Concept: Multiple Sclerosis (MS) - Defining in Administrative Data

Introduction

Multiple sclerosis (MS) is an immune-mediated inflammatory and neuro-degenerative disease of the central nervous system (brain, optic nerves, and spinal cord). Its effects are diverse and may include problems in balance, vision, communication, memory, and movement, as well as more general symptoms such as pain and fatigue. The course of disease varies between people. Disease modifying therapies are available which reduce relapses and disability progression but there is no cure. Rehabilitation and pharmacologic therapies can help to manage symptoms. ( Reich et al., 2018 )

This concept identifies the data sources used to define MS, describes a general method for defining MS, and then identifies several research projects and describes the specific method used to define MS in each project. A list of references and links to the research articles are also provided.

Data Sources for Defining MS

The following data sources, along with the type of variables, have been used in defining MS from the administrative data in the Manitoba Population Research Data Repository, and other similar Repositories across Canada:


• Hospital Abstracts data using International Classification of Disease (ICD) diagnosis codes;
  
• Medical Services/ Medical Claims data using ICD diagnosis codes; and
  
• Drug Program Information Network (DPIN) data (Manitoba's comprehensive prescription drug system) using specific drug identification numbers (DIN) or anatomical therapeutic classification (ATC) numbers.

General Method for Defining MS

Over time, different methods have been used to define MS in different research projects. These methods can vary based on the following considerations:


• the time frame of the research (e.g. different ICD coding systems being used);
  
• the role of demyelinating conditions in identifying a date of disease onset for determining a diagnosis date / incidence;
  
• the use of MS-related prescription claims, if the data are available;
  
• the time frame within which the number of diagnoses codes / prescription claims must appear in the records (e.g. within three years); and
  
• the required number of hospital visits and/or physician visits with an MS diagnosis code, or drug prescription claims (if this data is available and used in the definition).

Marrie RA, Yu N, et al. (2010), investigated different combinations of hospital and physician claims data with an MS diagnosis over different time frames, and validated the algorithm by reviewing 2,000 self-reported questionnaires, and a review of 400 randomly selected medical charts from the 2,000 who completed the questionnaire. The following describes two general methods to identify MS and date of diagnosis, with and without prescription drug claims.

The following International Classification of Disease (ICD) diagnosis codes are used to identify a diagnosis of MS in both the hospital abstracts data and the physician services / physician claims data over time:

Disease	ICD-9 code	ICD-10-CA code
Multiple Sclerosis	340	G35

Patients may present with symptoms suggesting the disease before a confirmed MS diagnosis. Thus, for all patients meeting the case definition of MS, the date of diagnosis may be identified as the earlier of:


• the first date of medical contact with a diagnosis for MS (ICD-9: 340 or ICD-10: G35), or
  
• the first date of diagnosis for a demyelinating condition:
  
  

  Demyelinating Condition	Physician Claims ICD-9 Codes	Hospital Claims ICD-9 Codes	Hospital Claims ICD-10 Codes
  Optic neuritis	377	377.3	H46
  Acute transverse myelitis	323, 341	323.82, 341.2	G37
  Acute disseminated encephalomyelitis	323	323	G36.9
  Demyelinating disease of CNS unspecified	341	341.9	G37.8
  Other acute disseminated demyelination	NA	NA	G36
  Neuromyelitis optica	341	341.0	G36.0

  NA = code not available in ICD-9

The identification of a demyelinating condition allows for a more accurate date of incidence based on the disease onset rather than a confirmed MS diagnosis, which may be delayed.

General Algorithm to Identify MS without Prescription Claims:


• Individuals with 3 or more MS specific records in physician and/or hospital data (ICD-9 code 340 or ICD-10 code G35).
  
• 3 in any combination of hospital/physician claims at any time (e.g. 3 hospital, 1 hospital + 2 physician, 3 physician, etc.).
  
• Claims must be on separate days (e.g. if > one MS-specific claim on any given day, this can only count as one claim towards the criteria. Hospitalizations must not overlap, and physician visits must not overlap with hospitalizations to avoid double counting.
  
• Use of the primary ICD code in the physician claims database.
  
• Use all discharge diagnoses in the hospital claims database.

General Algorithm to Identify MS with Prescription Claims:


• Individuals with 3 or more MS specific records in physician claims and/or hospital claims data, and/or prescription claims using the list of MS-specific medications in the following table:
  
  NOTE: Medication lists should be reviewed and updated as new MS-specific medications are approved for use or discontinued.
  
  

  Drug Identification Number (DIN)	Product Description
  02169649	Betaseron
  02337819	Extavia
  02444399	Plegridy
  02444380	Plegridy
  02444372	Plegridy
  02444402	Plegridy
  02237770	Avonex
  02269201	Avonex
  02281708	Rebif
  02277492	Rebif
  02237317	Rebif
  02237319	Rebif
  02237320	Rebif
  02318253	Rebif
  02318261	Rebif
  02318288	Rebif
  02233014	Copaxone
  02245619	Copaxone
  02441446	Teva-Glatiramer Acetate
  02456915	Copaxone
  02460661	Glatect
  02286386	Tysabri
  02365480	Gilenya
  02404508	Tecfidera
  02420201	Tecfidera
  02416328	Aubagio
  02418320	Lemtrada
  02511355	Kesimpta
  02467224	Ocrevus
  02470179	Mavenclad

MS Research Projects

The following is a list of MS research projects that have defined MS, along with any significant considerations in the MS definition.

1. Health Inequities in Manitoba: Is the Socioeconomic Gap in Health Widening or Narrowing Over Time? Martens P et al., 2010.

MCHP identified cases of MS meeting the following criteria: at least three hospital visits, three physician visits, or a combination of these, where each visit included a diagnosis code for MS (ICD-9-CM code 340 or ICD-10-CA code G35) within a period of six years.

2. Differences in the burden of psychiatric comorbidity in MS vs the general population. Marrie RA, Fisk JD, Tremlett H, et al., 2015.

MS cases were identified as those with >= 3 hospital or physician claims for MS (ICD-9/10 codes 340/G39) using an administrative case definition validated in Manitoba and Nova Scotia. ( Marrie RA, Yu N, et al., 2010 ; Marrie RA, Fisk JD, Stadnyk KJ, et al., 2013 ).

For each person with MS, they assigned the date of their first demyelinating disease claim as the date of diagnosis.

3. Rising incidence of psychiatric disorders before diagnosis of immune-mediated inflammatory disease. Marrie RA, Walld R, et al., 2017.

They applied a validated case definition ( Marrie RA, Yu N, et al., 2010 ) to the administrative dataset to identify all Manitobans aged 18 years and older with MS from 1984 to 2013. They assigned the date of diagnosis as the date of the first health claim for MS.

MS was identified using the ICD-9 / ICD-10 codes: 340 / G35, and the following ATC codes to identify MS prescriptions:

ATC Code	Product Description
L03AX13	Copaxone, Teva-Glatiramer Acetate, Glatect
L03AB07	Rebif, Avonex
L03AB08	Betaseron, Extavia
N07XX09	Tecfidera
L04AA23	Tysabri
L04AA27	Gilenya, *-Fingolimod
L04AA31	Aubagio, *-Teriflunomide
L04AA34	Mabcampath, Lemtrada

*- denotes multiple products ending in <drug name>

4. Establishing the Incidence and Prevalence of Multiple Sclerosis in Saskatchewan. Al-Sakran LH et al., 2018.

Two MS case definitions were investigated in this research. The first definition required >= 3 hospital, physician, or prescription claims. The second definition, released by the Canadian Chronic Disease Surveillance System (CCDSS), required >= 1 hospitalization or >= 5 physician claims within 2 years. Hospital transfers and re-admissions within 1 day of a discharge date were considered as one hospitalization episode and collapsed into a single hospital claim.

The definitions were validated using a reference standard of 200 patients with clinically definite MS from the provincial MS Clinic in Saskatchewan and 200 patients without an MS diagnosis from the Inpatient Rehabilitation Center database.

As a complementary analysis, they also tested several other case definitions to allow for comparability with other Canadian studies or for situations where a potentially more sensitive or more specific definition may be warranted.

They used a 5-year run-in period with no other claims for MS or demyelinating conditions to ensure that the identified cases were incident. However, because the Canadian Chronic Disease Surveillance System (CCDSS) recommends an 8-year run-in period to detect incidence in MS, a sensitivity analysis using an 8-year run-in period was also conducted to evaluate the effect of different run-in periods on incidence estimates. The first definition resulted in 2,226 incident cases of MS and 1,903 cases were identified with the CCDSS definition.

5. Incidence and prevalence of MS in children. A population-based study in Ontario, Canada. Marrie RA, O’Mahony J et al., 2018.

They developed several candidate administrative case definitions by varying the number of hospital and physician claims required, and time period over which these claims were accumulated to be classified as having MS.

To avoid double-counting hospitalizations related to transfers between facilities for continued care, overnight hospitalizations beginning within + or - 1 day of another hospital discharge with the same primary or secondary diagnosis codes were considered part of the same hospitalization.

Multiple physician claims on the same day with a diagnosis of MS were counted as a single claim. They excluded physician claims that overlapped with hospitalizations to avoid double counting.

They specifically investigated 2 case definitions that were previously validated and applied in adult populations in Canada. The first, used by the Canadian Chronic Disease Surveillance System (CCDSS) and validated in Ontario, required >= 1 hospitalization or >= 5 physician claims within 2 years. The second case definition required >= 3 hospital or physician claims using all available years of data, and hereinafter is referred to as the Marrie definition. Data was validated using a reference cohort of 124 children evaluated for possible demyelinating disease.

Incident cases were those who met the case definition for MS, had an index date in the fiscal year of interest, and had at least a 5-year period with no demyelinating disease claims preceding the index date.

The CCDSS definition had a sensitivity of 81.1%, specificity of 100%, PPV of 100%, and NPV of 86%. The Marrie definition had a sensitivity of 89.2%, specificity of 100%, PPV of 100%, and NPV of 91.5%. For more information on the results of this study, please see the abstract link in the References section below.

6. Lower prevalence of multiple sclerosis in First Nations Canadians. Marrie RA, Leung S, et al., 2018.

They compared the incidence and prevalence of multiple sclerosis (MS) between First Nations (FN) and non-FN populations in Manitoba.

The first case definition for MS was designed to be sensitive and required >= 3 claims ever, where eligible claims had an ICD-9/ ICD-10 code for MS (340/G35) or included a prescription for an MS-specific disease-modifying therapy. This was the primary case definition for the study (referred to as definition 1). A more specific second definition required >= 3 claims for individuals within <= 3 years of data and >= 7 hospital or physician claims for individuals within > 3 years of data.

They conducted complementary analyses using alternative case definitions to allow comparisons of the findings with those of other definitions, such as the CCDSS definition, which requires >= 1 hospital or >= 5 physician claims for MS in 2 years.

Incidence and prevalence

After identifying all persons with MS who met the case definition, they identified the first health claim for MS/demyelinating disease and assigned the claim date as the date of diagnosis.

MS / Demyelinating Disease	ICD-9 code	ICD-10 code
Multiple Sclerosis (MS)	340	G35
Other acute disseminated demyelination	NA	G36
Demyelinating disease of CNS unspecified	341.9	G37.8
Acute disseminated encephalomyelitis	323	G36.9
Optic neuritis	377.3	H46
Acute transverse myelitis	323.82	G37


NA = code not available in ICD-9

Individuals who met the case definition were classified as prevalent cases from the date of the first demyelinating disease claim until the date of death or emigration from Manitoba. To identify a case as incident, they required at least 5 years of administrative data without any claims for demyelinating disease preceding the first demyelinating disease claim.

7. Cancer Incidence and Mortality Rates in Multiple Sclerosis: A Matched Cohort Study. Marrie RA, Maxwell C, et al., 2021.

The objective of this study was to determine whether cancer risk differs in people with and without multiple sclerosis (MS). They compared incidence rates and cancer-specific (breast, colorectal, and 13 other cancers) mortality rates in MS and matched cohorts using population-based data sources.

The validated administrative MS case definition required >= 3 hospital or physician claims for MS based on ICD-9 or ICD-10-CA diagnostic codes (340/G35) using all available years of data. For each person with MS, the earliest ICD-9 or ICD-10 claim for demyelinating disease was designated as the index date.

8. Incidence and prevalence of paediatric-onset multiple sclerosis in two Canadian provinces: a population-based study representing over half of Canada’s population. Yusuf FLA et al., 2023.

They identified individuals with PoMS (Paediatric-onset Multiple Sclerosis) using a previously validated algorithm ( Marrie RA, O’Mahony J, et al., 2018 ) which cases were required to have >= 3 MS-specific hospital or physician claims on separate days, with the earliest demyelinating/MS-specific claim (‘MS onset’) occurring <= age 18 years. The date of the first demyelinating / MS specific claim indicated the index date.

Multiple sclerosis-specific and demyelinating disease related International Classification of Diseases-9/10 (ICD-9/10) from hospital abstracts and physician claims codes used to identify pediatric-onset MS cases and their respective index date (i.e., first recorded demyelinating disease code).

MS / Demyelinating Disease	ICD-9 Code	ICD-10 Code
Multiple Sclerosis (MS)	340	G35
Optic neuritis	377.3	H46
Acute transverse myelitis	323.82, 341.2	G37.3
Acute disseminated encephalomyelitis	323	G36.9
Demyelinating disease of CNS unspecified	341.9	G37.8
Other acute disseminated demyelination	NA	G36
Neuromyelitis optica	341.0	G36.0


NA = code not available in ICD-9

Cautions / Limitations

The following cautions / limitations were mentioned in different studies:

• A complementary analysis revealed that, in general, sensitivity was higher for those definitions requiring fewer claims and those with a longer, or unlimited timeframe.
  
• Individuals who do not have frequent contact with the health care system may be missed, resulting in an underestimation of the disease.
  
• Underestimation may also have been augmented due to billing practices; although many physicians who receive alternate payments (i.e. non-fee-for-service) submit shadow (dummy) claims which are captured in health administrative data, the physicians at the provincial MS clinic did not shadow bill during the study period. Therefore, not all encounters for MS may have been captured.
  
• Comprehensive prescription drug use is not captured across all provinces, therefore it may not be possible to use this data source equitably as part of the case definition in multi-provincial studies across Canada.
  
• If prescriptions are to be used, the list of medications should be reviewed to include the most recent medications and note any changes in previous medications identified to ensure they are still relevant.