A recent study showed that patients with isolated REM sleep behavior disorder had a high risk of neurodegenerative parkinsonian disorders or dementia based on electronic health records.
In a retrospective cohort study, a statistical model using electronic health records (EHR) and international classification of disorders (ICD) codes demonstrated moderate positive predictive value in identifying individuals with REM sleep behavior disorder (RBD) that have a high probability for actual isolated RBD (iRBD). Compared with the no iRBD group, the iRBD group had a higher risk of neurodegenerative parkinsonian disorders or dementia (NPD).
Overall, these findings support the feasibility of using statistical models that utilize EHR data to accurately predict iRBD in an outpatient setting.1 In the 1130 cases with RBD ICD code between 2011 and 2021, investigators reported that 499 of the records had secondary causes of RBD. For the remaining ones (n = 628), determination based on EHR review was that 168 (26.8%) of the cases did not have iRBD. Presented at the
Investigators conducted the EHR search at a single healthcare system to identify outpatient cases who received ICD9 or ICD10 RBD code. The data for each case was manually reviewed by the researchers. Authors excluded the cases with secondary causes of RBD and the remaining ones were classified as no iRBD or actual iRBD. They also identified incident cases of neurodegenerative parkinsonian disorders or dementia and calculated the positive predictive value of presence of RBD ICD code. The researchers compared cumulative incidence of neurodegenerative parkinsonian disorders or dementia with death as a competing event in those with versus without iRBD. Furthermore, authors used the least absolute shrinkage and selection operator (LASSO) regression to build a prediction model for iRBD and had the model validated in an independent dataset.
READ MORE:
Additional findings showed that the iRBD group had a higher risk of neurodegenerative parkinsonian disorders or dementia (subdistribution HR, 10.4; 95% CI, 2.5-43.1) compared with the no iRBD group. Notably, authors reported that the LASSSO prediction model for iRBD had an Area Under the Receiver Operating Characteristic Curve of 0.844 (95% CI, 0.806-0.880). Prior research shows that iRBD carries an increased risk for neurodegenerative parkinsonian disorders or dementia, but clinicians have had difficulty with accurately screening for it in the patient community. Thus, healthcare data provided by EHR may offer the opportunity to identify large numbers of cases in iRBD among outpatients.1
According to a previous review study published in the Journal of Neurology, results showed a substantial delay before neurologist are consulted about iRBD.1 The literature also revealed that patients were unaware of their nocturnal signs and the symptoms described could mimic or masquerade other sleep disorders. In addition, general providers did not recognize the significance of the symptoms or consider the diagnosis of iRBD since it was considered a lesser-known condition. Therefore, authors of the review recommended that neurologists and other clinicians diagnosing iRBD should conduct questionnaires to help them make a more "probable diagnosis" but this approach should not be used as the only diagnostic tool for iRBD.
As patients live longer worldwide, the number of those at risk of neurodegenerative diseases is to dramatically increase. Thus, the detection of iRBD presents an important opportunity for clinicians to intervene early in the condition, through therapies or risk modification, to adjust its trajectory or even prevent the development of neurodegenerative diseases in the future. Authors from the prior review noted that there is an increasing interest in pharmaceutical clinical trials for iRBD that is further increasing the criticalness of diagnosis and rasing awareness of the key role neurologists play to accurately diagnosis this disorder early on.
