During an 18 month period, 206 consecutive patients, referred to the Sleep Disordered Breathing Clinic of a University Hospital in Northern Portugal, who had been diagnosed with OSAS and recommended for CPAP treatment, followed a standardized protocol that included a general clinic evaluation by a sleep specialist, a home sleep study and a psychological evaluation that included an assessment of quality of life (SAQLI), anxiety and depression (HADS) and subjective somnolence (Epworth scale) (Table 1). Patients repeated the same tests after 1–3 months of CPAP therapy. None of the participants had a psychiatric disorder or any other sleep disorder. The study was approved by the hospital research ethics committee and written informed consent was obtained from all participants before participation in the study.

The sample was composed of 167 males and 39 female OSAS patients, between 27 and 75 years old, with a mean age of 52 years (SD=10). Nineteen of the patients dropped out at the time of the second assessment. Out of 187, 131 patients adhered to CPAP treatment and 54 patients were non-adherent (two patients did not have registered compliance due to download problems with their CPAP memory cards and therefore were not followed in the second phase).

43% of the patients had four years of education and were employed (61%). They presented an average BMI of 33. Almost 55% of patients had hypertension and 11% had cardiovascular diseases. 71% of all OSAS patients had an AHI higher than 30 apneas plus hypopneas per hour of estimated sleep time, after an overnight sleep study using a five-channel recording device (Sleep Screen, Viasys ™ Healthcare). This device produces a computerized recording of variations in oronasal airflow (measured by nasal cannula), body position, wrist oximetry, pulse rate and arterial oxygen saturation (measured by finger pulse oximetry). Sleep technicians carried out a manual analysis of the recordings, by counting apnea (episodes of ≤10% of previous airflow with at least 10s of duration) and hypopnea episodes (episodes showing 20–50% of the previous airflow, with at least 10s of duration together with a 4% dip in oxygen saturation), dividing the total number of these episodes by the sleep time in hours, obtaining a manual respiratory disturbance index according to established criteria.3 Patients were prescribed CPAP by an experienced sleep physician and instructed in the proper use by an experienced therapist, to whom patients had access for 7 days per week to resolve any problems that might occur during use.

SAQLI evaluates four domains of quality of life critical to this population and an additional domain to evaluate the effect of treatment: daily functioning (11 items); social interactions (13 items); emotional functioning (11 items); symptoms (5 items); treatment related symptoms (5 items, used in the second evaluation moment). The low end of the scale reveals the greatest impairment1,2 and the high end of the scale reveals the least impairment.6,7

The SAQLI has five domains and questions are equally weighted. In terms of scoring, each question ranges from 1 (greatest impairment) to 7 (least impairment). For each of the first four domains – A–D [A: daily functioning (11 items); B: social interactions (13 items); C: emotional functioning (11 items); D: symptoms (5 items)] – the mean score was obtained by dividing the sum of all the items, within the domain, by the number. The total score is calculated by summing the four domains (A–D) and then dividing by four. Thus, each of the domain scores and the total SAQLI has a range of 1–7, making the scores easier to interpret.15 After treatment, domain E is used and its mean score should be subtracted from the other domains totals, after being recoded – as 0 if the initial score is 7 and as 6 if the initial score is 1. For the symptoms domain and treatment related symptoms domain (domain E), patients may select up to 5 symptoms and the mean score is obtained by summing the scores for the five symptoms and dividing by five (even when the patient selects fewer than five symptoms). Section F measures the impact of treatment related symptoms (domain E) and the benefits of treatment (as reflected by any improvement in quality of life – domain A, B, C and D), drawing a vertical line through a 10cm visual analog scale. Domain E is divided by the impact score of the improvement in quality of life of domain A, B, C and D to provide a “weighting factor” result (with a maximum of 1.0) that is multiplied by the mean score for treatment related symptoms.15 The product is then subtracted from the total of mean scores from the first four domains giving the minimum important difference with a cut off point of 0.5 when a 7 item likert scale is used.15 Higher scores indicated better quality of life.

SAQLI presents a high internal consistency as reflected by the Cronbach's alpha coefficients of 0.88–0.92.15 To determine the measurement properties of SAQLI the authors presented, in 2002, new data.20 Repeatability of the SAQLI at one and two weeks showed a reliability (intraclass correlation) coefficient of 0.92 and did not change between the two weeks. Moderate to strong correlations were observed between SAQLI and other quality of life measures.

For the development of the Portuguese version of SAQLI, the translation and cultural adaptation group (TCA Group)’s guidelines for translation and adaptation of self-report instruments were followed.29 After receiving, from the authors, the authorization and the original articles with the development and measurement properties of SAQLI, the original English version was first translated into Portuguese by two independent native Portuguese speakers. The back translation was performed by an outsourced translator, who was not linked to the research group and did not know the original instrument and its first translations. The discrepancies between the original English version, the translated version into Portuguese and the back translation version were analyzed and only minor differences were found. The final Portuguese version of SAQLI was applied to five OSAS patients, native in Portuguese language to test for comprehension and suggestions. The main suggestion had to do with the format of the instrument, namely to change the assembled color coded cards to colored symbols in the patient's format sheet, which was done.

OSAS patients were invited to participate into the study after the overnight sleep study and the prescription of CPAP by their sleep physician. The information on demographic and clinical variables was obtained by self report and a case record clinic form. SAQLI was administrated by a psychologist. Prior to the start of the interview, the psychologist ensured that the patient fully understood the seven-point likert scales, i.e. the numbers 1–7 represented all possible answers. First SAQLI was used to measure the patient's baseline QOL.15 After that, the psychologist emphasized that the low end of the scale7 represented the least impairment and the high end of the scale1 represented the greatest impairment, while explaining the symbol in the patient's format SAQLI sheet. For all participants, the psychologist read the instructions and the questions while the patient filled in the responses on the questionnaire sheet.

The second moment of the study occurred 1–3 months after CPAP therapy. SAQLI, with all domains plus Domain E (treatment related symptoms) was administered.15 The electronic records of CPAP compliance were analyzed after downloading the therapy details from the CPAP software. Adherence was defined as a CPAP use ≥4h/night, following the cut-off point state in the majority of studies and base on a frequency over >70% of the total treatment days.30,31 After intervention, SAQLI was given to the same OSAS patients.

Internal consistency was assessed by Cronbach's alpha. Content validity was analyzed through Pearson correlations of each item with the total of the respective subscale and with the total score, before and after CPAP therapy. A varimax rotation was used to perform the factor analysis.

The evaluative properties of SAQLI were examined, namely the extent the instrument captures changes in quality of life over time.32 This was tested using paired t-tests to detect statistically significant differences in scores in OSA patients that adhere to CPAP treatment over the study period. Pearson correlation coefficients between SAQLI interdomain before and after CPAP treatment and clinical variables were also performed. Construct validity with interdomain correlations between SAQLI and HADS (Hospital Anxiety and Depression Scale) were also performed.

Construct validity was assessed by performing factor analysis with varimax rotation method.33 The suitability of factor analysis by checking the existence of significant correlations, between the items, was confirmed by Kaiser–Meyer–Olkin (KMO=0.896) and by the Sphericity Bartlett test (QQ=4533; gl=780; p<0.001). Four factors with eigenvalues greater than 1.0 were extracted accounting for 49, 25% of total variance. Factor loadings are shown in Table 2.

Factor A (daily functioning domain) explained 10.9% of total variance and covered all items. The Factor B (social interactions domain) explained 8.03% of total variance and covered almost items excepted for B8 (need to be left alone) and B12 (inadequate and/or infrequent sexual intimacy). These items were included in Factor C. The Factor C (emotional functioning domain) explained 17.82% of total variance and covered all items. Factor D (symptoms domain) covered all five symptoms items and explained 12.49% of variance.

Cronbach's alpha coefficients of internal reliability were above the standard (0.70) for all domains: daily functioning 0.88; social interaction 0.84; emotional functioning 0.87; and symptoms 0.89. Cronbach alpha for total SAQLI was 0.94 with item correlations ranging from 0.30 to 0.70 (Table 3).

Pearson correlation coefficients between SAQLI interdomains before and after CPAP treatment and clinical variables are presented in Table 4. The four domains before CPAP treatment and total SAQLI after CPAP treatment were highly correlated: daily functioning (r=0.572, p<0.001), social interaction (r=0.312, p<0.001), emotional functioning (r=0.675, p<0.001) and symptoms (r=0.401, p<0.001). Total score of SAQLI is highly correlated before and after CPAP treatment (r=0.590, p<0.001). AHI is highly correlated with Symptom domain (r=0.247, p<0.001) and slightly negative correlated with treatment symptoms (r=−0.113, p<0.005) and positively with total SAQLI score (r=0.183, p<0.005) after CPAP treatment.

To test for external validity, interdomains correlations between SAQLI and HADS were performed. All SAQLI domains were negatively correlated with anxiety and depression scales. The emotional functioning domain of SAQLI had the highest correlation with anxiety (r=−0.657, p<0.001) and depression (r=−0.637, p<0.001); then the daily functioning domain correlated with anxiety (r=−0.450, p<0.001) and depression (r=−0.526, p<0.001); following the social functioning domain correlated with anxiety (r=−0.422, p<0.001) and depression (r=−0.471, p<0.001); finally symptoms domain correlated with anxiety (r=−0.444, p<0.001) and depression (r=−0.416, p<0.001).

After CPAP treatment there was a significant decrease in somnolence with only 20.9% of patients with Epworth sleepiness scale>10.

According to SAQLI formula (multiplication of weighting factor by the mean score for treatment related symptoms), a minimum important difference superior to 0.5 was found in the adherent group before and after CPAP therapy, suggesting a change in quality of life. That did not occur in the non-adherent patients. A significant difference was obtained in each domain before and after treatment intervention and between total scores of SAQLI in both moments, in the adherent patients. In non-adherent patients, slightly significant differences were obtained for each domain of quality of life, but not for total quality of life. Table 5 shows the mean and standard deviation, evaluated before and after CPAP therapy and the differences between two moments for both adherent and non-adherent patients.