Abstract
Background
Obstructive sleep apnea (OSA) is associated with increased mortality and readmissions in patients with heart failure (HF). The effect of in-hospital diagnosis and treatment of OSA during decompensated HF episodes remains unknown.
Methods
A single site randomized controlled trial in hospitalized patients with decompensated HF (n=150) who were diagnosed with OSA during the hospitalization. All participants received guideline directed therapy for HF decompensation. Participants were randomized to an intervention arm which received positive airway pressure (PAP) therapy during the hospitalization (n=75) and a control arm (n=75). The primary outcome was discharge left ventricular ejection fraction (LVEF).
Results
LVEF changed in the PAP arm from 25.5 (10.4) at baseline, to 27.3 (11.9) at discharge. In the control group, LVEF was 27.3 (11.7) at baseline and 28.8 (10.5) at conclusion. There was no significant effect on LVEF of in-hospital PAP compared to control (p = 0.84) in the intention to treat analysis. On treatment analysis in the intervention arm showed a significant increase in LVEF in participants who used PAP ≥ 3 hours per night (n=36, 48%) compared to those who used it less (p=0.01). There was a dose effect with higher hours of usage associated with more improvement in LVEF. Follow-up of readmissions after 6-months post discharge revealed a >60% decrease in readmissions for patients who used PAP ≥3 hours/night compared to those who used it less than 3 hours/night (p<0.02) and compared to controls (p<0.04).
Conclusion
In-hospital treatment with PAP was safe but did not significantly improve discharge LVEF in patients with decompensated HF and newly diagnosed OSA. Exploratory analysis showed that adequate use of PAP was associated with higher discharge LVEF and decreased 6 months readmissions.
INTRODUCTION
Heart failure (HF) is a leading cause of morbidity and mortality(1). A significant portion of the human and economic burden of HF is associated with hospitalizations due to clinical decompensation(2). We previously found that patients hospitalized with acutely decompensated heart failure (ADHF) have a high prevalence of previously undiagnosed and untreated moderate to severe obstructive sleep apnea (OSA)(3). Furthermore, in these ADHF patients, previously undiagnosed OSA is an independent predictor of post-discharge mortality(4) and readmissions(5). It remains unknown whether immediate identification and treatment of OSA during ADHF hospitalization could have a positive impact on the outcome of heart failure hospitalization(6).
Long term treatment of OSA with positive airway pressure (PAP) in stable patients with heart failure with reduced ejection fraction (HFrEF) was associated with improved sympathetic nerve activity, and cardiac afterload resulting in improved systolic function(7–9). During hospitalizations for ADHF, PAP may be used for the treatment of pulmonary edema and respiratory failure regardless of OSA status(10, 11). However, PAP is not currently part of the standard therapy for decompensated heart failure(12). Furthermore, routine surveillance for OSA in asymptomatic patients with chronic or acute heart failure is not part of the current standards of practice(13). In terms of ADHF, there has been only one published small randomized controlled trial (RCT) evaluating the identification and treatment of OSA during ADHF episodes and it showed that in-hospital treatment was safe and resulted in a small improvement in discharge left ventricular ejection fraction (LVEF)(14). On this background, we sought to evaluate a systematic approach to the identification and treatment of OSA during hospitalizations with ADHF in an adequately powered trial. We hypothesized that treatment of newly diagnosed OSA in hospitalized patients with ADHF could improve cardiac function at hospital discharge relative to the current standard of care.
METHODS
Study Design and Oversight
This was a single center randomized controlled trial evaluating the effect of a management approach that includes diagnosing and treating OSA immediately during ADHF hospitalizations on discharge LVEF. The trial was conducted at the Ohio State University (OSU) and was sponsored by the National Heart, Lung, and Blood Institute (NHLBI). A data safety monitoring plan (DSMP) was established and conducted in accordance with the relevant NIH guidelines and the OSU regulations and policies. The protocol was approved by the OSU institutional review board (IRB) (IRB 2008H0011) and registered in the national clinical trials database (NCT00679549). All participants provided informed consent. A consort checklist is included in the supplement.
Participants
Patients with HFrEF who were hospitalized for ADHF were eligible for participation. ADHF was defined as a chief complaint of dyspnea and elevated left ventricular pressure, as indicated by at least one sign and one symptom of volume overload (i.e., crackles, consistent chest radiograph, increased left ventricular end-diastolic diameter, elevated B type natriuretic peptide (BNP) levels). Only patients with LVEF of ≤ 45% on an echocardiogram performed during the admission were eligible for enrollment.
In order to be eligible for randomization, patients had to have a projected length of stay of ≥ 3 days on the morning following the sleep study. This was based on our pilot trial using a similar design demonstrating an effect of PAP on LVEF within 3 days (14). Full inclusion and exclusion criteria are detailed in the supplement. The exclusion criteria included the following:
Patients who were already using PAP for sleep-disordered breathing (SDB);
Patients with predominantly central sleep apnea (CSA) as defined by central apnea index (CAI)> 50% of the Apnea index (AI) on the inpatient sleep study;
Hemodynamic instability, defined as a mean arterial BP < 55 mm Hg while not receiving vasopressors, or concurrent use of vasopressors;
Respiratory insufficiency, defined by PaO2/FIO2 ratio < 250 mm Hg; and
Renal failure requiring renal replacement therapy.
Interventions
Sleep studies
Hospitalized ADHF patients underwent attended sleep studies (Stardust II, Respironics, Inc.) that measure respiratory efforts, oxygen saturation, nasal flow and pulse rate. We have already reported on our validation of this method of in-hospital sleep testing in ADHF patients and its high positive predictive value (>90%) for OSA(3) and CSA(5) against polysomnography done in the post-discharge stable outpatient state. Other studies also support a similar predictive value for cardiorespiratory polygraphy for the diagnosis of OSA in patients with HF(15–17). Trained night shift nurses attended the in-hospital sleep studies and noted visual sleep and wakefulness, and any interruptions to sleep improving the accuracy of the test. Respiratory event scoring and classification of OSA and CSA were according to standard clinical criteria(18) using an oxygen desaturation cut off of 3% for hypopneas. OSA was defined as AHI>15 events per hour with >50% of the apneas being obstructive. Details of the sleep study protocol and criteria are provided in the supplement and a previous publication (19).
Intervention arm: In-hospital Treatment of OSA
Participants who were randomized to the intervention arm received nocturnal auto-adjusting PAP (APAP) [BIPAP auto; Respironics], with an algorithm that targets only upper airway obstruction. The treatment was started on the night following randomization and continued past hospital discharge. The device pressure was set at 5 to 15 cm H2O on the first night and was subsequently fixed on the second night based on review of the device download. The difference between inspiratory and expiratory pressures was always kept at < 4 cm H2O. The study coordinator visited patients daily to review the mask fitting and download the data from the treatment device. Daily review of the device download and adjustment of the settings continued throughout the hospitalization to ensure effectiveness and safety of treatment.
Control arm: Standard of care
Patients randomized to the control arm received a 30-minute education session on sleep hygiene and OSA. The coordinator visited patients in the control arm daily to record their symptoms and vitals and reinforce the importance of follow up. All patients in this arm were provided outpatient follow up at the OSU sleep clinic according the standards of practice. All patients in the control arm received guideline directed medical therapy for ADHF.
Enrollment and Randomization
Randomization was administered using a computer software designed by the OSU-Center for Biostatistics specifically for the trial. The research coordinator received the assigned randomization from the server after entering baseline characteristics and performing eligibility checks on the participant.
Study Outcomes
Demographics and baseline outcome measures were collected prior to randomization. The primary outcome measure was LVEF as evaluated by echocardiography and determined using the biplane Simpson method. Echocardiography was performed at baseline before randomization (day 0) and on the morning 3 nights after randomization (allowing for 3 nights of APAP in the treatment group as previously demonstrated (14). A secondary outcome measure was based on PAP’s effect on sympathetic tone. For this, we obtained plasma norepinephrine in the morning of the third day post randomization.
Blinding and Data Safety
Demographics and laboratory values were entered using the double-entry method by two data assistants blinded to participants’ assignment. Echocardiographic recordings were anonymized and interpreted in batches stripped of any time stamp after closure of the study. Multiple layers of blinding and data security are described in the supplement
Power and Sample Size
The targeted sample size was 75 patients per randomization group. This allowed for 10% missing outcome data. With this number, the primary test for PAP’s effect on the LVEF outcome would have greater than 84% power to detect a treatment effect size of 0.5 standard deviations, using a single--sided test in the expected direction at alpha=0.025. We use 0.025 for the single-sided test to make the test as conservative as a double sided test at 0.05. Effect sizes from prior trials of PAP ranged from 0.5 to 0.7 standard deviations on this outcome(9, 14).
Statistical Analysis
The primary comparisons between the two groups were conducted using the intention-to-treat principle. A closed testing gatekeeping approach was planned for hypothesis testing a primary and then a secondary outcome (14). In this approach the secondary outcome, norepinephrine, can be tested formally only when then the primary, LVEF, is significant. If the LVEF null hypothesis is rejected then norepinephrine is tested at α=0.025 also, preserving the familywise type I error rate for these two outcomes at α=0.025 (20).
The primary hypothesis test for PAP’s effect on LVEF was conducted using analysis of covariance, with the baseline LVEF value as a covariate to increase precision of the treatment effect estimate. Missing data were assumed to be missing at random conditional on baseline data (MAR) for the primary analysis. As a sensitivity analysis, multiple imputation using the fully conditional specification method was performed in SAS Proc MI as an alternative way of accounting for missing data. All patient characteristics reported in Table 1, along with baseline and conclusion LVEF, were included in the multiple imputation model.
Table 1.
Baseline Patient Characteristics
| Characteristic (Day 0) | Treatment (n=75) | Control (n=75) |
|---|---|---|
| Age | 57.2 (11.3) | 56.8 (13.5) |
| Sex (male) | 53 (70.7%) | 54 (72.0%) |
| African American | 18 (24.0%) | 17 (22.7%) |
| BMI | 31.4 (7.2) | 32.4 (9.0) |
| SBP | 118.8 (20.5) | 112.7 (20.6) |
| DBP | 69.7 (14.4) | 64.6 (12.0) |
| Cardiomyopathy type Ischemic | 40 (53.3%) | 47 (62.7%) |
| Non-Ischemic | 35 (46.7%) | 28 (37.3%) |
| CAD | 39 (52.0%) | 46 (61.3%) |
| Dyslipidemia | 42 (56.0%) | 43 (57.3%) |
| Hypertension | 51 (68.0%) | 53 (71.6%) |
| ICD | 35 (47.3%) | 33 (44.0%) |
| Atrial fibrillation | 30 (40.0%) | 36 (48.7%) |
| Baseline right ventricular dysfunction | 5 (6.7%) | 5 (6.9%) |
| COPD | 9 (12.0%) | 12 (16.0%) |
| Diabetes | 38 (50.7%) | 36 (48.0%) |
| Smoking | 23 (31.1%) | 31 (41.3%) |
| Admission Beta-blockers | 63 (84.0%) | 66 (88.0%) |
| Admission ACE-I | 51 (68%) | 51 (68%) |
| AHI events/hour | 41.0 (21.4) | 37.7 (16.8) |
| OAI events/hour | 21.6 (14.5) | 18.8 (13.3) |
| CAI events/hour | 6.4 (8.4) | 5.4 (8.9) |
| HI events/hour | 14.1 (11.8) | 14.6 (11.0) |
| LOS, median (IQR) | 8 (5–11) | 8 (6–13) |
| Epworth Sleepiness Scale | 10.3 (4.9) | 9.7 (4.8) |
Values are mean (SD) or n (%) unless otherwise noted; BMI: Body mass index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; HR: Heart rate; CAD: coronary artery disease; MI: myocardial infarction; CABG: history of coronary artery bypass grafting; ICD: implantable cardioverter-defibrillator; PHTN: COPD: chronic obstructive pulmonary disease; ACE-I; angiotensin converting enzyme inhibitors; BUN: blood urea nitrogen; AHI: apnea hypopnea index; OAI: obstructive apnea index; CAI: Central apnea index; HI: hypopnea index; LOS: Length of stay. There were no differences in any of the baseline characteristics.
On-treatment analysis
Additional planned analyses examined the relationship between PAP dose measured by usage time and the change in LVEF. Two PAP usage cutoffs were defined and tested for association with increased LVEF: ≥ 3 hours/night and ≥ 4 hours/ night. These cut-offs were chosen as representative of minimally adequate usage and correspond to the average usage found in recent large RCTs evaluating PAP in outpatients with cardiovascular disease or heart failure (21, 22). Two sets of comparisons were made: 1) patients who used the device ≥ cutoff (3 or 4 hours), vs. those who used < cutoff (3 or 4 hours) in the Intervention PAP arm and 2) treatment arm patients who used PAP ≥ cutoff vs. the combined group of treatment patients who used <cutoff and the control arm. Because these usage groups were not determined by randomization, potential imbalance in baseline characteristics were accounted for in the planned exploratory analyses. Inverse probability of treatment weighting (IPTW) using propensity scores was used to estimate the difference between usage groups. Propensity scores were estimated using logistic regression models with baseline values for AHI, LVEF, age, BMI, gender, CHF, CAD, hypertension, arrhythmia, diabetes, BUN, and BNP. Missing values at baseline were accounted for using multiple imputation methods. The analysis of covariance models with baseline LVEF covariate were then repeated incorporating these weights. Note that in all our exploratory analysis, double sided p-values are reported.
Post-discharge readmissions analysis
The effect of PAP treatment in the immediate post-discharge period on 6-months readmission was evaluated in an exploratory analysis. A 6-months readmissions period was chosen based on our previous study demonstrating a relation between OSA and 6 months readmissions in similar ADHF patients(5). Rates of readmissions within 6 months were analyzed using negative binomial regression on readmission counts with a follow-up time offset as we previously reported(5). Intention-to-treat analysis was performed comparing randomized groups in patients with the prespecified follow-up time available. Readmission rates were also analyzed in association with PAP usage based on a 3 months post-discharge device download, using the 3 and 4-hour thresholds discussed above. Similar to the LVEF analyses, propensity score based IPTW weights were applied to the negative binomial models for readmissions. Separate models were fit using those above the threshold versus those below within the treatment arm and again with controls combined in a group with non-adherent patients from the treatment group. Additional details are provided in the supplement.
RESULTS
Characteristics of Participants
Figure 1 details the disposition of all study participants. The trial enrolled 150 patients; 75 randomized into each arm. The most common reason for not meeting the inclusion criteria in patients with ADHF who had OSA was an LVEF ≥ 45% or a projected length of hospital stay of < 3 nights on the day after the sleep study (Fig 1 and supplement).
Figure 1: Flow Diagram of All Study Participants.
ADHF: Acutely decompensated heart failure; CSA: central sleep apnea; ESRD: End stage renal disease; GDMT: Guideline directed medical therapy; HD: Hemodialysis; LVAD: Left ventricular assist device; OSA: Obstructive sleep apnea; SDB: Sleep disordered breathing
Respiratory and Treatment Efficacy Parameters in the Intervention Arm
Mean (SD) usage of the treatment device during the hospitalization was 3.6 (2.5) hours per night. Average IPAP was 10.3 (2.3); average EPAP was 7.8 (2.4). Mean (SD) usage during the hospitalization was 3.6 (2.5) hours per night (range = 0.1 to 11.9). Efficacy of PAP was confirmed by an average treatment AHI of 7.85 (5.8) (n=54) events per hour on the day of the conclusion echocardiogram (>80% decrease from baseline AHI (41.0 (21.4)).
Effect of In-Hospital Treatment of OSA on LVEF
Intention to treat analysis
In the Intervention PAP arm, mean (SD) LVEF was 25.5 (10.6) at baseline and 27.3 (11.9) (n=69) at discharge. In the control group, mean (SD) LVEF was 27.3 (11.3) at baseline and 28.8 (10.5) (n=66) at conclusion. See Table 2. There was no significant difference in the LVEF change in this primary analysis (treatment effect = 0.27, p=0.42). The 95% CI on the PAP effect was (−2.35, 2.89), which excludes the effect size we were expecting (about 5 on the LVEF scale).. Sensitivity analysis using multiple imputations for missing data bias also resulted in a non-significant treatment effect (p=0.45). Note that the 10% missing data was exactly as expected in the power analysis
Table 2.
Effect of in-hospital PAP treatment on discharge LVEF: Intention to Treat Analysis
| Time | Treatment | Control | ||
|---|---|---|---|---|
| LVEF | n | LVEF | n | |
| Baseline | 25.5 (10.4) | 75 | 27.3 (11.7) | 75 |
| Inpatient Conclusion | 27.3 (11.9) | 69 | 28.8 (10.5) | 66 |
LVEF: Left ventricular ejection fraction; Mean (SD)
On treatment analysis
Tables 3 presents the analysis comparing patients within the treatment arm. Table 4 presents the analysis comparing the patient in the treatment arm who used the device above the cut-off to the combined group of the remaining participants. All comparisons were significant based on the double-sided p-values in Tables 3 and 4. Figure 2 depicts the dose relationship between higher usage and LVEF improvement. As these usage comparisons are not based on randomization, they are subject to confounding bias. We addressed this using the propensity score matching method outlined above. Using the Control group patients in the second comparison was also designed to address the sensitivity of results to confounding bias.
Table 3.
Effect of Hours of PAP Use on Discharge LVEF- Analysis within the Treatment Arm
| PAP Use Cutoff | Three Hours/night * | Four Hours/night ** | ||
|---|---|---|---|---|
| Patients with ≥3 hs/night (n=35) | Patients with <3 hs/night (n=34) | Patients with ≥4 hs/night (n=21) | Patients with <4 hs/night (n=48) | |
| Baseline LVEF | 25.4 (10.6) | 25.3 (10.8) | 25.9 (11.1) | 25.1 (10.5) |
| Conclusion LVEF | 29.8 (12.5) | 24.7 (10.7) | 31.4 (13.3) | 25.5 (10.8) |
LVEF: Left ventricular ejection fraction; Mean (SD)
ANCOVA results: LVEF was higher in patients who used the device above the 3 hour cutoff, p=0.01
ANCOVA results: LVEF was higher in patients who used the device above the 4 hour cutoff, p=0.01
Table 4.
Effect of Hours of PAP Use on Discharge LVEF- Analysis of All Participants
| PAP Use Cutoff | Three Hours/night* | Four Hours/night** | ||
|---|---|---|---|---|
| Patients with ≥3 hs/night (n=35) | Remaining Participants (n=100) | Patients with ≥4 hs/night (n=21) | Remaining Participants (n=114) | |
| Baseline LVEF | 25.4 (10.6) | 26.4 (11.1) | 25.9 (11.1) | 26.2.1 (10.9) |
| Conclusion LVEF | 29.8 (12.5) | 27.4 (10.7) | 31.4 (13.3) | 27.4 (10.7) |
Analysis comparing patients in the treatment arm with PAP use above the set cutoff to a combined group of patients in the control arm along with treatment arm patients with PAP use below the PAP use cut off. Participants with non-missing conclusion LVEF were included.
LVEF: Left ventricular ejection fraction; Mean (SD)
ANCOVA results: EF was higher in the treatment subgroup (≥3 hours/night), p=0.02
ANCOVA results: EF was higher in the treatment subgroup (≥4 hours/night), p=0.04
Figure 2: Effect of Hours of Nightly PAP Use on LVEF Change during Hospitalization.
LVEF: Left ventricular ejection fraction; PAP: Positive airway pressure
To explore possible mechanisms for the findings of the on-treatment analysis, we examined the changes in LV end-systolic volume (LVESV), LV end diastolic volume (LVEDV) and stroke volume (SV). There were significant decreases in both LVEDV and LVESV with a larger decrease in LVESV in the more adherent groups with more than 3 and 4 hours supporting a potentially larger effect on afterload (supplement).
Effect of in-hospital PAP Treatment on Plasma Norepinephrine
Because we did not obtain significant results for the primary outcome LVEF, we cannot test this hypothesis formally. In addition, our exploratory analysis showed a large p=0.9 for this comparison. Further details are provided in the supplement.
Effect of in-hospital OSA Treatment on Post Discharge Readmissions
In the intention to treat analysis, readmission rate did not differ between the treatment group (n=62) and the control group (n=67) at 30 days (rate ratio (RR) = 1.07 (95% CI = 0.63, 1.81), p=0.82). Determination of usage was based on a 3 months download. Efficacy of the treatment was also confirmed as a function of residual AHI (supplement). Among the 62 treatment group patients, 25 (40.3%) had confirmed usage on the device download with average use ≥3 hours per night and 18 (29.0%) were confirmed to have ≥4 hours. Table 5 presents the results of the readmission models based on outpatient device usage. Readmission rates within 6 months were significantly lower in patients with higher PAP usage. Using the adherence threshold of average outpatient CPAP usage of ≥3 hours per night, readmission rates were 67% lower for adherent versus non-adherent (RR = 0.33 (95% CI = 0.14, 0.76), p=0.01) and 69% lower for adherent versus non-adherent using a threshold of ≥4 hours (RR = 0.31 (95% CI = 0.11, 0.83), p=0.02). Comparisons of adherent versus non-adherent combined with controls yielded similar results. Interestingly, discharge LVEF was associated with the readmission rate within 6 months. For every 5 unit increase in discharge LVEF, the readmission rate decreased by 15% [rate ratio (RR) = 0.85 (95% CI =0.75, 0.97), p=0.02]
Table-5.
On Treatment Analysis of 6-months Readmissions:
| Outpatient PAP Usage Threshold | Comparison | Rate Ratio (95% CI) | p-value |
|---|---|---|---|
| Three hours/night | Intervention arm only patients ≥ threshold (n=25) vs. patients < threshold (n=37) | 0.33 (0.14, 0.76) | 0.01 |
| Intervention arm patients ≥ threshold (n=25) vs. All remaining participants (remaining PAP patients+ controls) (n=104) | 0.43 (0.20, 0.91) | 0.03 | |
| Four hours/night | Intervention arm only patients ≥ threshold (n=18) vs. patients < threshold (n=44) | 0.31 (0.11, 0.83) | 0.02 |
| Intervention arm patients ≥ threshold (n=18) vs. All remaining participants (remaining PAP patients+ controls (n=111) | 0.37 (0.15, 0.95) | 0.04 |
Participants were excluded from readmission analysis if less than 14 days of follow-up were available or the patient died within 30 days of initial discharge. Rate of readmissions within 6 months were analyzed using negative binomial regression on readmission counts with an offset for follow-up time. The choice of 6 months period was based on previous studies. Device download was obtained at 3 months follow up.
Adverse Events and Other Conditions of Interest
There were no cases of treatment failure (residual AHI >15 events per hour). There were no significant adverse events reported to the IRB in relation to the treatment device. Thirty day and 3 months mortality were similar between the arms.
DISCUSSION
In this RCT, we evaluated a management approach to OSA in hospitalized patients with decompensated HFrEF. There was no significant effect for in-hospital treatment of OSA on the primary outcome of discharge LVEF. On treatment analyses, focused on recorded device usage, demonstrated a relationship between increased hours of use of the PAP device and increased discharge LVEF. Furthermore, exploratory analyses demonstrated a decrease in 6 months readmissions in participants who used the device more than 3 hours/night in the initial 3 months post-discharge further supporting potential clinical benefit for this management approach.
Expectedly, sleep and circadian rhythm are quite disrupted during hospitalizations with limited sleep at night and naps during daytime, particularly during decompensated heart failure. Therefore, PAP usage during the hospitalization does not connote long-term adherent behavior. Yet, interestingly, the average use of PAP device during the acute phase observed in our trial compares well with long-term usage reported in large a RCT done in ambulatory OSA patients (21).
This study is the largest RCT evaluating the effect of treating OSA during hospitalization for ADHF. Only one previous pilot study, done by our group, addressed the same population and enrolled 46 patients hospitalized with ADHF. This trial demonstrated a small improvement in discharge LVEF unlike the findings of the current trial. The upper value of the confidence interval on the effect estimate was smaller than what we anticipated in the power analysis. Older studies examined the effect of PAP on LVEF in stable outpatients with HFrEF. These studies were smaller and included longer periods of intervention during which an atypical very high adherence to the treatment device was recorded in the intervention arm(8, 9). Reconciling the results of this study with the previous small positive trials is not difficult to accomplish. The current study has a pragmatic design, larger sample size, short intervention time, and a hospitalized patient population with severe decompensated HF. Nevertheless, the exploratory subgroup analysis of usage demonstrated an effect for PAP device on LVEF similar to these previous trials.
A recent RCT of PAP in OSA(21) patients did not show a significant effect on cardiovascular health outcomes such as mortality. However, this trial included a limited number of patients who otherwise have stable HF patients (less than 50 patients with NYHA class 1 and 2)(21). There were several possible explanations put forward for the lack of CPAP effect on cardiovascular outcomes(23, 24). Despite these negative findings, several human and animal studies have consistently demonstrated an effect of eliminating OSA with PAP on biomarkers of cardiovascular health such as blood pressure and ejection fraction(7). This study and more recent larger RCTs(21, 22) support a conclusion that inadequate usage of PAP based therapies is a serious limitation to the treatment of OSA.
From a clinical standpoint, the question of whether in-hospital treatment of SDB during hospitalization would modify ADHF outcomes is an important one(7, 25, 26). Given the relatively limited therapeutic options for ADHF(27) and the poor clinical outcomes(28), evaluation of interventions that target comorbidities, as in this trial, carries the potential for improving discharge outcome. HFrEF patients who are hospitalized with AHDF and have previously unrecognized SDB have up to a 70% risk of mortality or readmission in the 6 months post-discharge(4, 5). Therefore, an intervention that targets post discharge outcomes of ADHF must be implemented during or shortly after the hospitalization if it were to impact this exceedingly high post discharge mortality and readmission. On this background, the rationale, findings, and the limitations of this study are best considered. The challenges of conducting diagnostic testing for SDB during ADHF episodes and of administering a therapeutic intervention with established limitation and acceptance as discussed previously(3, 6) were addressed in this trial with the use of testing protocols, an education intervention, and daily device download and review as described above and in the supplement.
This study did not demonstrate a significant improvement in discharge LVEF with CPAP. However, the study supports that initiation of treatment for newly diagnosed OSA during hospitalization for ADHF is safe and might be associated with small improvement in cardiac function only when device therapy is used above the described cutoffs. The exploratory analysis demonstrating a possible decrease in readmissions may have clinical significance and may justify further and larger investigation of this approach. Identification of OSA during hospitalizations can be a useful management approach for this important comorbidity in a high-risk patient population with decreased ability for follow up and can provide an ideal opportunity for education.
Supplementary Material
Acknowledgments
This study was funded by a grant from the National Heart, Lung, and Blood Institute (R21HL092480). Respironics, Inc provided the treatment devices as an in-kind contribution. Dr. Winfried Randerath received research grant money and speaking fees from Philips Respironics and Resmed.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Institutional review board (IRB) number: 2008H0011
Clinical Trials Registration number: NCT00679549
Consort check list: attached in supplement
References
- 1.Writing Group M, Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, et al. Heart Disease and Stroke Statistics-2016 Update: A Report From the American Heart Association. Circulation.2016;133(4):e38–360. [DOI] [PubMed] [Google Scholar]
- 2.Ambrosy AP, Fonarow GC, Butler J, Chioncel O, Greene SJ, Vaduganathan M, et al. The global health and economic burden of hospitalizations for heart failure: lessons learned from hospitalized heart failure registries. J Am Coll Cardiol. 2014;63(12):1123–33. [DOI] [PubMed] [Google Scholar]
- 3.Khayat R, Jarjoura D, Patt B, Yamokoski T, and Abraham WT. In-hospital Testing for Sleep-disordered Breathing in Hospitalized Patients With Decompensated Heart Failure: Report of Prevalence and Patient Characteristics. Journal of Cardiac Failure. 2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Khayat R, Jarjoura D, Porter K, Sow A, Wannemacher J, Dohar R, et al. Sleep disordered breathing and post-discharge mortality in patients with acute heart failure. Eur Heart J. 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Khayat R, Abraham W, Patt B, Brinkman V, Wannemacher J, Porter K, et al. Central sleep apnea is a predictor of cardiac readmission in hospitalized patients with systolic heart failure. J Card Fail. 2012;18(7):534–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Khayat R, Small R, Rathman L, Krueger S, Gocke B, Clark L, et al. Sleep-disordered breathing in heart failure: identifying and treating an important but often unrecognized comorbidity in heart failure patients. J Card Fail. 2013;19(6):431–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Javaheri S, Barbe F, Campos-Rodriguez F, Dempsey JA, Khayat R, Javaheri S, et al. Sleep Apnea: Types, Mechanisms, and Clinical Cardiovascular Consequences. J Am Coll Cardiol. 2017;69(7):841–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mansfield DR, Gollogly NC, Kaye DM, Richardson M, Bergin P, Naughton MT. Controlled trial of continuous positive airway pressure in obstructive sleep apnea and heart failure. Am J Respir Crit Care Med. 2004;169(3):361–6. [DOI] [PubMed] [Google Scholar]
- 9.Kaneko Y, Floras JS, Usui K, Plante J, Tkacova R, Kubo T, et al. Cardiovascular effects of continuous positive airway pressure in patients with heart failure and obstructive sleep apnea. N Engl J Med. 2003;348(13):1233–41. [DOI] [PubMed] [Google Scholar]
- 10.Vital FM, Ladeira MT, Atallah AN. Non-invasive positive pressure ventilation (CPAP or bilevel NPPV) for cardiogenic pulmonary oedema. Cochrane Database Syst Rev. 2013(5):CD005351. [DOI] [PubMed] [Google Scholar]
- 11.Pirracchio R, Resche Rigon M, Mebazaa A, Zannad F, Alla F, Chevret S. Continuous positive airway pressure (CPAP) may not reduce short-term mortality in cardiogenic pulmonary edema: a propensity-based analysis. J Card Fail. 2013;19(2):108–16. [DOI] [PubMed] [Google Scholar]
- 12.Writing Committee M, Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr., et al. 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. Circulation. 2013;128(16):e240–327. [DOI] [PubMed] [Google Scholar]
- 13.Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr., Colvin MM, et al. 2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Failure Society of America. Circulation. 2017;136(6):e137–e61. [DOI] [PubMed] [Google Scholar]
- 14.Khayat RN, Abraham WT, Patt B, Pu M, Jarjoura D. In-hospital treatment of obstructive sleep apnea during decompensation of heart failure. Chest. 2009;136(4):991–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Paulino A, Damy T, Margarit L, Stoica M, Deswarte G, Khouri L, et al. Prevalence of sleep-disordered breathing in a 316-patient French cohort of stable congestive heart failure. Arch Cardiovasc Dis. 2009;102(3):169–75. [DOI] [PubMed] [Google Scholar]
- 16.Oldenburg O, Lamp B, Faber L, Teschler H, Horstkotte D, Topfer V. Sleep-disordered breathing in patients with symptomatic heart failure: a contemporary study of prevalence in and characteristics of 700 patients. Eur J Heart Fail. 2007;9(3):251–7. [DOI] [PubMed] [Google Scholar]
- 17.Quintana-Gallego E, Villa-Gil M, Carmona-Bernal C, Botebol-Benhamou G, Martinez-Martinez A, Sanchez-Armengol A, et al. Home respiratory polygraphy for diagnosis of sleep-disordered breathing in heart failure. Eur Respir J. 2004;24(3):443–8. [DOI] [PubMed] [Google Scholar]
- 18.Medicine AAoS. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. AASM, editor. Westchester, IL: American Academy of Sleep Medicine; 2007. [Google Scholar]
- 19.Khayat RN, Jarjoura D, Patt B, Yamokoski T, Abraham WT. In-hospital testing for sleep-disordered breathing in hospitalized patients with decompensated heart failure: report of prevalence and patient characteristics. J Card Fail. 2009;15(9):739–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Dmitrienko A, Tamhane AC, Wang X, Chen X. Stepwise gatekeeping procedures in clinical trial applications. Biom J. 2006;48(6):984–91. [DOI] [PubMed] [Google Scholar]
- 21.McEvoy RD, Antic NA, Heeley E, Luo Y, Ou Q, Zhang X, et al. CPAP for Prevention of Cardiovascular Events in Obstructive Sleep Apnea. N Engl J Med. 2016;375(10):919–31. [DOI] [PubMed] [Google Scholar]
- 22.Cowie MR, Woehrle H, Wegscheider K, Angermann C, d’Ortho MP, Erdmann E, et al. Adaptive Servo-Ventilation for Central Sleep Apnea in Systolic Heart Failure. N Engl J Med. 2015;373(12):1095–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Javaheri S, Campos-Rodriguez F. Outcomes of Positive Airway Pressure for Sleep Apnea. JAMA. 2017;318(20):2042–3. [DOI] [PubMed] [Google Scholar]
- 24.Martinez-Garcia MA, Campos-Rodriguez F, Javaheri S, Gozal D. Pro: continuous positive airway pressure and cardiovascular prevention. Eur Respir J. 2018;51(5). [DOI] [PubMed] [Google Scholar]
- 25.Kahwash R, Khayat RN. A Practical Approach to the Identification and Management of Sleep-Disordered Breathing in Heart Failure Patients. Sleep Med Clin. 2017;12(2):205–19. [DOI] [PubMed] [Google Scholar]
- 26.Tsai M, Khayat R. Sleep Apnea in Heart Failure. Curr Treat Options Cardiovasc Med. 2018;20(4):33. [DOI] [PubMed] [Google Scholar]
- 27.Teerlink JR, Alburikan K, Metra M, Rodgers JE. Acute decompensated heart failure update. Curr Cardiol Rev. 2015;11(1):53–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gheorghiade M, Vaduganathan M, Fonarow GC, Bonow RO. Rehospitalization for heart failure: problems and perspectives. J Am Coll Cardiol. 2013;61(4):391–403. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.