This was a retrospective, observational, case-crossover study in patients with a diagnosis of OSAHS receiving CPAP between January 2008 and February 2012 in the Sleep-Disordered Breathing Unit (SDBU) of the Hospital Universitari Sant Joan de Reus, who had been followed up for at least 1 year. The “case period” was defined as the 365 days following the clinical diagnosis of OSAHS and the “control period” was the preceding 365-day period. Of the 1790 patients seen in the SDBU due to suspected OSAHS, 373 patients with moderate or severe OSAHS treated with CPAP, with evaluable case report forms (CRF) at the beginning of treatment and at the end of the 1-year follow-up period were finally included (Fig. 1). The study was approved by the hospital's Ethics and Clinical Research Committee (code 15-03-26/3aclaobs1).

Fig. 1.

Flow chart of study design. *OSAHS diagnosis was performed by in-hospital CRPG in 34.3% of cases, on an outpatient basis in 33.2%, and by standard PSG monitored in the hospital in 32.4%. CRPG: cardiorespiratory polygraph; NIV: non-invasive ventilation; PSG: polysomnography; OSAHS: sleep apnea-hypopnea syndrome.

(0.14MB).

Exclusion criteria were: age younger than 18 years, psychophysical inability to complete questionnaires or refusal to answer them, Cheyne–Stokes syndrome, indication for uvulopalatopharyngoplasty, significant nasal obstruction preventing use of CPAP, need for invasive mechanical ventilation, and indication for bariatric surgery.

Diagnosis of OSAHS was determined by means of standard polysomnography (PSG) monitored in the hospital, or cardiorespiratory polygraphy (CRPG) either overnight in the SDBU or on an outpatient basis (Table 1). The techniques and the analysis were applied according to standardized guidelines.18,19 CPAP treatment was indicated in patients with a diagnosis of severe OSAHS and in patients with moderate OSAHS associated with cardiovascular disease or daytime sleepiness. Between 3 and 10 days after diagnosis, patients spent 1 night in the SDBU for titration of the effective pressure by automatic CPAP (auto-CPAP)20 (Table 1). All autotitration recordings for determining the optimal pressure were performed in the hospital to avoid unevaluable studies and to emphasize compliance, defined as at least 4h a night. This period was measured using the generator timer.

Each patient's clinical history, cardiovascular risk factors (diabetes mellitus, AHT, dyslipemia), cardiovascular events (CVE) (stroke, ischemic heart disease, heart failure, arrhythmias, other), cancer, use of healthcare resources (diagnostic tests, medical visits, emergency room visits, and hospitalizations), loss of work productivity, and OSAHS-related symptoms during the year before and the year after diagnosis were recorded on a CRF. All data were obtained from electronic clinical records.

Diagnoses given to the patient during their lifetime were recorded during the control period, with the exception of CVE and diagnoses of cancer, which were collected from the previous year only, and any new diagnoses were recorded during the case period.

To evaluate HRQoL, patients completed the Spanish version of the EuroQol-5D (EQ5D) questionnaire21 before and after receiving CPAP, consisting of a visual analog scale (VAS) from 0 to 100, where 100 is the best quality of life imaginable, and a utility index from 0 to 1 (where 1 represents perfect utility or quality of life, the annual equivalent of which is 1 quality-adjusted life year [QALY], and 0 represents the poorest quality of life imaginable, equivalent to death and to 0 annual QALYs).22,23 Patients also completed the Spanish version of the Epworth sleepiness scale (ESS) to evaluate their degree of daytime sleepiness.24

Total costs generated during the year before and the year after diagnosis were calculated. Total costs refer to the sum of direct healthcare costs, direct non-healthcare costs, and indirect costs.25,26 Healthcare costs correspond to those of the Catalan Health Service (CatSalut) for the year 2013. In the first year of treatment, the following were included: OSAHS diagnosis (€358.52); hospital admission for CPAP calibration (€2223.05) and annual CPAP maintenance by the supplier awarded the contract to provide these services through a public tender (€335.8), representing a total of €2917.37 per patient. The cost of the OSAHS diagnostic test is comprehensive our area, irrespective of whether PSG or CRPG are performed on an inpatient (in the SDBU) or outpatient basis; moreover, admission for nasal CPAP calibration in this area costs the same as a standard hospital stay. These items and the incremental cost-effectiveness ratio (ICER) are listed in Table 1. According to economic assessment manuals, we considered a clinical practice as cost-effective when its ICER was lower than a certain threshold of efficiency. In the Spanish healthcare system, the threshold of efficiency is defined as €30,000/QALY27 or £20,000–£30,000/QALY according to the Institute for Health and Clinical Excellence (NICE).28

The clinical situation, quality of life, degree of sleepiness, and costs generated by the patient during the control and case periods were compared and contrasted between compliers and non-compliers. The paired t test was used to analyze the same sample of individuals in the control and case periods. Statistically significant differences between both periods were examined. Non-parametric techniques (e.g., Wilcoxon test) were used for variables that did not have a normal distribution. The limit for statistical significance was set at a P value of <.05.

The ICER was calculated by comparing costs and QALYs between the previous year and the first year post-treatment, and between the previous year and the second year post-treatment. The costs of diagnosis and CPAP calibration were excluded from the costs of the second year of treatment. Since data were not collected after the second year, several hypothetical scenarios were proposed (sensitivity analyses), assigning a resource use of 10% higher or lower than that registered for the first year, while maintaining the QALY level constant.

Data were analyzed using SPSS version 19.0 (SPSS Inc., Chicago, IL, USA).

A total of 373 patients were included (15.5% women, mean age 56.3±10.2 years). Clinical and anthropometric characteristics are shown in Table 2. A total of 65% had a diagnosis of AHT and 35% had a history of respiratory disease. The diagnosis of OSAHS was performed by in-hospital CRPG in 34.3% of cases, on an outpatient basis in 33.2%, and by standard PSG monitored in the hospital in 32.4%. The mean apnea-hypopnea index (AHI) was 54.3±22.3 and the percentage of time during which SaO2 was less than 90% (CT90) was 27.3%±23.3%.

Mean CPAP pressure was 8.9±1.9cm H2O. After 1 year of treatment, median daily use was 5.3±1.7h. After 1 year of follow-up, 309 patients (82.8%) were categorized as compliers and 64 (17.2%) as non-compliers (Table 2), with no statistically significant baseline clinical differences between the groups, except for a greater prevalence of smoking and lower SaO2 among the non-compliers, compared to the compliers (P<.05).

Clinical course and quality of life of patients in the case period compared to the control period are shown in Table 3. The percentage of patients with CVE fell from 15.8% to 5.1% in the overall sample (P<.001) and from 14.9% to 0% among the compliers (P<.001), while it increased, insignificantly, among non-compliers.

In total, 38.9% of patients reported some form of side effect (31.4% of compliers and 75% of non-compliers, P<.001), the most common being nasal congestion (23.3%), pharyngeal dryness (7.5%), and skin irritation (6.7%).

Self-perceived health according to the EQ5D VAS increased by a mean of 5 points (P<.001), and patients gained a mean of 0.05 QALY (P<.001). Moreover, the score on the Epworth scale fell by a mean of 10 points (P<.001). In all cases, compliers had significantly better results than non-compliers (P<.001), who showed a loss of 0.04 QALY (Table 3).

Among the non-compliers (n=64), a subgroup analysis was performed according to hours of compliance (0 to <1h; 1 to <2h; 2 to <3h, and 3 to <4h). The analysis found no statistically significant differences between the subgroups, either in presence, absence or type of side effects, or in the mean QALY result.

CPAP treatment led to a significant reduction (P<.001) in the number of medical visits, emergency visits, hospitalizations and sick days off work by patients in active employment (Table 4). The use of resources by severe and moderate patients was compared (Table 4). After 1 year of treatment, the mean costs of compliers were significantly lower than those of non-compliers: €3357 vs €5959 (Table 5). Mean costs of severe and moderate patients were also compared (Table 5).

From the CatSalut perspective, which only includes the costs incurred by the health service itself, ICER after the first year of CPAP treatment was €51,147/QALY in all study patients, and €32,962 in compliers. From a social perspective, which also includes costs incurred by the patient and by society, the ICER fell to €47,334 for the whole population and to €26,976 for compliers (Table 6).

In the second year, the costs of OSAHS diagnosis and CPAP calibration were zero. The sensitivity analysis shows that with a 10% increase in costs, the ICER for CatSalut and for society would be €1544/QALY and €118/QALY, respectively, for the overall study population. If costs are reduced by 10%, the cost per QALY in both cases becomes negative. In any stage, the use of CPAP is a dominant strategy compared to non-use (Table 7).