Diagnostic accuracy of an algorithm identifying US veterans with Inclusion Body Myositis from the Corporate Data Warehouse

doi:10.21203/rs.3.rs-8387113/v1

Download PDF

Research Article

Diagnostic accuracy of an algorithm identifying US veterans with Inclusion Body Myositis from the Corporate Data Warehouse

https://doi.org/10.21203/rs.3.rs-8387113/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objectives

Methods

Of 732 previously identified veterans with IBM, 107 were randomly selected for manual record review using a keyword search for “Inclusion Body Myositis” or “IBM” via the tvf_TIU_FullTextSearch function. Rheumatology and Neurology notes from January 1, 2011 to June 1, 2024 were extracted and data tokenized using R packages tidytext and tokenizers. Incomplete records were supplemented with Voogle data.

Results

83 (77.5%) of 107 veterans had definite and 7 (6.5%) suspected clinical IBM. There were 4 cases with HIV-associated myopathy and 5 with inherited myopathies. Only 7 individuals were concurrently identified among a separate cohort of 1,136 veterans with a diagnosis of polymyositis based on administrative data. Detailed muscle biopsy pathology reports were available for 40/107 veterans of whom 30 (75%) noted findings consistent with IBM and 30 (75%) met ENMC 2024 diagnostic criteria for IBM. Based on ENMC 2011 criteria, 2 (5%), 12 (30%), and 25 (62.5%) veterans met clinicopathologic, clinical, and probable IBM definitions, respectively. Using Griggs criteria, 2 (5%) and 15 (37.5%) individuals met definite and probable IBM thresholds. The minimum positive predictive value (PPV) was 74.4% for an IBM clinical diagnoses and 88.0% for cases meeting diagnostic criteria, with specificity ≥ 90% for most groups.

Conclusions

The algorithm demonstrates robust performance and excellent specificity in identifying Veterans with IBM, comparable to approaches for systemic lupus and other IIMs.

Rheumatology

Neurology

Epidemiology

Idiopathic inflammatory myopathy

inclusion body myositis

veterans

• The proposed algorithm, utilizing a combination of billing data and specialty clinic visits, has robust performance and excellent specificity in identifying veterans with underlying IBM

• These data support the utility of using available big data resources to facilitate future IBM research.

Inclusion body myositis (IBM) is the most common idiopathic inflammatory myopathy (IIM) in adults over 50 years of age and has only one International Classification of Diseases (ICD) code across ICD-9 and ICD-10^1–7. Despite this, little is known about the performance of administrative data in accurately identifying individuals with IBM. Similarly, as IBM can share features with other subsets of IIM – most commonly polymyositis (PM) – its diagnosis cannot be routinely confirmed based on muscle biopsy features alone and requires the integration of clinical and pathologic data^8–10. European Neuromuscular Center (ENMC) 2024 IBM diagnostic criteria have been updated from the previously published 2011 ENMC criteria, incorporating atypical IBM features with predominantly proximal muscle weakness as well as anti-cN1a autoantibody (also termed NT5c1a or cytosolic 5’-nucleotidase 1A) testing^11–14. Although the Griggs criteria for IBM are still utilized and were the first standardized guidelines focusing on muscle biopsy findings, they were found to have less sensitivity, especially in identifying early disease^15,16. In this study, a previously published algorithm (Fig. 1) employing administrative and billing data from Veterans Affairs (VA) Corporate Data Warehouse (CDW) to identify veterans with IBM was compared to expert clinical IBM diagnoses and the aforementioned diagnostic criteria¹⁷. This work also assessed the performance of these formal diagnostic guidelines in routine clinical care.

Data Sources and Setting

Data were previously extracted from the VA CDW (Federal Register: 79 FR 4377) using a published algorithm (Fig. 1)¹⁷. Briefly, all adult veterans 18 years and older with an outpatient ICD-9 or ICD-10 visit code for IBM or entry for same in their problem list between January 1, 2011 and December 31, 2021 were identified. A minimum of 2 visit codes for IBM from neurology or rheumatology clinics at least 30 days apart were required to confirm a diagnosis. Date of first IBM ICD code or problem list entry was inferred to be the date of IBM diagnosis. Data on survival status were censored as of 06/01/2024. A separate CDW cohort, similarly based on blling and administrative data to identify veterans with PM, was used to derive true and false negative cases. PM was chosen as the main disease comparison as it is the most commonly occurring and most difficult to differentiate IBM mimic^8–10.

Human and Animal Rights

This study complies with the Declaration of Helsinki and was reviewed and approved by the Pittsburgh VA Research and Development committee on 1/23/2023 and granted informed consent exemption status based on the Revised Common Rule/2018 (Project #1707389). No animals were used in this study.

Individual Random Case Review

Using a random number generator, 107 veterans were selected from 732 individuals identified by the algorithm as having IBM. The Text Integration Utility (TIU) domain was used to obtain raw text data from all rheumatology and neurology clinical notes from January 1, 2011 to June 1, 2024 in the VA electronic health record system (CPRS) by means of a keyword search for ‘Inclusion Body Myositis’ or ‘IBM’ using the tvf_TIU_FullTextSearch function provided by VINCI in Microsoft SQL Server Management Studio (Ver. 20.1). Note text underwent natural language processing (NLP) and tokenized using “tidytext” and “tokenizers” R packages for efficiency. Extracted data related to muscle biopsy findings, laboratory testing, clinical motor strength examination as well as past medical and IBM disease histories were prioritized, as required to determine an IBM diagnosis by formal diagnostic criteria^11,12,15. Prior use of immunosuppressive therapy, creatine phosphokinase (CPK) levels, and autoantibody testing were queried along with a prior history of PM (mis)diagnosis and presence of IBM diagnostic mimics. Incomplete records were manually supplemented with data from Voogle.

Statistical Analysis

Continuous variables are displayed as mean and standard deviation. Categorical variables are displayed as frequency with percentage. Differences among groups greater than 3 were compared with Kruskal-Wallis tests for continuous variables and Chi-square test for categorical variables. Pairwise comparisons were performed with two-tailed t tests for continuous and Chi-square tests for categorical variables with a significance level set at p < 0.05. Bonferroni correction was instituted as necessary for multiple comparisons. All interim and final data analyses were performed using R statistical computing software (version 4.1.2, R Foundation for Statistical Computing) provided by VA Informatics and Computing Infrastructure (VINCI).

Patient and Public Involvement

No patients were involved in this study.

Veterans with clinical diagnosis of IBM

Randomly selected cases compared favorably with no statistically significant different features compared to the previously published cohort (Table 1)¹⁷. Selected veterans were overwhelmingly male, White, non-Hispanic or Latino, with a mean (±SD) age of 54.1 (±9.4) and a mean (±SD) follow-up of 5.3 (±9.4) years with 38 (36.9%) persons deceased as of the censor date. Among 107 individuals, 83 (77.5%) had a definite IBM clinical diagnosis based on the opinion(s) of their treating specialist(s), while 7 (6.5%) had suspected IBM (Table 2). No patients with a CPK exceeding 12 times the upper normal were identified, and all cases exhibited weakness for at least one year or more. Four veterans had HIV-associated myopathy indistinguishable from IBM and were not included in further analyses. Other non-IBM diagnoses included hereditary motor and sensory neuropathy with proximal dominance (HMSN-P), phosphorylase kinase alpha 1 (PHKA1) deficiency, suspected muscular dystrophy, suspected Pompe’s disease, steroid myopathy, myofibrillar myopathy, and Kennedy disease (n = 1 each). Regarding concurrent autoimmune disorders, one veteran had positive anti-MDA5 (melanoma differentiation-associated gene 5) and SSA (Ro) autoantibodies and was also excluded from the cohort. Four other individuals had concurrent autoimmune processes including vitiligo (n = 1), undifferentiated connective tissue disease (n = 1), rheumatoid arthritis (RA) coexisting with celiac disease and ulcerative colitis (n = 1), and discoid lupus erythematosus (n = 1). These veterans were kept in the cohort due to the very low likelihood of their other disease(s) causing myopathy^18–20.

Table 1

**Comparison of baseline patient characteristics between original IBM veteran cohort and randomly selected cases for review.** P value represents t test for continuous variables and Pearson’s χ² for categorical variables, α = 0.05.
	Case Review	Original Cohort	p
n	107	732
Age, yr - median (SD)	54.1 (9.4)	55.5 (8.4)	0.147
Follow-up Period, yr – mean (SD)	5.3 (9.4)	5.5 (3.4)	0.828
Status at End of Follow-up Period
Alive – no. (%)	69 (67.0)	416 (56.8)	0.134
Deceased or Censored – no. (%)	38 (36.9)	316 (43.2)	0.134
Sex
Male - no. (%)	103 (96.3)	708 (96.7)	0.803
Female - no. (%)	4 (3.7)	24 (3.3)	0.803
Race
White - no. (%)	78 (72.9)	514 (70.2)	0.567
Black or Other - no. (%)	29 (27.1)	218 (29.8)	0.567
Ethnicity
Not Hispanic or Latino - no. (%)	98 (91.6)	653 (89.2)	0.453
Hispanic or Latino or Other - no. (%)	9 (8.4)	79 (10.8)	0.453

Table 2

Summary of clinical information for IBM cases selected for review. *Values may exceed 100% due to use of more than one therapy.
	Case Review
n	107
Clinical IBM Diagnosis
Definite – no. (%)	76 (71.0)
Suspected – no. (%)	7 (6.5)
Alternate Diagnosis – no. (%)	24 (22.5)
Eating Difficulties
Dysphagia – no. (%)	34 (31.8)
Choking – no. (%)	5 (5.6)
Ambulation Difficulties
Falls	28 (26.2)
Assistive Device Use
Cane	15 (14.0)
Walker	23 (21.5)
Power Wheelchair	21 (19.6)
Tests Available for Review
Electromyography – no. (%)	41 (38.3)
Muscle biopsy – no. (%)	40 (37.4)
Magnetic Resonance Imaging – no. (%)	6 (5.6)
Anti-cN1a antibody – no. (%)	12 (11.2)
Myositis-specific Antibody Testing – no. (%)	8 (7.5)
Prior Diagnosis of Polymyositis – no. (%)	12 (11.2)
Prior Immunosuppressive Therapy* – no. (%)	40 (37.4)
Glucocorticoids – no. (%)	25 (62.5)
Methotrexate – no. (%)	19 (47.5)
Azathioprine – no. (%)	8 (20.0)
Mycophenolate – no. (%)	2 (5.0)
Hydroxychloroquine – no. (%)	2 (5.0)
Calcineurin Inhibitor – no. (%)	1 (2.5)
Intravenous Immunoglobulin – no. (%)	11 (27.5)
Rituximab – no. (%)	2 (5.0)
Other Biologic – no. (%)	2 (5.0)

Randomly selected cases reflected a spectrum of morbidity from IBM: almost one-third (34/107, 31.8%) of veterans with IBM had dysphagia and 5 (5.6%) reported choking symptoms. Similarly, 28 (26.2%) individuals noted impaired ambulation and falls, necessitating the use of assistive devices including canes (15, 14.0%), walkers 23 (21.5%), and power wheelchairs (21, 19.6%).

In terms of available diagnostic data, electromyography (EMG) and muscle biopsy were the most frequent studies available (41, 38.3% and 40, 37.4%, respectively). Relatively few veterans had muscle magnetic resonance imaging (MRI) (6, 5.6%) or testing for anti-cN1a (also termed NT5c1a or cytosolic 5’-nucleotidase 1A) antibodies, other myositis-specific antibodies (MSAs), or anti-SSA (Ro) and SSB (La) antibodies, all of which have previously been associated with IBM^{13,14,21–24}. The latter finding echoed a similar observation in the parent cohort¹⁷.

Although 12 (11.2%) individuals were treated as PM at some point prior to their IBM diagnosis, the algorithm identified only 7 IBM cases out of 1,136 veterans comprising a separate CDW PM cohort, based on blling and administrative data, confirming a very low false negative rate. A significant number of veterans had prior immunosuppressive therapy (40, 37.4%), with glucocorticoids (25, 62.5%), methotrexate (19, 47.5%), and intravenous immunoglobulin (IVIg) being the most common. This was not limited solely to those (mis)diagnosed with PM or those with other autoimmune diseases. Two veterans with concurrent RA received adalimumab and etanercept, respectively, but it was not clear if this was related to their muscle or joint manifestations.

Among veterans with a clinical IBM diagnosis, the algorithm displayed excellent sensitivity (Se) and specificity (Sp) (Se 92.2% and 97.9%, Sp 92.8% and 97.3%) while maintaining a reasonable positive predictive value (PPV) (77.6% and 74.4% for definite and definite & suspected cases, respectively) and robust negative predictive value (NPV) (99.4% for both groups) (Table 3).

Table 3

**Algorithm performance in identifying veterans with IBM compared to expert clinical diagnosis and published diagnostic criteria.** Sens. – sensitivity, spec. – specificity, PPV – positive predictive value, NPV – negative predictive value.
	n	Sens.	Spec.	PPV	NPV
Clinical IBM Diagnosis
Definite	76	92.2%	97.9%	77.6%	99.4%
Definite & Suspected	83	92.8%	97.3%	74.4%	99.4%
ENMC 2024
Common & Uncommon	30	70.0%	90.9%	95.5%	52.6%
ENMC 2011
Probable	25	73.3%	70.0%	88.0%	46.7%
Clinically Defined	12	26.7%	80.0%	80.0%	26.7%
Clinico-pathologically Defined	2	3.3%	90.0%	50.0%	23.7%
Griggs
Possible	15	43.3%	90.0%	92.9%	34.6%
Definite	2	3.3%	90.0%	50.0%	23.7%

Veterans meeting formal diagnostic criteria for IBM

Of the 40 veterans with available detailed muscle biopsy pathology reports, 30 (75%) met ENMC 2024 diagnostic criteria for IBM encompassing both common (25, 83.3%) and uncommon (5, 16.7%) disease presentations¹¹. This compared to 25 (62.5%), 12 (30%), and 2 (5%) individuals fulfilling ENMC 2011 criteria for probable, clinically defined, and clinico-pathologically defined disease, respectively¹². Only 15 (37.5%) and 2 (5%) veterans fulfilled Griggs criteria for possible and definite IBM, respectively¹⁵. Thus, the algorithm performed the best with ENMC 2024-defined cases with Se of 70.0%, Sp of 90.9%, and PPV of 95.5% and worst with those meeting ENCM 2011 clinico-pathologically defined and Griggs definite IBM criteria (Se of 3.3%, Sp of 90.0%, and PPV of 50% for both). Despite poor sensitivity, even the latter categories exhibited excellent specificity, demonstrating the algorithm’s low rate of false positive cases at the expense of total case number.

This work is a continuation of a previously published project using big data to identify veterans with IBM¹⁷. While a cohort of 732 individuals with IBM was established, a formal algorithm-based validation of either a clinical or criteria-based diagnosis was not reported. Using a randomly selected subset of 107 cases from the original cohort with a spectrum of disease severity, the work herein demonstrated excellent performance in identifying definite and suspected clinical diagnoses of IBM with Se and Sp measures above 90% and negative predictive values exceeding 99%. Furthermore, this algorithm demonstrated robust Sp and PPV (> 90%) among veterans using the most recent ENMC 2024 diagnostic criteria¹¹. Among IBM mimics, only HIV infection occurred at a high enough frequency (4 cases, 3.7%) to include as a rule out condition in subsequent algorithm iterations. While other myopathies were identified as false positive cases in this cohort, these represented disparate entities that would be difficult to exclude - although one could consider adding the ICD-10 code G12.1 (‘other inherited spinal muscular atrophy’) to a list of exclusion diagnoses in the future. None of these cases ultimately met formal IBM diagnostic criteria. The performance of this algorithm compares favorably to other autoimmune diseases, including systemic lupus erythematosus and other IIM subsets^25,26. These results are encouraging with such a difficult to diagnose entity such as IBM, which can be challenging to distinguish from PM in early cases and requires a comprehensive combination of clinical and pathological data beyond the interpretation of a muscle biopsy.

Of note is the observation that many pathology features were absent from routine muscle biopsy reports: while most noted endomysial inflammation and rimmed vacuoles, many omitted details regarding MHC class I expression, protein accumulation staining, or electron microscopy studies confirming characteristic filaments as required by the more stringent subsets of IBM diagnostic criteria. Less difficulty was encountered with extraction of clinical information such as muscle strength testing, CK values, or disease duration, although some clinical notes omitted full documentation of muscle strength, missing either lower or upper extremities (e.g. missing leg strength in wheelchair-dependent veteran). These factors also contributed to the significant discrepancy in number of cases meeting more stringent criteria subsets - especially for ENMC 2011. Thus, the algorithm performed best among cases meeting ENMC 2024 criteria, as these were the least clinically restrictive and offered an important benefit in allowing for cases with atypical presentation, which were seen in 16.7% of the selected cohort.

Conversely, the low observed level of anti-cN1a and SSA (Ro) and SSB (La) testing was surprising. This was likely attributable to a combination of factors including cohort censor date occurring before ENMC 2024 criteria publication and suspicion that serologic testing was preferentially ordered and resulted via non-electronic means (e.g. paper slip for outside testing and PDF lab report scanned into the medical record system) and, thus, not captured in the CDW. This observation is supported by an equally low SSA (Ro) and SSB (La) antibody testing rate in the parent cohort¹⁷.

A notable challenge of this study was the definition of IBM cases. To address this barrier, a strategy of including multiple options such as expert clinical diagnosis and the use of updated diagnostic criteria, all of which require muscle biopsy data in addition to clinical information, was employed. The latter significantly decreased the number of eligible veterans, as many biopsies were performed outside the VA system at academic institutions resulting in detailed pathology reports being available only for a subset of the selected cohort. Consise reports and third-party references such as ‘biopsy-proven IBM’ or ‘inflammatory muscle biopsy’ were considered inconclusive. Additionally, in the fraction of IBM veterans with concurrent autoimmune disease where it was difficult if not impossible to determine which process was ultimately responsible for their muscle disease, patients were kept in the cohort due to the very low likelihood of their other disease(s) causing myopathy based on published literature^18–20.

Another possible drawback was the use of PM as the only disease comparator to IBM – this was chosen due to previous unpublished work identifiying veterans with this diagnosis and its’ common occurrence as a frequent IBM mimic. While the algorithm was not validated against other diagnoses such as muscular dystrophy or HIV myopathy, these diagnoses were encountered at a low frequency during manual case review, suggesting that they would not appreciably affect the results or conclusions of this work. Further supportive of this is that no alternate diagnoses in the selected cohort met formal IBM diagnostic criteria.

Finally, it should be acknolowedged that the current approach does not identify all possible veterans with IBM nor was that the goal of the original cohort. The algorithm was designed to maximize true positive IBM cases at the expense of overall cohort size for the purpose of future research. Despire the low prevalence of IBM, this goal was succefully accomplished, illustrating the importance and role of big data resources like the CDW in rare disease research.

IBM

Inclusion body myositis

IIM

Idiopathic inflammatory myopathy

ICD

International Classification of Disease

polymyositis

ENMC – European neuromuscular conference

Veterans Affairs

CDW

Corporate Data Warehouse

CPK

creatine phosphokinase

rheumatoid arthritis

EMG

electromyography

MRI

magnetic resonance imaging

sensitivity

specificity

PPV

positive predictive value

NPV

negative predictive value

MSA

Myositis specific antibody.

Data sharing statement

The data used in this study is maintained by the VA CDW and VINCI, is not property of the authors, and is not publicly available. Thus, we are unable to share it with others. Those wishing to access raw data may do so by following the policies laid out by their respective owners.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosures

VML is employed part-time as follows: Staff Rheumatologist, Department of Medicine, Department of Veterans Affairs Medical Center, Pittsburgh, PA.

Funding:

None

Acknowledgements

The author would like to thank Drs. Dana Ascherman, MD, Chester V. Oddis, MD, Rohit Aggarwal, MD, MS and Siamak Moghadam-Kia, MD, MPH as well as other members of the University of Pittsburgh Myositis Center for their input and feedback regarding this study and thorough review of the manuscript. In addition, I would also like to acknowledge Dr. Steuart Richards, MD, Chief of Pittsburgh VA Rheumatology, as well as Rachel Socrates, MA, Claire Raible, MPH, and members of the Pittsburgh VA IRB committee as well as CDW and VINCI staff who helped with this project. This material is the result of work supported with resources and the use of facilities at the VA Pittsburgh Healthcare System, Pittsburgh, PA.

Badrising UA, Maat-Schieman M, van Duinen SG et al (2000) Epidemiology of inclusion body myositis in the Netherlands: a nationwide study. Neurology Nov 14(9):1385–1387. 10.1212/wnl.55.9.1385
Phillips BA, Zilko PJ, Mastaglia FL (2000) Prevalence of sporadic inclusion body myositis in Western Australia. Muscle Nerve Jun 23(6):970–972. 10.1002/(sici)1097-4598(200006)23:6%3C970::aid-mus20%3E3.0.co;2-i
Needham M, Corbett A, Day T, Christiansen F, Fabian V, Mastaglia FL (2008) Prevalence of sporadic inclusion body myositis and factors contributing to delayed diagnosis. J Clin Neurosci Dec 15(12):1350–1353. 10.1016/j.jocn.2008.01.011
Felice KJ, North WA (2001) Inclusion body myositis in Connecticut: observations in 35 patients during an 8-year period. Medicine (Baltimore) Sep 80(5):320–327. 10.1097/00005792-200109000-00006
Keshishian A, Greenberg SA, Agashivala N, Baser O, Johnson K (Sep 2018) Health care costs and comorbidities for patients with inclusion body myositis. Curr Med Res Opin 34(9):1679–1685. 10.1080/03007995.2018.1486294
Control CD (2013) Prevention, Control CfD, Prevention. International classification of diseases, ninth revision, clinical modification (ICD-9-CM)
Statistics NCfH (2021) International classification of diseases, 10th revision, clinical modification (ICD-10-CM). Hyattsville, MD
Greenberg SA (2019) Inclusion body myositis: clinical features and pathogenesis. Nat Rev Rheumatol May 15(5):257–272. 10.1038/s41584-019-0186-x
Dimachkie MM, Barohn RJ (2014) Inclusion body myositis. Neurol Clin. Aug ;32(3):629–46, vii. 10.1016/j.ncl.2014.04.001
Dalakas MC (1991) Polymyositis, dermatomyositis and inclusion-body myositis. N Engl J Med Nov 21(21):1487–1498. 10.1056/NEJM199111213252107
Lilleker JB, Naddaf E, Saris CGJ et al (2024) 272nd ENMC international workshop: 10 Years of progress - revision of the ENMC 2013 diagnostic criteria for inclusion body myositis and clinical trial readiness. 16–18 June 2023, Hoofddorp, The Netherlands. Neuromuscul Disord. Apr ;37:36–51. 10.1016/j.nmd.2024.03.001
Rose MR, Group EIW (2013) 188th ENMC International Workshop: Inclusion Body Myositis, 2–4 December 2011, Naarden, The Netherlands. Neuromuscul Disord. Dec ;23(12):1044–55. 10.1016/j.nmd.2013.08.007
Amlani A, Choi MY, Tarnopolsky M et al (2019) Anti-NT5c1A Autoantibodies as Biomarkers in Inclusion Body Myositis. Front Immunol 10:745. 10.3389/fimmu.2019.00745
Ikenaga C, Findlay AR, Goyal NA et al (2021) Clinical utility of anti-cytosolic 5'-nucleotidase 1A antibody in idiopathic inflammatory myopathies. Ann Clin Transl Neurol Mar 8(3):571–578. 10.1002/acn3.51294
Griggs RC, Askanas V, DiMauro S et al (1995) Inclusion body myositis and myopathies. Ann Neurol Nov 38(5):705–713. 10.1002/ana.410380504
Lloyd TE, Mammen AL, Amato AA, Weiss MD, Needham M, Greenberg SA (2014) Evaluation and construction of diagnostic criteria for inclusion body myositis. Neurology 83(5):426–433
Liarski VM (2025) Retrospective analysis of US veterans with inclusion body myositis: initial findings from the Veterans Affairs Corporate Data Warehouse. Mil Med Res Jan 28(1):6. 10.1186/s40779-025-00592-5
Aguila LA, Lopes MRU, Pretti FZ et al (2014) Clinical and laboratory features of overlap syndromes of idiopathic inflammatory myopathies associated with systemic lupus erythematosus, systemic sclerosis, or rheumatoid arthritis. Clin Rheumatol 33(8):1093–1098
Martínez-Cordero E, Aguilar León DE, Andrade Ortega L (2001) Association of polymyositis with rheumatoid arthritis. Rheumatol Int 20(3):119–123
Mastaglia FL, Garlepp MJ, Phillips BA, Zilko PJ (2003) Inflammatory myopathies: clinical, diagnostic and therapeutic aspects. Muscle Nerve: Official J Am Association Electrodiagn Med 27(4):407–425
Halilu F, Christopher-Stine L (2022) Myositis-specific Antibodies: Overview and Clinical Utilization. Rheumatol Immunol Res Mar 3(1):1–10. 10.2478/rir-2022-0001
Basuita M, Fidler LM (2022) Myositis Antibodies and Interstitial Lung Disease. J Appl Lab Med Jan 5(1):240–258. 10.1093/jalm/jfab108
Dobloug GC, Antal EA, Sveberg L et al (2015) High prevalence of inclusion body myositis in Norway; a population-based clinical epidemiology study. Eur J Neurol Apr 22(4):672–e41. 10.1111/ene.12627
Shelly S, Mielke MM, Mandrekar J et al (2021) Epidemiology and Natural History of Inclusion Body Myositis: A 40-Year Population-Based Study. Neurology May 25(21):e2653–e2661. 10.1212/WNL.0000000000012004
Bernatsky S, Linehan T, Hanly JG (2011) The accuracy of administrative data diagnoses of systemic autoimmune rheumatic diseases. J Rheumatol Aug 38(8):1612–1616. 10.3899/jrheum.101149
Hannah JR, Gordon PA, Galloway J et al (2022) Validation of methods to identify people with idiopathic inflammatory myopathies using hospital episode statistics. Rheumatol Adv Pract 6(3):rkac102. 10.1093/rap/rkac102

The authors declare no competing interests.

Download PDF

Version 1

posted

You are reading this latest preprint version

Diagnostic accuracy of an algorithm identifying US veterans with Inclusion Body Myositis from the Corporate Data Warehouse

Status:

Version 1

Abstract

Objectives

Methods

Results

Conclusions

Figures

Key Points

Introduction

Patients and Methods

Data Sources and Setting

Human and Animal Rights

Individual Random Case Review

Statistical Analysis

Patient and Public Involvement

Results

Veterans with clinical diagnosis of IBM

Veterans meeting formal diagnostic criteria for IBM

Discussion

Abbreviations

Declarations

Data sharing statement

Funding

Disclosures

Funding:

Acknowledgements

References

Additional Declarations

Status:

Version 1