 1 - 6 months), and when making predictions, users need to manually map between the clinical trial models of these two systems. Also, reporting in the current system is strictly regulated, and data reuse can usually only be done outside external systems such as Excel. Moreover, many existing systems do not obtain certifications or comply with CFR Part 11 - that is, they do not comply with what clinical trial software generally complies with - so there is a risk of some non-conformities or inappropriate proposals, and important supply decisions may be made based on such inappropriate proposals.

[0011] Generally, when a prediction system models the required supply during a pharmaceutical clinical trial, it is necessary to consider a large number of variables that are correlated with each other. For example, the patient's clinic visit schedule and the drug administration schedule must be considered. If there are changes in the clinic visit schedule of the clinical trial protocol that the patient may follow during the clinical trial period, those changes are also included in the considerations. In addition, it is necessary to consider the temporal and spatial characteristics of the number of patients, such as the patient enrollment time, the location and start time of the clinical trial facility. Furthermore, when making predictions, it is necessary to consider the logistics of the supply network, which includes matters such as the means of delivering inventory materials to the clinical trial facility, the shipping frequency of inventory materials, and the time required for the inventory materials to reach the clinical trial facility. Also, country-specific restrictions (such as labeling requirements) and restrictions regarding inventory materials such as the production cost per unit and its price, the quality retention period, and the expiration date of each lot must be considered.

[0012] Resupply algorithms (and predictors) typically handle two types of demand: predictable demand and unpredictable demand. Predictable demand encompasses the necessary quantities that existing known patients will consume at any future time, and is demand for which the type of treatment and dosage level are known in advance. On the other hand, unpredictable demand encompasses the necessary quantities that unknown (non-existing) patients will consume ("uncorrelated unpredictable demand"), known patients for whom treatment assignment by randomization has not yet been performed, or known patients in a clinical trial design where the overall treatment process has been assigned but future dosing variations are anticipated ("correlated unpredictable demand").

[0013] Typically, predictable demand can be appropriately handled by an excellent RTSM predictor. However, unpredictable demand is demand that is unknown beforehand, so only the most robust and powerful predictors can handle it. Regarding unpredictable demand, a simulator can be used to obtain statistically realistic upper limits within a certain time frame, and the resulting upper limits can be used to generate safety stock or stack buffer stock (stock placed at a clinical trial facility in advance even if the potential for future use is not clear) from the floor to the ceiling. However, usually, such buffer stock and settings have been manually set through a labor-intensive process.

[0014] There are two important time frames in RTSM. The first is sometimes called the "short window," which is usually equal to or slightly longer than the shipping lead time from the supply point to the target clinical trial facility. For example, by setting all or almost all of the expected needs within the short window to be covered by buffer stock or safety stock, stockouts and medication shortages for patients can be prevented. However, it may not be desirable to ship for replenishment every time the drug is consumed. Therefore, the second time frame, sometimes called the "long-ahead window" or "long window," instructs the system on the period for which stock should be held until the next shipment is required. For example, the long window may be set to 30 days.

[0015] If an excellent prediction tool is always synchronized with an excellent RTSM system (by constantly transmitting data), the prediction tool can continuously provide advice to the company as the clinical trial requester for setting parameter values as described above. However, this can only happen in such situations. The complexity of many aspects of this process is solely based on a deep understanding of the resupply algorithm and its parameters, and it was not manageable for many clinical supply experts (or other persons in charge of this duty).

[0016] The supply prediction system according to the present invention can improve some or all of the above-mentioned defects.

[0017] Referring to the drawings here, the same or similar features (functions) among the drawings are represented by the same reference numerals. FIG. 1 is a schematic diagram of an exemplary system 10 for predicting and managing clinical trials such as randomized pharmaceutical clinical trials. The system 10 can include a supply prediction system 12 and a randomization and trial supply management (RTSM) system 14. Either or both of the supply prediction system 12 and the RTSM system 14 can perform electronic communication with a plurality of user computing devices 16 (only one of which is illustrated in FIG. 1).

[0018] Generally, the prediction system 12 and the RTSM system 14 are considered to be involved in different aspects of the planning and implementation of pharmaceutical clinical trials. As will be described in more detail below, generally, the RTSM system 14 can provide guidelines and conduct clinical trials, while the prediction system 12 can formulate clinical trial plans. Here, the formulation of a clinical trial plan is carried out before complete information about the clinical trial is available, or for a clinical trial in a stage where complete information is still not available.

[0019] In various embodiments, the RTSM system 14 includes one or more computing devices and can perform many operations for running and directing pharmaceutical clinical trials, such as randomizing patients, blinding, drug administration, and automatic drug replenishment. The RTSM system 14 can communicate electronically with one or more user computing devices 16, receive inputs from users (e.g., clinical trial parameters), and provide outputs to users (e.g., orders for one or more drugs, status of patients participating in a clinical trial, information about each clinical trial facility).

[0020] The RTSM system 14 can be configured to communicate and exchange information with a variety of users (i.e., communicate with one or more user computing devices of each of the various users). For example, in a clinical trial, the RTSM system 14 can provide an input / output interface for a clinical trial sponsor 18 (such as a pharmaceutical company). In each embodiment, the clinical trial sponsor 18 can input into the RTSM system 14 the complete parameters of a clinical trial or the complete parameters of a stage in a clinical trial. Such a complete set of parameters can include, for example, actions regarding patients such as outpatient visits, lots and lot operations, inventory and inventory operations at the clinical trial facility level or site level, shipments and shipment operations, temperature excursions and their management, returns of dispensed or undispensed drugs and approvals and parameters related to drug returns, inventory disposal status, responses to warnings and notifications, and addition and parameterization of clinical trial facilities, supply sites, and countries of implementation, setting of prediction parameters by means of reliability dials and long-term time frame dials, setting of subject enrollment rates in a group of clinical trial facilities, setting of users and their access rights, setting of subject cohorts, loading of kit-randomization lists, unblinding of patients and kits, etc. Next, the RTSM system 14 can determine which restocking actions are necessary and output them to the clinical trial sponsor. Also, as will be described later, the RTSM system 14 can track and record the status of a clinical trial based on data from one or more supply sites, clinical trial facilities, physicians, and patients, and make that information available to the clinical trial sponsor.

[0021] The RTSM system 14 can perform electronic communication with one or more clinical trial facilities 20, such as a hospital or a clinic. For example, the RTSM system 14 can output instructions, such as dosing instructions for a specific patient, to the clinical trial facility 20. Also, the RTSM system 14 can receive data such as patient information and dosing records from the clinical trial facility 20 and hold data indicating which patient took which drug related to the clinical trial at which trial facility and when.

[0022] Furthermore, the RTSM system 14 can also communicate with one or more distribution facilities 22. The distribution facility 22 is a facility such as a supply base or a warehouse used to distribute drug lots and placebos used in clinical trials to the clinical trial facility 20. The RTSM system 14 can issue, for example, a shipping instruction to the distribution facility 22. In each embodiment, the RTSM system 14 can automatically generate, automatically transmit, or automatically generate and transmit such a shipping instruction based on a supply plan. Also, the RTSM system 14 can receive data such as shipping records and inventory information from the distribution facility 22 and hold data indicating the distribution quantity, remaining inventory quantity, and location of the inventory materials for the drugs used in the clinical trial.

[0023] Also, one or more physicians 24 can perform electronic communication with the RTSM system 14. The RTSM system 14 can output information about the drugs used in the clinical trial and information about the clinical trial to such a physician 24 so that the physician can disclose information about the clinical trial to the patient if necessary. Also, the RTSM system 14 can receive subject registration information (such as the number of patients recommended by the physician to participate in the clinical trial, the characteristics of those patients, and specific information of the registered patients) from such a physician and hold data indicating the number of registrants and the characteristics of the registered patients.

[0024] The supply prediction system 12 can include a processor 28 and a non-transitory computer-readable memory 30. When executed by the processor 28, the memory 30 contains instructions that cause the processor 28 to execute one or more methods, algorithms, processes, etc. of the present invention. The memory 30 can store one or more modules in the form of instructions executable by the processor 28 (such as software, etc.). Exemplary modules will be described below.

[0025] In each embodiment, the memory 30 can store a supply plan formulation module 32. Based on this supply plan formulation module 32, the supply prediction system can formulate a general supply plan for the drugs for clinical trials and the placebo used in the clinical trials before complete information of the clinical trial (or one stage in the clinical trial) becomes available. Then, the supply prediction system 12 outputs this supply plan for the clinical trial to the clinical trial requester 18, enabling the clinical trial requester 18 to appropriately set the budget for the clinical trial and ensure an implementation facility, a logistics channel, etc. suitable for the clinical trial. To realize such information exchange, the supply prediction system 12 can conduct electronic communication with one or more user computing devices 16 including the user computing device 16 of each clinical trial requester 18.

[0026] In each embodiment, the memory 30 can also store a specification interpretation module 34. The user can submit information indicating the parameters of the clinical trial (e.g., via a user computing device that may be associated with the clinical trial requester as described above), and the supply prediction system 12 can receive it. The specification interpretation module 34 can process and analyze the received information to extract and identify the clinical trial parameters. Then, based on the clinical trial parameters received from the user, the supply plan formulation module 32 can generate a supply plan.

[0027] The prediction system 12 can provide an electronic user interface for use by clinical trial requesters, one or more clinical trial facilities, one or more supply bases, one or more physicians, and one or more patients. For example, in each embodiment, the prediction system 12 can provide an electronic user interface with one or more dials, and the user can set certain constraints on the supply plan calculated by the prediction system by adjusting the dials. Based on the user's operation on an input means such as such a dial, the supply prediction system 12 can receive user input as described below.

[0028] In some embodiments, the prediction system 12 can receive from the user a reliability level indicating, for example, how much confidence should be placed in the predicted total demand (e.g., whether it is predictable demand, correlated unpredictable demand, uncorrelated unpredictable demand, etc.). In some embodiments, the user can also present a plurality of different reliability levels, in which case the prediction system can calculate a supply plan prediction for each of those reliability levels. In each embodiment, the prediction system can receive from the user a number of reliability levels set for each of various clinical trial parameters, such as, for example, a first reliability level for the total demand, a second reliability level for correlated unpredictable demand, etc.

[0029] Also, the prediction system can receive from the user one or more clinical trial durations. For example, the user can present the length of the clinical trial period, the length of one or more specific stages of the clinical trial, etc. In some embodiments, the user can also present any different duration lengths. In that case, the prediction system can calculate a supply plan prediction for each of these different duration lengths.

[0030] The supply prediction system 12 can also communicate electronically with the RTSM system 14. During a clinical trial, the supply prediction system 12 can retrieve and read data on each clinical trial from the RTSM system 14. To perform such information retrieval, the supply prediction system 12 can communicate with the RTSM system 14 via the application programming interface (API) 36 of the RTSM system 14. Then, based on the plan comparison module 36, the supply prediction system 12 can compare the information on the status of the read clinical trial with the predicted supply plan, and if the status of the clinical trial deviates significantly from the predicted supply plan, it can generate a warning to the clinical trial requester.

[0031] In one embodiment, one or both of the RTSM system 14 and the supply prediction system 12 can be implemented on a Software-as-a-Service (SaaS) basis. Accordingly, at least one of the supply prediction system 12 and the RTSM system 14 can be realized in the form of one or more servers, databases, etc. that are electronically accessible to users via the Internet. In addition to or instead of this, one or both of the supply prediction system 12 and the RTSM system 14, or one or more aspects or functions of the supply prediction system 12 and the RTSM system 14, can also be implemented, for example, on a local server for the clinical trial requester, or as a local copy stored on each user computing device such as the clinical trial requester, the clinical trial facility, the supply base, etc.

[0032] Although the RTSM system 14 and the supply prediction system 12 are illustrated and described as separate systems, in each embodiment, the RTSM system 14 and the supply prediction system 12, or some of their functions, can also be implemented with one system or common computing resources.

[0033] According to one or more aspects of the present invention, it is possible to provide a supply prediction system that is superior to known supply prediction methods and systems. First, known supply prediction systems are typically used to calculate an overall prediction for an entire clinical trial. And such prediction systems typically require information on a large number of clinical trials to formulate an overall plan. As a result, in the very early stage of the clinical trial planning process where relatively little information is considered to be available regarding the clinical trial, there is a possibility that known prediction systems may not be very useful, and many assumptions had to be made to use known prediction systems at this stage. On the other hand, the supply prediction system according to the present invention can be configured in each embodiment to calculate a supply plan or a part thereof based on any information available to the user, and further, the supply prediction system according to the present invention can identify specific milestones (such as a date, the availability of specific information, etc.) that can trigger the recalculation of the supply plan or trigger the calculation of new aspects of the supply plan and notify the user, thereby enabling the user to change the plan during the clinical trial so as not to cause delays in clinical trials accompanied by high cost increases. The supply prediction system according to the present invention can provide the user with a method of flexibly transmitting different levels of information at the timing when the information becomes clear. Therefore, the supply prediction system according to the present invention enables the improvement of logistics during the implementation of pharmaceutical clinical trials by providing a more flexible and robust supply plan prediction, and enables more accurate prediction of the required amount of drugs, thereby enabling the avoidance of high-cost waste and the prediction of situations that become problems such as out-of-stock. For example, according to the supply prediction system according to the present invention, overage (the drugs required in a clinical trial exceeding the actual patient needs to cover uncertainties and inefficiencies in the supply chain) can be reduced by 20%.

[0034] Second, since a conventional prediction system is usually a separate system from the RTSM system, it is necessary to obtain information from the RTSM system through inefficient file transfers. Such file transfers usually require manual intervention by the user, and there is a possibility that more data exchange than necessary is performed during the transfer, both of which may be less efficient than the level desired by the user, and there is also a risk of inaccurate mapping between the prediction system and the RTSM system. On the other hand, the supply prediction system according to the present invention can be integrated into the RTSM system via an API published by the RTSM system in each embodiment. As a result, the supply prediction system according to the present invention does not perform inefficient file transfers, but can collect specific information required by the prediction system by issuing requests and commands to the RTSM system via the API, without requiring manual intervention. Therefore, according to the present invention, the functions and efficiency of a computer-implemented prediction system can be improved. Also, since data mapping can be almost automated, errors are less likely to occur than when performing data mapping through manual intervention.

[0035] Third, as described above, a conventional supply prediction system usually requires specific types of information for plan generation. Also, a conventional prediction system usually requires such information in a specific single format. On the other hand, the prediction system according to the present invention can accept each aspect of clinical trial information in a plurality of formats (for example, such as one or more text files and one or more spreadsheets) in each embodiment, and can process and analyze the information thus submitted to extract and identify clinical trial parameters that form the basis of the supply plan.

[0036] Figure 2 is a flowchart showing an exemplary operation method 40 of a clinical trial supply prediction system. Method 40, or one or more aspects of method 40, can be executed by a supply prediction system such as the supply prediction system 12 of FIG. 1. The method 40 of FIG. 2 includes various relatively high-level functions of the clinical trial supply prediction system. Details of some of these functions will be described later with reference to FIGS. 4, 5, 6, and 7.

[0037] Continuing to refer to FIG. 2, method 40 can include step 42 of receiving one or more specification files from a user. The user can be, for example, a clinical trial requester (such as a pharmaceutical company). The specification file may include one or more file formats. For example, in one embodiment, the specification file can include one or more text files and one or more spreadsheet files. And in each embodiment, the text input can be in a free (natural) form rather than in response to a structured prompt. The specification file may include information available to the clinical trial requester regarding the pharmaceutical clinical trial proposed by the clinical trial requester. Such information includes the start time and duration of the clinical trial, the desired number of patients, the expected value of the subject enrollment rate of patients in the clinical trial, the presence or absence of screening introduction and the failure rate of the screening, descriptions of patient treatment (in particular, drug administration, any events that can occur such as missed doses by patients and dose changes by patients, and what temporal and probabilistic correlations exist between those events), the number of clinical trial facilities, a list of countries where the trial is conducted, groups of clinical trial facilities, regions, supply bases, descriptions of the supply network including descriptions of supply lines and lead times, descriptions of the various drugs to be tested in the clinical trial, descriptions of each dose to be tested in the clinical trial, and drug supply constraints such as the expiration date of the drug, the residual quality retention period, and the maximum lot size (i.e., the maximum release frequency).

[0038] Method 40 may further include step 44 which includes converting and combining specification files and identifying clinical trial parameters included in the specification files by methods such as extraction. For an exemplary method for implementing step 44, reference is made to FIGS. 4 and 5 for description.

[0039] Method 40 may further include step 46 which includes formulating a clinical trial supply plan based on the clinical trial parameters identified from the specification. For an exemplary method for implementing step 46, reference is made to FIG. 5 for description. In step 46, one or more clinical trial supply plans can be determined. For example, multiple clinical trial supply plans can be determined for each of different reliability levels associated with the clinical trial period, or different reliability levels associated with the number of patients required for the clinical trial, or the number of variations of the dosage to be tested in the clinical trial.

[0040] Method 40 may further include step 48 which includes generating one or more graphic displays of the clinical trial supply plan and the inputs used for its calculation. FIGS. 3A and 3B show exemplary graphic displays 60, 70 of the clinical trial supply plan or the inputs used for its calculation. Referring to FIG. 3A, the first graphic display 60 may include a plurality of nodes 62 (in FIG. 3A, three nodes 62 1 , 62 2 , 62 3 are shown. For clarity of the figure, some of the nodes 62 are not labeled with reference numbers.). And each node 62 represents an action in the clinical trial represented by the graphic display 60, such as randomization of the subject group or administration of a drug. The graphic display may further include branches 64 extending from one or more of each node. Each branch 64 can indicate the result of the action represented by the node 62 at the root of the branch 64. For example, from a start node 62 1 representing randomization of the patient group, three branches 64 1 , 64 2 , 64 3 may extend. The first branch 641 shows that 50% of the randomized patient group is assigned to part ABC of the clinical trial, and the second branch 64 2 shows that 25% of the patient group is included in the placebo part of the trial, and the third branch 64 3 shows that 25% of the patient group is included in part XYZ of the trial.

[0041] Referring to FIG. 3B, the second graphic display 70 can include a plurality of nodes 72 (in FIG. 3B, three nodes 72 1 ,72 2 ,72 3 are shown. For clarity of the figure, some of the nodes 72 are not labeled with reference numbers.). And each node 72 represents a location or a collection of locations in a clinical trial, such as a facility involved in a clinical trial, such as a supply base or a clinical trial facility. For example, the first node 72 1 represents a supply base to the United States (the service area may include Canada), and the second node 72 2 represents a group of subjects in the United States, and the third node 72 3 represents a group of low-dose subjects in the United States. The graphic display can further include one or more branches 74 extending from each node 72. Each branch 74 can indicate a distribution channel for the drug for the portion included in the node 72 at the root of the branch 74. For example, from the node 72 representing a group of subjects in the United States, three branches 74 2 may extend. The first branch 74 1 ,74 2 ,74 3 When extending. The first branch 74 1 indicates that a part of the supply designated for the group of subjects in the United States reaches the group of low-dose subjects 72 in the United States 3 , the second branch 74 2 indicates that a part of the supply designated for the group of subjects in the United States reaches the group of medium-dose subjects in the United States, and the third branch 74 3 indicates that a part of the supply designated for the group of subjects in the United States reaches the group of high-dose subjects in the United States.

[0042] Referring back to FIG. 2, method 40 can further include step 50 of sending a clinical trial supply plan to a user. Also, sending the clinical trial supply plan to the user can include sending data of the supply plan including one or more graphic displays to the user computing device for display on the user computing device. Additionally or alternatively, sending the clinical trial supply plan to the user may include sending the supply plan or one or more portions thereof to the user in a tabular format, a text format, other formats, or any combination thereof. Also, for example, the transmission can be made in the form of a hosted web page delivered to the user computing device. Additionally or alternatively, the transmission can also be made in other electronic forms such as an email format or a downloadable electronic file.

[0043] Method 40 can further include step 52 of monitoring the status of the clinical trial over time after the start of the clinical trial. Also, monitoring the status of the clinical trial can include periodically and automatically obtaining data of each clinical trial from the RTSM system and comparing the data with the clinical trial plan. And in step 52, any or all aspects of the clinical trial can be compared with the supply plan prediction. For example, comparing the inventory of the investigational drug at each clinical trial site with the supply plan prediction to determine whether a change in the order schedule or a change in the distribution schedule is necessary or essential to ensure that each clinical trial site has sufficient inventory to provide to the subjects within the facility.

[0044] Method 40 can further include step 54 of warning the user when the status of the clinical trial deviates from the clinical trial supply plan. Examples and more detailed content of steps 52 and 54 will be described with reference to FIG. 7.

[0045] FIG. 4 is a flowchart showing an exemplary method 80 for identifying parameters of a proposed pharmaceutical clinical trial. Method 80, or one or more aspects of method 80, can be performed by a supply prediction system such as supply prediction system 12 of FIG. 1, for example.

[0046] Method 80 can include step 82 of receiving from the user a spreadsheet file and one or more text files. These files can include information available to the user at that time regarding the clinical trial to be conducted. The text file can be, for example, an HTML file. The receipt of these spreadsheet files and text files can be done, for example, by file upload via a website provided by the supply prediction system. In each embodiment, the supply prediction system can also provide an interface. The user can freely enter text describing the clinical trial proposed by the user from this interface (this entered free-form text corresponds to the "one or more text files" in step 82 or can be converted into the "one or more text files" of step 82), and can further upload one or more spreadsheets via this interface.

[0047] Method 80 can further include step 84 of converting the spreadsheet into one or more text-formatted tables. In each embodiment, the spreadsheet can be converted, for example, into one or more HTML-formatted tables, or into any format that the submitted text-formatted file has.

[0048] Method 80 can further include step 86 of inserting a table in text form into one or more text files, and step 88 of performing natural language processing on the one or more text files to identify parameters of a clinical trial. An example of natural language processing that can be performed in step 88 will be described with reference to FIG. 5. For an explanation of natural language processing, see U.S. Patent Application Publication No. 2017 / 0154168 (Title of Invention: Generating RTSM System for Clinical Trials, Inventor: Edward A. Tourtellotte, Assigned to 4G Clinical), which is shown in more detail, and all the contents of the same are incorporated herein by reference. By using natural language processing (NLP), the user side can describe a clinical trial in ordinary English regardless of how detailed information is available at that time, and the prediction system side can identify detailed items available at the current time of the clinical trial from such a description and perform subsequent processing.

[0049] FIG. 5 is a schematic diagram of an exemplary electronic specification interpreter 34 that performs natural language processing. In each embodiment, the specification interpreter 34 can execute method 80 of FIG. 4, particularly step 88. The specification interpreter 34 can include a master interpreter 92, one or more specialized interpreters 94, one or more information extractors 96, one or more specialized natural language processing (NLP) tools 98, and one or more core NLP libraries 100. Also, in each embodiment, the specification interpreter 34 can further include additional components.

[0050] Core NLP libraries such as Core NLP library 100 can include, as information of natural languages (such as English, for example), vocabulary and syntax, corpora, taggers, tokenizers, stemmers, lemmatizers, learning algorithms, and the like. And the specification interpreter 34 can examine the specification input by the specification creator using this Core NLP library 100. For example, the specification interpreter 34 can perform word or stem recognition, flagging of spelling mistakes, grammar and semantic analysis, detection of logical operations and data definitions, proposal of alternative terms, or any combination thereof based on the Core NLP library 100. Also, in some cases, the Core NLP library 100 may be provided by a third party in the form of part of a development tool kit for a specific programming language (such as Java or Python, for example) (such as the Natural Language Tool Kit (NLTK), for example).

[0051] Specialized NLP tools such as specialized NLP tool 98 can provide extended functions specialized for each system to the Core NLP library 100. For example, the specialized NLP tool 98 can include additional information regarding, for example, words, phrases, technical terms, syntax, system-specific words and phrases, keywords, operators, etc., which can be widely used in a specific field (such as clinical trials, for example). Also, the specialized NLP tool 98 can be equipped with advanced language analysis functions, especially among natural language processing functions, such as tokenization, stemming, grammar, collocation, language corpora, sentence structure, and semantic analysis. Also, the specialized NLP tool 98 can be equipped with an artificial neural network.

[0052] In addition, using a specialized NLP tool 98, the specification interpreter 34 can also examine the specifications input by the specification creator. For example, the specialized NLP tool 98 enables the specification interpreter 34 to recognize system-specific words, flag spelling mistakes, propose alternative terms, or any combination thereof. The specialized NLP tool 98 may be provided by a third party or developed internally. Also, an information extractor such as the information extractor 96 can extract specific information from the specification. One type of information to be extracted can be a matrix (such as table structure information, for example). When it is more convenient to organize data in matrix form rather than in natural language, the specification may be created using a matrix. For example, using a matrix, it is possible to define events during a hospital visit such as a patient's hospital visit schedule and the associated medication plan, as well as any data suitable for definition in a matrix, such as the type of patient kit, the relationships between supply chains, and lead times.

[0053] As another type of information to be extracted, sentences of a specific style such as definitions can also be cited. A definition may, for example, have a structure of "[subject of definition (noun being defined)][verb presentation][item or list of items]", but the structure can be made different from the above or more flexible so that the necessary information can be extracted while allowing writing in English in a free style. Also, other sentence styles can be defined as needed.

[0054] The master interpreter 92 can receive the specifications of a clinical trial. Also, the master interpreter 92 can perform a high-level analysis on the specifications and summarize the interpretations according to a series of logics that span multiple themes. Note that one theme may include one or more predefined clinical trial topics, and these topics may span one or more sections of the specifications. A specific interpretation theme corresponds to one specialized interpreter 94, and each specialized interpreter 94 is configured to analyze the information related to the specific theme across the entire specification (for one or more sections). And the master interpreter 92 can call a plurality of different specialized interpreters to process a plurality of different clinical trial themes of the entire specification.

[0055] As the specialized interpreter 94, interpreters such as a clinical trial level interpreter, a randomization interpreter, an action interpreter, a drug supply interpreter, a manufacturing interpreter, a subject registration interpreter, etc. can be provided as needed. The clinical trial level interpreter can interpret the high-level information in the specification. The high-level information includes certain clinical trial level information such as, for example, the total number of patients for whom screening and randomization are desired, the randomization scheme and method, and other necessary parameters. The randomization interpreter can interpret the randomization information in the specification. The action interpreter can interpret the event action information in the specification. The event action includes information such as what drug administrations can occur in relation to a single patient visit, for example. The drug supply interpreter can interpret the drug supply information in the specification. For example, the drug supply information may include a supply network, shipping rules, dispensing kit information, etc. The manufacturing interpreter can interpret various manufacturing steps and intermediate lead times during manufacturing to create a kit that can be administered to patients from the substances under clinical trial. The subject registration interpreter can interpret the participation rate of patients in clinical trials at various levels (such as clinical trial level and regional level, etc.). Above, the specialized interpreters are listed, but this list does not cover all. Additional specialized interpreters can be added as needed.

[0056] In some cases, the master interpreter 92 can also serve as the interface to the specification interpreter 34. For example, the master interpreter 92 can receive a specification and split its interpretation by section or theme. Subsequently, the master interpreter 92 can call specialized interpreters 94 suitable for analyzing each section or theme. And each specialized interpreter 94 can call one or more information extractors 96 as needed to extract information such as matrices and definitions from each section or theme of the specification. And each information extractor 96 can call one or more specialized NLP tools 98 as needed. Subsequently, each specialized NLP tool 98 can call one or more core NLP libraries 100 to assist in processing each section or theme of the specification.

[0057] And the output result of the core NLP library 100 can be returned to the specialized NLP tool 98 that called the core NLP library 100. The output result of the specialized NLP tool 98 can be returned to the information extractor 96 that called the specialized NLP library 100. The output result of the information extractor 96 can be returned to the specialized interpreter 94 that called the information extractor 96. The output result of the specialized interpreter 94 can be returned to the master interpreter 92 that called the specialized interpreter 94.

[0058] In some embodiments, the output of the specification interpreter 34 may include a basic data structure and the logic for using it. And the basic data structure and the logic for using it can include necessary or related structures such as logical containers. For example, this data structure can hold information related to a patient's clinic visit schedule and medication, information required when using the data (such as BMI used at the time of medication), and any data desired to be collected or addressed during a clinical trial, such as data related to supply chain organizations, shipments of supplies, patient kits, or any combination thereof. Also, if necessary, the logical container can hold special processing instructions for conforming the prediction system to the specification.

[0059] FIG. 6 is a flowchart showing an exemplary method 110 for obtaining a clinical trial supply prediction. Method 110, or one or more aspects of method 110, can be executed by a supply prediction system such as supply prediction system 12 of FIG. 1, for example.

[0060] Method 110 can include step 112 of calculating a statistical distribution of demand for each of a plurality of candidate patients. A given demand can include the amount of drug required at one or more points during patient treatment, the dose per administration, the amount of placebo, or any combination thereof. Also, the distribution of demand can be a discrete distribution determined, for example, by a randomization ratio. However, in other embodiments, other types of statistical distributions can be used to determine patient demand. Also, the results of this step of calculating patient demand can be made into a time series. In that case, at each point, the probability that demand reaches a particular amount can increase cumulatively or over a given period. For example, the cumulative demand from new patients for kit KT-12mg after one month can be 0 with a probability of 40%, 1 with a probability of 30%, 2 with a probability of 25%, and 3 with a probability of 5%.

[0061] Method 110 can further include step 114 of calculating subject enrollment patterns at a plurality of clinical trial facilities. In each embodiment, the subject enrollment pattern for each clinical trial facility can be calculated based on a Monte Carlo simulation. As a result of this step of simulating subject enrollment, a combination of the most likely number of enrolled subjects can be obtained for each of the plurality of clinical trial facilities.

[0062] Method 110 can further include step 116 of calculating a supply-demand profile for each clinical trial facility based on the subject registration pattern obtained by simulation and the demand profile of each patient. That is, when the subject registration rate is represented as a Poisson distribution that can be estimated as described above based on the patient-level distribution, step 116 can include calculating the composite statistical distribution of demand for all patients who may be registered at each specific clinical trial facility.

[0063] Method 110 can further include step 118 of calculating a supply buffer for each clinical trial facility. In each embodiment, the supply buffer can be calculated according to user input. For example, as described above, the user can operate one or more dials to set a confidence level. The lower the confidence level set by the user, the higher the supply buffer can be set, and conversely, the higher the confidence level set by the user, the lower the supply buffer can be set.

[0064] Method 110 can further include step 120 of calculating the demand at the facility group level based on the facility demand profile and the facility supply buffer. The demand at this facility group level can include, for example, the demand for each supply material type (such as one or more drug dosages and placebo) at each clinical trial facility. In one embodiment, the demand at the facility group level can also be obtained by adding the buffer set for each clinical trial facility and calculating the demand for each facility.

[0065] Method 110 may further include step 122 of calculating regional demand based on the demand of a group of facilities, step 124 of calculating the supply point demand based on the regional demand, and step 126 of calculating the total supply demand based on the supply point demand. In each embodiment, steps 122, 124, and 126 may include not only simple arithmetic calculations but also statistical calculations. Such statistical calculations may include, for example, adding the demands of each group of facilities within a region from the perspective of expected value or statistical distribution to determine the total demand of the region, adding the demands of each region included in the service area of a specific supply point to determine the total demand of the supply point, and the like.

[0066] Method 110 may further include step 128 of determining the appropriate timing and quantity of manufacturing and packaging over the course of a clinical trial and determining the execution trigger for the re-forecasting calculation of the supply plan. The execution trigger can be any one or a combination of minimum frequency, calculations based on the expiration date of the drug, or external constraints given by the user such as the availability of the drug. Additionally, it can be determined based on operations research techniques such as Constraint Programming, Neural Network, and Mixed Integer Programming.

[0067] Figure 7 is a flowchart showing an exemplary method 130 for monitoring a clinical trial based on the supply forecast of the clinical trial. This method, or one or more aspects of this method, can be executed by a supply forecasting system such as the supply forecasting system 12 in FIG. 1.

[0068] Method 130 can include step 132 of connecting (e.g., electronically) to the RTSM system, and step 134 of transmitting one or more commands to the RTSM system via an API published by the RTSM system. The commands may request a specific type of data. This eliminates the need to send and receive all files between the RTSM system and the prediction system, thus enhancing the efficiency of both the RTSM system and the prediction system. For example, the commands may request the number of patients enrolled at one or more clinical trial sites, one or more groups of sites, or one or more regions. Additionally or alternatively, the commands may request inventory information available at one or more clinical trial sites, one or more supply bases, etc. Additionally or alternatively, the commands may request a record of information regarding the status of a clinical trial, such as a patient's drug dosage or patient information.

[0069] Method 130 can further include step 136 of receiving the supply status of one or more drugs during a clinical trial in response to the commands transmitted in step 134. As described above, since the RTSM system API is used, when providing supply status information from the RTSM system to the prediction system, it is not necessary to send and receive all files, and the requested data can be directly sent and received. Moreover, there is no need to aggregate the data as one or more files or manually select and open such files.

[0070] Method 130 can further include step 138 of comparing, for each clinical trial, the supply status of the clinical trial with the supply plan of the clinical trial calculated in the past. As described above, the prediction system can compare any or all aspects of a clinical trial, such as the current inventory levels at one or more supply bases or one or more clinical trial sites, the number of subjects enrolled at one or more clinical trial sites, etc., with a predicted supply plan that the prediction system calculated in the past and that is considered to be used by the clinical trial sponsor to fulfill responsibilities in terms of the logistics and finance of the clinical trial.

[0071] Method 130 can further include step 140, which includes determining that the supply situation of the clinical trial is different from the supply plan. For example, from the difference between the clinical trial situation and the supply plan, situations such as out-of-stock at the clinical trial facility or supply base, or the inability to complete the clinical trial within the desired time due to insufficient numbers of subject enrollees at one or more clinical trial facilities may occur. In such cases, it can be determined that the clinical trial situation is different from the supply plan. For example, step 140 can include determining that the supply situation of the clinical trial is different from the supply plan prediction when the difference from the supply plan is greater than a threshold ratio, such as a few percent (e.g., 5%). Of course, any appropriate threshold ratio can be used.

[0072] Method 130 can further include step 142, which includes generating an electronic warning and sending it to the user when it is determined that the supply situation of the clinical trial is different from the supply plan. In each embodiment, the electronic warning can be sent via a message such as an email or a text message. For example, in an embodiment where the prediction system provides an electronically accessible interface, the warning can be sent to the user account associated with the user within that interface so that the user can view it when accessing the interface. The warning can include, for example, a notification informing that the clinical trial situation deviates from the prediction plan. In addition to or instead of this, the warning can also include corrective actions such as correcting the quantity of one or more shipments to one or more clinical trial facilities, correcting the production quantity of one or more supply materials, correcting the upper limit of subject enrollment at one or more clinical trial facilities, etc. In one embodiment, the prediction system can automatically generate documents such as orders, shipping instructions, and instructions to clinical trial facilities, incorporate the automatically generated documents into the warning, and send them to the users of the prediction system so that the users can execute corrective actions as needed.

[0073] FIG. 8 is a schematic diagram of an exemplary computing system having a general-purpose computing system environment 150. Examples of the general-purpose computing system environment 150 include any device capable of executing instructions stored, for example, in a non-transitory computer-readable medium, such as a desktop computer, a laptop computer, a smartphone, a tablet, and the like. In the following description of the present specification and the drawings, an example having one computing system 150 is used for explanation. However, those skilled in the art will understand that various tasks described below can also be implemented in a distributed environment in which a plurality of computing systems 150 are connected via a local network or a wide area network. In such a distributed environment, the above-described executable instructions can be associated with one or more of the plurality of computing systems 150, executed by the one or more computing systems 150, or both. And such a computer environment or each part thereof can include one or more of the supply prediction system 12, the RTSM system 14, and one or more user computing devices 16 of FIG. 1.

[0074] Continuing to refer to FIG. 8, typically, a computing system environment 150 can include, as its most basic configuration, at least one processing unit 152 and at least one memory 154 that can be connected to the processing unit 152 via a bus 156. Depending on the exact configuration and type of the computing system environment, the computer-readable memory 154 can be volatile memory (such as RAM 160), non-volatile memory (such as ROM 158, flash memory, etc.), or a combination of volatile and non-volatile memory. Also, the computing system environment 150 can further have additional features and functions. For example, the computing system environment 150 can further include additional storage devices (removable storage devices, non-removable storage devices, or both). Such additional storage devices can include, but are not limited to, magnetic disks, optical disks, tape drives, flash drives, or any combination thereof. Such additional memory devices can be made accessible to the computing system environment 150 via, for example, a hard disk drive interface 162, a magnetic disk drive interface 164, an optical disk drive interface 166, etc. Of course, by connecting each of the above devices to the system bus 156, it becomes possible to read from and write to the hard disk 168, read from and write to the removable magnetic disk 170, and read from and write to the removable optical disk 172 such as an optical medium like a CD-ROM or DVD-ROM. The drive interface and the computer-readable medium associated with each of them can be used to perform non-volatile storage of data for the computing system environment 150, such as computer-readable instructions, data structures, program modules, etc. Furthermore, those skilled in the art will understand that other types of computer-readable media capable of storing data can also be used for the same purpose as described above.Examples of such media devices include magnetic cassettes, flash memory cards, digital video discs, Bernoulli cartridges, random access memory, nanodrives, memory sticks, other read / write memories and read-only memories, and any method or technology for storing information such as computer-readable instructions, data structures, data such as program modules, etc., but are not limited thereto. Any such computer storage medium can also form part of a computing system environment 150.

[0075] And a number of program modules can be stored in one or more memories and media devices. For example, the ROM 158 can store a basic input / output system (BIOS) 174 incorporating basic routines useful for information transmission between elements within the computing system environment 150 at startup and the like. Similarly, computer-executable instructions can also be stored using the RAM 160, the hard disk drive 168, and peripheral memory devices. Such computer-executable instructions include an operating system 176, one or more application programs 178 (such as an application for executing the methods and processes of the present invention, for example, a web browser, etc.), other program modules 180, and program data 182. Further, if necessary, computer-executable instructions can also be downloaded into the computing system environment 150 via, for example, a network connection.

[0076] An end user can input commands and information into the computing system environment 150 via input devices such as a keyboard 184 and a pointing device 186. Although not shown, other input devices may include a microphone, a joystick, a game pad, a scanner, etc. Usually, input devices such as these are connected to the processing unit 152 via a peripheral interface 188 connected to a bus 156. Also, an input device can be directly or indirectly connected to the processing unit 152 via an interface such as a parallel port, a game port, a FireWire, a Universal Serial Bus (USB), etc. To display the information output from the computing system environment 150, a display device such as a monitor 190 can be connected to the bus 156 via an interface such as a video adapter 192. Also, the computing system environment 150 can include other peripheral output devices not shown, such as speakers and printers, in addition to the monitor 190.

[0077] Also, the computing system environment 150 can utilize a logical connection to one or more computing system environments. Communication between the computing system environment 150 and a remote computing system environment can be performed bidirectionally via additional processing devices such as a network router 202 responsible for selecting a network path. And communication with the network router 202 can be performed via a network interface component 204. Thus, it will be understood that, for example, within an environment connected by a wired or wireless network such as the Internet, the World Wide Web, a LAN, etc., a program module or a part thereof depicted for the computing system environment 150 can be stored in one or more memory storage devices of the computing system environment 150.

[0078] Moreover, the computing system environment 150 can also include localization hardware 206 for identifying the location of the computing system environment 150. In each embodiment, the localization hardware 206 can include, for example, a GPS antenna, an RFID chip, an RFID reader, a WiFi antenna, and other computing hardware that can be used to capture and transmit signals that can be used to identify the location of the computing system environment 150, but these are merely illustrative.

[0079] As described above, specific embodiments of the present invention have been described. However, it should be understood that the scope of the claims of the present invention is not intended to be limited to these embodiments, except as specifically recited in the claims. Rather, the present invention is intended to cover alternative forms, modifications, and equivalents within its scope, and these may also be included within the spirit and scope of the present invention. Further, in the foregoing detailed description of the invention, numerous specific details have been set forth in order to provide a thorough understanding of each described embodiment. However, it will be apparent to those skilled in the art that systems and methods consistent with the disclosure of the present invention can be implemented without including these specific details. On the other hand, well-known methods, procedures, components, and circuits have been omitted from the detailed description in order to avoid unnecessarily obscuring various aspects of the present invention.

[0080] Part of the detailed description of the above invention includes descriptions related to expressions that symbolize operations on data bits in a computer or digital system memory, such as procedures, logical blocks, processes, etc. Such descriptions and expressions are the most effective means used by those skilled in the data processing field when communicating their work content to other practitioners in the same field. Generally speaking, in this specification, procedures, logical blocks, processes, etc. are considered to be a series of self - consistent steps or instructions leading to a desired result. And steps require physical operations on physical quantities. Although not always the case, usually, this physical operation takes the form of electronic data or magnetic data that can perform operations such as storage, transmission, combination, comparison, etc. in an electronic computing device such as a computer system. For reasons of convenience and general usage, for various embodiments of the present invention, such data is referred to as bits, values, elements, symbols, characters, terms, numbers, etc.

[0081] However, it should be noted that these terms should be interpreted as referring to physical operations and physical quantities, and are merely convenient labels for further interpretation from the perspective of terms commonly used in the art. As is clear from the description in this specification, unless otherwise specified, throughout the description of this embodiment, descriptions using terms such as "determine / specify", "output", "transmit", "record", "locate", "store", "display", "receive", "recognize", "utilize", "generate", "provide", "access", "check", "notify", "distribute", etc. are understood to refer to the operations and processes of an electronic computing device such as a computer system that manipulates and transforms data. Data is represented as a physical (electronic) quantity in the registers and memory of a computer system and is either converted into other data represented as a physical quantity in the memory or registers of the same computer system, or into other data similarly represented as a physical quantity in other information storage devices, transmission devices, or display devices that are conceivable by those skilled in the art and described in this specification.

Claims

1. A method implemented on a computer, comprising: a step in which the computer predicts a supply plan for a clinical trial, wherein the computer obtains a plurality of parameters of the clinical trial, the parameters being related to a plurality of patients, a plurality of implementation facilities, and a plurality of actions related to the clinical trial, wherein the computer specifies, for each patient included in the plurality of patients, a statistical distribution of respective demands based on the plurality of actions related to the clinical trial, wherein the computer specifies, for each implementation facility included in the plurality of implementation facilities, a respective supply-demand profile, the supply-demand profiles of the plurality of implementation facilities being specified based on the statistical distribution of the demands of the plurality of patients, wherein the computer calculates the supply plan for the clinical trial based on the supply-demand profile, and a step in which the computer re-predicts the supply plan for the clinical trial based on variations in the treatment processes of the respective patients in the plurality of patients, and a step in which the computer transmits the re-predicted supply plan to a user.

2. The method according to claim 1, wherein the plurality of parameters of the clinical trial include an expected value of the subject registration rate of patients in the clinical trial, a failure rate of patient screening, randomization of the plurality of patients, explanations regarding treatment of the plurality of patients, and changes in the dosage administered to the plurality of patients.

3. The method according to claim 1, wherein the plurality of parameters of the clinical trial include the respective outpatient schedules and treatment processes of the plurality of patients.

4. The method according to claim 1, wherein the plurality of patients include a specific patient, and the statistical distribution of the demand for the specific patient indicates the amount of drug determined to be required, the dosage per administration, the amount of placebo, or any combination thereof, at one or more time points during the treatment of the specific patient.

5. The method according to claim 1, wherein the plurality of actions related to the clinical trial include screening of one or more expected patients, assignment of respective treatment processes to the plurality of patients, administration of drugs to one or more patients, dropout of one or more patients from the trial, change in the dosage of drugs administered to one or more patients, or any combination thereof.

6. The step of predicting the supply plan includes predicting a first supply plan based on information available at a first point in time, and the step of re-predicting the supply plan includes predicting a second supply plan based on information available at a second point in time. The method according to claim 1.

7. The variation in the treatment process of each of the plurality of patients includes a variation in the dosage administered to the plurality of patients. The method according to claim 1.

8. The supply-demand profile of the plurality of implementation facilities is further specified based on each registration pattern in the plurality of implementation facilities. The method according to claim 1.

9. A supply prediction system, a processor, a computer-readable medium storing instructions that, when executed by the processor, cause the system to implement the method according to any one of claims 1-8 A supply prediction system including.

10. A method implemented on a computer, wherein the computer predicts a supply plan for a clinical trial, wherein the computer obtains a plurality of parameters of the clinical trial, the parameters being related to a plurality of patients, a plurality of implementation facilities, and a plurality of actions related to the clinical trial, wherein the computer identifies a statistical distribution of each demand for each patient included in the plurality of patients based on the plurality of actions related to the clinical trial, wherein the computer identifies a supply-demand profile for each implementation facility included in the plurality of implementation facilities, the supply-demand profile of the plurality of implementation facilities being identified based on the statistical distribution of the demand of the plurality of patients, wherein the computer calculates the supply plan for the clinical trial based on the supply-demand profile including steps, wherein the computer re-predicts the supply plan for the clinical trial when a specific event is reached, wherein the computer transmits the re-predicted supply plan to the user including methods.

11. The plurality of parameters of the clinical trial include an expected value of the subject registration rate of patients in the clinical trial, a failure rate of patient screening, randomization of the plurality of patients, a description of the treatment of the plurality of patients, and a change in the dosage administered to the plurality of patients. The method according to claim 10.

12. The method according to claim 10, wherein the plurality of parameters of the clinical trial include the outpatient schedule and treatment process of each of the plurality of patients.

13. The method according to claim 10, wherein the plurality of patients includes a specific patient, and the statistical distribution of the needs for the specific patient indicates the amount of drug required, the dosage per administration, the amount of placebo, or any combination thereof, at one or more time points during the treatment of the specific patient.

14. The method according to claim 10, wherein the plurality of actions regarding the clinical trial includes screening of one or more scheduled patients, assignment of each treatment process to the plurality of patients, administration of drugs to one or more patients, withdrawal of one or more patients from the trial, change in the dosage of drugs administered to one or more patients, or any combination thereof.

15. The step of predicting the supply plan includes predicting a first supply plan based on information available at a first time point, and the step of re-predicting the supply plan includes predicting a second supply plan based on information available at a second time point. The method according to claim 10.

16. The method according to claim 10, wherein the step of re-predicting the supply plan of the clinical trial is based on the variation in the treatment process of each of the plurality of patients.

17. The method according to claim 16, wherein the variation in the treatment process of each of the plurality of patients includes variation in drug administration to the plurality of patients.

18. The method according to claim 10, wherein the specific item includes that a specific date or specific information becomes available.

19. The method according to claim 10, wherein the supply-demand profile of the plurality of implementation facilities is further specified based on the respective registration patterns in the plurality of implementation facilities.

20. A supply prediction system, a processor, a computer-readable medium storing instructions that, when executed by the processor, cause the system to implement the method according to any one of claims 10 - 19 A supply prediction system comprising.

Images