Concept: MCHP Research Registry

 Printer friendly

Concept Description

Last Updated: 2012-04-02

Introduction
    This concept describes the MCHP Research Registry, an extremely important component of the Manitoba Centre for Health Policy (MCHP) Population Health Research Data Repository (Repository). The main source of information for the MCHP Research Registry comes from the Manitoba Health Insurance Registry system. This concept will provide information that distinguishes between the two data sources and describes the importance of developing and maintaining the MCHP Research Registry.
Manitoba Health Insurance Registry versus the MCHP Research Registry
    It is important to differentiate between the Manitoba Health Insurance Registry and the MCHP Research Registry. The Manitoba Health Insurance Registry, maintained and managed by Manitoba Health, is a population-based registry that collects demographic and insurance coverage information that enables the tracking of Manitoba residents for longitudinal, intergenerational analyses and demographic history creation.

    The Data Acquisition Team at MCHP receives a copy of the Manitoba Health Insurance Registry (also known as the Master Registry or the Population Registry) twice a year. Various adjustments/extensions are made to the data by the Data Acquisition Team before the Research Registry is available for approved research projects at MCHP. This resource is known as the MCHP Research Registry. The MCHP Research Registry is a key resource for the research conducted at the Centre and is central to the use of the Population Health Research Data Repository (Repository).
The Manitoba Health Insurance Registry
    Highlights of the coverage details for the Manitoba Health Insurance Registry include:

    • all individuals who have been insured by the Manitoba Health Services Insurance Plan at some point since 1970 are included in the Manitoba Health Insurance Registry (including dates of insurance coverage);
    • individuals who are federally funded (military personnel, and individuals of the RCMP) are excluded from the Manitoba Health Services Insurance Plan and thus not included in the Registry (while ineligible). Spouses and family members of these individuals are part of the Manitoba Health Insurance Registry;
    • individuals who migrated out of the province to another province remain covered by this insurance plan for up to three months after the move. At that point their registration will be inactivated, which means they will remain active in the database for up to three months after they have left the province;
    • individuals who migrate into the province from another province are not eligible for coverage until residential status is achieved (three months after the move into the province), meaning that they will not appear in the Registry until after three months of being in the province.
    For each registrant, the following types of information are collected and maintained in the Manitoba Health Insurance Registry:

    • Family Registration Numbers (REGNOs)
    • Personal Health Information Numbers (PHINs)
    • demographic characteristics (e.g.: date of birth and sex)
    • location of residence (e.g.: postal code and/or municipal code)
    • family composition (relationship to family head), and
    • insurance coverage and cancellation dates

    The following information focuses on three important data elements maintained in the Manitoba Health Registry. This information is derived from emails from Patrick Nicol on December 16, 2008 and November 2, 2010.
1) Registration Number (REGNO)

    REGNOs, are a 6-digit number used to identify the family structure of related registrants. REGNOs were first randomly assigned by Manitoba Health in 1966.

    Assignment of REGNOs:

    • There were various methods used to initially assign REGNOs in 1966:

      • Registered family heads paid a monthly premium
      • Employers registered employees (and their dependents)
      • Enrolment was "grandfathered from the predecessor Hospital Insurance Plan (1958-1966)"
      • Voluntary enrolment
      • Other methods not documented at MCHP (e.g.: tax rolls, census, schools, providers offices, etc.)

    • Currently, REGNOs are issued when/if:

      • a resident turns 18 years of age (since 1992) (prior to 1992, REGNOs were issued at age 19)
      • an individual migrates into the province and establishes residence status
      • an individual is divorced or separated
      • a registrant or guardian requests a new REGNO
      • a resident becomes a ward of the province

    • REGNOs can be cancelled or changed over time:

      • When the last person in a family "cancels" (e.g.: deceased, leaves the province), the REGNO is cancelled and kept unused for up to two years. At that time, the REGNO is put back into the pool for reassignment (information from Fred Toll).

      • A REGNO can change when/if:

        • an individual is married
        • an individual is adopted
        • an individual requests a changed REGNO
        • an individual is absent from the province for more than two years (and returns to a different family or if prior REGNO has been reassigned)

    Note: Although the REGNO is available in the Manitoba Health Insurance Registry, the use of the REGNO is restricted in the MCHP Research Registry.
2) Personal Health Identification Number (PHIN)
    The Manitoba Health Insurance Registry contains Personal Health Identification Numbers (PHINs), unique 9-digit numeric identifiers assigned by Manitoba Health to every person registered for health insurance in Manitoba.

    Assignment of PHINs:

    • Beginning January 1, 1984, all individuals who had been registered with the Manitoba Health Services Insurance Plan at some point prior to 1984 (regardless of whether they still were covered) were assigned a PHIN. This assignment was essentially random although the system saw families in a crude geographical pattern (by REGION, by MUNCODE) roughly going counter-clockwise (from SE corner) around the province. In each zone, the order was by REGNO - which has no particular bias for age or group by 1984.
    • Individuals who were born in Manitoba or migrated into the province after January 1, 1984, were assigned the next available sequential PHIN.
    • PHINs only change if duplication is detected OR in certain legal circumstances involving wards of the province and/or adoption.
    • A cancelled PHIN is never reissued except to the original resident. This may occur when someone leaves the province and then returns as a resident at a later date (information from Fred Toll).
3) Place of Residence
    Every individual eligible for coverage by Manitoba Health is assigned a municipal code (MUNCODE), in order to capture their place of residence. The following is an explanation of steps taken by Manitoba Health to assign and verify addresses prior to sending MUNCODEs and Postal Codes to MCHP. (This information comes from Manitoba Health as of November 2001.)

    Assignment of MUNCODEs:

    Municipal codes are assigned to each individual at the time of registration or address update to capture their place of residence.

    • Non-First Nation residents are automatically assigned a municipal code from the Postal Code Master Table based on their self-reported mailing address. The operator can override the auto-assigned MUNCODE if the mailing address and the residential address of the resident are not the same.
    • First Nation residents are assigned an A-code of the band they are registered to, which is used as their MUNCODE, regardless of where they live.

    Postal Code Validation:

    • Manitoba Health maintains an historical Postal Code Master Table of all postal codes ever used in Manitoba.
    • If the registration system encounters a postal code that does not pass validation (i.e.: is not found in the Postal Code Master Table) the operator will go through a process to validate the code with Canada Post and request that it be appended to the Postal Code Master Table.
    • If a registrant provides a postal code but has a municipality not previously associated with that postal code, again the operator will request that the municipal code be added to that postal code record as an alternate municipal code. There is no advance validation of postal codes performed by Manitoba Health through Canada Post.
    • New or inactive postal codes are dealt with as they are encountered.
    • Every two months, Information Services runs a job to validate all postal codes on the registration file against the Postal Code Master Table and generates a report with registrations that have an inactive postal code. Those maintaining the Manitoba Health Insurance Registry will use the report in attempt to update the postal code information in the Registry.

    Where possible, the information is validated, but it is virtually impossible to audit each and every registration for valid information. Change of address is generally accomplished over the telephone and again the onus is on the resident to provide a correct postal code and, if applicable, the municipality in which their residence is located.

    Manitoba Health checks self-reported addresses by various methods and sources of information including:

    • verifying the services provided to registrants if unusual activity if detected;
    • using other sources of information, such as phone books, property rolls, Manitoba Public Insurance, physician records and hospital discharge abstracts, to specify whether an individual is within Manitoba. An individual who is not found is removed from the registry and given the label "Cannot Locate". (It may take 2 years or more to declare "Cannot Locate");
    • verifying individual cases with the Manitoba Immunization Monitoring System (MIMS) which includes all Manitoba residents;
    • using Drug Program Information Network (DPIN) deductible applications (which are automatically verified with Canada Revenue Agency);
    • Long Term Care program; and
    • Office of the Public Trustee

    Please see the Postal Codes and Coding Methods in Manitoba Health Claims Data concept for more information on postal codes.
Updates to the Manitoba Health Insurance Registry
    In addition to self-reported updates from registrants, Manitoba Health uses other methods to update information in the Manitoba Health Insurance Registry (the following is based on information received from Manitoba Health in November 2001):

    1. Vital Statistics (VS)

      Checks and updates to the Registry are made through a regular data exchange with Manitoba Vital Statistics. For births and deaths, processes are setup to match the registration number on the VS record to that on the Manitoba Health registration file. The VS records must have a valid registration number and pass rules of eligibility before the registration file is updated.

    2. Municipal Audits

      In the past, Manitoba Health requested that municipalities help confirm their residents by manually comparing a Manitoba Health Registration report to the municipal tax roll. The reports included name, address and birthdate. However this process was discontinued due to restrictions on the release of personal health information as governed by the Personal Health Information Act (PHIA).

    3. Age of Majority (18 Years)

      When a registrant turns 18 years of age, they are assigned their own registration number with the same address, municipal code and postal code as that of the registration they were previously under. This process occurs once a month for all registrants who turned 18 in the previous month. It then becomes the responsibility of the person receiving the new registration card to inform Manitoba Health if the address information is incorrect. The family head of the original registration number is sent a revised registration card as well.

    4. Population Reports: RHA Population Report Data

      Upon the implementation of the Regional Health Authorities (RHAs), the need to situate residents geographically became imperative, particularly for First Nation residents. It was decided that because a significant number of First Nation residents are living "off reserve", assignment of municipal code should be based on the postal code. A process was created to assign MunOfPostal and create RHA.RETPOPDT (June 1 and December 1 of each year) which is the source file for all RHA population reports. The job reads the "old" RETPOPDT that contains the municipal codes assigned through Registration, selects only First Nation residents (A-code municipalities), and matches the postal code field to that on the On Reserve Postal Code table. If the postal code does not match that assigned to the registered A-code, the MunOfPostal becomes the Primary Municipal Code assigned to that postal code record on the Postal Code Master file. This would give the registrant an "off reserve" municipality and they would be counted in the general population of that municipality. Note that this is ONLY done for Treaty Status Indians living in a STRICTLY URBAN LOCATION. Those in rural or mixed urban/rural areas are still counted as living on their home reserve, whether they live there or not.
      Example 1 - A person from Birdtail Sioux First Nation (AOI;ROMOBO) moves to Fairford First Nation (A23;ROCOJO) and retains their original Manitoba Health registration number, their new municipal code will become 181, Gypsumville.

      If a resident registered to one band moves to another band, their registration will receive the "off reserve" municipality associated with their new postal code.

      Exception: If they change their Manitoba Health registration number to that of someone registered to the band to which they are moving, they are assigned the A-code of their new band.

      Example 2 - If a person from Birdtail Sioux First Nation (AOI;ROMOBO) moves to Fairford First Nation (A23;ROCOJO) and informs Manitoba Health that they have married someone registered to Fairford First Nation and wish to be put under his/her registration number, the new MunOfPostal will become A23.
      Although the assignment of MunOfPostal is solely dependent upon the A-code and postal code it can be indirectly influenced by the registration number. The system dictates that all people under one registration must have the same mailing address, municipal code and postal code. Therefore, if a non-First Nation changes their registration number to that of a First Nation person living on reserve, the non-First Nation is counted in the band population. The reverse is also true.
Impact of the Manitoba Health Insurance Registration System on Other Databases
    The Manitoba Health Insurance Registration file has a direct influence on the hospital discharge abstract and medical services/physician claims systems. When Manitoba Health receives an abstract or claim, it goes through an eligibility check against the registration system. When the patient passes the eligibility process, the record is added to the respective systems current-year file (hospital statistical claim file and the bi-weekly medical payment file). If one is comparing geographical utilization between hospital and medical there may arise some disparities in the postal code/municipal code combinations. Medical claims are generally processed quickly in comparison to hospital discharge abstracts. It may occur that a medical claim for a patient is processed within 2 weeks of having the medical service. Meaning, the claim was run through eligibility and added to the file with the postal code/municipal code information from the registration system at that point in time. If that same patient had a hospital encounter around the same time as the medical service, that hospital discharge abstract may not be received at Manitoba Health for up to 12 months. At that time, the abstract is run through eligibility against a registration file that has gone through several months of updated resident information. The medical claim may have patient information from before a change-of-address, and the same patient may have a different address a few months later. Why would we be getting a different set of postal code/municipal code combinations between the registration file and the hospital discharge abstracts and medical services/physician claims files ?

    One possible reason is that the registration database undergoes daily updates and changes. For example, if a new resident has a hospital stay before they register it could take up to 3 months before that abstract is run against the registry. If a population count file was run before the person registered (1) , the postal code/municipal code will not include data for that particular resident. They may then register, their hospital discharge abstract will be processed after their registration is accepted (sometimes 2 to 3 months after), and the resulting hospital discharge abstract will have the resident's postal code/municipal code combination.
    Note: (1) - The population run would include residents who were effectively covered 6 months before that run (December population is run the following June and the June population is run the following December).

Physician "Dummy" Claims Data
    Starting 1989/99, two sets of "dummy" cases (a unique combination of PHIN, REGNO, SEX, and BIRTHDATE) have been introduced to the Manitoba Health Insurance Registry. These fictitious cases allow physicians to bill for time on-call (especially anaesthetists) through the medical services/physician claims system (AKA the Medical Statistical File) without triggering Manitoba Health's manual patient identification checks. In this case, there are no real patients to which these on-call allowances can be attributed.

    These two fictitious registrations have no meaning at MCHP and should be dropped from all relevant population, cohort selection and follow-up projects involving the MCHP Research Registry (as of Jan 2002). See the Physician "Dummy" Claims Data documentation for more information (internal access only).

    (Thanks to Deborah Malazdrewicz at Manitoba Health)
The MCHP Research Registry
    The MCHP Research Registry is a key resource in the MCHP Population Health Data Repository and contains data as far back as 1970. The main source of information for the MCHP Research Registry comes from the Manitoba Health Insurance Registry files. At regular intervals (twice yearly), Manitoba Health supplies "snapshots" of these files to MCHP, allowing the Research Registry to be kept up-to-date. As of January 2011, 65 separate archival "snapshots" have been processed. The files include the encrypted PHIN and REGNO of each individual along with their demographic, residential and health insurance coverage information.

    The MCHP Research Registry introduces extensions of the Manitoba Health Insurance Registry data that are completed by additional programming at MCHP. MCHP analyzes the data in the Manitoba Health Insurance Registry and other sources to provide (information from personal communication with Pat Nicol, July 25, 2011):

    • extension of scrambled PHIN to 1970-1983;
    • marital status by individual by date (1970 onwards);
    • family structures (spouse, siblings, etc.);
    • mortality verification and cause of death (from Vital Statistics data);
    • longitudinal continuity of PHIN coverage;
    • ability to track registrants lost to follow-up; and
    • residential mobility within Manitoba and in/out of province.
    Information regarding births, obstetrical stays and deaths can be retrieved from hospitalization data. Death reports can also be found in the Long Term Care database. Deaths noted in the hospitalization data are resolved with those reported in the Manitoba Vital Statistics Mortality Database (separate annual files) using probabilistic record linkage techniques. All details regarding marital status and family composition are derived from these snapshots and incorporated into the Research Registry at the individual-level.

    The MCHP Research Registry is population-based and allows the ability to identify individuals who have no contact with the health care system and their reason for lack of contact. Three common reasons are:

    • "Well" individuals living in the province who simply have no interaction with the health care system
    • Loss to follow-up (see Follow-Up Issues concept for more information) - this includes individuals who become ineligible for health care coverage, which occurs when people leave the province, join the RCMP, etc.
    • Death of an individual.

    The MCHP Research Registry requires a significant continuing investment, but provides a tremendous overall saving when the information is used in many studies. A massive programming effort maintained over many years has joined the "snapshot" files from Manitoba Health together such that individual histories can be constructed over the entire period of the database. In addition, software has been developed to facilitate longitudinal follow-up, residential mobility, migration, family structure, and mortality. This effort results in the creation of the MCHP longitudinal population research registry.
De-identified Data
    De-identified individual level data is "information about an individual that has been modified or from which identifying or potentially identifying information has been removed in a way that minimizes the likelihood that an individual's identity can be determined by any reasonably foreseeable method. Methods of de-identifying information can include scrambling or encrypting identifying or potentially identifying information" (from section 1.01 (d) within An Agreement Respecting Access to Manitoba Health Information at the Manitoba Centre for Health Policy (University of Manitoba) for Research Being Conducted by University Researchers Within The Secure Data Environment of MCHP - http://umanitoba.ca/admin/vp_admin/ofp/legal/media/MCHP_UofM_Researchers_2010.doc - accessed December 8, 2010).

    • PHINs are de-identified data fields in the MCHP Research Registry. They are provided to MCHP from Manitoba Health in an encrypted form. The encrypted PHIN allows MCHP to define individual-level associations over time.
    • When a PHIN cannot be found for an individual, MCHP uses REGNO-SEX-DOB-INITIAL and patterns of coverage to develop a unique MCHP-specific PHIN (DOB = date of birth; INITIAL = initial of first given name). These MCHP generated PHINs are distinguishable from the encrypted Manitoba Health PHINs.
    • MCHP receives only MUNCODEs and Postal Codes from Manitoba Health. This data allows for a variety of geographical analyses (e.g.: by RHA or community areas).
    • MCHP does not receive names, street addresses or other possibly identifying information in the data received from Manitoba Health.
Common Uses of the MCHP Research Registry
    In Roos et al. (1999) , they describe examples of the type of work that can be done using the MCHP Research Registry. These include:

    • track the movement of Manitoba residents within the province, and in and out of the province;
    • specify family types and family characteristics;
    • define cohorts by parent-child or sibling (including twins and multiple births) relationships;
    • compare data with those generated by disease-specific registries; and
    • providing date-specific values for insurance coverage, marital status, residence and other characteristics to databases that do not report them.

    Please see Table 1. The research registry: common tasks in Roos LL et al. (1999) for a more detailed description of applications of the research registry.

    One of the most important uses of the MCHP Research Registry is that it can be "linked" to other administrative databases available in the MCHP Population Health Research Data Repository, thus providing a wealth of comprehensive information that would otherwise be unavailable. The "linkage" can be accomplished because the encrypted Personal Health Identification Numbers (PHINs) in the MCHP Research Registry are available in other databases, such as hospital separations and physician claims, allowing for "linkage" between separate data sources. In this manner, a comprehensive history for an individual can be compiled.

    Figure 1 - Manitoba Centre for Health Policy Research Repository illustrates the types of data sources that can be linked with the MCHP Research Registry.
Quality of the MCHP Research Registry
    The quality and thoroughness of the MCHP Research Registry is enhanced by information from additional sources. For example, increased accuracy regarding date and cause of death is achieved through the inclusion of reports of date of death in Manitoba facilities and from Vital Statistics Agency (Manitoba) mortality data. Integration of hospitalization data increases the thoroughness of the MCHP Research Registry since it reports births, deaths, and obstetrical stays (e.g. verification of important mother-child relationships, siblings, etc.).

    • Canadian Census versus MCHP Research Registry

      Multiple checks between the 1986, 1991, and 1996 Canadian Census and the MCHP Research registry have shown the following:
      1) Differences between registries and census diminished overtime. Overall, registry numbers always exceeded the census figures.
      2) Differences were greatest for the age group from 15 - 24 (for both genders).

      The table, Correspondence between registry and census data provides the actual numbers from this comparison. Further information regarding systems for updating and reviewing the quality of registry data can be found in A research registry: uses, development, and accuracy , Roos and Nicol (1999) .
    • MCHP Research Registry vs. Manitoba Health Statistical Bulletins

      The MCHP Research Registry has been compared to Manitoba Health Statistical Bulletins to evaluate any differences in the reporting of live births by Manitoba Health and MCHP. It was found that there is between -2.8% and +3.8% difference between these two sources. For more information, please see the Live Births and Comparison of Data Sources concept.

    • Comparison of Multiple Data Sources

      Multiple data sources have been used to develop population estimates for 5 year intervals from 1986 to 2006. The numbers come from Manitoba Health Statistical Reports, Manitoba Health Snapshots, MCHP Research Registry and the Canadian Census. In this comparison, the MCHP Research Registry numbers are closer to the Census numbers than the Manitoba Health "snapshot" numbers. Please see the Population Estimates and Comparison of Data Sources concept for the actual numbers used in the comparison of multiple data sources.
Confidentiality of Personal Health Information with the MCHP Research Registry
    MCHP has taken several security and research-related steps to ensure that the MCHP Research Registry can be used without jeopardizing the confidentiality of any individual's health information:

    • Prior to sending the Manitoba Health Insurance Registry and other health related data sets to MCHP, Manitoba Health encrypts the Personal Health Identification Numbers (PHINs) contained in the data to ensure that individuals cannot be identified by their real number.
    • Only the postal code and municipal code sections of the address are sent to MCHP. These codes enable MCHP to define a variety of geographical regions and communities for research purposes.
    • Identifying information, such as patient name and detailed address, is removed by Manitoba Health prior to the records being sent to MCHP to maintain privacy.
    • Physical access to the Registry is tightly controlled and monitored. Only projects that have passed several screens and reviews (ethical, peer, and the Manitoba Health Access and Confidentiality committee) are allowed to access the data.
    • Any research using Registry data must be credible and must contribute to the expansion of knowledge for the public good.
    • MCHP has implemented several internal system and procedural controls to ensure that confidentiality is always maintained; these are actively monitored.

    For more information on privacy, please read our MCHP Pledge of Privacy: http://www.umanitoba.ca/faculties/medicine/units/community_health_sciences/departmental_units/mchp/privacy.html .
SPDS SAS Code - Registry Examples
    Two recent examples of SAS code providing details on coding with the Registry files in SPDS are provided in the section below. These SAS code examples relate to a general introduction to the Registry file in SPDS and how to identify changes in residence using the Registry file (internal access only).

Related concepts 

Related terms 

Links 

References 

  • Kozyrskyj A, Brown TER, Mustard CA. Community pharmacist perceptions of a provincial drug utilization database. Can Pharm J 1998;131(8):24-29.(View)
  • Kozyrskyj AL, Mustard CA. Validation of an electronic, population-based prescription database. Ann Pharmacother 1998;32(11):1152-1157. [Abstract] (View)
  • Roos LL, Nicol JP. A research registry: uses, development, and accuracy. J Clin Epidemiol 1999;52(1):39-47. [Abstract] (View)
  • Roos LL, Mustard CA, Nicol JP, McLerran DF, Malenka DJ, Young TK, Cohen MM. Registries and administrative data: organization and accuracy. Med Care 1993;31(3):201-212. [Abstract] (View)
  • Roos NP, Roos LL, Brownell M, Fuller EL. Enhancing policymakers' understanding of disparities: Relevant data from an information-rich environment. Milbank Quarterly 2010;88(3):382-403. [Abstract] (View)
  • Roos NP, Black C, Roos LL, Frohlich N, De Coster C, Mustard C, Brownell MD, Shanahan M, Fergusson P, Toll F, Carriere KC, Burchill C, Fransoo R, MacWilliam L, Bogdanovic B, Friesen D. Managing health services: how the Population Health Information System (POPULIS) works for policymakers. Med Care 1999;37(6 Suppl):JS27-JS41. [Abstract] (View)
  • Roos NP, Shapiro E. Revisiting the Manitoba Centre for Health Policy and Evaluation and its population-based health information system. Med Care 1999;37(6 Suppl):JS10-JS14. [Abstract] (View)

Keywords 

  • methods


The best way to keep up to date with what's happening at MCHP is to subscribe to our news list. We'll periodically send you notice of recently released reports, events, and other material from MCHP. If you're interested, please enter your email address in the space below. You can update your information or unsubscribe at any time.

 

* Email
* First Name
* Last Name
* Business
* = Required Field
This email list is privately maintained for MCHP by iContact.