AU2011218632B2 - Multi-source longitudinal patient-level data encryption process - Google Patents

Abstract

C:\NRPortbADCCVASU 9 B144 I U 44-/2WZ01 I An amplifier device for amplifying RF communication signals compatible with IEEE 802.11 standards (100). An amplifier circuit is housed within a cylindrical metal housing 5 that provides excellent heat dissipation and requires no mounting hardware for installing the device within a communication system (10). The metal housing with an integrated lightning arrestor creates a watertight seal to protect the amplifier circuit from damage when installed outdoors and the amplifier circuit is compatible with both 2.4 GHz and 5.8 GHz frequency applications. %j Standardized Longitudinal S Prescription Database ( Attributes V 0i Ego 0IZ Cm UU z0 aT. C) Pro essS) c 21/In

Description

P/00/0II Regulation 3.2 AUSTRALIA Patents Act 1990 COMPLETE SPECIFICATION STANDARD PATENT (ORIGINAL) Name of Applicant(s): IMS Software Services, Ltd of 660 West Germantown Pike, Plymouth Meeting, Pennsylvania 19462-0905, United States of America Actual Inventor(s): KOHAN, Mark E. AND WOLFE, Clinton J. Address for Service: DAVIES COLLISON CAVE, Patent Attorneys, of 1 Nicholson Street, Melbourne, Victoria 3000, Australia Invention Title: "Multi-source longitudinal patient-level data encryption process" The following statement is a full description of this invention, including the best method of performing it known to us: CANRPonbl\DCC\MASU841169_1.DOC - 298/1 H:\Id\Iitenvovcl\NRPortbl\DCC\TLD\6879315_.docx-14/10/2014 MULTI-SOURCE LONGITUDINAL PATIENT-LEVEL DATA ENCRYPTION PROCESS Cross Reference to Related Applications 5 This application claims the benefit of U.S. provisional patent application Serial No. 60/568,455 filed May 5, 2004, U.S. provisional patent application Serial No. 60/572,161 filed May 17, 2004, U.S. provisional patent application Serial No. 60/571,962 filed May 17, 2004, U.S. provisional patent application Serial No. 60/572,064 filed May 17, 2004, 10 and U.S. provisional patent application Serial No. 60/572,264 filed May 17, 2004, all of which applications are hereby incorporated by reference in their entireties herein. Field of the Invention 15 The present invention relates to a process for assembling a longitudinally-linked database from individual patient healthcare transaction data records; and a system for longitudinally-linking individual patient healthcare transaction data records obtained from multiple data suppliers. For example, the present invention relates to the management of personal health information or data on individuals. The invention in particular relates to the 20 assembly and use of such data in a longitudinal database in manner, which maintains individual privacy. Background of the Invention 25 Electronic databases of patient health records are useful for both commercial and non commercial purposes. Longitudinal (life time) patient record databases are used, for example, in epidemiological or other population-based research studies for analysis of time-trends, causality, or incidence of health events in a population. The patient records assembled in a longitudinal database are likely to be collected from a multiple number of 30 sources and in a variety of formats. An obvious source of patient health records is the modern health insurance industry, which relies extensively on electronically communicated patient transaction records for administering insurance payments to medical H:\tId\lIenvoven\NRPoibi\DCC\TLD\6563341 .doc-14/O/2014 - lA service providers. The medical service providers (e.g., pharmacies, hospitals or clinics) or their agents (e.g., data clearing houses, processors or vendors) supply individually identified patient transaction records to the insurance industry for compensation. The patient transaction records, in addition to personal information data fields or attributes, 5 may contain other information concerning, for example, diagnosis, prescriptions, treatment or outcome. Such information acquired from multiple sources can be valuable for longitudinal studies. However, to preserve individual privacy, it is important that the patient records integrated to a longitudinal database facility are "anonymized" or "de identified". A data supplier or source can remove or encrypt personal information data fields or attributes (e.g., name, social security number, home address, zip code, etc.) in a patient transaction record before transmission to preserve patient privacy. The encryption or standardization of certain personal information data fields to preserve patient privacy is now mandated by statute and government regulation. Concern for the civil rights of individuals has led to government regulation of the collection and use of personal health data for electronic transactions. For example, regulations issued under the Health Insurance Portability and Accountability Act of 1996 (HIPAA), involve elaborate rules to safeguard the security and confidentiality of personal health information. The HIPAA regulations cover entities such as health plans, health care clearinghouses, and those health care providers who conduct certain financial and administrative transactions (e.g., enrollment, billing and eligibility verification) electronically. (See e.g., http://www.hhs.gov/ocr/hipaa). Commonly invented and co-assigned patent application Serial No. 10/892,021, "Data Privacy Management Systems and Methods", filed July 15, 2004 (Attorney Docket No. AP35879), which is hereby incorporated by reference in its entirety herein, describes systems and methods of collecting and using personal health information in standardized format to comply with government mandated HIPAA regulations or other sets of privacy rules. For further minimization of the risk of breach of patient privacy, it may be desirable to strip or remove all patient identification information from patient records that are used to construct a longitudinal database. However, stripping data records of patient identification information to completely "anonymize" them can be incompatible with the construction of the longitudinal database in which the stored data records must be linkable patient by patient. Consideration is now being given to integrating "anonymized" or "de identified". patient records from diverse data sources in a longitudinal database, where the data sources may employ different encryption techniques that can hinder or -2- H:\tid\1ntnvocn\NRPortbl\DCC\TLD\68793 15_1.docx-14/10/2014 -3 prohibit accurate longitudinal linking patient records. Attention is in particular directed to efficient processes for integrating multi-sourced longitudinal data in a manner that ensures the protection of an individual patient's privacy and complies with industry privacy regulations. 5 It is generally desirable to overcome or ameliorate one or more of the above described difficulties, or to at least provide a useful alternative. Summary of the Invention 10 According to the present invention, there is provided a process for assembling a longitudinally-linked database from individual patient healthcare transaction data records, the process comprising; at a central longitudinal database facility (LDF), 15 (a) receiving data records including at least one patient non-identifying attribute and at least one individually encrypted patient-identifying attribute, the at least one patient-identifying attribute first encrypted using a first encryption key specific to the LDF and further encrypted with a second encryption key unique to a data source so that original 20 unencrypted patient identification information in the at least one patient (b) identifying attribute remains secret with respect to other data sources; (c) partially decrypting the received data records using a decryption key complementary to the second encryption key so that the at least one patient-identifying attribute retains only the encryption by the first 25 encryption key specific to the LDF; (d) using an attribute-matching algorithm to assign an LDF identifier (ID) to the encrypted data records; and (e) linking the encrypted data records ID by ID, whereby the longitudinally linked data base is formed. 30 According to the present invention, there is also provided a system for longitudinally linking individual patient healthcare transaction data records obtained from multiple data H:\lld\Intenvovc\NRPorltbl\DCC\TLD\6879315_ldocx-14/10/2014 -3A suppliers, the system comprising: at a data supplier location, a first component configured to: acquire data records having patient-identifying attributes and non identifying attributes; and 5 doubly encrypt the patient-identifying attributes in the data records with a first encryption key specific to a central facility (LDF) and a second encryption key specific to the data supplier; at the LDF, a second component configured to: receive doubly-encrypted data records from the multiple data suppliers; 10 partially decrypt the received data records using a decryption key complementary to the second encryption key so that the patient-identifying attributes retain the encryption by the first encryption key specific to the LDF; and perform an additional layer of encryption; and a third component configured to: 15 assign an LDF identifier (ID) to the encrypted data records by matching attributes in the encrypted data records; and link the encrypted data records ID by ID, whereby a longitudinal database is formed. 20 Preferred embodiments of the present invention provide systems and processes for assembling de-identified patient data records in a longitudinal database. Preferred embodiments of the present invention is to provide processes for encrypting multi-sourced patient data records to overcome data source variances in individual encryption techniques and in the content of data records. The present systems and processes allow de-identified H:\lId\hiievovnc\NRPon1bl\DCC\TLD\656334 1_1.doc-i4/08/2014 -3B data records received from multiple data sources or suppliers to be assembled in a longitudinal database for market research and other analysis. Preferred embodiments of the inventive systems and processes ensure patient privacy consistent with industry and other regulations concerning patient privacy. 5 An exemplary system for assembling the longitudinal database may include three sequential and tightly integrated components. A first component, which may be implemented at each of the multiple data source or vendor locations, is configured to perform critical data encryption of patient or healthcare data records generated or collected 10 by the data source. Patient-identifying attributes in the data records are encrypted and coded in these processes. The patient-identifying attributes may be encrypted using two or more suitable encryption keys. Each attribute is encrypted using a longitudinal encryption key that only resides at the data source location and a second data source specific encryption key unique to each data source involved in the encryption process. Prior to 15 encryption, the process standardizes the input data to ensure consistent encryption. The encrypted data records are transmitted securely to a common "central facility," which may receive data records from multiple data sources or vendors for assembly in the longitudinal database. 20 A second component of the exemplary system is implemented at the common central facility, which receives encrypted data records from the multiple data source or vendor locations. The various data sources or vendors may have encrypted the transmitted data records differently using the longitudinal encryption key and the data source specific encryption key. Accordingly, the second system component is configured to further 25 process the received data records into standard encrypted formats. This process includes partial decryption of the data source specific encryption to a longitudinally encrypted level, and may further include re-encryption of the received data records using a "central facility" encryption key (e.g., a token based key) to add an additional layer of encryption protection on the data before use. The data records are processed at the second component into a common encrypted format that allows the data records to be linked longitudinally (at the third system component). The second component processes are designed to be irreversible so that the processed data record attributes or fields cannot be decrypted under any circumstance to reveal the original patient-identifying attribute values. A third and final system component, which also may be implemented at the central facility or other convenient longitudinal database facility, assembles the processed data records in the longitudinal database. This component applies matching algorithms, which may configurable according to data source specifics, against the processed data to determine suitable fictitious or pseudo "longitudinal patient" identifiers which can be used to label or tag the encrypted data records (processed at the second component ). The two-component encrypted data records may, for example, be associated or tagged with a unique de-identified central facility "Longitudinal Patient Identifier, (LI)". Suitable statistical matching algorithms, which may be customizable, are employed to associate the LIs with the encrypted data records. The LIs can be used to link the respectively tagged data records longitudinally - i.e. LI by LI, to assemble the longitudinal database. The longitudinal databases, which are assembled by the inventive system and processes, may address pharmaceutical and healthcare industry demand for multi-sourced longitudinal data analysis while protecting individual patient privacy. The inventive data encryption processes using multiple encryption keys, allow data sources (suppliers or vendors) to provide patient-level data to a longitudinal database facility, without fear of compromising HIPAA regulations or other privacy requirements. In addition, the inventive two-component encryption processes restrict the ability of a user or intruder to discover the identity of any individual patient from the encrypted data records. Thus, a longitudinal database facility can safely analyze data records received from multiple, independent data sources without risk of breaching individual patient privacy. -4- H:\tid\lniienvoven\NRPoitbl\DCC\TLD\656334 I _I.doc-I4A)8/2014 -5 Brief Description of the Drawings Preferred embodiments of the present invention are hereafter described, by way of non limiting example only, with reference to the accompanying drawings, in which: 5 Figure 1 is a block diagram illustrating system components and processes for assembling a longitudinal database from multi-sourced patient data records; Figures 2a- 2d illustrate exemplary data supplier input and output file formats conforming to software applications that are used to process patient data records; 10 Figure 3 illustrates exemplary structure and formats of various encryption keys deployed in the software applications for processing patient data records; Figures 4a-c illustrate exemplary data source audit file formats and audit reports generated by the software applications for processing patient data records; Figures 5a-c illustrate exemplary audit file formats and audit reports generated by the 15 software applications for processing patient data records; and Figure 6 illustrates exemplary formats for some characters and standardized data fields in files that are processed by the software applications; While the present invention will now be described in detail with reference to the Figures, it 20 is done so in connection with the illustrative embodiments. Detailed Description of Preferred Embodiments of the Invention Preferred embodiments of the present invention provide solutions for integrating patient 25 data records, which may be acquired from multiple sources into a longitudinal database. The patient data records are integrated into the longitudinal database, at least on a statistical basis, individual patient-by-patient, without any risk of breaching of any individual patient's privacy.

An exemplary solution involves multiple data processing steps in which patient data records are encrypted at least twice using multiple encryption keys. The solution may be implemented in a data processing system, which spans or involves one or more entities (e.g., a data source, vendor or supplier ("DS"), a longitudinal database facility ("LDF"), a third party implementation partner ("IP"), and an encryption key administrator. The IP, who also may serve as the encryption key administrator, may be authorized to operate at a LDF component site. At a first processing step of the solution, each DS prepares data records for secure transmission to the LDF. The DS standardizes and encrypts selected data fields (e.g., patient-identifying attributes and/or other standard attribute data fields) in the patient data records to convert the patient data records into a first "anonymized" format. Each DS uses two keys (i.e., a vendor-specific key and a common longitudinal key associated with a specific LDF) to doubly-encrypt the selected data fields. The doubly-encrypted data records are transmitted to the LDF component site, where the IP can further process the data records. The received data records are further processed at the LDF component site into a second anonymized format, which is designed to allow the data records to be linked (on a statistical basis) individual patient-by-patient without recovering the original unencrypted patient identification information. For this purpose, the doubly-encrypted data fields in the patient records received from a DS are partially decrypted using the specific vendor key (such that the doubly-encrypted data fields still retain the encryption by the common longitudinal key). A third key (e.g., a token based key) may be used to further prepare the now-singly encrypted (i.e. common longitudinal key encrypted) data fields for use in a longitudinal database. Longitudinal identifiers (lDs) or dummy labels that are internal to the LDF, may be used to tag the data records so that they can be linked individual ID by ID in the longitudinal database without knowledge of original unencrypted patient identification information. Statistical matching algorithms are used to associate the IDs with the data records. FIG. I shows components 11 Oa, 11 Ob and 11 Oc of an exemplary data processing system 100, which may be used for assembling multi-sourced patient data records into a longitudinal database 200 at an longitudinal data base facility 130. Component 1 10a relates to data processing activities at one or more data source locations 116 (e.g., data suppliers or vendors, pharmacies, hospitals, etc.) -6- At these locations, raw transaction data records (e.g., patient prescription records) are prepared for transmission to a LDF. The raw data records may contain patient identifying attributes 112 and other attributes (e.g., prescription attributes 114) that are non-identifying infonnation. Component 110 includes a two-key encryption application, which is used to doubly-encrypt and code patient-identifying attributes 112 in the raw data records. This double-encryption may be accomplished using a longitudinal facility encryption key "K I" and a data supplier's private encryption key "K2". The resulting doubly-encrypted data attributes 11 2e are merged with corresponding prescription attributes 114 in the data records by a suitable merging application (e.g., application 118b). The merged data records may be assembled as doubly-encrypted data records in a data file 121 or other form suitable for secure transmission to longitudinal database facility (e.g., LDF 130). Components 1 Ob and 1 Oc relate to data processing activities at LDF 130. Component 11 Ob is configured to remove data source specific characteristics from the data record. This may be achieved, for example, by partial decryption of the doubly-encrypted data attributes in a data record received from an individual data source. A suitable decryption application 122 is included for this purpose in component IOb. The data records in file 121 received at LDF 130 are first processed by a split process application 126 to separate prescription attributes 114 from the doubly-encrypted data attributes 112e. The doubly-encrypted data attributes 112e are then processed by decryption application 122. Decryption application 122 may use a public key K2' (which is complementary to the data supplier's private encryption key K2) to partially decrypt the doubly-encrypted data attributes 112e, and thereby recover singly-encrypted data attributes 112a. The singly-encrypted data records correspond to raw data record attributes that are singly encrypted by longitudinal facility encryption key K1. Component 110b also may include another encryption application 124 to further encrypt the now singly-encrypted data attributes 112a in a manner specific to particular LDF 130. Encryption application 124 may, for example, use a token-based key K3 to encrypt data attributes 112a to generate LDF-encrypted attributes 112f. Component 1Oc at LDF 130 includes a data acquisition process 1 15a, which issued to segregate prescription attributes 114 into two categories (1) a set of standardized prescription attributes 114a, and (2) a set of matching attributes 114b. -7- The latter attributes may be used in a linking process application I 15b by which encrypted data records 112f are associated with or assigned unique longitudinal patient identifiers (IDs). Component 1 10c also may include a longitudinal ID cross reference table or database 113, which can be referenced by application 11 5b during the assignment process. Cross-reference database 113 may include a list of IDs and sets of corresponding data attribute values. Linking process application 115b may include suitable matching algorithms, which assign unique longitudinal patient identifiers (IDs) to the encrypted data records on a statistical and/or deterministic basis by matching attributes in the encrypted data records with reference attribute values. Exemplary matching algorithms are described in co-invented and co-pending U.S. patent application No. , filed on even date. The assigned IDs may be stored in a ID database 117. Component 1 IOc may further include a merge process application 119, which merges standardized prescription records 114a by assigned IDs 117 into a longitudinal database 200. The probabilistic and/or deterministic matching algorithms of linking process application 11 5b are designed to utilize only a small set of encrypted patient identifying data attributes for effective assignment of a longitudinal patient identifier (ID). In an exemplary implementation, the assignment of a longitudinal patient identifier (ID) may be based on a select set of a few encrypted patient-identifying data attributes (e.g., up to only eight attributes such as a patient's date of birth, cardholder identification, record number, zip code, first name, last name, street address, and an industry standard patient identifier. The industry standard patient identifier may, for example, be a patient identifier according to the National Council for Prescription Drug Programs (NCPDP) standards). The selected set of data attributes in a data record may be first placed in a standard format using any suitable standardization algorithm and then stored, for example, in database or file 112. The selected set of data attributes is then twice encrypted using a two-key encryption application l I8a, which includes a longitudinal encryption key Ki and vendor specific encryption key K2. In particular, each data attribute in the selected set is independently encrypted twice at the data source 116. First, the universal longitudinal encryption key KI (which is available only at data suppler sites but not at LDF 130) is used to encrypt the data attributes. Then, the vendor specific key K2, is used to again encrypt the data attributes. The set of twice -8encrypted data attributes and the prescription attributes in the data record are transmitted to LDF 130 in a suitable electronic file (e.g., in file 121). LDF 130 may arrange with a third party implementation partner (IP) to assist data suppliers or sources in defining and implementing the acquisition, encryption and transmission processes. Further, LDF 130 may include a secure processing environment 133 to receive and process the transmitted data file 112. Secure processing environment 133 may be accessible only to the third party IP, to minimize or avoid the risk of breach of patient privacy by LDF personnel or processes. The secure processing environment 133 may include a list of the vendor specific keys K2-K2' corresponding to all data sources involved in supplying data to LDF 130. These keys may be used by the IP to partially decrypt the twice encrypted set of data attributes in secure processing environment 133. Secure processing environment 133 also may include suitable counting processes to account for and audit records for each data source 116. Similar counting processes may be conducted in components IOb and I 1c to provide traceable audit records of all activities. The traceable audit records may be useful for identifying and resolving data quality issues that may occur. In another view, the present invention can be described from a perspective of software implementations of data processing system 100. Various software applications may be utilized in components 1 Oa-11 Oc of data processing system 100 to conduct different data processes (e.g., processes 118a, 118b, 122, 124, 126, 115a, 115b, and 119). The software applications may be designed so that the processes can be operated individually or separately by the following primary parties: a LDF, various data suppliers who communicate various healthcare transactions to the LDF, a key administrator responsible for generation and management of encryption keys used at both the data suppliers' and LDF's environments, and an implementation partner (IP) who may be responsible for installation / maintenance of the execution environments. The IP and the key administrator may be the same party. The separate software applications may be designed to meet specific private initiative privacy standards, government regulations, and industry standard data formatting requirements. The separate software applications are operated in a integrated manner (referred to herein collectively as the "Software System") across data processing system 100. -9- The Software System and its applications provide various functions at various stages of steps in the inventive double encryption/matching solution for constructing a longitudinal database with formatted data records. The functions may be called or operated by an authorized user (e.g., one or more of the four parties: LDF, DS, IP and/or key administrator). The Software System through its encryption applications, enables multiple data suppliers to provide an LDF with data records that have patient attributes in an industry accepted secure encrypted format. The encrypted attributes are in placed in a format that enables the LDF to link multi-sourced transaction records by individual. The format of the encrypted attributes is such that it does not allow the LDF to learn the identity of the individual(s). However, the format allows the LDF to link the records by individual using suitable statistical matching applications without having to learn any individual's identity. The Software System also may include other applications or routines for performing other functions, e.g., standardizing data, reformatting or acquiring data attributes, and generating audit counts or audit reports The Software System includes at least a LDF encryption application and a DS encryption application, and at least one key administration (IP) application.. The DS encryption application may include one or more routines that are designed to acquire data records or attributes, standardize the data attributes including HIPAAtization of data attributes if requested, and doubly encrypt patient-identifying information. The DS encryption application may include routines for creating encryption keys and to encrypt the encryption keys. A secure storage is provided for the encryption keys. The DS application also may include routines for auditing and secure auditing data management (audit data to be written to a file), and audit reporting. Further, the application also may include routines for reference file integration (HIPAA, zip codes/ first name standardization file etc.). The LDF encryption application allows decryption and encryption of data supplier's encrypted attributes using a data supplier key and a LDF encryption key, respectively. A secure storage is provided for the encryption keys. The LDF application also may include routines for auditing and secure audit data management (e.g., writing audit data to a file), and audit reporting. Secure audit data management advantageously allows parties (e.g., a LDF) to demonstrate or verify that the audit - 10 files have not been modified. Secure auditing data management may include the capability to detect modification and/or tampering of the audit files or reports. The LDF application may be deployed on suitable hardware security modules (e.g., an nCipherHSM). The key administration application may include routines for generation of encryption keys, a solution for secure storage solution for encryption keys, and processes for deploying the encryption keys to data suppliers, LDF, and IP. FIG. 3 shows exemplary formats and characteristics of the encryption keys (e.g., K1, K2 and K3). In an implementation of the Software System, the software applications and encryption keys may be installed by the three parties (DS, LDF and IP) by themselves or with the help of external technical support. The Software System may be designed to be operate on a variety of hardware or software environments or platforms. The data supplier applications, in particular, may be designed to be operated to be deployed across a variety of platforms that are in common use. The reporting routines in the applications may involve customized GUIs. The applications may include subroutines to fill in blanks (e.g., write a 'B' character to an attribute's data field) if the input data received has missing data fields (blank, spaces, zeros). Similarly, the applications may mark (e.g., write an 'I' character to an attribute's content) if the data fields received are invalid. (See e.g., FIG. 6). The Software System may be configured so that the LDF will not have access to the LDF encryption application software, encryption keys, or any encryption hardware. In suitable implementations of the Software System, the LDF encryption application and keys may be maintained and operated in a secure hardware platform that can only be accessed using a smart card or other restricted access mechanism. A suitable hardware platform may be nCipher's nShield Hardware Security Module. In preferred embodiments of the Software System, the encryption applications and other applications may be designed to be sufficiently generic to accept a variety of input file formats. For example, some data suppliers may choose to send the patient-identifying attributes and the non-identifying transaction attributes as a separate file, while other data suppliers may choose to send both together in a single file. The encryption applications may be designed to operate on either format. -11 - The individual data supplier file formats may be agreed upon by the data supplier and the LDF beforehand. Upon receipt of data from a data supplier, the LDF may map the received encrypted attributes into the LDF Data Encryption application's standard format. Since common data suppliers usually provide data files that contain the transaction data and patient data in the same record, separation / merging of encrypted and unencrypted data attributes is not necessary and may be optional in most instances. The Software System applications may be configured so that data attributes that need to be encrypted are acquired and encrypted, and then added to other non-encrypted transaction attributes at the end of a data record. The data file, data record, and report formats at input and/or output stages of various applications in the Software System may be standardized. The file or data records may, for example, have a defined structure with fixed length data fields. FIGS. 2a-d show exemplary formats for Data Encryption Application (DS) Input Record and Output Record, Data Encryption Application (LDF) Input File and Output Files. FIGS. 4a-4c show exemplary Data Supplier Audit file formats, and FIGS 5a-c show exemplary LDF Audit file formats. With reference to FIG. 2a, the data records in data supplier encryption input file 2001 may have a standardized fixed length, e.g., 1 to 634 bytes long. Each data field in an input data record is reserved for a designated data attribute (e.g., prescription or patient-identifying attribute). A select set of designated attributes may be intended for encryption (e.g., eight data fields at bytes 501-632, which contain sensitive private information such as patient's date of birth, zip code, etc.). With reference to FIG. 2b, data supplier encryption output file 2002 shows encrypted versions of the selected set of eight designated attributes at bytes 512-923. Output file 2002 also contains attributes, which identify the data supplier and the date of processing (e.g., bytes 471-481). These attributes may be useful for audit or tracing purposes. Output file 2002 may contain optional attributes, which correspond to HIPAA versions of the patient zip code and birthdate, for example, at bytes 926-938. With reference to FIG. 2c, the data records in LDF encryption input file 2003 include the data supplier encrypted set of eight designated attributes (at bytes 44 -455). Input file 2003 also includes optional data fields identifying the gender of the patient (byte 1), a data supplier identifier (bytes 2-5), a processing date -12- (bytes 6-13) and matching attributes (bytes 14-43). The matching attribute data fields may be used to store suitable parameters (e.g., ID) that can be used to match encrypted attribute contents back to a data supplier's unique transaction attributes. FIG. 2d shows the format of an exemplary output file 2004 in which the set of the data supplier encrypted set of eight designated attributes (in file 2003) have been further encrypted by the LDF encryption application (e.g., a token based application). The System Software and its applications provide various functions at various stages of the inventive double encryption/matching solution for constructing a longitudinal database with formatted data records. The functions may be called or operated by an authorized user (i.e., one or more of the involved parties, LDF, DS or IP/key administrator). The Software System may include specific functions for acquiring data attributes, standardizing data attributes, making the data HIPAA Privacy compliant, and encrypting the data attributes, and requesting audit count reports at a data supplier location ( e.g., component 1 10a), and specific functions for acquiring data-supplier encrypted data attributes, and encrypting the data attributes and requesting audit count reports at an LDF location. The Software System also may include specific functions for key administration, key security, and audit storage. Exemplary properties and features of the various functions of a System Software configuration are shown in tabular form in Appendices A-K. The foregoing merely illustrates the principles of the invention. Various modifications and alterations to the described embodiments will be apparent to those skilled in the art in view of the teachings herein. It will thus be appreciated that those skilled in the art will be able to devise numerous techniques which, although not explicitly described herein, embody the principles of the invention and are thus within the spirit and scope of the invention. -13i:\Ild\ntIm'vovn\NKPolbl\DUCVT LD\68793 15 . docx-14/10/2014 - 13A Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", and variations such as "comprises" and "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps. 5 The reference in this specification to any prior publication (or information derived from it), or to any matter which is known, is not, and should not be taken as an acknowledgment or admission or any form of suggestion that that prior publication (or information derived from it) or known matter forms part of the common general knowledge in the field of 10 endeavour to which this specification relates.

Appendix A Data Supplier Encryption Function: Acquire Data Attributes Summary The operational processes at the data supplier end acquire the required attributes (that are to be encrypted / standardized) from the data supplier's data sources and convert them into a format that is accepted by the encryption process. Primary Actor(s) Data Supplier, Implementation Partner Secondary None Actor(s) Pore-conditions Configuration file containing: * The list of attributes (and offsets) that are sent as input to the encryption application o The configuration parameters required by the application. Post-conditions The input data attributes are passed to the standardization and HIPAA privacy compliance process Exceptions Error message encountered will be written to the command line and application log file Dependent Use N/A Case Description 1. The encryption application will analyze the configuration file and understand 1.1. Record type where the encrypted information exists (Some data supplier provide multiple record types in one file. For example a header, footer and detail record). The process will only utilize the record e indicated b this parameter. 1.2. Input/Output file format like name of the attribute that will be sent in with their start and end positions from beginning of record. 1.3. Gender standardization request - If the data supplier requests for gender standardization this value is set to "yes 1.4. DOB Format indicators as given below A - CCVYMMDD B - MMDDCCYY C - MM/DD/CCYY D - CCYY/MM/DD 1.5. Patient Gender Format indicators as given below -14- A - (M=Male, F=Female, O=Other) B - (l=Male, 2=Female, 3=Unspecified) 1.6. HIPAA attributes - If HIPAA attributes are requested by the data supplier, the encryption application should produce the HIPAAtized value of patient zip code and patient date of birth. 1.7. Data supplier ID 1.8. Location of data supplier longitudinal key file and data supplier encryption key file. 1.9. Location of input and output data feed file. 1.10. Location of First Name Standardization file 1.11. Location of HIPAA zip code reference file 1.12. Location of Audit and secure audit file The following kinds of scenarios should be handled by the acquire attributes section. 1. Data Supplier Feed format file as in Appendix A 2. Format file with no transactional information and only patient information as in Appendix A. 3. Feed format file with no transactional information and selective patient information. 4. Feed file with selective patient information - 15 - ARDendix B Data Supplier Encryption Function: Standardize Data Attributes Summary The operational processes at the data supplier end acquire the required attributes (that are to be encrypted / standardized) from the data suppliers data sources and convert them into a format that is accepted by the encryption process. These attributes are then subject to standardization process. Primary Actor(s) Application processes that acquire data attributes, Standardize Attributes process a (application process) Secondary Implementation Partner, Data Supplier Actor(s) Pre-conditions 1. Data attributes to be encrypted / standardized have been acquired by the 'Acquire Attribute' process and is made available to the encryption application 2. Patient Gender standardization parameter is set-up for the data supplier 3. Record Number standardization parameter is set-up for the data supplier. 4. Attribute mapping (where attributes reside in the input file) has been set up Post-conditions The input data attributes are standardized and passed into the HIPAA privacy compliance process Exceptions No exception messages. All errors are recorded in the Audit Encryption Record counts Table Dependent Use Case Description Attributes are passed into the Standardize Data Attributes process Process will standardize the Patient Date of Birth Attribute as follows: 2.1. If Patient Date of Birth Format Indicator equals "B, "C or "D", the application will convert the date into indicator "A" format (CCYYMMDD) 2.2. If the Patient Date of Birth Format Indicator does not equal "A" or "B" or "C" or "D" then the encryption application will cease processing, inform the administrator and terminate. No records will be processed. 2.3. If the Patient Date of Birth attribute is missing (blank, spaces or zeros), I will be added to the Missing Patient Date of Birth count and 'B' will be written to the attributes contents. All - other Patient Date of Birth validation will be -16skipped. 2.4. If the year is not valid (valid if age of the patient is less than or equal to 150 years or if the patient's year of birth is less than + 1 year from the current year), I will be added to the invalid Patient DOB Year count and T written to the Patient Date of Birth attribute. All other Patient Date of Birth validation steps will be skipped. 2.5. If the month is not valid (valid if between I and 12), 1 will be added to the invalid Patient DOB Month count and T is written to the Patient Date of Birth attribute. All other Patient Date of Birth validation steps will be skipped. 2.6. If the Day is not valid (valid if between I and 31), 1 will be added to the invalid Patient DOB Day count and T is written to the Patient Date of Birth attribute. 2.7. A check is also made to ensure that the day value does not exceed the maximum value for a month. If the month day combination is not valid (including February in a Leap year), 1 will be added to the invalid patient DOB count and "I" is written to the Patient Date of Birth attribute. All other DOB validations steps will be skipped. Process will standardize the Patient Gender Attribute 3.1. Job accepts the Patient Gender Input Parameter prior to execution 3.1.1. If the Patent Gender input parameter indicates the need to be standardized the patient gender the following steps are executed. Otherwise, the Patient Gender is skipped in the process and no output is produced. 3.2. If Patient Gender Format Indicator equals "A", the application will convert the Gender attributes contents into indicator "B" format (l=Male, 2=Female, 3=Unspecified). The Patient Gender Format Indicator will be configured by the implementation partner and available in a configuration file. 3.3. If the Patient Gender Format Indicator does not equal "A" or "B" the encryption application will cease processing, inform the administrator and terminate. No records will be processed. 3.4. If the Patient Gender is missing (if blank or - 17 space), I will be added to the Missing Patient Gender count and 'B' is written to the Patient Gender attribute. All other Patient Gender validation steps will be skipped. 3.5. If the Patient Gender is not valid (valid if 1, 2, 3), 1 will be added to the invalid Patient Gender Count and 'I' is written to the Patient Date of Birth attribute Process will standardize the Cardholder Id Attribute 4.1. Data contents will be left justified. 4.2. Leading zeros and spaces will be removed. Remaining contents left justified 4.3. All special symbols (-*&I\\][{};:'.,<") will be removed. Remaining contents left justified to fill gaps 4.4. Remaining right most bytes space filled 4.5. If the Cardholder Id is missing (all blanks, spaces, zeros), I will be added to the Missing Cardholder Id count and 'B' is written to the Cardholder Id attribute. All other Cardholder Id validation will be skipped. 4.6. If the Cardholder Id is invalid (all one character or one number, e.g. 11111111 or aaaaaaaa), I will be added to the Invalid Cardholder Id count and 'I' is written to the Cardholder Id attribute. 4.7. All contents will be standardized to all upper case to make sure that the encrypted values are same for comparison. Process will standardize the Record Number 5.1. Job accepts the Record Number Input Parameter prior to execution 5.1.1. If the Record Number input parameter indicates the need to standardized the Record Number the following steps are executed. Otherwise, the Record Number is skipped in the process and no output is produced. 5.2. Data contents will be left justified 5.3. Leading zeros and spaces will be removed. Remaining contents left justified 5.3.1. Leading zeros in a Record Number will be stripped off uniformly across data suppliers for standardization purposes even if they are a valid piece of the record number. 5.4. All special symbols (-*&I\\][{};;'.,<") will be removed. Remaining contents left justified to fill gaps 5.5. Remaining right most bytes space filled -18- 5.6. If the Record Number is missing (all blanks, spaces, zeros), I will be added to the Missing Record Number count and 'B' is written to the Record Number attribute. All other Record Number validation will be skipped. 5.7. If the Record Number is invalid (all one character or number), I will be added to the Invalid Record Number count and 'I' is written to the Record Number attribute. 5.8. Contents will be standardized to all upper case Process will standardize the Patient Zip Code 6.1. If the Patient Zip Code is missing (all blanks, zeros), I will be added to the Missing Patient Zip Code count and 'B' is written to the Patient Zip Code attribute. All other Patient Zip Code validation will be skipped. 6.2. If the Patient Zip Code is invalid Patient Zip Code count and 'I' is written to the Patient Zip Code attribute. All other Patient Zip Code validation will be skipped. A Patient Zip Code will be invalid if it contains e Contains one or more special characters Process will standardize the Patient Name 7.1. For the Patient First Name and Patient Last Name attributes the following standardization will be performed (Refer to Appendix D for valid input character list) 7.1.1. Data contents will be left justified 7.1.2. Leading zeros, blanks and spaces will be removed. Remaining contents left justified 7.1.3. All special symbols (-*&"\][{};:'.,<") will be removed. Remaining contents left justified to fill gaps 7.1.4. Remaining right most bytes space filled 7.1.5. Contents will be standardized to all upper case 7.2. If the Patient First Name is missing (all blanks, zeros, spaces), I will be added to the Missing Patient First Name count and 'B' is written to the Patient First Name attribute. All other Patient First Name validations will be skipped. 7.3. If the Patient Last Name is missing (all blanks, zeros, spaces), I will be added to the Missing Patient Last Name count and 'B' is written to the Patient Last Name attribute. All other Patient Last Name validations will be skipped -19- 7.4. If the Patient First Name is invalid (all numbers, all the same character), I will be added to the Invalid Patient First Name count and 'I' is written to the Patient First Name attribute. All other Patient First Name validations will be skipped. 7.5. If the Patient Last Name is invalid (all numbers, all the same character), I will be added to the Invalid Patient Last Name count and 'I' is written to the Patient Last Name attribute. All other Patient Last Name validations will be skipped 7.6. The Patient First Name contents will be compared to the Common First Name field in the First Name Standardization file. If a match is found, the Standard First Name contents will be copied to the Patient First Name field Process will standardize the NCPDP Patient Id Attribute 8.1. If no NCPDP Patient Id Qualifier is provided in the input file, it should be assumed to contain "99" for the remainder of the processing 8.2. If the NCPDP Patient Id Qualifier Indicator does not equal "01", "02", "03", "04", "05") "99" or spaces then 1 is added to the Invalid Patient Id Qualifier count and "I" is written to the Patient Id attribute. All other Patient Id validation steps are skipped. 8.3. For the NCPDP Patient Id attribute the following standardization will be performed: 8.3.1. Data contents will be left justified 8.3.2. Leading zeros, blanks and spaces will be removed. Remaining contents left justified 8.3.2.1. Leading zeros in a patient id will be stripped off uniformly across for standardization purposes for data suppliers even if they are a valid piece of the cardholder's id. 8.3.3. All special symbols (-*&I\\][{};:'.,<") will be removed. Remaining contents left justified to fill gaps 8.3.4. Remaining right most bytes spaced filled 8.3.5. Contents will be standardized to all upper case. 8.4. If the NCPDP Patient Id is missing (all - 20 blanks, zeros or spaces), I will be added to the Missing Patient Id count and 'B' is written to the Patient Id attribute. All other Patient Id validations will be skipped 8.5. If the NCPDP Patient Id is invalid (all the same numbers or all the same character), 1 will be added to the Invalid Patient Id count and 'I' is written to the Patient Id attribute. All other Patient ID validations will be skipped. Process will standardize the Patient Street Address Attributes 9.1. Data contents will be left justified 9.2. Leading spaces will be removed. Remaining contents left justified 9.3. All special symbols (-*&/\\][{};:'.,<") will be removed. Remaining contents left justified to fill gaps 9.4. If the Patient Street Address is missing (all blanks, spaces, zeros), 1 will be added to the Missing Patient Street Address count and 'B' is written to the Patient Street Address attribute. All other Patient Street Address validation will be skipped. 9.5. If the Patient Street Address is invalid; I will be added to the Invalid Patient Street Address count and 'I' is written to the Patient Street Address attribute. Patient Street Address will be considered invalid if 9.5.1. Street Address does not contain any numbers 9.5.2. Contains more than 15 numeric digits in a row The process will read the Patient Street Address Attribute from left to right one digit at a time until a numeric value is found. The process continues reading until a non-numeric digit is found. The numeric value found from the starting and ending point will be written to the Patient Street Address attribute. For example if "1245 Main Street, Apartment 403" was provided, "1245" will be written to the Patient Street Address at the conclusion of the Patient Street Address standardization process. The maximum number of digits to be accepted in a row will be 15. Standardized Attributes are passed to the HIPAA Privacy Compliance process -21- Appendix C Data Supplier Encryption Function: HIPAA Privacy Compliance Summary After standardization, HIPAA privacy compliance process is invoked which creates HIPAA compliant attributes for Patient Date of Birth and Patient Zip Code attributes. This standardization process is only invoked if the data supplier requests for the same in the encryption application. Primary Standardize Attributes process (application process), Actor(s) HIPAA Privacy Compliance process a application process Secondary Implementation Partner, Data Supplier Actor( s) Pre-conditions All input attributes have been standardized by the Standardize Attributes application process Post- HIPAA input parameter has been setup conditions Exceptions A HIPAA compliant Patient Date of Birth and Patient Zip Code attributes are produced Dependent No exception messages. All errors are recorded in the Use Case Audit Encryption Record counts Table - 22 - Description Jobs accepts the HIPAA input parameter prior to execution 1.1. If the HIPAA input parameter is set up to exclude the HIPAA processing the remainder of the steps are skipped. Process accepts the Patient Date of Birth and Patient Zip Code attributes from the Standardize Attributes process To create the HIPAA Patient Date of Birth attribute the following processes are performed against the Patient Date of Birth Field 3.1. For the HIPAA Patient Date of Birth attribute "01" is written to the Day and Month components of the date 3.2. For the HIPAA Patient Date of Birth Year component the year is copied from the Patient Date of Birth attribute 3.2.1. If the Patient Date of Birth attribute contains 'B' or 'I' move "1800" to the HIPAA patient Date of Birth year 3.3. If the year of birth calculates to an age of over 88 years old, move "0000" to the HIPAA Patient Date of Birth year component. 3.3.1. Add I to the HIPAA Patient Year Over 88 counter within the Audit Encryption Counts table To create the HIPAA Patient Zip Code attribute the following processes are performed against the Patient Zip Code Attribute 4.1. If the Patient Zip Code is '' (invalid), move 'I' to the HIPAA Patient Zip Code 4.2. If the Patient Zip Code is 'B' (missing), move 'B' to the HIPAA Patient Zip Code 4.3. If the Patient Zip code is Valid (not missing or. invalid) 4.3.1. Move the Patient Zip Code to the HIPAA Patient Zip code 4.3.2. Move "00" to the right most two digits of the HIPAA Patient Zip Code 4.3.3. Compare the left most remaining three digits of the HIPAA Patient Zip Code to the HIPAA Zip Code Reference Table. If a match is found move "00000" to the HIPAA Patient Zip Code attribute 4.3.4. Add I to the HIPAA Under 20K Zip counter within the Audit Encryption Counts table Pass control to the Encrypt Data Attribute process - 23 - Appendix D Data Supplier Encryption Function: Encrypt Data Attributes Summary After the hipaatization process, attributes to be encrypted are passed to the encryption process. Primary HIPAA Privacy Compliance process (application Actor(s) process), Encrypt Data Attributes process (application process) Secondary Implementation Partner, Data Supplier Actor(s) Pre-conditions Data Attributes have been processed by the Standardize Data Attributes and HIPAA privacy compliance processes Post-conditions The input data attributes are encrypted. The encrypted attributes are available as output from the encryption process Exceptions No exception messages. All errors are recorded in the Audit Encryption Record counts Table Dependent Use Case Description 1. Process accepts attributes from the HIPAA Privacy Compliance Process 2. The following attributes will be encrypted e Patient Data of Birth * Cardholder Id " Record Number (ifrequested) e Patient Zip Code * Patient First Name e Patient Last Name e NCPDP Patient Id * Patient Street Address 3. To encrypt an attribute the application will 4. Acquire the attributes contents 4.1. If the attribute's content is 'I' or 'B' (identified to be missing or invalid) move the contents directly into the corresponding encrypted attribute. Skip the remaining encryption steps for the attribute 4.2. Acquire the encrypted Data Supplier Longitudinal Encryption Key 4.3. Decrypt the Data Supplier Longitudinal Encryption Key into its raw format 4.4. Check for integrity of the Data Supplier Longitudinal Encryption Key. If the matching key is not found, the application will log an error and exit 4.5. Encrypt the attributes contents using the -24 - Data Supplier Longitudinal Encryption Key (un-encrypted) 4.6. Acquire the encrypted Data Supplier Encryption Key 4.7. Decrypt the Data Supplier Encryption Key into its raw format 4.8. Check for integrity of the Data Supplier Encryption Key. If the matching key is not found, the application will log an error and exit 4.9. Encrypt the longitudinal encrypted results using the Data Supplier Encryption Key (un encrypted) 4.10. Repeat this step until all the input data attributes have been encrypted 5. After all attributes for a record have been encrypted add I to the Data Output Record count in the Audit Encryption Record Counts file -25 - Appendix E Data Supplier Encryption Function: Request Audit Counts Reports Summary The data encryption process creates audit counts for reporting / analysis. The reporting process, when requested, creates a report containing aggregated data encryption process audit counts Primary The operational process requesting a report Actor(s) Secondary Implementation Partner, Data Supplier Actor(s) Pre-conditions Data encryption application has processed creating audit counts for reporting / analysis Report parameters have been defined Post- Standard report of audit counts is produced for analysis conditions Exceptions Default values used if input parameters are invalid Dependent Use Case Description Implementation Partner requests Audit Count information for analysis User defines the execution parameters 1. Start Date (CCVYMMDD) - Audit information prior to the date specified will not be included. Defaults to current date minus one month 2. End Date (CCYYMMDD) - Audit information after the date specified will not be included. Defaults to current date 3 Aggregation Level 3.1. Single (S) - Data aggregated by individual count into a single aggregated number (default). Aggregated counts for the entire time period 2.3.2. By Day (D) - Data aggregated by unique date and present for each date encountered. Aggregated counts by day for the entire time period 2.4. Reporting Option 2.4.1. Print copy (P) - Results displayed in a report format for printing / faxing 2.4.2. File copy (F) - Results written to a standard, predefined file location for further analysis j transmission (default). File format is identical to the Audit Encryption Counts table and enhance to hold potentially larger aggregated counts 2.5. If any parameter is invalid the default value is used User executes report -26 - Appendix F LDF Encryption application Functions: Acquire Attributes Summary This process will acquire attributes from the data supplier encryption process after the data feed split process. The split process will split the data supplier feed into transactional records and patient records. Primary LDF, Implementation Partner Actor(s) Secondary Actor(s) Pre- New Data feed is available for processing at LDF conditions Post conditions Exceptions Error messages encountered will be written to the log file Dependent N/A Use Case Description The encryption application will process the configuration file and understand the list of attributes that will be encrypted. Refer Appendix A for LDF input file format. The LDF encryption application should be able to handle the following scenarios 1. Data file containing patient information. 2. Data file containing limited patient information. -27 - Appendix G LDF Encryption Function: LDF Data Encryption Summary The encryption application will decrypt the records received after the acquire attribute process using the data supplier longitudinal key followed by encryption using the LDF longitudinal key. Primary Encryption Application (application process), Hardware Actor(s) Security Module (nCipher) Secondary Implementation Partner Actor(s) Pre- 1. Encrypted Attributes are identified and separated conditions before passing it on the encryption application. (Scope of implementation partner). 2. Data Supplier encryption key, LDF Longitudinal Encryption Key and the LDF Data Encryption software has been successfully installed and secured in a restricted access environment managed by the 3. Execution Type parameter set prior to execution. 4. Attribute mapping of the encrypted attributes on the input file have been set-up. Trigger File containing all encrypted attributes have been identified and separated. Pot- None conditions Frequency Agreed upon between LDF and Data Supplier Exceptions Exception message 1 if the file is not complete. All errors are recorded in the Audit Encryption Record counts Table Special The matching attribute in the data file will serve as Requirement a link to the corresponding health care records. It s will be passed through the. process and remain linked to the individual transaction information. The encryption application runs in a secure isolated hardware platform to which only the implementation partner and the key administrator have access. Audit counts containing the number of encrypted records Description Encryption application process accepts records once they are identified and separated. The attributes that need to be decrypted are e Patient Data of Birth " Cardholder Id e Record Number e Patient Zip Code -28- * Patient First Name * Patient Last Name e Patient Id * Patient Street Address The process will accept the Execution Type parameter. 2.1. The valid execution types will be 2.1.1. Blank - routine execution 2.1.2. R - Re-execution due to processing error 2.1.3. D - Process identified the current feed to be a duplicate of one already processed Within the hardware security module; 3.1. -Extract and decrypt the data supplier encryption key and the LDF longitudinal key from the hardware security module and load it in secure access area. 3.2. For each encrypted record received 3.2.1. Add one to the LDF Data Encryption Record Received count for the identified data supplier. 3.2.2. Each time a new Supplier Number / Data Supplier Processing Date is encountered a new LDF Audit Encryption count record is written to the secure and non-secure file. 3.2.2.1. The Supplier number is populated with the Supplier Number on the Encrypted data record 3.2.2.2. The Data Supplier's Processing Date is populated with the Data Supplier's Processing date received. 3.2.2.3. The LDF Processing Date is populated with the current system date 3.2.2.4. Populate Execution type field as appropriate. 3.2.2.5. All other count fields are initialized to zero 3.3. Extract attribute by attribute from the file. 3.4. Application will validate each encrypted attribute to match the size / data type as expected (from the Data Supplier Encryption application). Validation includes: 3.4.1. Proper length 3.4.2. Attributes are fully populated - 29 - 3.4.3. Attributes are populated with a encrypted code or populated with 'I' or 'B' (invalid or missing data submitted) 3.4.4. If the encrypted data attributes are not properly formatted exception message 1 will be displayed and the process is terminated (until the input file errors can be resolved) 3.5. Perform decryption on the data attribute using the data supplier encryption key 3.5.1. Determine if the attribute can be decrypted. If yes perform decryption. 3.5.2. Attributes populated with 'P or '8' (invalid or missing data submitted) will not be decrypted 3.6. For each record decrypted add 1 to the Attribute Unencrypted Count field within the LDF Audit Encryption Counts file 3.7. Encrypt the attribute using the LDF longitudinal key. 3.8. Attributes populated with 'I' or 'B' (invalid or missing data submitted) will not be encrypted 3.9. Add 1 to the Attribute LDF Encrypted Count field within the LDF Audit Encryption Counts file. 3.10. Write the re-encrypted record to the LDF Encrypted Record file / table For each record encrypted add 1 to the Record Release Count field within the LDF Audit Encryption Counts file / table. 3.11. On completion of decryption destroy data supplier encryption key and the LDF longitudinal key from the memory space of the Hardware Security Module. 3.12. Record is passed into the LDF Data Encryption Process -30- Appendix H LDF Encryption Application Function: LDF Audit Count Reporting Summary This use describes the process of generating reports from both secure and non-secure logs Primary Implementation Partner Actor Secondary None Actor(s) Pre- 0 LDF data encryption application has processed conditions creating audit counts for reporting/analysis 0 Report execution parameters have been set-up Trigger None Post- Standard report of audit counts produced for analysis conditions Frequency Not Applicable Exceptions Default values used if input parameters are invalid Special None Requiremen ts Description 1. Implementation Partner requests Audit Count information for analysis 2. User defines the execution parameters 2.1. Date to be utilized -User specifies whether to use the data supplier processing date or the LDF processing date. 2.2. Start Date (CCVYMMDD) - Audit information prior to the date specified will not be included. Defaults to current date minus one month. 2.3. End Date (CCVYMMDD) - Audit information after the date specified will not be included. Defaults to current date. The end date can be either a data supplier processing date or the LDF encryption application processing date 2.4. Data Supplier - Data supplier id to execute report against. Blank assumes all data suppliers (default) 2.5. Aggregation Level 2.5.1. Single (S) - Data aggregated by data supplier by unique count into a single aggregated number (default) 2.5.2. By Day (D) - Data aggregated by data supplier by unique date and present for each date encountered 2.6. Reporting Option 2.6.1. Print copy (P) - Results displayed in a report format for printing / faxing -31- 2.6.2. File copy (F) - Results written to a standard, predefined file location for further analysis / transmission (default). File format is identical to the Audit Encryption Counts table and enhance to hold potentially larger aggregated counts 2.7. If any parameter is invalid the default value is used 3. User executes report 3. User executes report Reports Audit Data Encryption Report -32- Appendix I Key Administrator Function: Generate Encryption Keys Summary The key administrator serves as a trusted third party and generates various encryption keys for use at the data supplier and LDF ends. Primary Key generation process Actor Secondary Key Administrator Actor(s) Pre- The key generation application has been installed at the conditions key administrator's end Post- A 128-bit encryption key is generated conditions Exceptions Dependent Use Case Description 1. The key administrator invokes the key generation process 1.1. A unique 128-bit encryption key is generated 2. After generation of the key, the process invokes the key storage process for securing the key that is generated 3. The key is encrypted for distribution (compliant with the expectations of the Data Supplier encryption application and the LDF encryption application) -33- Appendix J Other Function: Secure key storage Summary The encryption keys, LDF Longitudinal Key and Data Supplier Encryption Key, should be stored securely. For secure storage of these keys, the keys are stored in an encrypted form. The password for decryption of the keys is known j can be derived only by the application. Primary The application process that implements secure key Actor storage Secondary Key Administrator, Implementation Partner, Data Actor(s) Supplier Pre- 1. At the key administrator end, the Key conditions Administrator has generated the encryption keys using the key generation process and encrypted them. 2. At the data supplier end, the Key Administrator has securely communicated the encrypted LDF Longitudinal Key and Data Supplier Encryption Key to the Implementation Partner at the data supplier end Post- The key generated using the key generation process is conditions encrypted Exceptions Dependent Use Case Description 1. The password for encryption of the encryption keys is buried into / derived by the secure key storage logic 2. The encryption keys are encrypted using the password got from the above step -34- Appendix K Other Function: Secure audit storage Summary The integrity of the audit counts produced by the standardization and encryption process will have to be preserved at both the data supplier end and LDF end Primary The application process that implements secure audit Actor storage Secondary Implementation Partner, Data Supplier, LDF Actor(s) Pre- The audit counts are produced by the encryption and conditions standardization processes Post- The audit is secured by preserving its integrity conditions Exceptions Dependent Use Case Description 1. Process reads the audit counts file 2. Process ensures integrity of the audit counts file, say by calculating a checksum over the file -35-

Claims (24)

1. A process for assembling a longitudinally-linked database from individual patient healthcare transaction data records, the process comprising; 5 at a central longitudinal database facility (LDF), (f) receiving data records including at least one patient non-identifying attribute and at least one individually encrypted patient-identifying attribute, the at least one patient-identifying attribute first encrypted using a first encryption key specific to the LDF and further encrypted 10 with a second encryption key unique to a data source so that original unencrypted patient identification information in the at least one patient (g) identifying attribute remains secret with respect to other data sources; (h) partially decrypting the received data records using a decryption key complementary to the second encryption key so that the at least one 15 patient-identifying attribute retains only the encryption by the first encryption key specific to the LDF; (i) using an attribute-matching algorithm to assign an LDF identifier (ID) to the encrypted data records; and (j) linking the encrypted data records ID by ID, whereby the longitudinally 20 linked data base is formed.

2. The process of claim 1, wherein the step of linking the encrypted data records ID by ID, further comprises encrypting the attributes using a third LDF encryption key. 25

3. The process of claim 1, wherein the step of receiving data records, further comprises processing the acquired data records to place their data fields in a standard content format.

4. The process of claim 1, wherein the step of receiving the data records, further 30 comprises separating the patient-identifying attributes and the non-identifying attributes in the data records, and steps (c) and (d) further comprise encrypting the separated patient- H:\lld\lnenvoven\NRPobl\DCC\TLD\68793 15_Ldoc,-14/10/2014 - 37 identifying attributes.

5. The process of claim 2, further comprising the step of merging the encrypted patient-identifying attributes with the separated non-identifying attributes in the data 5 records.

6. The process of claim 1, wherein the step of using an attribute-matching algorithm, further comprises processing the received data records to separate the encrypted patient identifying attributes and the non-identifying attributes in the data records, and the step of 10 linking the encrypted data records ID by ID, further comprises partially decrypting the separated patient-identifying attributes.

7. The process of claim 6, wherein the step of linking the encrypted data records ID by ID, further comprises merging the partially decrypted patient-identifying attributes with 15 the separated non-identifying attributes in the data records.

8. The process of claim 1, wherein the attribute-matching algorithm assigns an ID to an encrypted data record based on a statistical match of a select set of data attributes. 20

9. The process of claim 8, wherein the select set of data attributes comprises at least one of a patient's date of birth, cardholder identification, record number, zip code, first name, last name, street address, and an industry standard patient identifier.

10. The process of claim 1, wherein the attribute-matching algorithm assigns an ID to 25 an encrypted data record by referencing a cross database of IDs and corresponding attributes.

11. The process of claim 1, wherein the step of using an attribute-matching algorithm and the step of linking the encrypted data records ID by ID, are performed in a secure 30 environment that limits unauthorized access to patient-identifying attribute information in the data records. H:\tld\InIerwovcu\NRPorlbi\DCC\TLD\6879315 Ldocx.-14/10/2014 - 38

12. A system for longitudinally-linking individual patient healthcare transaction data records obtained from multiple data suppliers, the system comprising: at a data supplier location, a first component configured to: 5 acquire data records having patient-identifying attributes and non identifying attributes; and doubly encrypt the patient-identifying attributes in the data records with a first encryption key specific to a central longitudinal database facility (LDF) and a second encryption key specific to the data supplier; 10 at the LDF, a second component configured to: receive doubly-encrypted data records from the multiple data suppliers; partially decrypt the received data records using a decryption key complementary to the second encryption key so that the patient-identifying attributes retain the encryption by the first encryption key specific to the LDF; and 15 perform an additional layer of encryption; and a third component configured to: assign an LDF identifier (ID) to the encrypted data records by matching attributes in the encrypted data records; and link the encrypted data records ID by ID, whereby a longitudinal database is 20 formed.

13. The system of claim 12, wherein the first component is further configured to separate the patient-identifying attributes and the non-identifying attributes in the data records and to encrypt the separated patient-identifying attributes. 25 1

14. The system of claim 13, wherein the first component is further configured to merge the encrypted patient-identifying attributes with the separated non-identifying attributes in the data records. 30

15. The system of claim 12, wherein the first component is further configured to process the acquired data records to place their data fields in a standard format. -:\lIa~IhUkCIW0'CINKronl-1I\U I LUbK -N3 I [I.dOCX-14/0l/2014 - 39

16. The system of claim 12, wherein the second component is further configured to separate the encrypted patient-identifying attributes and the non-identifying attributes in the received data records, and to partially decrypt the separated patient-identifying 5 attributes.

17. The system of claim 16, wherein the second component is further configured to merge the partially decrypted patient-identifying attributes with the separated non identifying attributes in the data records. 10

18. The system of claim 12, wherein the third component is further configured to assign an LDF identifier (ID) to the encrypted data records based on a statistical match of a select set of data attributes. 15

19. The system of claim 18, wherein the select set of data attributes comprises at least one of a patient's date of birth, cardholder identification, record number, zip code, first name, last name, street address, and an industry standard patient identifier.

20. The system of claim 18, wherein the third component is further configured to 20 assign an LDF identifier (LI) to the encrypted data records by referencing a cross database of LIs and corresponding attributes.

21. The system of claim 12, wherein the second component is configured for operation in a secure environment that limits unauthorized access to patient-identifying attribute 25 information in the data records.

22. The system of claim 12, whose functions are implemented using software applications. 30

23. A process for assembling a longitudinally-linked database from individual patient healthcare transaction data records, substantially as hereinbefore described with reference H:\IId\In1Ienoven\NRPorlb\DCC\TLD\6879315_ Ldocx-14/10/2014 - 40 to the accompanying drawings.

24. A system for longitudinally-linking individual patient healthcare transaction data records obtained from multiple data suppliers, substantially as hereinbefore described with 5 reference to the accompanying drawings.