The IDeaS initiative: pilot study to assess the impact of rare diseases on patients and healthcare systems.

BACKGROUND: Rare diseases (RD) are a diverse collection of more than 7-10,000 different disorders, most of which affect a small number of people per disease. Because of their rarity and fragmentation of patients across thousands of different disorders, the medical needs of RD patients are not well recognized or quantified in healthcare systems (HCS). METHODOLOGY: We performed a pilot IDeaS study, where we attempted to quantify the number of RD patients and the direct medical costs of 14 representative RD within 4 different HCS databases and performed a preliminary analysis of the diagnostic journey for selected RD patients. RESULTS: The overall findings were notable for: (1) RD patients are difficult to quantify in HCS using ICD coding search criteria, which likely results in under-counting and under-estimation of their true impact to HCS; (2) per patient direct medical costs of RD are high, estimated to be around three-fivefold higher than age-matched controls; and (3) preliminary evidence shows that diagnostic journeys are likely prolonged in many patients, and may result in progressive, irreversible, and costly complications of their disease CONCLUSIONS: The results of this small pilot suggest that RD have high medical burdens to patients and HCS, and collectively represent a major impact to the public health. Machine-learning strategies applied to HCS databases and medical records using sentinel disease and patient characteristics may hold promise for faster and more accurate diagnosis for many RD patients and should be explored to help address the high unmet medical needs of RD patients.

Current and Future Projections of Amyotrophic Lateral Sclerosis in the United States Using Administrative Claims Data.

BACKGROUND: Various methodologies have been reported to assess the real-world epidemiology of amyotrophic lateral sclerosis (ALS) in the United States. The aim of this study was to estimate the prevalence, incidence, and geographical distribution of ALS using administrative claims data and to model future trends in ALS epidemiology. METHODS: We performed a retrospective analysis of deidentified administrative claims data for >100 million patients, using 2 separate databases (IBM MarketScan Research Databases and Symphony Health Integrated DataVerse [IDV]), to identify patients with ALS. We evaluated disease prevalence, annual incidence, age- and population-controlled geographical distribution, and expected future trends. RESULTS: From 2013 to 2017, we identified 7,316 and 35,208 ALS patients from the MarketScan databases and IDV, respectively. Average annual incidence estimates were 1.48 and 1.37 per 100,000 and point prevalence estimates were 6.85 and 5.16 per 100,000 and in the United States for the MarketScan databases and IDV, respectively. Predictive modeling estimates are reported out to the year 2060 and demonstrate an increasing trend in both incident and prevalent cases. CONCLUSIONS: This study provides incidence and prevalence estimates as well as geographical distribution for what the authors believe to be the largest ALS population studied to date. By using 2 separate administrative claims data sets, confidence in our estimates is increased. Future projections based on either database demonstrate an increase in ALS cases, which has also been seen in other large-scale ALS studies. These results can be used to help improve the allocation of healthcare resources in the future.

Use of machine learning techniques in the development and refinement of a predictive model for early diagnosis of ankylosing spondylitis.

OBJECTIVE: To develop a predictive mathematical model for the early identification of ankylosing spondylitis (AS) based on the medical and pharmacy claims history of patients with and without AS. METHODS: This retrospective study used claims data from Truven databases from January 2006 to September 2015 (Segment 1) and October 2015 to February 2018 (Segment 2). Machine learning identified features differentiating patients with AS from matched controls; selected features were used as inputs in developing Model A/B to identify patients likely to have AS. Model A/B was trained and developed in Segment 1, and patients predicted to have AS in Segment 1 were followed up in Segment 2 to evaluate the predictive capability of Model A/B. RESULTS: Of 228,471 patients in Segment 1 without any history of AS, Model A/B predicted 1923 patients to have AS. Ultimately, 1242 patients received an AS diagnosis in Segment 2; 120 of these were correctly predicted by Model A/B, yielding a positive predictive value (PPV) of 6.24%. The diagnostic accuracy of Model A/B compared favorably with that of a clinical model (PPV, 1.29%) that predicted AS based on spondyloarthritis features described in the Assessment of SpondyloArthritis international Society classification criteria. A simplified linear regression model created to test the operability of Model A/B yielded a lower PPV (2.55%). CONCLUSIONS: Model A/B performed better than a clinically based model in predicting a diagnosis of AS among patients in a large claims database; its use may contribute to early recognition of AS and a timely diagnosis.