WO2009129168A2

WO2009129168A2 - Dynamic provisioning system and method for providing patient specific dose and dose interval information

Info

Publication number: WO2009129168A2
Application number: PCT/US2009/040333
Authority: WO
Inventors: Bonny Lewis Bukaveckas; Shivi Kansal; Zhongming Zhao
Original assignee: Virginia Commonwealth University
Priority date: 2008-04-16
Filing date: 2009-04-13
Publication date: 2009-10-22
Also published as: WO2009129168A3

Abstract

The use of algorithms developed from the clinical literature and patient care guidelines, as well as from new biomedical knowledge is facilitated by the use of remote computers, such as personal digital assistants (PDAs), which are synchronized with a central computing site. All of these algorithms are stored in a central database which can be accessed by the remote computers, and the remote computers can selectively load one or more of the algorithms for use on a computer when, for example, prescribing a dose or dose interval for a specific patient. The dynamic nature of the invention allows empiric, de-identified clinical data from patients being treated and/or receiving therapy during research to be used in a feedback loop to improve the algorithms.

Description

DYNAMIC PROVISIONING SYSTEM AND METHOD

FOR PROVIDING PATIENT SPECIFIC DOSE AND

DOSE INTERVAL INFORMATION

DESCRIPTION

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention generally relates to personalized dosing tools and, more particularly, to a dynamic provisioning process and system where a clinical provider can be provided with calculated clinical information (e.g., dose) using a personal digital assistant (PDA) or other computing device having one or more stored algorithms that consider demographic and genetic characteristic information.

Background Description

Chronically administered drugs with a narrow therapeutic index pose significant hurdles in patient care. Such drugs, if given in amounts which are too high or in amounts which are too low, may lead to severe adverse consequences including death. Coumadin (referred to as warfarin) is one example; however, there are a variety of other drugs in this category including without limitation drugs having psychiatric applications. The proper dose of these drugs is dependent on demographic attributes of the patient (e.g., sex, age, height, weight, race, other medical conditions or medications, etc.) as well as genetic information.

Personalized dosing tools are known. For exemplary purposes, this invention describes an improved personalized dosing tool in the context of improving warfarin dose selection and in calculating the dose interval. However, it will be recognized the dynamic provisioning concept described herein may be applied to a number of drugs which have a narrow therapeutic index. The article by E. A. Sconce, T. I. Khan, H. A. Wynne, P. Avery, L.

Monkhouse, B. P. King, P. Wood, P. Kesteven, A. K. Daly, and F. Kamali entitled "The impact of CYP2C9 and VKORCl genetic polymorphism and patient characteristics upon warfarin dose requirements: proposal for a new dosing regimen", Blood, Oct. 1, 2005, 106(7):2329-33, the complete contents of which are herein incorporated by reference, describes a validated warfarin dosing algorithm which uses age, height and genotypes of patients as follows:

Dose = 0.628 - 0.0135 (age-years) - 0.240(CYP2C9*2) - 0.370(CYP2C9*3) - 0.241(VKORCl) + 0.0162(height-cm),

where CYP2C9 is cytochrome P450 2C9, the specific enzyme primarily responsible for the metabolism of warfarin, and VKORCl is vitamin K epoxide reductase, the gene that makes a protein to help control clotting and is the key molecular target of warfarin. The article by B. L. Bukaveckas entitled "Personalized Medicine:

Anticoagulation Treatment", Pfeiffer Foundation, 2006, reported on a dose change interval finding equation. In short, the average half-life of warfarin is 36 hours in healthy volunteers; with a range of 20 to 80 hours. These values were used to estimate the period of time required for five half-lives to pass. Five half-lives is a rough rule for the time when the pharmacodynamic effect of a medication should be assessed. In this case, the pharmacodynamic effect of interest in the international normalized ratio (ESfR). 5*20 hours = 4 days 5*36 hours = 7.5 days 5*80 hours = 16 days

www.warfarindosing.org is a site maintained by Brian Gage at Washington University in St Louis, Missouri. This site provides dose calculations using genetics.

SUMMARY OF THE INVENTION

It is therefore an object of the present invention to provide a tool for use by clinicians to dynamically use and improve algorithms such as taught by Sconce et al. and Bukaveckas and others.

According to the present invention, the use of algorithms developed from the clinical literature, and patient care guidelines, as well as from new biomedical knowledge is facilitated by the use of remote computers, such as personal digital assistants (PDAs), which are synchronized with a central computing site. All of these algorithms are stored in a central database which can be accessed by the remote computers, and the remote computers can selectively load one or more of the algorithms for use on a computer when, for example, prescribing a dose or dose interval for a specific patient. The dynamic nature of the invention allows empiric, de-identified clinical data from patients being treated and/or receiving therapy during research to be used in a feedback loop to improve the algorithms.

More specifically, the invention is a system for the dynamic provisioning of patient specific calculated clinical information including one or more of dose and dose change intervals which is implemented on one or more remote computers, such as PDAs, and a central database. The one or more remote computers which include a mechanism for inputting both demographic information and genetic characteristic information for a patent, a mechanism for calculating patient specific information from the inputted demographic information and genetic characteristic information including one or more of dose and dose change intervals, said mechanism for calculating employing one or more algorithms for computing said one or more of dose or dose change intervals, a mechanism for storing limited medical records associated with said patient, and a mechanism for outputting de-identified clinical data based on said limited medical records and said patient specific calculated information. The central database serves as a repository for de- identified clinical data output from the one or more remote computers. The central database stores one or more algorithms for computing one or more of dose and dose change intervals based on one or more of clinical literature, clinical guidelines, and biomedical knowledge, at least one of said one or more algorithms being updated or modified based on said de- identified clinical data, and includes a mechanism for outputting said one or more algorithms for computing one or more of dose and dose change intervals. A communications linkage between the central database and the one or more remote computers which permits outputting of said de-identified clinical data from said one or more remote computers to said central database, and uploading of said one or more algorithms form said central database to said one or more remote computers.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and other objects, aspects and advantages will be better understood from the following detailed description of a preferred embodiment of the invention with reference to the drawings, in which:

Figure 1 is a block diagram illustrating the dynamic provisioning system according to a preferred embodiment of the present invention; Figures 2A and 2B, taken together, are a flow diagram illustrating the process with screen components according to a preferred embodiment of the present invention;

Figure 3A is an example of a screen on the PDA GUI prompting the user to enter patient data for a New Patient;

Figure 3B is an example of a screen on the PDA GUI showing the dosing results as determined for the data entered for the New Patient;

Figure 3C is an example of a screen on the PDA GUI prompting the user to enter patient identification for Open Patient; Figure 3D is an example of a screen on the PDA GUI showing recorded dose and INR for "Clinical" data; and

Figure 3 E is an example of a screen on the PDA GUI showing additional information for "Research" data.

DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT OF THE INVENTION

For exemplary purposes, a software based system and method to aid clinical care providers to incorporate genetic information into the management of patients which are prescribed Coumadin (warfarin) is presented for illustrative purposes. The software based system and method is referred to as the "SmartWarf™" tool.

The SmartWarf™ tool is a clinical tool designed to assist clinicians in personalized anti-coagulation therapy in patients, hi particular, the SmartWarf™ tool allows the clinician to apply the results of genetic testing to patient care. A Graphical User Interface (GUI) has been developed and installed on Palm TX Personal Digital Assistant (PDA) devices running Palm Garnet 5.4.8 Operating System. This GUI has been developed using the Code Warrior Integrated Development Environment (IDE). The SmartWarf™ tool supports the following processes:

• Create Patient - New patient's information is entered in this form. Based on the information provided, maintenance dose information is calculated using the above equations. • Open Patient - This form is used to append additional information about the care of an existing patient. If you need to access existing patient information, you will need to open the patient. To use the SmartWarf™ tool, the user clicks on the "SmartWarf™" icon on in the Applications area of the PDA desktop, on the main page of the PDA. Then the main form of the SmartWarf™ tool will open, and the user selects either the Create Patient form or the Open Patient form, or to Exit from the menu according to user's requirements. The Create Patient process is used begin a new clinical record in the SmartWarf™ tool database. After selecting Create Patient from the menu of the main form, the following steps are followed:

• User types in the first name of the patient.*

• User types in the last name of the patient.*

• User types in the medical record number of the patient.*

• User enters the date of birth of the patient, in MM/DD/YYY format in the DOB field.*

• User types in the height of the patient in inches.*

User selects the CYP2C9 and VKORCl genotype result of the patient from among a discrete choice list of options *1*1, *1*2, *1*3, *2*2, *2*3, *3*3 and AA, GA and GG respectively.* • User enters a numeric value for the target INR of the patient, t

• User clicks on the Save GUI.

Note that all of the steps indicated by an asterisk (*) may be automated, through wireless interface with the patient's electronic limited medical record. In the step indicated by (t), logical restrictions may be placed to ensure values are within the realm of clinic possibility. Additionally, information may be linked which provides succinct information, taken from the prescribing information, to aid clinical provides in the decision making process to determine the correct dose.

After clicking Save, the next form that opens displays the dose and dose interval, calculated using the algorithm which has been programmed.

• The user then enters the dose given to the patient and a free text field allows the clinician to record any comments, up to 250 characters in length.

• User clicks on Save. Once this final Save command is executed, all the patient information that has either been obtained from the electronic limited medical record (EMR) or entered by the clinician will be saved in the SmartWarf™ tool PDA database.

The Open Patient form can be used for follow-up clinical management, or to support a research project in which the patient is a consented participant. To enter additional clinical data or to enter additional research data for an existing patient in the database, first the user must retrieve the patient record from the database. The following steps are used to open an existing patient. The user selects Open Patient from the main SmartWarf™ tool menu and then follows the steps outlined below.

Clinical data -

• User enters the last name of the patient and then clicks on the Clinical GUI. • A search form will open, showing all patients' names in the

"Clinical" database for the device being used which share that last name. Displayed are the medical record numbers and dates of birth for all patients in this category. There is an "OK" GUI next to each patient in the list. • The user clicks on the Ok button corresponding to the name of patient for whom the clinical data has to be entered in this encounter.

• User clicks on Save.

Research data -

• User enters the last name of the patient and then click on the Research GUI.

• A search form will open, showing the patient's name, MR# and date of birth.

• Click on the appropriate name of patient for whom the research data has to be entered. • A search form will open, showing all patients' names in the

"Research" database for the device being used which share that last name. Displayed are the medical record numbers and dates of birth for all patients in this category. There is an "OK" GUI next to each patient in the list. • The user clicks on the Ok button corresponding to the name of patient for whom the research data has to be entered in this encounter.

• User clicks on the Main Page GUI to save the research data.

The SmartWarf™ tool was developed using the Code Warrior IDE. It will be recognized by those skilled in the art that the code and software can be varied considerably depending on the application. The "C" programming language was used as the programming language. The following functions provide support the SmartWarf™ tool: 1. PilotMain: It is the entry point into the SmartWarf™ tool.

2. RomVersionCompatible: This function checks that a read only memory (ROM) version meets the minimum requirement.

3. AppStart: Initializes the application 4. AppEventLoop: This function is the event loop for the application.

5. AppHandleEvent: This function loads the form resources and sets the event handler for the form loaded.

6. GetObjectPtr: This function returns a pointer to an object in the current form.

7. SWFormHandleEvent: This function is the event handler for the main form. It calls the SWFormDoCommand function (9).

8. SWFormlnit: Initializes the main form

9. SWFormDoCommand: Performs the menu commands specified for instance: Create patient, Open patient or Exit. On selecting

Create Patient or Open Patient, AppEventLoop function is activated and on selecting exit, all forms are closed.

10. CPFormHandleEvent: This function is the event handler for Create Patient form. This function saves the information entered in create patient form to the SmartWarf™ tool database.

11. CPFormlnit: Initializes the create patient form

12. DoB: Displays the date dialog box and calls the num_of_days function. It also converts the date of birth selected into age to be used for calculating the dose and dose interval. 13. num_of_days: Calculates the number of days in the month.

14. genetype: It contains the cases for selecting the CYP2C9 and VKORCl genotype.

15. CPlFormHandieEvent: This function is the event handler for the form that calculates and then displays the dose and dose interval. This function saves the information entered in this form. 16. CPlFormlnit: It initializes the form that displays calculated dose and dose interval.

17. OPFormHandleEvent: This function is the event handler for the open patient form. This function converts the last name entered into an array and stores into a global array.

18. OPFormlnit: It initializes the open patient form.

19. MPFormHandleEvent: This function is the event handler for the form that displays the patient's full name, date of birth and MR# for the patient's last name that has been searched. 20. MPFormlnit: It initializes the form that displays the patient's full name, date of birth and MR# for the patient's last name that has been searched.

21. RDIFormHandleEvent: This function is the event handler for the recommended dose and dose interval form. Also, the dose and the interval entered in the PDA database.

22. RDIFormlnit: It initializes the recommended dose and dose interval form.

23. RPFormHandleEvent: This function is the event handler for the research patient form. In this function the research data is entered and saved.

24. RPFormlnit: It initializes the research patient form.

25. AppStop: It de-initializes the application. It also saves the current state of application and close all open forms.

With reference to Figure 1, the system may be comprised of the graphical user interface above running on a PDA OS personal digital assistant (PDA) 1 and a back end MySQL database running on a secure personal computer (PC) 2. The resulting system is dynamic in the sense that dosing data collected during use will then be used to refine dosing algorithms that will be periodically updated on the handheld when it is synced with the database. As shown in Figure 1, the PDA 1 supports the SmartWarf™ tool GUI 11 which receives input of identifiable clinical data 12 from the PDA database 13 and from the end user 3. The end user may be a physician, pharmacist, nurse or other health provider, and the data input by the end user is derived from the patients 4. The PDA database 13 may be loaded from an external electronic limited medical record (EMR) database 5 via the EMR interface 14. In addition, The PDA 1 is synchronized with the secure PC 2 via a synchronization device 6. The secure PC 2 includes the SmartWarf™ tool database 21 as well as a database 22 of clinical literature and guidelines. These two databases are accessed by an expert system 23 to generate new biomedical knowledge and output decision tools. In the synchronization process, the PDA 1 sends de-identified clinical data to the secure PC 2 via the synchronization device 6, and the secure PC sends new GUI with, for example, updated dosing algorithm. The PDA 1 arrives at an initial dose and all the associated information is stored in a temporary database in the device's memory until it is synchronized on the secure PC 2. Upon being synchronized, the data is uploaded to the central MySQL database. Genotype and clinical information can be collected and stored centrally. When syncing the PDA, the clinician will be notified of any available updates to the dosing algorithm and will have the option of using whichever algorithm they choose. Multiple dosing algorithms can be stored on the PDA. The available algorithms can made to be selected from a dropdown list, and the list can be edited as new algorithms become available. The "PDA Desktop" includes an open applications programming interface (API) for developing arbitrary custom synchronization logic that allows the device to be integrated with any resource available to the desktop. This custom synchronization logic is known as a conduit. A conduit had to be developed that could implement the synchronization of the handheld's data with the central database that would house all the device users data. Microsoft's Visual Studio 2002 integrated development environment was deployed, along with PDA source's Conduit Development Kit 4.03. A conduit is currently being developed that provides two way synchronization of the handheld data and a remote database (such as (MySQL) providing the central database repository.

Useful features of the Smart Warf™ tool include the ability to apply pharmacogenetic tests directly to patient care in the dosing and management of warfarin pharmacotherapy. This is done using a novel, self-improving process whereby use of the SmartWarf™ tool contributes data to a centralized database. This database in turn is used to improve the predictive value of the calculations made by the SmartWarf™ tool. And finally those improved calculation methods are returned to the contributor via the "sync" mechanism of the mobile device. It is this self-improving feature that is novel. There are no current technologies for warfarin pharmacogenetics. Therefore, the advantage of the SmartWarf™ tool is that it enables clinical providers to simultaneously provide clinical care and to contribute to clinical knowledge, which in turn improves clinical care. It is an excellent example of using technology to move knowledge and discovery from bedside to bench and back to the bedside. Figures 2 A and 2B, taken together, illustrate the process steps used with the SmartWarf™ tool, as an example of the practical application of the invention. The process begins at START button 21 in Figure 2 A. The user is presented with a GUI on the PDA desktop. The first step in the process is the user selects the SmartWarf™ icon from the PDA desktop in process block 22. A choice is presented to the user; specifically, is this a

New Patient or an Open Patient? The user makes the selection in decision block 23. If the selection is New Patient, the process goes to process block 24; otherwise, the process goes to connector 25 and Figure 2B. If New Patient is selected, the user is prompted in process block 24 to enter patient data in a screen displayed on the PDA GUI, an example of which is shown in Figure 3A. Once the patient data is entered, the user can then select the SAVE button. When the SAVE button is selected, as determined in decision block 26, the user is presented with the results of the algorithm as calculated by the SmartWarf™ tool, example of which is shown in Figure 3B. At this point, the process ends at button 28. This is the end of the first data collection for this patient. When the user next synchronizes the PDA with the desktop software, the information that they have entered since the last sync will be transferred to their local database. The upload of de- identified patient data from the local SmartWarf™ tool database to the central SmartWarf™ tool database can be automated or a manual process, similar to the process used for anti-virus software updates.

Returning to decision block 23, if the user selected Open Patient, the process goes to connector A and Figure 2B. The user is then prompted in process block 29 to enter the patient's name using a screen, an example of which is shown in Figure 3C. The user selects the correct patient from the list in the SmartWarf™ local database displayed based on the information provided by the user. The user is then prompted to select either "Clinical" data or "Research" data. If the user selects "Clinical" data, as determined in decision block 30, the user can then record in process block 31 dose and INR for an unlimited number of visits for this patient. An example of this is illustrated in Figure 3D. If the user selects "Research" data, the user can record in process block 32 additional information for an unlimited number of visits for this patient. An example of this is illustrated in Figure 3E. Once the information is recorded for either "Clinical" or "Research" data, the process goes to connector 33 which returns to Figure 2 A, and the process ends. When the user next synchronizes the PDA with the desktop software, the information that they have entered since the last sync will be transferred to their local database. The upload of de- identified patient data from the local database to the central database can be automated or a manual process, similar to the process used for anti-virus software updates.

It is assumed that many other users have been contributing to the Smart Warf™ tool central database 21. With the information that is available from these records, statistical regression can be performed on the information by the expert system 23 to evaluate improved dosing algorithms and monitoring intervals. For example, suppose the current algorithm does not include a correct factor for ethnicity, and our example patient is Hispanic. This reflects the limited diversity of the population used to derive the original algorithm. But if data is collected from a sufficient number of Hispanic patients, an adjustment to the SmartWarf™ tool algorithm can be made by the expert system 23. When the user next synchronizes the PDA 1 with the secure PC 2, the new algorithm that has now become available since the last sync will be transferred from the SmartWarf^{1 M} tool central site to the PDA local database in the form of a software update. The download of the new and improved software from the central database to each user's local database can be automated or a manual process, similar to the process used for anti-virus software updates. The same process can be used to increase the accuracy of the Monitoring Interval. While the invention has been described in terms of a single preferred embodiment, those skilled in the art will recognize that the invention can be practiced with modification within the spirit and scope of the appended claims.

Claims

Having thus described our invention, what we claim as new and desire to secure by Letters Patent is as follows:

L A system for the dynamic provisioning of patient specific calculated clinical information including one or more of dose and dose change intervals, comprising: one or more remote computers which include a mechanism for inputting both demographic information and genetic characteristic information for a patent, a mechanism for calculating patient specific information from the inputted demographic information and genetic characteristic information including one or more of dose and dose change intervals, said mechanism for calculating employing one or more algorithms for computing said one or more of dose or dose change intervals, a mechanism for storing limited medical records associated with said patient, and a mechanism for outputting de-identified clinical data based on said limited medical records and said patient specific calculated information; a central database which serves as a repository for said de- identified clinical data output from said one or more remote computers, stores one or more algorithms for computing one or more of dose and dose change intervals based on one or more of clinical literature, clinical guidelines, and biomedical knowledge, at least one of said one or more algorithms being updated or modified based on said de-identified clinical data, and a mechanism for outputting said one or more algorithms for computing one or more of dose and dose change intervals; and a device appropriate communications linkage between said central database and said one or more remote computers which permits outputting of said de-identified clinical data from said one or more remote computers to said central database, and uploading of said one or more algorithms from said central database to said one or more remote computers.

2. The system of claim 1 wherein at least some of said one or more remote computers are personal data assistants, and wherein said communications linkage includes a synchronization device.

3. The system of claim 1 wherein said one or more remote computers include in said mechanism for calculating said patient specific calculated information, a mechanism for selecting an algorithm from amongst said one or more algorithms.

4. The system of claim 1 wherein said one or more remote computers include a mechanism for selectively controlling a timing of uploading of said one or more algorithms from said central database with said communications linkage.

5. A method of dynamically provisioning of patient specific calculated clinical information including one or more of dose and dose change intervals for a chronically administered drug having a narrow therapeutic index, comprising the steps of: using at least one remote computer for inputting both demographic information and genetic characteristic information for a patent, calculating patient specific information from the inputted demographic information and genetic characteristic information including one or more of dose and dose change intervals, said calculating step employing one or more algorithms for computing said one or more of dose or dose change intervals, storing limited medical records associated with said patient, and outputting de-identified clinical data based on said limited medical records and said patient specific calculated information; using a central database for storing said de-identified clinical data output from said at least one remote computer, and at least a one or more additional remote computers, storing one or more algorithms for computing one or more of dose and dose change intervals based on one or more of clinical literature, clinical guidelines, and biomedical knowledge, at least one of said one or more algorithms being updated or modified based on said de-identified clinical data, and outputting said one or more algorithms for computing one or more of dose and dose change intervals; and providing a communications linkage between said central database and said remote computer which permits outputting of said de-identified clinical data from said remote computer to said central database, and uploading of said one or more algorithms form said central database to said remote computer.

6. A method of dynamically provisioning of one or more of dose and dose change intervals for Coumadin which are patient specific, comprising the steps of: using at least one remote computer for inputting both demographic information and genetic characteristic information for a patent, calculating patient specific calculated information from the inputted demographic information and genetic characteristic information including one or more of dose and dose change intervals for Coumadin for said patient, said calculating step employing one or more algorithms for computing said one or more of dose or dose change intervals for Coumadin, storing limited medical records associated with said patient, and outputting de-identified clinical data based on said limited medical records and said patient specific calculated information; using a central database for storing said de-identified clinical data output from said at least one remote computer, and at least a one or more additional remote computers, storing one or more algorithms for computing one or more of dose and dose change intervals for Coumadin based on one or more of clinical literature, clinical guidelines, and biomedical knowledge, at least one of said one or more algorithms being updated or modified based on said de-identified clinical data, and outputting said one or more algorithms for computing one or more of dose and dose change intervals for Coumadin; and providing a communications linkage between said central database and said remote computer which permits outputting of said de-identified clinical data from said remote computer to said central database, and uploading of said one or more algorithms form said central database to said remote computer.

7. The method of claim 6 wherein said remote computer provides dose information which is patient specific.

8. The method of claim 6 wherein said remote computer provides dose change interval information which is patient specific.