With the current state of evidence, there is no doubt that sleep apnea-hypopnea syndrome (SAHS) is today a public health problem due to the high prevalence that it reaches in the general population (estimated at 4% of men and 2% of women1) as well as its negative effects on health, especially those related to cardiovascular morbidity and mortality, neuropsychiatric alterations and the loss of quality of life of the individuals with SAHS.2–4

One of the less frequently evaluated aspects in the literature is the impact that SAHS has on health-related quality of life (HRQL). To evaluate this impact, validated questionnaires are required, especially for the specifics of the disease in question. Such surveys provide great sensitivity for detecting the changes produced after therapeutic alterations and are able to discern the modifications in HRQL caused by respiratory disorders during sleep or of other chronic diseases.5 The most commonly used and evaluated generic questionnaire is the SF-36, which has been validated in Spanish6 and has been demonstrated to be a very adequate tool for its use in research and clinical practice.7,8 Among the specific questionnaires for evaluating the quality of life of patients with SAHS are the Sleep Apnea Quality of Life Index (SAQLI),9 which has been translated, adapted and validated in Chinese and Lithuanian.10,11 The Quebec Sleep Questionnaire (QSQ) was created by Lacasse et al. in French, their native tongue, and later translated to English for its use in patients with SAHS, demonstrating some excellent psychometric qualities in the English-language version.12,13 Spanish, with 329 million speakers, is currently the second most widely spoken language after Chinese (1.213 billion), followed by English (328 million).14 Although the QSQ has been successfully translated and adapted to Spanish,15 it has not been demonstrated whether this version has the psychometric qualities necessary for its use in patients with SAHS. Given that until now there has been no specific questionnaire for SAHS validated in Spanish in the literature, the objective of our study was to analyze the reliability and validity of the Spanish version of the Quebec Sleep Questionnaire HRQL survey in SAHS patients.

During the study period, 160 consecutive patients were remitted to the specialized outpatient sleep consultation due to clinical suspicion of SAHS. Twenty patients were excluded, of which 15 presented significant unstable comorbidities (4 hypoventilation-obesity, 6 severe COPD, 2 congestive heart failure, 2 severe arrhythmia, 1 after-effects of poliomyelitis); 2 patients were excluded due to important cognitive disorders and 3 refused to take part in the study. Out of the 140 patients who were finally included, 121 (86.4%) were diagnosed with SAHS, whose data were used for the analysis of internal consistency and validity of the QSQ survey. The mean age of these patients was 57±13, with an Epworth scale of 9±4. The mean AHI was 36±20h−1 (range: 5–89) and the body mass index (BMI) was 28.1±3.2kgm−2. Of all the patients studied, 82 received treatment with CPAP, following SEPAR guidelines. Sixty-one patients tolerated and complied with a minimum treatment of 4h daily. Out of the remaining 21 patients treated with CPAP, 10 abandoned the study, 9 did not tolerate the treatment and 2 patients were lost to follow-up (Fig. 1). We found no statistically significant differences between the patients who abandoned the study compared with those who did not abandon for the variables evaluated. Table 1 shows the results for the general description of the QSQ survey (mean, standard deviation, ranges, percentage of patients with ceiling effect and floor effect).

Fig. 1.

Methodological flowchart of the study. *Following SEPAR guidelines. **Validation of the questionnaire. 14 patients were randomly chosen for the test-retest reliability study. ***A use of ≥4h is considered good tolerance. +31 patients were randomly chosen for the sensitivity-to-change analysis. COPD: chronic obstructive pulmonary disease; SAHS: sleep apnea-hypopnea syndrome; CPAP; continuous positive airway pressure.

(0.25MB).

Predictive Validity

The group of patients with mild–moderate SAHS was compared with those who presented severe SAHS. Statistically significant differences were observed according to the severity in the domains of daytime sleepiness (P=.006), daytime symptoms (P=.012), nighttime symptoms (P=.0001) and emotions (P=.002). However, no differences were found in the social interaction domain (P=.074) (fig. 2).

Fig. 2.

Quebec Sleep Questionnaire score (mean±standard deviation) according to SAHS severity by groups (mild–moderate vs severe). EDS: excessive daytime sleepiness; DS: daytime symptoms; NS: nighttime symptoms; EM: emotions; SI: social interaction. * P<.05. Values tabulated as mean (standard deviation).

(0.05MB).

According to our results, the Spanish version of the QSQ presents adequate internal consistency, validity and repeatability for its use in patients with SAHS. Likewise, this questionnaire showed sensitivity to change produced by CPAP treatment for the domains referring to SAHS symptoms, both diurnal as well as nocturnal.

The internal consistency of the tool was high, and the coefficients obtained in each of the domains surpassed 0.7, generally considered acceptable. In the construct validity analysis, we observed a similar structure in three of the five domains, although without finding considerable similarities in the emotion and social interaction domains. Using a factorial analysis, the study showed, after analyzing the sediment chart, that the selection of five factors was similar to the distribution of the original questionnaire in three of them. These factors (1, 4 and 5) explained 36.5% of the variance. In the first factor, referring to daytime symptoms, 7 of the 10 items of this domain were included. The third factor included both items of the emotions domain as well as items of the social interaction domain. The fourth was mainly made up of items corresponding to nighttime symptoms. All the items that were grouped in the fifth factor corresponded with the daytime sleepiness domain. Therefore, there is a good correlation between factors 1, 4 and 5 with the domains of daytime symptoms, nighttime symptoms and daytime sleepiness, respectively, without there being a clear correlation of factors 2 and 3 with the emotions and social interaction domains.

Upon studying the concurrent validity of the test, we observed that there was a strong correlation between the daytime sleepiness domain of the QSQ and the Epworth test, as well as between the daytime symptoms domain of the QSQ and physical function, vitality, social function and the physical role of SF-36. The nighttime symptoms domain likewise presented a moderate correlation with the previously mentioned parameters and with the score of the FOSQ. This does not surprise us because, as we have commented earlier, the items corresponding with these domains are grouped similarly to the original questionnaire, and this has demonstrated similar correlations with the mentioned questionnaires.12 In the original questionnaire, the Pearson's correlation coefficient between the daytime sleepiness domain of the QSQ and the Epworth test was −0.64, and in our case it is −0.59. Likewise, the daytime symptoms domain of the original QSQ and those referring to the physical function, vitality and physical role of the SF-36 were 0.35, 0.85 and 0.72, respectively. These values were very similar to those found in our study, which were 0.52, 0.78 and 0.71. In the same manner as in the original questionnaire, the emotions and social interaction domains presented a strong correlation with their homologous domains of the SF-36 questionnaire, with correlation coefficients of 0.52 and 0.48, respectively.

Regarding the predictive validity of the test, as we have shown in our study there is an observed worsening of the HRQL in patients who present severe SAHS compared with those who present either a mild or moderate degree of disease. This deterioration in the quality of life was observed in all the domains of the questionnaire, except in the social interaction domain. The interpretation of this situation could be based on the fact that the score obtained in the social interaction domain can often be influenced by other factors that also influence quality of life, such as depression, invalidating physical problems or even adverse episodes that alter the daily life of patients.

Retest reliability was studied using the intraclass correlation coefficient of each of the domains, which demonstrated either moderate or good agreement in each.

In the sensitivity to change analysis after CPAP treatment, statistically significant differences were observed in the daytime and nighttime symptoms domains while nearly clinically significant in the daytime sleepiness domain. Nevertheless, the clinical significance that the original publication stated was not reached.12 We believe that the explanation of this phenomenon is due to the fact that the patients included in the initial questionnaire validation presented more marked SAHS symptoms, especially greater hypersomnia (mean Epworth test=14), while in our series the patients were more paucisymptomatic (mean Epworth test=7.6). Therefore, the possibility for significant clinical improvement in the original group after CPAP was greater than that of the series of the present study. Thus, our results support the use of the QSQ survey after the application of CPAP treatment, especially in the analysis of the changes in patient symptoms.

As for the limitations of the study, the following should be mentioned: the sample size is relatively small compared with other larger studies that comparatively analyzed the capacity for change in patients who had complied with treatment and those who had not as well as the size effect.22 On the other hand, a new sleep study was not done in order to verify the disappearance of the apnea at the end of the study. Lastly, no information was collected about the side effects of the CPAP treatment.

In conclusion, our validation study indicates that the Spanish version of the QSQ is a valid tool for determining HRQL while providing adequate psychometric characteristics for its use in patients with SAHS. Likewise, the QSQ is sensitive to the changes induced by treatment in the domains referring to SAHS symptoms.

Grants were received from the Valencia Pulmonology Foundation (Fundación Valenciana de Neumología, 2008) and Gasmedi 2000 (2008).