ARTICLE OPEN ACCESS

The prevalence of MS in the United StatesA population-based estimate using health claims data

Mitchell T Wallin MD MPH William J Culpepper PhD Jonathan D Campbell PhD Lorene M Nelson PhD

Annette Langer-Gould MD PhD Ruth Ann Marrie MD PhD Gary R Cutter PhD Wendy E Kaye PhD

Laurie Wagner MPH Helen Tremlett PhD Stephen L Buka ScD Piyameth Dilokthornsakul PharmD PhD

Barbara Topol MS Lie H Chen DrPH and Nicholas G LaRocca PhD on behalf of the US Multiple Sclerosis

Prevalence Workgroup

Neurologyreg 201992e1029-e1040 doi101212WNL0000000000007035

Correspondence

Dr Wallin

mitchellwallinvagov

AbstractObjectiveTo generate a national multiple sclerosis (MS) prevalence estimate for the United States byapplying a validated algorithm to multiple administrative health claims (AHC) datasets

MethodsA validated algorithm was applied to private military and public AHC datasets to identify adultcases of MS between 2008 and 2010 In each dataset we determined the 3-year cumulativeprevalence overall and stratified by age sex and census region We applied insurance-specificand stratum-specific estimates to the 2010 US Census data and pooled the findings to calculatethe 2010 prevalence of MS in the United States cumulated over 3 years We also estimated the2010 prevalence cumulated over 10 years using 2models and extrapolated our estimate to 2017

ResultsThe estimated 2010 prevalence of MS in the US adult population cumulated over 10 years was3092 per 100000 (95 confidence interval [CI] 3081ndash3101) representing 727344 casesDuring the same time period the MS prevalence was 4501 per 100000 (95 CI 4481ndash4516)for women and 1597 (95 CI 1587ndash1606) for men (femalemale ratio 28) The estimated2010 prevalence of MS was highest in the 55- to 64-year age group A US north-southdecreasing prevalence gradient was identified The estimated MS prevalence is also presentedfor 2017

ConclusionThe estimated US national MS prevalence for 2010 is the highest reported to date and providesevidence that the north-south gradient persists Our rigorous algorithm-based approach toestimating prevalence is efficient and has the potential to be used for other chronic neurologicconditions

RELATED ARTICLES

ArticleValidation of an algorithmfor identifying MS cases inadministrative healthclaims datasets

Page e1016

Views and ReviewsA new way to estimateneurologic diseaseprevalence in the UnitedStates Illustrated with MS

Page 469

MORE ONLINE

PodcastDr Stacey Clardy talks withDr Mitchell Wallin abouthis paper on the prevalenceof MS in the United States

NPuborgbsqqt1

From the Department of Veterans Affairs Multiple Sclerosis Center of Excellence (MTW WJC) Georgetown University School of Medicine (MTW) Washington DC University ofMaryland (WJC) Baltimore University of Colorado (JDC PD) Aurora Stanford University School of Medicine (LMN BT) CA Southern California Permanente Medical Group(AL-G LHC) Pasadena Departments of Internal Medicine and Community Health Sciences (RAM) Max Rady College of Medicine Rady Faculty of Health Sciences University ofManitoba Winnipeg Canada University of Alabama at Birmingham (GRC) McKing Consulting Corp (WEK LW) Atlanta GA Faculty of Medicine (Neurology) and Centre for BrainHealth (HT) University of British Columbia Vancouver Canada Brown University (SLB) Providence RI and National Multiple Sclerosis Society (NGL) New York NY

Go to NeurologyorgN for full disclosures Funding information and disclosures deemed relevant by the authors if any are provided at the end of the article

United States Multiple Sclerosis Prevalence Workgroup (MSPWG) coinvestigators are listed in appendix 2 at the end of the article

The Article Processing Charge was funded by the National Multiple Sclerosis Society

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 40 (CC BY-NC-ND) which permits downloadingand sharing the work provided it is properly cited The work cannot be changed in any way or used commercially without permission from the journal

Copyright copy 2019 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology e1029

Chronic disease morbidity is challenging to assess within theUnited States because it lacks a unified health system andlimited infrastructure exists for identifying and trackingpatients across their lifespan Options for determining in-cidence and prevalence estimates include surveys registriesor administrative health claims (AHC) datasets Each methodhas strengths and limitations however the increasing avail-ability of large AHC datasets has made this approach efficientand cost-effective1

Multiple sclerosis (MS) is the most common progressiveneurologic disease of young adults worldwide23 Currentestimates suggest that 300000 to 400000 individuals are af-fected in the United States but this is based largely on revi-sions of estimates from older data2ndash6 These estimates do notreflect the changing demographics of the United States orpotential changes in the ascertainment of MS due to mod-ifications in the diagnostic criteria and new treatment optionsMoreover studies in neighboring Canada have reported steepincreases in the prevalence of MS over the past few decadesacross several provinces7ndash9

Because of the challenges in estimating MS prevalence for theUnited States the National Multiple Sclerosis Society(NMSS) formed the Multiple Sclerosis Prevalence Work-group made up of scientists and policy advocates with thegoal of producing a scientifically sound and economicallyfeasible national MS prevalence estimate By applying a vali-dated case algorithm for MS to multiple large AHC datasetswe aimed to generate a robust national MS prevalence esti-mate for the adult population stratified by age sex andregion

MethodsSetting and data sourcesThe United States includes 48 contiguous states Hawaii andAlaska The 48 contiguous states range from asymp2452deg N to4928deg N latitude and from asymp6695deg W to 12477deg W longi-tude The US population encompassing all 50 states is steadilygrowing having increased from 3093 million in 2010 to 3258million in 201710 In the United States health insurance maybe obtained from several private or public (government)sources and a proportion of the population is uninsured Weacquired several AHC datasets representing the US privateand government-sponsored insurance programs reasoningthat nearly all persons with MS except the uninsured NativeAmericans using the Indian Health Service and the in-carcerated would receive health services through one of theseprograms Each included the adult population (ge18 years)

and their health care use for the years 2008 to 2010 Thebreakdown of the population within specific health insuranceprograms varies by income sex disability and age group1

Private insuranceIn 2016 most adults lt65 years of age (73) obtained theirhealth care coverage from private insurance plans11 Becausethe AHC datasets available for the private insurance sectorvary in terms of geographic coverage and the types of pro-viders represented we accessed 3 datasets Optum (OP)Truven Health (TH) and Kaiser Permanente SouthernCalifornia (KPSC) These 3 private health datasets representa broad sample of all such plans in the US insurance marketTwo of these datasets (OP and TH for 2008ndash2010) were usedin the prevalence calculations and collectively capture asymp35of the privately insured in the United States

Public insuranceLow-income adults and those with particular disabilities mayobtain health care coverage through government-fundedMedicaid plans In 2016 96 of US adults ge65 years of agewere enrolled in government-funded Medicare11 In thepublic sector the Centers for Medicare amp Medicaid Servicesdatasets captured all eligible persons (100) enrolled inMedicare or Medicaid across the United States (gt50 millionindividuals)12 All active-duty military enrollees receive theirhealth care from the Department of Defense and based oncurrent eligibility criteria asymp30 to 40 transfer to theDepartment of Veterans Affairs (VA) health care andbenefit system when they leave active duty To assess themilitary sector we used the VA database which included allpersons enrolled in the VA health care system Collectivelythese datasets provided health care information for gt125million persons

The AHC datasets varied with respect to the informationcaptured Therefore we developed a common data dictionaryand variable list for this study These included a denominatorfile for all enrollees including dates of insurance eligibilitysex year of birth and geographic region of residence Becauserace and ethnicity were not available in all data sources theywere not considered in this analysis We also accessed data onhealth care encounters in the inpatient and outpatient set-tings as well as prescription drug claims Each inpatient andoutpatient encounter included ge1 diagnostic codes recordedwith the ICD-9 system as well as the date of the encounter Inthe ICD-9 system MS is uniquely identified by the 340 codeFor inpatient encounters we used the date of admissionPrescription drug claims included the name of the medicationand the date of release

GlossaryAHC = administrative health claims CI = confidence interval ICD-9 = International Classification of Disease-9th revisionKPSC = Kaiser Permanente Southern CaliforniaMEPS =Medical Expenditure Panel SurveyMS =multiple sclerosisNMSS =National Multiple Sclerosis Society OP = Optum TH = Truven Health VA = Department of Veterans Affairs

e1030 Neurology | Volume 92 Number 10 | March 5 2019 NeurologyorgN

Diagnostic algorithm for MSAs described elsewhere13 we developed and tested severalalgorithms to identify people with MS using AHC datasetscompared with physician-adjudicated MS cases as the refer-ence standard The optimal algorithm in terms of sensitivityspecificity and simplicity required the accumulation of ge3MS-related hospitalizations outpatient visits or pre-scription release encounters for an MS disease-modifyingtherapy in any combination within a 1-year period Forprescription drug claims we considered only disease-modifying therapies approved by the US Food and DrugAdministration for MS by 2010 including the interferonbetas glatiramer acetate natalizumab and fingolimod14

To avoid misclassification claims for natalizumab were notincluded if the individual also had an ICD-9 code for in-flammatory bowel disease another disorder for which thismedication is approved

When tested among individuals with at least 1 MS claimthe sensitivity of the MS algorithm was 86 to 92 and thespecificity was 66 to 83 depending on the dataset13

When tested in a Canadian population that included indi-viduals with and without any MS claims (ie generalpopulation) the sensitivity was 960 and the specificitywas 99513

Prevalence estimatesFor the TH OP VA and KPSC datasets enrollees who alsohad Medicare coverage were removed from both thenumerators and the denominators within each dataset toprevent double counting The annual prevalence withina given dataset was demarcated as all those who met the MScase definition divided by the annual population at risk de-fined as all enrollees ge18 years of age at the beginning of thecalendar year and with health plan eligibility for a total of 6months within the calendar year Because individuals with MSmay have variable contact with the health system once anenrollee met the case definition and remained eligible for carehe or she was considered a case thereafter Applying the al-gorithm to each dataset we determined the prevalence at theend of the 3-year study period by identifying all persons whomet the case definition in any 1 of the 3 study years who werestill alive and eligible for care in the last year of the studyperiod (2010) and dividing this by the population at risk in2010 These 3-year prevalence estimates were stratified by sexage (18ndash24 25ndash34 35ndash44 45ndash54 55ndash64 67ndash74 and ge75years) and US Census region (North East South and West)They were directly age and sex standardized to the 2010 USCensus10 Confidence intervals (CIs) were calculated for thefinal total number of cases using binomial CIs (plusmn196 timesradic(NPQ where P and Q are the proportions of cases andnoncases and N is the estimated US population in 2010) The95 CIs were then adjusted for the rate per 100000 and theinflation factors were calculated with a fixed-effects modelVerification of the prevalence estimates was performed foreach dataset by 2 independent reviewers (WEK and LW)for quality control

To obtain a national US prevalence estimate for MS we un-dertook several analytic steps First we treated estimates fromOP and TH as random samples drawn from the same un-derlying population (the commercially insured) so their age-and sex-stratified estimates were pooled with the use ofa random-effects model to represent the US private insurancepopulations Because KPSC was included in the generalpopulation denominator in the West region it was not in-cluded in these calculations Sensitivity analyses examiningthe effects of including or excluding KPSC from the Westregion were conducted and the findings did not differ sig-nificantly (data not shown) Second we used data from theUS Census to determine the total size of the US population ineach age and sex stratum and the proportion with privatepublic and military health insurance coverage15 MedicaidMedicare and military veteran prevalence estimates fullycaptured these populations1015 The stratum-specific estimatewas multiplied by the total insured US population in thatstratum to determine the number of individuals affected In2010 16 of the US population was uninsured but theproportion uninsured varies across conditions1116 To ac-count for the uninsured MS population we used data fromboth the Sonya Slifka cohort study17 and the NMSS whichreported a 50 uninsured rate within the MS populationbefore the initiation of the Affordable Care Act Thus thenumber of individuals affected in each stratum was summedinflated by 50 to account for the uninsured across all strataand then divided by the total US population to generatea summary prevalence estimate for the United States

The term cumulative prevalence applies to our case findingapproach within datasets in that once an individual meets theMS case definition for a given year that person is counted asa case for subsequent years through 2010 if he or she remainsactive in the health plan This method of case ascertainmenteffectively represents a limited-duration (3-year) prevalenceUltimately the prevalence estimate of interest is lifetimeprevalence which is the proportion of a population that atsome point in life (up to the time of assessment) has de-veloped MS In chronic predominantly relapsing diseasessuch as MS that start in early adult life individuals may forgocontact with the health system for extended periods Thuslong periods of observation (minimum 10 years) are neededto approach lifetime prevalence in the assessment of AHCdatasets as described previously for systemic lupus eryth-ematosus18 and as widely recognized in the cancer literature19

As noted elsewhere13 by using AHC datasets available fromIntercontinental Marketing Services the VA and the provinceof Manitoba over the period of 2000 to 2016 we determinedthe proportion of cases missed by using a 3-year vs 10-yearcumulative prevalence estimate On the basis of these find-ings undercount adjustment factors for the 10-year cumula-tive prevalence were required and were estimated to rangefrom 137 (lower bound 95 CI 113ndash166) to 147 (upperbound 95 CI 123ndash176) We applied these factors to deriveestimates for the 2010 prevalence of MS cumulated over 10years13

NeurologyorgN Neurology | Volume 92 Number 10 | March 5 2019 e1031

Complementary analysisThe cumulative prevalence of MS grows at variable rates andeventually levels off when an algorithm is applied to a givenhealth system AHC dataset713 We have shown that the av-erage annual growth in the MS prevalence rate between 2010and 2017 for 2 AHC datasets was 23y13 This growth ratecan be applied to the 2010 prevalence estimates to obtaina more recent prevalence figure

SAS version 94 (SAS Institute Inc Cary NC) and SPSSversion 22 (IBM Inc Armonk NY) were used to conduct thestatistical analyses

Standard protocol approvals registrationsand patient consentsThe study was approved by the Institutional Review Board atthe VA Medical CenterndashBaltimoreUniversity of MarylandMedical Center KPSC Colorado Multiple Institutional Re-view Board Stanford University and Quorum Review Stan-dard contracts and data use agreements were obtained for theanalysis of all datasets

Data availabilityThe datasets for this study were purchased and are owned bythe NMSS There are no current sharing agreements and dataare held under a data use contract with the NMSS

ResultsThe characteristics of the AHC datasets during the studyperiod are summarized in table 1 in total they captured 125million persons ge18 years of age The 3 private insurancedatasets collectively captured 58 million individuals approx-imately half of the privately insured US adult population (age18ndash64 years) Collectively the public (government) in-surance sources captured all 68 million individuals nationwidewho are insured through these plans

The prevalence estimates reported refer to the adult populationThe age- and sex-stratified prevalence estimates for MS in 2010cumulated over 3 years (2008ndash2010) and number of cases fromthese datasets are displayed in tables e-1 through e-5 (availablefrom Dryad httpsdoiorg105061dryadpm793v8) Theprevalences for OP and TH are remarkably similar while the VAsex-stratified estimates are highest among the datasets Denomi-nator data from the 2010 US Census are shown in table e-6(available fromDryad httpsdoiorg105061dryadpm793v8)

The annual cumulative prevalence of MS in the United Statesfor 2008 to 2010 using the MS algorithm is displayed for the 2private health insurance datasets (OP and TH) and the 3government insurance datasets (Medicaid Medicare andVA) in figure 1 The average annual increase in prevalence

Table 1 Characteristics of AHC datasets used for the US MS prevalence estimate

Health claims database (URL)Adult enrollees2008ndash2010 n Health plan characteristics

Geographiccoverage Variables in database

OP (optumcom) 15 million Employer-based fee-for-servicepreferred provider or capitatedhealth plans United Health Care

All 50 states Demographic data hospitaladmissions outpatient claimsemergency room claims prescriptionmedications

TH Market Scan (marketscantruvenhealthcom)

40 million Mix of private health insurancecompanies

All 50 states Demographic data hospitaladmissions outpatient claimsbehavioral health claims emergencyroom claims prescriptionmedications

KPSC (sharekaiserpermanenteorg)

27 million Health maintenance organizationwith integrated clinical and hospitalnetwork

SouthernCalifornia

Demographic data hospitaladmissions outpatient claimsemergency room claims prescriptionmedications clinical records datalaboratory data imaging data

VA (vagov) 85 million National government health careprogram for military veteransoutpatient clinics medical centersrehabilitation facilities nursinghomes and service offices

All 50 states Demographic data hospitaladmissions outpatient claimsemergency room claims prescriptionmedications clinical recordslaboratory data imaging data

Medicaid (national) (cmsgov) 279 million State-based health insuranceprogram supplemented by federalfunding to provide for those inpoverty orwith designated disabilities

All 50 states Demographic data hospitaladmissions outpatient claimsemergency room claims prescriptionmedications

Medicare (national) (cmsgov) 312 million Federal health insurance program forthe elderly (ge65 y) and disabled

All 50 states Demographic data hospitaladmissions outpatient claimsemergency room claims prescriptionmedications

Abbreviations AHC = administrative health claims KPSC = Kaiser Permanente Southern California MS =multiple sclerosis OP = Optum TH = Truven HealthVA = Department of Veterans Affairs

e1032 Neurology | Volume 92 Number 10 | March 5 2019 NeurologyorgN

over this 3-year period for all datasets was 63 (SD 38)The 2010 MS prevalence estimate cumulated over 3 years forthe combined datasets was 19984 (95 CI 19983ndash19985)corresponding to 470053 people with MS

After adjustment for the uninsured and application of thelower-bound inflation factor to account for undercountingdue to the limited period of observation the estimated 2010prevalence for MS cumulated over 10 years was 2651 per100000 (95 CI 2643ndash2658) corresponding to 623437people with MS Similarly after adjustment for the uninsuredand application of the upper-bound inflation factor the 2010prevalence for MS cumulated over 10 years was 3092 per100000 (95 CI 3081ndash3101) representing a total of727344 people with MS The 2010 MS prevalence estimatescumulated over 10 years 95 CIs and number of cases in theUnited States stratified by age and sex (lower and higherestimates) are shown in table e-7 (available from Dryadhttpsdoiorg105061dryadpm793v8) Figure 2 showsthe lower and figure 3 displays the higher 2010MS prevalenceestimates cumulated over 10 years in the United Statesstratified by age and sex The overall femalemale prevalenceratio for 2010 was 28 The peak age-specific MS prevalencewas 55 to 64 years followed by 65 to 74 years

Tables 2 and 3 show the 2010 MS prevalence estimates cu-mulated over 10 years 95 CIs and number of cases in theUS by age sex and geography (lower and higher estimates)Figure 3 illustrates the higher estimate for 2010 cumulatedover 10 years by Census regions along with correspondingsex ratios The prevalence in the northern Census regions ofthe US (Northeast and Midwest) was statistically significantly

higher than in the southern Census region as evidenced bynonoverlapping 95 CIs

DiscussionWe report a current national prevalence estimate for MS inthe adult population by using a validated algorithm across 5large US AHC datasets which also accounted for the un-insured population and for the limited (3-year period) ofobservation Our estimates were based on a case-findingstrategy that identified MS cases annually combined with the2010 Census for denominator data Overall nearly 45 of theUS population was assessed including 100 of those withpublicly funded insurance In 2010 our higher 10-year prev-alence estimate was 3092 per 100000 population repre-senting 727344 adults affected by MS This higher 2010estimate is based on the adjustment (3-year vs 10-year) seenin a single health insurance payer system covering the entirepopulation The lower-level estimate is based on a datasetadjustment (3-year vs 10-year) for a government insurancecarrier for a segment of the population Our approachaccounted for the demographics of the national populationthe sporadic follow-up for a chronic disease with young-adultonset the different insurance providers within the health caresystem and the uninsured

When coupled with prior estimates of the prevalence of MS inthe US our findings suggest that there has been a steady risein the prevalence of MS over the past 5 decades that theprevalence of MS remains higher for women than men andthat a north-south geographic gradient still persists2 Theearliest published national US MS prevalence estimate for the

Figure 1 Average annual prevalence ofMS in the United States for 2008 2009 and 2010 per 100000 population for private(OP and TH black lines) and government (Medicaid Medicare and VA red lines) health insurance datasets

Multiple sclerosis (MS) cases acquired from the MSalgorithm in year 1 are displayed alongwith new casesadded cumulatively in subsequent years OP = OptumTH = Truven Health VA = Department of VeteransAffairs

NeurologyorgN Neurology | Volume 92 Number 10 | March 5 2019 e1033

adult and pediatric populations combined was 58 per 100000for the year 19764 corresponding to 123000 cases with a 17femalemale ratio That study used surveys of hospitalsphysicians and patients to estimate prevalence Subsequentresearchers used this prevalence of 58 per 100000 as a base-line and estimated the number of physician-diagnosed MScases in the United States as 300000 by 1990 after factoring inpopulation changes and more contemporary regional preva-lence estimates5 Using the National Health Interview Surveyinvestigators calculated a national MS prevalence rate of 85per 100000 for the period of 1989 to 199420 The corre-sponding femalemale ratio was reported as 261 muchhigher than the 1976 sex ratio4 Authors of a later studyreported a national annual period prevalence of 021 overa 10-year period (1998ndash2009) with cases with MS identifiedby at least 1 ICD-9 340 code in theMedical Expenditure PanelSurvey (MEPS)21 This corresponded to 572312 patientswith MS More recently a team with a more restrictive al-gorithm over 5 years (2008ndash2012) using the PharMetricscommercially insured claims database produced an annualperiod MS prevalence estimate of 149 per 100000 with403630 individual cases22 which is reasonably consistentwith our 3-year estimate in 2010 of 470053 cases The cor-responding femalemale ratio was 31 The higher prevalenceestimate of the MEPS database over a longer ascertainmentperiod than that of the PharMetrics study emphasizes theneed to account for undercounting with limited observationperiods This also accounts for the fact that our prevalencefigure is most in line with the 10-yearMEPS database estimate

but with the advantage of a formally validated algorithm anda broader demographic sample

We did not assess the prevalence of MS in children and thisshould be considered when our findings are compared tothose reported in other populations If we use our age-stratified rates for 2010 (low and high estimates) they fallwithin the range of the 2006 MS prevalence estimates inManitoba Canada7 for all 10-year age groups Similarly theage-stratified prevalence estimates for Northern IrelandUnited Kingdom in 2004 were slightly higher than our age-specific estimates for the following age deciles 25 to 34 25 to44 and 45 to 54 years23 MS prevalence in older age groupswas slightly higher in the 2010 US population Compared tothe 2008 age-stratified prevalence estimates from BritishColumbia Canada8 our US rates were variably higher(2ndash30) for all age categories Thus our age-stratifiedestimates are in line with recent North American and Euro-pean figures If we were to assume that the prevalence of MSin children were zero although this is a conservative as-sumption our 2010 prevalence estimates would range from2186 (low estimate) to 2345 (high estimate) per 100000population

Thus our prevalence results are consistent with recent reportsof MS prevalence from other regions that examined the entirepopulation Canada has observed dramatic increases in theprevalence of MS with an estimate of 2669 per 100000 inNova Scotia9 for the year 2010 and a prevalence of 1799 per

Figure 2 2010 Prevalence for MS cumulated over 10 years in the United States per 100000 population by age and sex

Higher and lower estimates adjusted to 2010 US Census based on combined datasets from the multiple sclerosis (MS) algorithm inclusive of the followingTruven Optum Department of Veterans Affairs Medicare and Medicaid (full data available for all age and sex estimates in data table e-7 available fromDryad httpsdoiorg105061dryadpm793v8) (A) Lower-estimate and (B) higher-estimate 2010MS prevalence in theUnited States per 100000 population

e1034 Neurology | Volume 92 Number 10 | March 5 2019 NeurologyorgN

100000 in the province of British Columbia in 20088 InEuropean regions the prevalence of MS has been variablewith northern regions having higher estimates For examplea 2005 national prevalence of 1545 per 100000 (EuropeanStandard Population) was reported from Denmark24 Overthe past decade France has had national prevalence estimatesgenerally lt947 per 100000 (standardized to French pop-ulation)25 South American prevalence studies are largely re-gional but a 2005 study from Panama produced a crudeprevalence of 52 per 10000026 substantially lower thanNorth American estimates

Prevalence is the product of the incidence rate and the averageduration of a condition Incidence rates for MS have beengenerally stable or have slightly increased over the past 4 to 5decades in white populations but are higher in selected racialgroups27ndash29 Therefore the rising prevalence estimates for MSacross the Western world (ie populations of predominatelyEuropean ancestry) largely reflect the aging of the populationwith improved survival7 In addition the diagnostic criteria forMS have evolved and an earlier diagnosis of MS is possiblewith the use of neuroimaging and this is likely contributing tothe increased prevalence observed3031

With the wide availability of electronic datasets within healthcare systems and new techniques to analyze data there hasbeen a rise in the number of morbidity and mortality studieson a global scale3 However the methods for identifying casesand other variables of interest have not been standardized To

ensure that our approach is transparent and to supportcomparisons to future work in the United States and otherjurisdictions we have reported the annual MS prevalencerates for our combined and individual datasets

A strength of our study is the validation of our algorithmagainst an available gold standard that is medical records inmultiple datasets and health systems13 Our approach alsoaccounted for the complexity of the current national healthcare system1 To address the need for accurate morbidity andmortality data for neurologic conditions a national surveil-lance system approach compiling electronic AHC datasetsand vital statistics would be a logical way forward The Neu-rological Disease Surveillance System that was authorized bythe US Congress in 2016 could adopt this relatively time- andcost-efficient approach32

On the basis of observed increases in prevalence with our VAand Intercontinental Marketing Services datasets after 2010we estimated that the prevalence of MS in 2017 cumulatedover 17 years would range from 3379 per 100000 population(n = 851749 persons with MS) to 3626 per 100000 pop-ulation (n = 913925 persons with MS) These 17-year cu-mulative estimates approach lifetime prevalence for MSwithin the bounds of our AHC datasets However theseestimates should be viewed with caution because they assumethat the factors that have contributed to the rising prevalenceobserved in the United States as of 2010 based on codingrecords have persisted to 2017 and that substantial changes in

Figure 3 High-estimate 2010 prevalence of MS in the United States per 100000 population (2010 US Census) by Censusregion with 95 CI and FM prevalence ratio

CI = confidence interval FM = femalemale MS = multiple sclerosis

NeurologyorgN Neurology | Volume 92 Number 10 | March 5 2019 e1035

the distribution and survival of the population at risk of MShave not occurred These assumptions include the continuedhigh longevity of patients withMS and the general populationstable incidence of MS and similar coding practices from2010 to 2017 Extrapolated estimates have been modeled forMS and systemic lupus erythematosus within the Canadianhealth care system1733 by assuming stable incidence whichhas been consistently reported across Canada In the UnitedStates recent data from the Centers for Disease Control andPrevention show a slight decrease in life expectancy for the USpopulation in 201634 and the demographic composition ofthe US population is also changing35 Future studies can usethese methods to reassess the prevalence of MS and to ex-amine how these factors affect the findings

There are limitations to our approach First we did not in-clude children the Indian Health Service the US prisonsystem or undocumented US residents in our prevalenceestimates However these segments of the population arerelatively small or in the case of children would contributefew cases and many individuals would be detected by otherhealth systems including the Medicare insurance program atsome point later in life Furthermore diagnosing pediatric MSis challenging the performance of our proposed algorithmwould need to be tested in this population given the recog-nized differences in relapse rates more prominent cognitiveimpairment that may affect health care use and reporteddifferences in performance of algorithms across the pediatricand adult populations in other chronic diseases36ndash38 This

Table 2 2010 Prevalence for MS cumulated over 10 years in the United States per 100000 population by age sex andgeography

Lower-estimate 2010 MS prevalence in the US per 100000 population

Age and sexgroups (y)

West Midwest Northeast South Total

Casesn

Prevalence(95 CI)

Casesn

Prevalence(95 CI)

Casesn

Prevalence(95 CI)

Casesn

Prevalence(95 CI)

Prevalence(95 CI)

Women 18ndash24 1276 360(336ndash384)

1506 468(440ndash496)

1716 639(603ndash675)

2652 471(450ndash493)

474(461ndash488)

Men 18ndash24 759 198(182ndash215)

599 179(162ndash197)

832 301(276ndash325)

1123 192(178ndash205)

210(201ndash218)

Women 25ndash34 8921 1794(1749ndash1838)

10254 2410(2354ndash2466)

9130 2583(2520ndash2647)

14769 1936(1898ndash1973)

2112(2088ndash2136)

Men 25ndash34 2935 568(543ndash592)

3442 800(768ndash833)

3548 1020(979ndash1060)

5618 745(721ndash768)

758(744ndash773)

Women 35ndash44 17235 3560(3497ndash3624)

23328 5377(5294ndash5460)

20554 5431(5342ndash5520)

33574 4286(4231ndash4341)

4553(4518ndash4588)

Men 35ndash44 6254 1267(1229ndash1305)

7476 1741(1694ndash1789)

6834 1878(1825ndash1931)

10473 1364(1333ndash1395)

1510(1491ndash1531)

Women 45ndash54 28460 5612(5533ndash5690)

35392 7052(6964ndash7140)

33093 7623(7525ndash7722)

47026 5612(5551ndash5673)

6312(6273ndash6351)

Men 45ndash54 8765 1750(1706ndash1794)

11889 2409(2357ndash2461)

10775 2605(2546ndash2664)

14107 1756(1721ndash1791)

2059(2036ndash2082)

Women 55ndash64 27055 6401(6309ndash6492)

31080 7451(7352ndash7550)

26874 7522(7414ndash7630)

39796 5640(5574ndash5706)

6560(6516ndash6603)

Men 55ndash64 9964 2488(2429ndash2546)

11753 2955(2891ndash3019)

10018 3048(2976ndash3119)

12451 1932(1891ndash1972)

2494(2466ndash2522)

Women 65ndash74 14304 5731(5618ndash5843)

17410 6849(6728ndash6971)

15373 7036(6902ndash7169)

23247 5178(5098ndash5257)

6005(5952ndash6058)

Men 65ndash74 5376 2412(2335ndash2489)

5972 2702(2620ndash2784)

4867 2654(2565ndash2744)

7232 1860(1809ndash1912)

2308(2272ndash2343)

Women ge75 4685 2047(1977ndash2117)

5838 2208(2140ndash2276)

6565 2769(2689ndash2850)

6987 1746(1697ndash1795)

2130(2097ndash2162)

Men ge75 1749 1104(1042ndash1166)

1583 956(899ndash1012)

1288 894(836ndash953)

2084 805(763ndash846)

922(896ndash949)

Total 137734 2541(2525ndash2557)

167521 3291(3272ndash3310)

151467 3518(3496ndash3539)

221138 2540(2527ndash2553)

2882(2874ndash2890)

Abbreviations CI = confidence interval MS = multiple sclerosisLower estimates adjusted to 2010USCensus on the basis of combined datasets from theMS algorithm inclusive of the following TruvenOptum Departmentof Veterans Affairs Medicare and Medicaid

e1036 Neurology | Volume 92 Number 10 | March 5 2019 NeurologyorgN

work was out of the scope of the present project but should bepursued in future studies Second those withMS not followedin the traditional health care system (eg alternative medicineor cash health care practices that bypass health insurancereimbursement) would be missed by our method This wouldresult in an underestimate of MS cases Third we had 10 yearsof data for the VA health care system and the province ofManitoba but shorter lengths of data for other health systemsto assess period effects on prevalence While a decade of datawould have been optimal for all health system datasets thehigh costs of obtaining gt3 years of data for all insurance poolswere prohibitive Finally we have not characterized the racialor ethnic demographics of our MS population in this reportbecause race and ethnicity were not uniformly collected in the

AHC datasets used Racial and ethnic differences in MS sus-ceptibility may be a factor contributing to the geographicdifferences in prevalence in US Census regions28 Strengths ofour approach included the large sample which captured one-third of the US population broad healthcare system repre-sentation and the use of a validated case-finding algorithmthat performed consistently across different health systems13

The US national MS prevalence estimate for 2010 is thehighest reported to date and provides a contemporary un-derstanding of the disease burden Our rigorous algorithm-based approach to estimate prevalence is efficient and can bereproduced in other health systems We would advocate forthis approach to be used for other chronic neurologic

Table 3 2010 Prevalence for MS cumulated over 10 years in the United States per 100000 population by age sex andgeography

Higher-estimate 2010 MS prevalence in the US per 100000 population

Age and sexgroups (y)

West Midwest Northeast South Total

Casesn

Prevalence(95 CI)

Casesn

Prevalence(95 CI)

Casesn

Prevalence(95 CI)

Casesn

Prevalence(95 CI)

Prevalence(95 CI)

Women 18ndash24 1369 386(361ndash412)

1616 502(472ndash533)

1841 686(647ndash725)

2846 506(482ndash529)

509(495ndash523)

Men 18ndash24 814 213(195ndash231)

642 192(174ndash211)

893 323(296ndash349)

1205 206(191ndash220)

225(216ndash234)

Women 25ndash34 9572 1925(1877ndash1973)

11003 2586(2526ndash2646)

9796 2772(2074ndash2940)

15846 2077(2037ndash2117)

2267(2241ndash2292)

Men 25ndash34 3149 609(582ndash635)

3694 859(824ndash893)

3807 1094(1051ndash1137)

6029 799(774ndash824)

814(798ndash829)

Women 35ndash44 18493 3820(3752ndash3889)

25030 5770(5681ndash5858)

22055 5827(5732ndash5922)

36024 4599(4540ndash4658)

4885(4848ndash4923)

Men 35ndash44 6711 1360(1319ndash1400)

8022 1869(1818ndash1919)

7333 2015(1958ndash2072)

11237 1464(1430ndash1497)

1621(1599ndash1643)

Women 45ndash54 30537 6021(5937ndash6105)

37975 7367(7473ndash7662)

35509 8180(8074ndash8285)

50458 6022(5957ndash6087)

6772(6730ndash6814)

Men 45ndash54 9404 1878(1831ndash1925)

12756 2584(2529ndash2640)

11562 2795(2732ndash2858)

15136 1884(1847ndash1921)

2209(2185ndash2234)

Women 55ndash64 29030 6868(6770ndash6966)

33349 8000(7888ndash8101)

28835 8071(7955ndash8187)

42701 6052(5980ndash6123)

7038(6991ndash7085)

Men 55ndash64 10691 2669(2607ndash2732)

12611 3170(3102ndash3239)

10750 3270(3193ndash3347)

13360 2073(2029ndash2116)

2676(2646ndash2706)

Women 65ndash74 15349 6149(6028ndash6270)

18680 7349(7219ndash7480)

16495 7549(7406ndash7692)

24945 5556(5470ndash5641)

6443(6386ndash6500)

Men 65ndash74 5769 2588(2505ndash2671)

6408 2900(2811ndash2988)

5222 2848(2752ndash2944)

7760 1996(1941ndash2051)

2476(2438ndash2514)

Women ge75 5027 2196(2121ndash2272)

6264 2369(2296ndash2442)

7043 2972(2885ndash3058)

7497 1873(1821ndash1926)

2285(2250ndash2320)

Men ge75 1877 1185(1118ndash1251)

1699 1026(965ndash1086)

1382 959(897ndash1022)

2236 864(819ndash908)

989(961ndash1018)

Total 147792 2727(2710ndash2744)

179749 3531(351ndash3552)

162523 3774(3752ndash3797)

237279 2726(2722ndash2740)

3092(3083ndash3101)

Abbreviations CI = confidence interval MS = multiple sclerosisHigher estimates adjusted to 2010 US Census on the basis of combined datasets from the MS algorithm inclusive of the following Truven OptumDepartment of Veterans Affairs Medicare and Medicaid

NeurologyorgN Neurology | Volume 92 Number 10 | March 5 2019 e1037

conditions Further work is needed to better understand thecurrent differences in MS prevalence by race and to evaluatepossible regional differences in health care use and diseasemorbidity for MS

AcknowledgmentThe authors thank the following coinvestigators and membersof US Multiple Sclerosis Prevalence Workgroup for theircontributions Albert Lo Robert McBurney Oleg MuravovBari Talente and Leslie Ritter The authors also thank thefollowing key staff members at the NMSS as contributors tothis project Cathy Carlson Timothy Coetzee Sherri GigerWeyman Johnson Eileen Madray Graham McReynolds andCyndi Zagieboylo

Study fundingThis study was funded by a grant from the NMSS (HC-1508-05693)

DisclosureMWallin has served on data safety monitoring boards for theNational Institutes of Neurological Disease and StrokendashNIHis a member of the NMSS Health Care Delivery and PolicyResearch study section and receives funding support from theNMSS and the Department of Veterans Affairs Merit ReviewResearch Program W Culpepper has research funding fromthe NMSS receives support from the VHA MS Center ofExcellence and is a member of the NMSS Health Care De-livery and Policy Research study section J Campbell hasconsultancy or research grants over the past 5 years from theAgency for Healthcare Research and Quality ALSAMFoundation Amgen AstraZeneca Bayer Biogen IdecBoehringer Ingelheim Centers for Disease Control andPrevention Colorado Medicaid Enterprise CommunityPartners Inc Institute for Clinical and Economic ReviewMallinckrodt NIH NMSS Kaiser Permanente PhRMAFoundation Teva Research in Real Life Ltd Respiratory

Appendix 1 Authors

Name Location Role Contribution

Mitchell T WallinMD MPH

Department of Veterans Affairs Medical CenterGeorgetown University Washington DC

Author Study concept and design critical review statistical analysesrevision of the manuscript for content approval of finalmanuscript

William J CulpepperPhD

Department of Veterans Affairs Medical CenterWashington DC University of Maryland Baltimore MD

Author Study concept and design statistical analyses critical reviewpreparation of draft manuscript approval of finalmanuscript

Jonathan CampbellPhD

University of Colorado Denver CO Author Statistical analyses critical review of the manuscript forcontent approval of final manuscript

Lorene M NelsonPhD

Stanford University Stanford CA Author Study concept and design critical review revision of themanuscript for content approval of final manuscript

Annette Langer-Gould MD

Kaiser Permanente Southern California Los AngelesCA

Author Study concept and design statistical analyses critical reviewrevision of the manuscript for content approval of finalmanuscript

Ruth Ann MarrieMD PhD

University of Manitoba Winnipeg Manitoba Canada Author Study concept and design critical review revision of themanuscript for content approval of final manuscript

Gary R Cutter PhD University of Alabama Birmingham Author Statistical analyses study concept and design revision of themanuscript for content approval of final manuscript

Wendy Kaye PhD McKing Consulting Corp Atlanta GA Author Statistical analyses critical review of the manuscript forcontent

Laurie Wagner MPH McKing Consulting Corp Atlanta GA Author Statistical analyses critical review of the manuscript forcontent

Helen Tremlett PhD University of British Columbia Vancouver BC Canada Author Study concept and design critical review of the manuscriptfor content approval of final manuscript

Steven L Buka ScD Brown University Providence RI Author Critical reviewof themanuscript for content approval of finalmanuscript

PiyamethDilokthornsakulPharmD PhD

University of Colorado Aurora CO Author Statistical analysis critical review of the manuscript forcontent

Barbara Topol MS Stanford University Stanford CA Author Statistical analysis critical review of the manuscript forcontent

Lie H Chen DrPH Southern California Permanente Medical GroupPasadena CA

Author Statistical analysis critical review of the manuscript forcontent

Nicholas G LaRocca National Multiple Sclerosis Society New York NY Author Study concept and design critical review of the manuscriptfor content approval of final manuscript

e1038 Neurology | Volume 92 Number 10 | March 5 2019 NeurologyorgN

Effectiveness Group and Zogenix Inc L Nelson receivesgrants from the Centers for Disease Control and PreventionNIH and NMSS and contracts from the Agency for ToxicSubstances and Diseases Registry She receives compensationfor serving as a consultant to Acumen Inc and is on the DataMonitoring Committee of Neuropace A Langer-Gould wassite principal investigator for 2 industry-sponsored phase 3clinical trials (Biogen Idec and Hoffman-LaRoche) and 1industry-sponsored observation study (Biogen Idec) Shereceives grant support from the NIH National Institutes ofNeurological Disease and Stroke Patient-Centered Out-comes Research Institute and the NMSS R Marrie is sup-ported by the Waugh Family Chair in Multiple Sclerosis andreceives research funding from Canadian Institutes of HealthResearch Research Manitoba Multiple Sclerosis Society ofCanada Multiple Sclerosis Scientific Foundation Consor-tium of MS Centers and NMSS She also serves on the Ed-itorial Board of Neurology G Cutter is a member of Data andSafety Monitoring Boards for AMO Pharmaceuticals ApotekHorizon Pharmaceuticals ModigenetechProlor MerckMerckPfizer Opko Biologics Neurim Sanofi-Aventis ReataPharmaceuticals ReceptosCelgene Teva PharmaceuticalsNational Heart Lung and Blood Institute (Protocol ReviewCommittee) and National Institute of Child Health andHuman Development (Obstetric Pharmacy Research Unitoversight committee) and on the Consulting or Advisoryboards for Atara Biotherapeutics Argenix Bioeq GmBHConsortium of MS Centers (grant) Genzyme GenentechInnate Therapeutics Klein-Buendel Inc Medimmune Med-day Novartis Opexa Therapeutics Roche Savara IncSomahlution Teva Pharmaceuticals Transparency Life Sci-ences and TG Therapeutics W Kaye receives funding fromthe Agency for Toxic Substances and Disease Registry theNMSS and the Association for the Accreditation of HumanResearch Protection Programs L Wagner receives fundingfrom the Agency for Toxic Substances and Disease Registryand NMSS H Tremlett is the Canada Research Chair forNeuroepidemiology and Multiple Sclerosis She receives re-search support from the NMSS the Canadian Institutes ofHealth Research the Multiple Sclerosis Society of Canadaand the Multiple Sclerosis Scientific Research Foundation In

addition in the last 5 years she has received research supportfrom the Multiple Sclerosis Society of Canada the MichaelSmith Foundation for Health Research and the UK MSTrust as well as travel expenses to attend conferences allspeaker honoraria are either declined or donated to an MScharity or to an unrestricted grant for use by her researchgroup S Buka receives research funding from the NIH andthe NMSS P Dilokthornsakul receives research funding fromPfizer Thailand and the Thai Traditional Medical KnowledgeFund B Topol receives grant support from the Centers forDisease Control and Prevention and the NMSS and contractsfrom the Agency for Toxic Substances and Diseases RegistryL Chen reports no disclosures relevant to the manuscript NLaRocca is employed full-time by the NMSS Go to Neurol-ogyorgN for full disclosures

Publication historyReceived byNeurology February 24 2018 Accepted in final form August17 2018

References1 Nelson LM Wallin MT Marrie RA et al A new way to estimate neurologic disease

prevalence in the United States illustrated with MS Neurology 201992469ndash4802 Evans C Beland SG Kulaga S et al Incidence and prevalence of multiple sclerosis in

the Americas a systematic review Neuroepidemiology 201340195ndash2103 GBD 2015 Neurological Disorders Collaborator Group Global regional and na-

tional burden of neurological disorders during 1990-2015 a systematic analysis for theGlobal Burden of Disease Study 2015 Lancet Neurol 201716877ndash897

4 Baum HM Rothschild BB The incidence and prevalence of reported multiple scle-rosis Ann Neurol 198110420ndash428

5 Anderson DW Ellenberg JH Leventhal CM Reingold SC Rodrigues M SilberbergDH Revised estimate of the prevalence of multiple sclerosis in the United States AnnNeurol 199231333ndash336

6 Mayr WT Pittock SJ McClelland RL Jorgensen NW Noseworthy JH Rodrigues MIncidence and prevalence of multiple sclerosis in Olmsted County Minnesota1985ndash2000 Neurology 2003611373ndash1377

7 Marrie RA Yu N Blanchard J Leung S Elliott L The rising prevalence and changingage distribution of multiple sclerosis in Manitoba Neurology 201074464ndash471

8 Kingwell E Zhu F Marrie RA et al High incidence and increasing prevalence ofmultiple sclerosis in British Columbia Canada findings from over two decades(1991ndash2010) J Neurol 20152622352ndash2363

9 Marrie RA Fisk JD Stadnyk KJ et al The incidence and prevalence of multiplesclerosis in Nova Scotia Canada Can J Neurol Sci 201340824ndash831

10 US Census Bureau Annual estimates of the resident population for the United States2010ndash2017 Available at censusgovdatadatasets2017demopopeststate-totalhtml Accessed July 14 2018

11 US Census Bureau American community survey Available at censusgovdatatablestime-seriesdemohealth-insuranceacs-hihtml Accessed July 14 2018

12 Leonard CE Brensinger CM Nam YH et al The quality of Medicaid and Medicaredata obtained fromCMS and its contractors implications for pharmacoepidemiologyBMC Health Serv Res 201717304ndash311

Appendix 2 Coinvestigators and members of the US Multiple Sclerosis Prevalence Workgroup

Name Location Role Contribution

Albert Lo MD PhD Brown University Providence RI Coinvestigator Study concept anddesign

Robert McBurneyPhD

Accelerated Cure Project Boston MA Coinvestigator Study concept anddesign

Oleg Muravov PhD Agency for Toxic Substances and Disease Registry Division of Health StudiesAtlanta GA

Coinvestigator Study concept anddesign

Bari Talente Esq National Multiple Sclerosis Society Washington DC Coinvestigator Study concept anddesign

Leslie Ritter National Multiple Sclerosis Society Washington DC Coinvestigator Study concept anddesign

NeurologyorgN Neurology | Volume 92 Number 10 | March 5 2019 e1039

13 Culpepper WJ Marrie RA Langer-Gould A et al Validation of a Standardized Algorithmfor Identifying MS Cases across Claims Datasets Neurology 201992e1016ndashe1028

14 Reich D Lucchinetti CF Calabresi PA Multiple sclerosis N Engl J Med 2018378169ndash180

15 Ruggles S Genadek K Goeken R Grover J SobekM Integrated public use microdataseries version 60 Minneapolis University of Minnesota 2015 Available at doiorg1018128D010V60 Accessed July 14 2018

16 Institute of Medicine Care Without Coverage Too Little Too Late WashingtonDC National Academies Press 2002 Available at doiorg101722610367Accessed July 14 2018

17 Minden SL Frankel D Hadden L Hoaglin DC Access to health care for people withmultiple sclerosis Mult Scler 200713547ndash558

18 Ng R Bernatsky S Rhame E Observation period effects on estimation of systemic lupuserythematosus incidence and prevalence in Quebec J Rheumatol 2013401334ndash1336

19 Capocaccia R Colonna M Corazziari I et al Measuring cancer prevalence in Europethe EUROPREVAL project Ann Oncol 200213831ndash839

20 Noonan CW Kathman SJ White MC Prevalence estimates for MS in the UnitedStates and evidence of an increasing trend for women Neurology 200258136ndash138

21 Campbell JD Ghushchyan V Brett McQueen R et al Burden of multiple sclerosis ondirect indirect costs and quality of life national US estimates Mult Scler Relat Disord20143227ndash236

22 Dilokthornsakul P Valuck RJ Nair KV Corboy JR Allen RR Campbell JD Multiplesclerosis prevalence in the United States commercially insured population Neurology2016861014ndash1021

23 Gray OM McDonnell GV Hawkins SA Factor in the rising prevalence of multiplesclerosis in the north-east of Ireland Mult Scler 200814880ndash886

24 Bentzen J Flachs EM Stenager E Bronnum-Hansen H Koch-Henriksen N Preva-lence of multiple sclerosis in Denmark 1950ndash2005 Mult Scler 201016520ndash525

25 Kingwell E Marriott JJ Gette N et al Incidence and prevalence of multiple sclerosisin Europe a systematic review BMC Neurol 201313128

26 Gracia F Castillo LC Benazdon A et al Prevalence and incidence of multiple scle-rosis in Panama (2000ndash2005) Neuroepidemiology 200932287ndash293

27 Koch-Henriksen Sorensen PS The changing demographics of multiple sclerosisepidemiology Lancet Neurol 20109520ndash532

28 Langer-Gould A Brara SM Beaber BE Zhang JL Incidence of multiple sclerosis inmultiple racial and ethnic groups Neurology 2013801734ndash1739

29 Wallin MT Culpepper WJ Coffman P et al The Gulf War era multiple sclerosiscohort age and incidence rates by race sex and service Brain 20121351778ndash1785

30 Polman CH Reingold SC Banwell B et al Diagnostic criteria for multiple sclerosis2010 revisions to the McDonald criteria Ann Neurol 201169292ndash302

31 Smith-Bindman R Miglioretti DL Larson EB Rising use of diagnostic medical im-aging in a large integrated health system Health Aff 2008271491ndash1502

32 21st Century Cures Act of 2016 Pub L No 114-255 section 2061 National Neu-rological Diseases Surveillance System 2016

33 Amankwah N Marrie RA Bancej C et al Multiple sclerosis in Canada 2011 to 2031results of a microsimulation modelling study of epidemiological and economicimpacts Health Promot Chronic Dis Prev Can 20173737ndash48

34 Kochanek KD Murphy SL Xu J Arias E Mortality in the United States 2016 NCHSData Brief 20171ndash8

35 Colby SL Ortman JM Projections of the Size and Composition of the US Pop-ulation 2014 to 2060 Current Population Reports P25-1143 Washington DC USCensus Bureau 2014

36 Guttmann A Nakhla M HendersonM et al Validation of a health administrative dataalgorithm for assessing the epidemiology of diabetes in Canadian children PediatrDiabetes 201011122ndash128

37 Ghezzi A Baroncini D Zaffaroni M Comi G Pediatric versus adult MS similar ordifferent Mult Scler Demyelin Disord 201725

38 Benchimol EI Guttmann A Griffiths AM et al Increasing incidence of paediatricinflammatory bowel disease in Ontario Canada evidence from health administrativedata Gut 2009581490ndash1497

e1040 Neurology | Volume 92 Number 10 | March 5 2019 NeurologyorgN

DOI 101212WNL0000000000007035201992e1029-e1040 Published Online before print February 15 2019Neurology

Mitchell T Wallin William J Culpepper Jonathan D Campbell et al claims data

The prevalence of MS in the United States A population-based estimate using health

This information is current as of February 15 2019

ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

ServicesUpdated Information amp

httpnneurologyorgcontent9210e1029fullincluding high resolution figures can be found at

References httpnneurologyorgcontent9210e1029fullref-list-1

This article cites 31 articles 9 of which you can access for free at

Citations httpnneurologyorgcontent9210e1029fullotherarticles

This article has been cited by 5 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectionscreening_in_epidemiologyScreening in epidemiology

httpnneurologyorgcgicollectionprevalence_studiesPrevalence studies

httpnneurologyorgcgicollectionmultiple_sclerosisMultiple sclerosis

httpnneurologyorgcgicollectionall_epidemiologyAll epidemiologyfollowing collection(s) This article along with others on similar topics appears in the

Errata

content93156882fullpdf or page

nextAn erratum has been published regarding this article Please see

Permissions amp Licensing

httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online

ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

ServicesUpdated Information amp

httpnneurologyorgcontent9210e1029fullincluding high resolution figures can be found at

References httpnneurologyorgcontent9210e1029fullref-list-1

This article cites 31 articles 9 of which you can access for free at

Citations httpnneurologyorgcontent9210e1029fullotherarticles

This article has been cited by 5 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectionscreening_in_epidemiologyScreening in epidemiology

httpnneurologyorgcgicollectionprevalence_studiesPrevalence studies

httpnneurologyorgcgicollectionmultiple_sclerosisMultiple sclerosis

httpnneurologyorgcgicollectionall_epidemiologyAll epidemiologyfollowing collection(s) This article along with others on similar topics appears in the

Errata

content93156882fullpdf or page

nextAn erratum has been published regarding this article Please see

Permissions amp Licensing

httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online

ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

CORRECTIONS

Out-of-pocket costs are on the rise for commonly prescribedneurologic medicationsNeurologyreg 201993688 doi101212WNL0000000000008353

In the article ldquoOut-of-pocket costs are on the rise for commonly prescribed neurologic medi-cations by Callaghan et al1 first published online May 1 2019 the 2004 out-of-pocket costsfor MS medications (mean [SD]) in the Abstract and Results should be $33 ($50) rather than$15 ($23) and the 2004 medianIQR in the Results should be $25 ($20ndash$32) rather than $14($10ndash$16) The authors regret the errors

Reference1 Callaghan BC Reynolds E Banerjee M et al Out-of-pocket costs are on the rise for commonly prescribed neurologic medications

Neurology 201992e2604ndashe2613

The prevalence of MS in the United StatesA population-based estimate using health claims dataNeurologyreg 201993688 doi101212WNL0000000000007915

In the article ldquoThe prevalence of MS in the United States A population-based estimate usinghealth claims data by Wallin et al1 first published online February 15 2019 the text regardingthe lower bound for MS prevalence in a paragraph in Results should read ldquoAfter adjustment forthe uninsured and application of the lower-bound inflation factor to account for undercountingdue to the limited period of observation the estimated 2010 prevalence for MS cumulated over10 years was 2882 per 100000 (95 CI 2874ndash2890) corresponding to 623437 people withMSrdquo This is correctly represented in table 2 The authors regret the error

Reference1 Wallin MT Culpepper WJ Campbell JD et al The prevalence of MS in the United States a population-based estimate using health

claims data Neurology 201992e1029ndashe1040

Epidemiology of NMOSD in Sweden from 1987 to 2013A nationwide population-based study

Neurologyreg 201993688 doi101212WNL0000000000008382

In the article ldquoEpidemiology of NMOSD in Sweden from 1987 to 2013 A nationwidepopulation-based study by Jonsson et al1 in figure 5 the incidence of NMOSD in Australiaand New Zealand should have been 0371000000 person-years (CI 035ndash039) The figureshould also have included A and B labels for the panels and a label for the first panelrsquos x-axisldquoIncidence rate (per 1000000 individuals)rdquo The authors and the editorial office regret theerrors

Reference1 Jonsson DI SveinssonO HakimR Brundin L Epidemiology of NMOSD in Sweden from 1987 to 2013 a nationwide population-based

study Neurology 201993e181ndashe189

688 Copyright copy 2019 American Academy of Neurology

Copyright copy 2019 American Academy of Neurology Unauthorized reproduction of this article is prohibited