Greater positive airway pressure adherence in OSA yields lower health care costs
Click Here to Manage Email Alerts
Key takeaways:
- The average cost of health care went down with long and consistent positive airway pressure use.
- There was a $832 difference in 6-month health care costs between those with high and low adherence.
Patients with obstructive sleep apnea who used positive airway pressure consistently for 4 hours or more per night over 3 years had lower health care costs than patients with infrequent use, according to results published in CHEST.
“While this study was a retrospective analysis and, therefore, we cannot prove there is a causative relationship between PAP adherence and lower health care utilization and cost, a potential causative relationship makes conceptual sense,” Dennis Hwang, MD, director of sleep medicine at Kaiser Permanente, told Healio. “Other research we have performed demonstrated a 70% higher rate of acute cardiovascular events (ie, heart attack and stroke) in patients with at least moderately severe OSA but 53% lower in those who use PAP at least 4 hours per night. These results are complementary in promoting a narrative of the importance for clinicians to dedicate effective strategies and effort to optimize PAP adherence for patients with OSA.”
Using data from the Tele-OSA clinical trial, Hwang and colleagues evaluated 543 patients (mean age, 50 years; 59% men; 42% non-Hispanic white; 40% Hispanic) with OSA and prescribed PAP to determine the relationship between PAP adherence over 3 years and health care costs via multivariable generalized linear models.
Researchers used electronic health record and claims data from their integrated health care system to determine average health care costs (2020 U.S. dollars) during 6-months periods. They considered 11 types of health care use, including those that were sleep related (office visits, equipment) and nonsleep related (eg, office visits, hospitalization, ED, pharmacy).
Over a span of 3 years, a little more than half (52%; 36% white) of patients had low PAP adherence (< 2 hours/night), with an average use of 0.5 hours a night. A quarter of patients (57% white) had high adherence (consistently ≥ 4 hours/night), with an average use of PAP for 6.5 hours a night, and 22% (39% white) had moderate adherence (2-3.9 hours/night or inconsistently ≥ 4 hours/night), with an average PAP use of 3.7 hours a night.
Researchers found that average PAP use significantly grew with time among patients from the high adherence group, whereas use of PAP among those with low and moderate adherence went down with time.
Average health care costs
When comparing health care costs between the adherence groups, researchers adjusted for several covariates: age, sex, race/ethnicity, BMI category, Medicaid enrollment, smoking, apnea-hypopnea index (AHI) category, hypertension, Charlson Comorbidity Index score, mild liver disease, preoperative referral to sleep clinic prior to randomization, no-show percent, log-transformed prior 6-month cost, time indicators representing the 6-month interval and adherence to antihypertensive, antidiabetic or statin medications.
Following these adjustments, patients with low adherence had average 6-month health care costs of $4,040, whereas those with high adherence had reduced average 6-month costs at $3,207 for a significant difference of $832 (95% CI, $127-$1,538) between the two groups.
Costs for patients with moderate adherence fell in between those observed for the low and high adherence groups ($3,638) and did not significantly differ from those of the low adherence group.
Individual costs, OSA severity
When individually assessing the 11 different types of health care use, researchers found that 6-month costs for those with high vs. low adherence significantly differed for sleep-related office visits ($67 vs. $70; P = .01), sleep-related durable medical equipment ($141 vs. $76; P < .001), hospitalizations ($171 vs. $2,076; P < .001), laboratory ($163 vs. $216; P = .004) and phone encounters ($42 vs. $59; P < .001).
Compared with the moderate adherence group, the high adherence group had significantly lower costs for nonsleep-related office visits ($829 vs. $995; P = .03), nonsleep-related durable medical equipment ($15 vs. $322; P = .04) and hospitalizations ($171 vs. $852; P = .01).
“Given that OSA is a chronic disease and the clinical benefit of PAP would theoretically be preventative by reducing the risk of chronic cardiovascular conditions such as hypertension, we were surprised that the lower associated health care utilization and cost in those with high PAP adherence was evident immediately within the first 6 months of PAP therapy initiation,” Hwang told Healio. “This finding does complement our other surprising finding that the lower cost in those with high adherence was primarily seen in the cost of hospitalization and other hospital services. Thus, the lower acute care clinical events can drive immediate potential cost savings.”
Notably, patients with moderate vs. low adherence had significantly reduced costs for other hospital services ($339 vs. $1,151; P = .02) but increased costs for sleep-related durable medical equipment ($127 vs. $76; P < .001).
Although moderate to severe OSA (AHI ≥ 15) was not linked to cost differences between the adherence groups in an a priori-specified stratified analysis, researchers found significant cost differences among PAP adherence groups when evaluating those with mild OSA (AHI 5-14.9). Similar to the main finding above, among patients with mild OSA, those with low adherence had higher costs than patients with high (cost difference, $1,431; 95% CI, $223-$2,640) or moderate (cost difference, $1,878; 95% CI, $657-$3,100) adherence.
“We have already performed a similar analysis on a much larger cohort of patients (n = 25,000) in Southern California which shows similar conclusions,” Hwang told Healio. “Additionally, we plan to perform the analysis over a longer time frame; utilize machine learning to identify specific patterns of usage that relate to outcomes so that we can better personalize PAP usage targets; and perform additional analyses for under-represented minorities in order to highlight potential disparities that may be amenable to specific intervention strategies to achieve greater equity of outcomes,” Hwang added.
Overcoming potential bias
Although it can be difficult to prevent biases from entering observational studies, this study by Hwang and colleagues demonstrates a good effort through its covariate adjustments, according to an accompanying editorial by Jennifer S. Albrecht, PhD, associate professor in the department of epidemiology and public health at the University of Maryland School of Medicine, and Vishesh K. Kapur, MD, MPH, director of sleep medicine for the division of pulmonary, critical care and sleep medicine at the University of Washington.
“To better address healthy adherer bias, future studies could incorporate additional measures of healthy behavior such as receipt of immunizations, appropriate screening (eg, colonoscopy, mammogram) and receipt of annual wellness visits,” Albrecht and Kapur wrote. “Accounting for other behaviors that are not captured adequately in the medical record, such as diet, physical activity and regular alcohol use, would also be useful.”
References:
- Albrecht JS, et al. CHEST. 2023;doi:10.1016/j.chest.2023.03.018.
- Mazzotti Diego, et al. Abstract 439. Presented at: SLEEP; June 10-13, 2021 (virtual meeting).
Perspective
Back to Top
The researchers examined the question of whether CPAP usage is associated with lower overall health care costs. This is an important topic as the costs for sleep-related durable medical equipment (CPAP machines, masks, etc) are not insignificant and sleep apnea is incredibly common. There is therefore great utility in examining the “business case” for treating sleep apnea with CPAP from a health care insurance and societal standpoint.
The authors found that those with higher CPAP usage had lower overall health care costs. It looks like the authors did a fairly good job controlling for potentially confounding factors. However, the association could simply be due to the fact that those who are less likely to use their CPAP machines were also less likely to be adherent with other aspects of their health care plan, such as taking their diabetes and high blood pressure medications. General nonadherence to care would be expected to increase health care costs due to developing secondary complications from chronic diseases, such as diabetic retinopathy or a hemorrhagic stroke from untreated hypertension.
In my clinical experience, patients who are nonadherent with CPAP tend to be globally nonadherent, and I would expect them to be sicker and require more health care resources over time. I’m not sure, based on the information provided, if the researchers completely controlled for these sorts of factors.
Another disparity between the “high adherence” and “low/medium adherence” groups appears to be race and ethnicity with Blacks and Hispanics overrepresented in the “low/medium adherence” groups. I think we need to know if these differences were statistically significant and, if so, the reasons behind this need to be explored.
What was really interesting is that there was a stronger statistical relationship between CPAP adherence and lower health care costs in the mild sleep apnea (AHI < 15) category as opposed to moderate and severe degrees of sleep apnea (AHI ≥ 15). This is important because insurances, including Medicare and Medicaid, often do not pay for CPAP equipment for patients with mild OSA. This study suggests that there may in fact be the highest “return on investment” with treating mild sleep apnea.
Collapse
Click Here to Manage Email Alerts