Polysomnographic phenotype as a risk factor for cardiovascular diseases in patients with obstructive sleep apnea syndrome: a retrospective cohort study

Automatic detection of respiratory events during sleep from Polysomnography data using Layered Hidden Markov Model

OBJECTIVE

Sleep apnea is a serious respiratory disorder, which is associated with increased risk factors for cardiovascular disease. Many studies in recent years have been focused on automatic detection of sleep apnea from polysomnography (PSG) recordings, however, detection of subtle respiratory events named Respiratory Event Related Arousals (RERAs) that do not meet the criteria for apnea or hypopnea is still challenging. The objective of this study was to develop automatic detection of sleep apnea based on Hidden Markov Models (HMMs) which are probabilistic models with the ability to learn different dynamics of the real time-series such as clinical recordings.

APPROACH

In this study, a hierarchy of HMMs named Layered HMM was presented to detect respiratory events from PSG recordings. The recordings of 210 PSGs from Massachusetts General Hospital's database were used for this study. To develop detection algorithms, extracted feature signals from airflow, movements over the chest and abdomen, and oxygen saturation in blood (SaO2) were chosen as observations. The respiratory disturbance index (RDI) was estimated as the number of apneas, hypopneas, and RERAs per hour of sleep.

MAIN RESULTS

The best F1 score of the event by event detection algorithm was between 0.22±0.16 and 0.70±0.08 for different groups of sleep apnea severity. There was a strong correlation between the estimated and the PSG-derived RDI (R2=0.91, p<0.0001). The best recall of RERA detection was achieved 0.45±0.27.

SIGNIFICANCE

The results showed that the layered structure can improve the performance of the detection of respiratory events during sleep.

Severity of spleep apnea syndrome and life-threatening tachyarrhythmias in patients with implantable cardioverter defibrillator

BACKGROUND

Sleep apnea syndrome (SAS) has been reported to be associated with a higher incidence of ventricular arrhythmias. The aim of this study was twofold: (1) to investigate whether in SAS patients receiving an implantable cardioverter defibrillator (ICD) the severity of SAS was associated with the occurrence of ventricular arrhythmias; (2) to assess whether changes in nocturnal apnoic/hypopnoic episodes may favor the occurrence of life-threatening arrhythmias, that is, sustained ventricular tachycardia (VT)/fibrillation (VF), requiring ICD intervention.

METHODS

We enrolled 46 patients with documented SAS at polysomnography (apnea/hypopnea index [AHI] > 5) who also had a left ventricle ejection fraction (LVEF) < 35% and, according to primary prevention indications, implanted an ICD (Boston Scientific Incepta) able to daily monitor apnoic/hypopnoic episodes occurring during sleep. Patients were followed at 3-month intervals.

RESULTS

At a mean follow-up of 18 months, 21 episodes of sustained VT/FV requiring ICD intervention were documented in eight patients (17.4%). Baseline AHI was significantly higher in patients with compared to those without ICD intervention. ICD interventions, however, were not preceded by any worsening of apnoic/hypopnoic episodes. The respiratory disturbance index (RDI) of the week during the event, indeed, was not different from that recorded during the previous 2 weeks (25.4 ± 11, 25.6 ± 10 and 25.1 ± 10, respectively; p = .9).

CONCLUSIONS

In patients with SAS who received an ICD for primary prevention of sudden death, those with ICD interventions showed a more severe form of the disease at baseline. ICD interventions, however, were not preceded by any significant changes in SAS severity.

Transforming electronic health record polysomnographic data into the Observational Medical Outcome Partnership's Common Data Model: a pilot feasibility study

Well-defined large-volume polysomnographic (PSG) data can identify subgroups and predict outcomes of obstructive sleep apnea (OSA). However, current PSG data are scattered across numerous sleep laboratories and have different formats in the electronic health record (EHR). Hence, this study aimed to convert EHR PSG into a standardized data format-the Observational Medical Outcome Partnership (OMOP) common data model (CDM). We extracted the PSG data of a university hospital for the period from 2004 to 2019. We designed and implemented an extract-transform-load (ETL) process to transform PSG data into the OMOP CDM format and verified the data quality through expert evaluation. We converted the data of 11,797 sleep studies into CDM and added 632,841 measurements and 9,535 observations to the existing CDM database. Among 86 PSG parameters, 20 were mapped to CDM standard vocabulary and 66 could not be mapped; thus, new custom standard concepts were created. We validated the conversion and usefulness of PSG data through patient-level prediction analyses for the CDM data. We believe that this study represents the first CDM conversion of PSG. In the future, CDM transformation will enable network research in sleep medicine and will contribute to presenting more relevant clinical evidence.

Recognizable Clinical Subtypes of Obstructive Sleep Apnea After Ischemic Stroke: A Cluster Analysis

BACKGROUND AND PURPOSE

Obstructive sleep apnea (OSA) increases risk of stroke recurrence and mortality in ischemic stroke patients. However, equivocal treatment effects warrant further categorization of post-stroke OSA for risk stratification and individualized treatment planning.

METHODS

The study recruited 232 ischemic stroke patients with moderate-to-severe OSA admitted for inpatient rehabilitation consecutively. Latent class analysis was performed based on sex, age, smoking, daytime sleepiness, depression, obesity, sedative use, atrial fibrillation, diabetes, dyslipidemia, hypertension, recurrent stroke and dysphagia. The augmentation index, a marker of arterial stiffness, was measured by applanation tonometry.

RESULTS

A three-cluster model provided the best fit. Cluster 1 (n=84, 36.2%) was older in age, predominantly female, with the highest hypopnea index and prevalence of atrial fibrillation. Moreover, patients in Cluster 1 had significantly higher augmentation index than those in Cluster 2. Cluster 2 patients (N=80, 34.5%) were of older age, predominantly male, with the highest prevalence of depression, the lowest prevalence of hypertension and had the most normal body mass index (BMI). Additionally, Cluster 2 had less nocturnal hypoxia as compared to Cluster 3. Cluster 3 (n=68, 29.3%) was the youngest in age, predominantly male, with the highest BMI, cumulative risk score, and prevalence of dyslipidemia of the three clusters.

CONCLUSION

Post-stroke OSA can be categorized into three clinical phenotypes. Patients in Clusters 1 and 3 both had elevated cardiovascular risk and treatment can be based on their distinct characteristics. Patients in Cluster 2 had relatively lower risk of cardiovascular events and the benefits of OSA treatment requires further study.