Categoric Definitions of PAP Adherence | ResMed India
                     PERFORMANCE OF CLAIMS-BASED 
ALGORITHMS FOR ADHERENCE TO 
POSITIVE AIRWAY PRESSURE 
THERAPY IN COMMERCIALLY 
INSURED PATIENTS WITH OSA
BACKGROUND: Positive airway pressure (PAP) therapy is first-line therapy for OSA, but consistent use is 
required for it to be effective. Previous studies have used Medicare fee-forservice claims data (eg, device, 
equipment charges) as a proxy for PAP adherence to assess its effects. However, this approach has not been 
validated in a US commercially insured population, where coverage rules are not standardized.
RESEARCH QUESTION: In a commercially insured population in the United States, how well do claims-based 
algorithms for defining PAP adherence correspond with objective PAP device usage? STUDY DESIGN AND 
METHODS: Deidentified administrative claims data of commercially insured patients (aged 18-64 years) with OSA 
were linked to objective PAP therapy usage data from cloud-connected devices. Adherence was defined based on 
device use (using an extension of Centers for Medicare & Medicaid Services 90-day compliance criteria) and from 
claims-based algorithms to compare usage metrics and identify potential misclassifications. 
RESULTS: The final sample included 213,341 patients. Based on device usage, 48% were adherent in the first 
year. Based on claims, between 10% and 84% of patients were identified as adherent (accuracy, sensitivity, and 
specificity ranges: 53%-68%, 12%-95%, and 26%-92%, respectively). Relative to patients who were claims-
adherent, patients who were deviceadherent had consistently higher usage across all metrics (mean, 339.9 vs 
260.0-290.0 days of use; 6.6 vs 5.1-5.6 d/wk; 6.4 vs 4.6-5.2 h/d). Consistent PAP users were frequently identified 
by claims-based algorithms as nonadherent, whereas many inconsistent users were classified by claims-based 
algorithms as adherent. 
INTERPRETATION: In aggregate US commercial data with nonstandardized PAP coverage rules, concordance 
between existing claims-based definitions and objective PAP use was low. Caution is warranted when applying 
existing claims-based algorithms to commercial populations.
OSA is the most common sleep-related breathing disorder among adults, currently estimated to affect > 54 
million adults in the United States and nearly 1 billion people worldwide.1 Consequences of untreated OSA 
include adverse cardiometabolic, neurocognitive, and psychiatric outcomes, and increased mortality.2,3 
Untreated OSA is also associated with a substantial economic burden, with estimated total costs of $149.6 billion 
in the United States in 2015.4 Positive airway pressure (PAP) therapy is considered first-line treatment for OSA5 
and is associated with a range of health6-9 and economic10 benefits. However, consistent use is required for PAP 
to be effective, and adherence varies widely. PAP usage is often reported to be low, with studies estimating up to 
83% of patients using PAP < 4 h per night.11,12 As a result of suboptimal PAP adherence, intent-to-treat analyses 
in OSA clinical trials can produce underwhelming results, even though on-treatment analyses consistently 
demonstrate a beneficial effect of PAP on health outcomes.
To accelerate discovery of new understanding regarding the effects of PAP therapy, researchers have sought to 
analyze data from additional sources beyond randomized clinical trials, which are time-consuming and expensive. 
One such data source is payer administrative claims that include demographic characteristics, disease 
information (eg, diagnostic codes assigned by health providers), procedure codes (eg, Current Procedural 
Terminology codes for diagnostic sleep tests), and treatment information (eg, insurance charges for PAP 
machines and supplies). Using these data, investigators have examined the effect of PAP by using PAP charges as 
a proxy for PAP use, applying various claims-based algorithms to operationalize PAP adherence.6,15-19 
Advantages of these approaches include large sample sizes to support generalizability and ensure adequate 
statistical power, ready data accessibility, and direct representativeness of the payer/policymaker perspective.
Study Design and Methods
Deidentified payer-sourced administrative medical claims data from > 100 US commercial health plans (Inovalon 
Insights LLC) was linked with objective PAP usage data from cloud-connected devices ( ResMed Corp). Claims data 
included demographics, clinical information, and information about equipment and supplies related to PAP 
therapy submitted for payment. PAP device data included detailed metrics around frequency and duration of use. 
Data from both sources were linked through a secure tokenization process, and the resulting database underwent 
Health Insurance Portability and their sleep test and 2 years after device setup (ie, the index date) to ensure 
capture of equipment fi lls and resupply during the study period. Exclusion criteria included evidence of PAP 
resupply, pregnancy, dialysis, or end-stage renal disease during the year prior to index, and evidence of central 
sleep apnea or nocturnal hypoventilation at any time during the study period. Categoric Definitions of PAP 
Adherence Device-Based Definition: Consistent with prior efforts,7,8,21-23 the first definition was based on an 
extension of the Centers for Medicare & Medicaid Services (CMS) criteria for PAP adherence, applied to the 
continuous usage metrics obtained directly from a patient’s device (eg, hours per day). For PAP coverage, CMS 
requires PAP device usage $ 4 h per night for $ 70% of nights within a 30-day period during the first 90 days 
after receiving PAP, and provider-documented symptomatic benefit. In this study, the year after the index date 
was divided into four 90-day quarters. PAP adherence was evaluated separately within each 90-day quarter. 
Patients were labeled as adherent, intermediate, or nonadherent to PAP therapy if they met CMS criteria in all 
four, one to three, or zero follow-up quarters, respectively. Because data were missing for the entire night if the 
device was not turned on, usage values for nights with missing data were imputed as 0. As described 
previously,24 the device-based definition of adherence produces groups with distinct PAP usage patterns. Claims-
based definitions of PAP adherence among patients with incident OSA were identified from published literature. 
Categoric Definitions of PAP Adherence
Device-Based Definition: Consistent with prior efforts,7,8,21-23 the first definition was based on an extension of 
the Centers for Medicare & Medicaid Services (CMS) criteria for PAP adherence, applied to the continuous usage 
metrics obtained directly from a patient’s device (eg, hours per day). For PAP coverage, CMS requires PAP device 
usage $ 4 h per night for $ 70% of nights within a 30-day period during the first 90 days after receiving PAP, and 
provider-documented symptomatic benefit. In this study, the year after the index date was divided into four 90-
day quarters. PAP adherence was evaluated separately within each 90-day quarter. Patients were labeled as 
adherent, intermediate, or nonadherent to PAP therapy if they met CMS criteria in all four, one to three, or zero 
follow-up quarters, respectively. Because data were missing for the entire night if the device was not turned on, 
usage values for nights with missing data were imputed as 0. As described previously,24 the device-based 
definition of adherence produces groups with distinct PAP usage patterns. Claims-based definitions of PAP 
adherence among patients with incident OSA were identified from published literature. All identified definitions 
were originally published using Medicare fee-for-service data. Claims-Based Definition 1: The first claims-based 
definition of PAP adherence from Bock et al18 was based on evidence of PAP treatment or related supply fi lls on 
or after day 91 from the index date. Those with one or more eligible claims within the time frame were 
considered adherent. All others were considered nonadherent. Claims-Based Definition 2: The second claims-
based definition16,17 was based on Medicare’s rent-to-own approach to PAP, in which PAP machines are rented 
(billed) monthly for 13 months then owned. Patients were considered highly adherent (denoted as adherent in 
this study) if they had > 12 device claims (indicating that they continued to use PAP throughout the entire rent-
to-own period), partially adherent (denoted as intermediate in this study) if they had four to 12 device claims 
(indicating that they met 90-day CMS adherence but discontinued PAP prior to conclusion of rentto-own period), 
and nonadherent if they had less than four device claims from the index date (ie, initiated PAP but did not meet 
the threshold for 90-day adherence per CMS).
Visualization of PAP Adherence
Device-Based Definition: Categoric adherence groups were well differentiated with distinct usage patterns across 
all use metrics (Figs 1A, 2A, 3A). Patients who were adherent had frequent and consistent device use, whereas 
patients who were nonadherent had lower use across all metrics. Variability was highest for patients who were 
intermediate. Claims-Based Definitions 1, 2, and 3: For each claimsbased definition, there was substantial 
overlap in device usage metrics across adherence groups (Figs 1B-D, 2B-D, 3B-D), with usage ranging from rare 
to frequent. In many cases, patients classified as adherent had infrequent use, in some cases as low as 0 to 1 h/d 
or 0 to 1 d/wk. Patients classified as intermediate or nonadherent often had high levels of usage, including those 
who used daily over the first year, or for an average of > 7 h/d. Accuracy: 
Relative to the objective device-based definition, claimsbased definitions of adherence generally demonstrated 
low accuracy (range, 53%-68%). Claims-based definition 1, which classified most patients as adherent.

Interpretation 
Administrative claims provide an essential data source to study OSA and other sleep disorders, including PAP 
therapy and other sleep treatments. Indeed, using Medicare claims, researchers have studied the impact of PAP 
adherence on cardiovascular, cerebrovascular, 30- day hospital readmissions, and economic outcomes.15-18 
However, when analyzing US commercial claims data, researchers must exercise caution in using claims-based 
algorithms to define PAP adherence in the absence of objective device data and/or known PAP coverage policies 
because it cannot be assumed without further validation that administrative claims correspond closely to actual 
device usage. In the context of nonstandardized PAP coverage policies, incorporation of objective device data 
represents an important opportunity to improve understanding of potential beneficial effects of PAP among 
individuals with commercial insurance.