Citations

Abstract Introduction Although diastolic heart failure (HFpEF) patients frequently have comorbid obstructive sleep apnea (OSA), which has been associated with poor prognosis, the impact of positive airway pressure (PAP) therapy remains unclear. This retrospective study sought to investigate the association between 1 year adherence to PAP therapy and healthcare resource utilization (HCRU) in patients with newly diagnosed OSA and comorbid HFpEF. Methods A national sample of administrative claims data linked to objective PAP therapy usage was used for this analysis. Eligible patients either had two healthcare encounters or one hospitalization with a diastolic heart failure ICD-10 diagnosis code the year prior to being diagnosed with OSA and initiated on PAP therapy. HCRU was measured as a composite outcome of hospitalizations and emergency room (ER) visits. Propensity score matching was used to create well-balanced groups with differing PAP adherence levels to assess the association with 1 year HCRU outcomes. Risk-adjusted models were used to generate the number needed to treat (NNT) that referred to the number of patients that need to become adherent to PAP from non-adherent to avoid 1 visit. Results 4,237 patients with OSA and comorbid HFpEF were included. With propensity score matching, 963 PAP adherent and 963 non-adherent patients remained. After 1 year of PAP therapy, adherent patients had fewer visits than non-adherent patients (1.16 visits per subject vs. 1.75, p < 0.001), driven by a 57% reduction in hospitalizations and a 36% reduction in ER visits for adherent patients. Total HCRU costs were lower in adherent patients than in non-adherent patients ($12,732 vs. $15,610, p< 0.01). Converting 1 patient from non-adherent to adherent would avoid 1.25 visits (either ER visit or hospitalization), with an NNT of 0.8. Conclusion HFpEF patients with OSA that are adherent to PAP therapy for 1 year have fewer hospitalizations and ER visits than patients that are not adherent. More importantly, HCRU costs were lower in adherent patients, highlighting the clinical and economic benefits of treating OSA with PAP therapy in HFpEF patients. Support (if any) Funding Source: ResMed

Abstract Introduction In-laboratory polysomnography (PSG) is the gold standard diagnostic test for sleep-disordered breathing (SDB). WatchPAT® is a type of home sleep apnea testing (HSAT) that utilizes peripheral arterial tone (PAT) signals and other sensors to evaluate for SDB. There is an abundance of data to support the use of a PAT-based HSAT for obstructive sleep apnea (OSA), though data to support its use in central sleep apnea (CSA) in the general population undergoing PSG is lacking. We therefore aimed to compare the agreement of a PAT-based HSAT and PSG in detecting central disordered breathing events in the general population undergoing PSG. Methods A prospective cross-sectional study was conducted in eligible adult participants with suspected sleep apnea who were scheduled to undergo diagnostic or split-night PSG at the Center for Sleep Medicine at Mayo Clinic in Rochester, MN, between May 2021 and October 2022. The PAT-based HSAT was applied during the night of PSG to obtain simultaneous recording. Subjects were excluded if they had chronic atrial fibrillation, permanent pacemaker, or finger deformity precluding finger probe use or if they were on alpha- antagonists or short-acting nitrates. The PAT-based HSAT data were automatically scored with the use of proprietary software, and the PSG data were manually scored by registered polysomnography technologists. All studies were reviewed by board-certified sleep medicine providers. Results A total of 147 subjects (53% female) were included in the final analysis after excluding 18 subjects due to malfunction of the PAT-based HSAT. The subjects had a mean age of 58±15 yrs and BMI of 34±9 kg/m². 95 (65%) subjects underwent split-night studies. OSA was diagnosed in 81% with a mean PSG AHI of 28±27 events per hour. CSA was diagnosed in 3% of subjects with a mean PSG AHI of 33±27 events per hour. Of the total 2120 central disordered breathing events reported by the PAT-based HSAT (mean pAHIc 3±5 events per hour), 365 (17%) correlated with central apneas and 78 (3.7%) correlated with central hypopneas found on PSG. Conclusion In our prospective study, the PAT-based HSAT demonstrated low positive predictive value in detecting central disordered breathing events when compared to PSG. Support (if any)

Abstract Introduction Interstitial lung disease (ILD) is a heterogenous group of parenchymal lung disorders with multiple etiologies. ILD often affects the quality of life and survival of patients. OSA is commonly diagnosed among ILD patients with reported prevalences of 50-90%, and a prevalence of moderate to severe OSA as high as 65%. There is also concern that patients with ILD may be less tolerant of PAP therapies. We aim to identify differences in PAP treatment response and adherence in patients with ILD and OSA compared with a group of matched OSA controls. Methods A retrospective cohort analysis was conducted on patients ≥ 18 years-old that were evaluated at Mayo Clinic Center for Sleep Medicine between July 2012 to June 2022. Inclusion criteria were patients with a diagnosis of ILD per ATS criteria also diagnosed with OSA by HSAT or PSG. Control OSA patients were matched to ILD/OSA patients by age, BMI, and AHI ranges. Exclusion criteria included comorbid conditions of asthma or COPD. Patients with early non-adherence to CPAP or those who didn't have sleep follow up were also excluded. Follow-up sleep visits were assessed to determine CPAP adherence, change in PAP pressure, addition of oxygen, and change in Epworth sleepiness scale. Results We compared 51 patients with ILD/OSA and 51 OSA matched-controls. Twenty-five ILD/OSA patients were diagnosed by PSG and 26 by HSAT. Median treatment follow-up for the ILD group was 7 months (IQR of 3 to 18 months). Baseline characteristics, including oxygen desaturation index (ODI), oxygen saturation, time spent below 89% during overnight oximetry, and baseline Epworth scale, were similar between ILD/OSA and non-ILD OSA patients. Importantly, there was no statistically significant difference in PAP adherence, median PAP levels, PAP pressure change, residual AHI, or change in Epworth after treatment Conclusion There was no significant difference in baseline OSA characteristics, treatment response, and PAP adherence in ILD patients with OSA compared with matched non-ILD OSA controls. Outcome assessment may be limited by sample size and the shorter follow-up period. Future studies should include longer-term follow-up with a larger ILD cohort to assess the impact of ILD progression on PAP adherence over time. Support (if any)

Abstract Introduction People with heart failure (HF) have frequent hospitalizations, and emergency department (ED) visits. We examined the effects of insomnia and rest-activity rhythms (RARs) on time to hospitalizations and ED visits after controlling for other known risk factors among participants in a randomized controlled trial of cognitive behavioral therapy for insomnia compared to heart failure self-management education (attention control). Methods The analysis included 168 HF patients who had available medical records for dates of hospitalization, ED visit, transplant, and death from a randomized clinical trial for insomnia of people with HF and insomnia. The known risk factors included age, sex, New York Heart Association (NYHA) Class, ejection fraction, and comorbid conditions. We used the insomnia severity index (ISI) and evaluated the strength of RAR using the ratio of the amplitude to the MESOR estimated with cosinor analysis with activity counts from a wrist-actigraph worn for at least five consecutive days. Cox proportional hazard models were used to examine ­insomnia and the RAR metrics with the known risk factors on time to first hospitalization, ED visit, transplant, and death. Results Eighty-five (50.1%) had at least one hospitalization, and ninety-one (54.2%) participants had an ED visit. Approximately 42% of the participants were admitted to the hospital due to cardiac events. Older age, male sex, higher NYHA, and greater comorbidity were associated with earlier hospitalization or ED visits. After controlling for those risk factors, stronger RAR was associated with later time to hospitalization, and ED visit with hazard ratios (HRs) of 0.26 (95% CI=[0.07, 0.91]) and 0.25 (95% CI=[0.07, 0.89]); higher insomnia severity (ISI ≥22) was associated with shorter time to ED visits, with an HR of 2.28 (95% CI=[1.19, 4.38]). Conclusion In addition to known demographic and clinical risk factors, insomnia severity and RAR predict hospitalizations and ED visits in people with HF. Further research is necessary to assess the biological mechanism of the associations and examine whether the effects of the risk factors could be modified by improving insomnia severity and RAR in people with HF. Support (if any) This study was financially supported by grants R01NR016191 and P20NR014126 from NIH/NINR.

Abstract Introduction Studies that have resulted in recent approval of hypnotics for chronic insomnia disorder (CID) have extremely restrictive inclusion criteria. We sought to determine how many patients seen in a sleep medicine clinic would be eligible to participate in these trials in order to assess the generalizability of these findings to clinical populations. Methods Six-hundred and twenty-eight patients were screened to identify 100 patients who were ≥18 years old, had a diagnosis of CID, and were prescribed a hypnotic medication at a sleep medicine clinic in a quaternary level healthcare center. Information regarding age, sex, body mass index, comorbid diagnoses, sleep schedule, medications, alcohol/tobacco/illicit drug use, caffeine intake, as well as Epworth Sleepiness Scale, Generalized Anxiety Disorder-7, Patient Health Questionnaire-9, and Insomnia Severity Index scores were abstracted. These data were compared against the inclusion and exclusion criteria for recent regulatory Phase III trials of orexin receptor antagonists (suvorexant, lemborexant, daridorexant). Results Of the 100 patients with CID (mean age=56.2 years; 57% female), only 3 patients met eligibility criteria for inclusion. A majority of patients (80%) would have been excluded due to a prior/concurrent trial of cognitive behavioral therapy for insomnia. If this criterion was set aside, only 11 patients would have been eligible to participate. The remaining top reasons for potential exclusion from the regulatory trials were: presence of medical comorbidities (36%), daytime napping (33%), and sleep apnea (30%). Conclusion A very small number of patients with CID seen in a clinical setting met eligibility criteria for clinical trials of hypnotic medications resulting in regulatory approval. This substantially limits the generalizability of findings from these trials and restricts our ability to determine the efficacy of these drugs in clinical populations. Future trials should enroll patients with less restrictive criteria to help determine the effectiveness of these medications in real world settings. Support (if any)

Abstract Introduction Cognitive behavioral therapy for insomnia (CBT-I) improves individual symptoms of fatigue and daytime sleepiness; however, little is known how CBT-I improves multiple co-occurring daytime symptoms. The purposes of this study are to (1) identify daytime symptom trajectories in adults with insomnia and stable chronic heart failure (HF) over one year, and (2) determine how symptom trajectories vary based on participation in CBT-I or HF self-management education (attention control condition). Methods We conducted a retrospective analysis of data from a randomized controlled trial that looked the sustained effects of CBT-I compared to HF self-management over one year (NCT02660385). We measured daytime symptoms (fatigue, depression, anxiety, dyspnea, pain, and daytime sleepiness) at baseline (pre-treatment) and 3, 6 and 12 months. Symptom cluster trajectory were identified using group-based trajectory modeling, and the association with CBT-I was examined using logistic regressions. Results We randomized 175 participants (M age=63.0 (12.9) years, 57% men). We found four daytime symptom cluster trajectories, (A) low improving symptoms (N = 67, 38.3%); (B) low psychological symptoms and high improving physical symptoms (N = 40, 22.8%); (C), high improving symptoms (N = 42, 24.0%); and (D) high not improving symptoms (N = 26, 14.9%). There was no statistically significant difference between CBT-I and HF self-management across 4 symptom trajectories (p = .2509). However, people who received CBT-I vs. HF self-management education had higher odds of belonging to trajectory A compared to B/C/D after controlling for baseline fatigue (OR = 3.27 95% Confidence Intervals [1.39, 7.68]). Conclusion Most participants experienced improved daytime symptom profiles over 12 months after both interventions. Participants in the CBT-I group experienced symptom improvement even when they started with low symptoms. Support (if any) National Institute of Nursing Research [R01NR016191 and P20NR014126].

Abstract Introduction Nightshifts represent a substantial proportion of time dedicated to working in the inpatient setting, particularly among ICU staff. Our institution implemented a 24/7 in-person ICU attending model more than a decade ago. With practice growth and more than 200 ICU beds across disciplines, there are now at least 8 ICU attendings in-hospital all night. Yet, little is known about attending staff perceptions and desires about night work, so we began an exploratory analysis. Methods An online survey of 13 multiple-choice, slider scale, and open-ended questions were sent via email to critical care staff physicians of different disciplines within our institution. Results 69 out of 150 physicians completed the survey (46% response rate). Half were between 35 and 45 years old; 29% were age 45 to 55 years old. About half self-identified as a morningness chronotype. Most (84%) work 3- or 4-night shifts consecutively. Most viewed night shift work as having a negative influence on their quality of life (48% “somewhat negative” and 22% “very negative”) and on their health (51% “somewhat negative” and 24% “very negative”). Most would choose to work less nightshifts (37%) or none (29%). During nightshift assignments, about half (53%) self-report an average of 5 hours of sleep per 24-hour period; 18% self-report ≤ 4hours. Only 6% reported 7 or 8 hours of sleep. Mitigating strategies utilized during series of night shifts include use of caffeine (78%), naps (66%) and use of sleeping aids (21%). Outside of the nightshift series, more than half (56%) reported no sleep difficulties while 21% reported onset or maintenance insomnia. Conclusion At our large tertiary hospital employing an in-house nightshift model across multiple ICUs, most staff ICU physicians have a negative perception of night shift work on their quality of life and health. The majority of respondents would choose to work less or no night shifts. How self-reports of limited sleep (average 5 hours) per 24-hour period compare to objectively measured sleep duration in this population remains to be determined in the next phase of this study. Support (if any)

Background: According to Tonnessen BH (2011), 1 iatrogenic arteriovenous (AV) fistulas in adults most commonly occur due to endovascular access and procedures. Rarely, AV fistulas have been reported in low birth weight neonates following repeating venipuncture. This complication is extremely uncommon in adults, but has been reported after routine venipuncture for blood transfusion. Case presentation: We report the case of an elderly female patient who presented to the office for evaluation of left upper extremity swelling, ecchymosis, and dilated vessels after routine venipuncture at an outpatient laboratory. She was subsequently found to have an acquired AV fistula from her left cephalic vein to a small branch artery. Conclusion: This case demonstrates the rare but relevant risk in routine venipuncture and may underscore the benefit of using ultrasound guidance in high-risk populations, such as patients with coagulopathies, or thin, fragile veins, like the elderly or neonates.

AbstractObjectiveThe purpose of this study was to assess differences in reported information about treatment, integration into care, and respect by self‐identified Black and White individuals with advanced prostate cancer in the United States.Patients and MethodsThis is a prospective cohort study of 701 participants (20% identifying as Black) enrolled in the International Registry for Men with Advanced Prostate Cancer at 37 US sites from 2017 to 2022. Participants were asked six questions from the Cancer Australia National Cancer Control Indicators about their experience with care at study enrollment. Prevalence differences by self‐reported race were estimated using marginal standardization of logistic‐normal mixed effects models (adjusted for age at enrollment and disease state at enrollment), and 95% CIs were estimated using parametric bootstrapping.ResultsMost participants reported a high quality of care for each question. Black participants generally reported higher care quality compared with White participants. Black participants reported more frequently that they were offered a written assessment and care plan (71%) compared with White participants (58%; adjusted difference, 13 percentage points; 95% CI, 4–23). Black participants also reported more frequently being given the name of nonphysician personnel who would support them (64%) than White participants (52%; adjusted difference, 10; 95% CI, 1–20). Prevalence differences did not differ by disease state at enrollment.ConclusionsBlack participants generally reported a higher quality of care compared with White participants. This study calls attention to the need to study potential mediating factors and interpersonal aspects of care in this population to improve survivorship.

Abstract Machine learning is a powerful tool that is increasingly being used in many research areas, including neuroscience. The recent development of new algorithms and network architectures, especially in the field of deep learning, has made machine learning models more reliable and accurate and useful for the biomedical research sector. By minimizing the effort necessary to extract valuable features from datasets, they can be used to find trends in data automatically and make predictions about future data, thereby improving the reproducibility and efficiency of research. One application is the automatic evaluation of micrograph images, which is of great value in neuroscience research. While the development of novel models has enabled numerous new research applications, the barrier to use these new algorithms has also decreased by the integration of deep learning models into known applications such as microscopy image viewers. For researchers unfamiliar with machine learning algorithms, the steep learning curve can hinder the successful implementation of these methods into their workflows. This review explores the use of machine learning in neuroscience, including its potential applications and limitations, and provides some guidance on how to select a fitting framework to use in real-life research projects.