As the field of sleep medicine advances and more patients are diagnosed with OSAS and other sleep disorders, primary care clinicians will need to be able to recognise, diagnose, and treat these conditions. Recognizing sleep complaints is often challenging for both patients and clinicians. Although a polysomnogram is required to establish the diagnosis of OSAS, the long waiting lists for PSG in sleep centres have created interest in screening tools for obstructive sleep apnoea. While each of these has its own strengths and limitations, they all basically focus on similar high-risk features, such as habitual snoring, witnessed apnoeas, a high BMI, male sex, and advanced age. One validated tool that can help identify those with OSAS is the BQ. BQ is useful in screening sleep apnoea in a primary care population and may be more convenient and less costly than PSG.
Our investigation showed that, for different AHI cut-off points, BQ showed a moderate to high sensitivity (76–84%), a specificity that was low to moderate (40–61%), and a low positive likelihood ratio (1.26 to 2.14). However, positive predictive value is high (80–94%), indicating that if BQ is positive, for example at an AHI ≥ 5–< 15, there is a high (94%) likelihood that a person would actually have sleep apnea. We found a higher sensitivity and specificity when AHI was between 15 and 30. This finding suggests that in primary care patients, the Berlin questionnaire can be helpful in detecting a high risk of having OSAS, especially if the OSAS is moderate or severe.
The BQ is widely used as a screening tool for OSAS. It was an outcome of the Conference on Sleep in Primary Care in April 1996 in Berlin, Germany. The BQ is a standardised, self-administered questionnaire developed for assessing subjects at high risk for OSAS. It is inexpensive, easy to administer and has acceptable test–retest reliability
[11, 15, 16]. This instrument has been used worldwide, such as in the USA
, and Nigeria
. In addition, it has been used for OSAS screening in surgical patients before anaesthesia
, in patients undergoing endoscopy
, as well as among orchestra members
 or to predict outcomes after the catheter ablation of atrial fibrillation
The predictive performance of the BQ varies among different patient populations. Our results are consistent with those reported by Netzer et al.
, who also validated the BQ in a primary care population. The sensitivity and specificity obtained by Netzer and colleagues were 54% and 97%, respectively, at a cut-off of AHI ≥ 15
. However, these sensitivity and specificity values have been criticised
 as being erroneous, and it has been suggested that the actual sensitivity and specificity for Netzer’s study were 95% and 48%, respectively. Furthermore, in the Netzer manuscript, specificity increased as the AHI threshold increased and sensitivity fell. In our study, the sensitivity and specificity does not change very much at AHI ≥ 5, 15, 30, compared to the previous study. This may be due to several reasons. First, the population could be different. Polysomnography was offered to both high risk and lower risk patients, but more patients at high risk, who had sleep symptoms, selectively consented to the overnight polysomnography. Secondly, in the Netzer study, portable monitoring was used to assess the validity of the risk grouping strategy, compared to attended overnight polysomnography used in our study, which is the gold standard for the diagnosis of OSAS.
Apart from the Netzer study, previous studies exploring the BQ as an instrument to detect sleep apnoea showed similar or lower sensitivity and specificity compared to our results. The performance of the BQ in our population is in accordance with results obtained in other studies undertaken in a sleep laboratory in Portugal (72% and 50% for AHI ≥ 5– < 15, 82% and 45% for AHI ≥ 15– ≤ 30, and 88% and 39% for AHI > 30)
, in a hypertension clinic (86% and 65% for AHI > 10)
, and in preoperative patients (69% and 56% for AHI ≥ 5– < 15, 79% and 51% for AHI ≥ 15– ≤ 30, and 87% and 46% for AHI > 30)
. A better sensitivity of 86% and a specificity of 95% at a cut-off of AHI ≥ 5 were found when a modified version of the Berlin questionnaire was used
 and in the Arabic version of the BQ (sensitivity of 97% and specificity of 90%)
On the other hand, the results of our study differ from some others in the literature, where the BQ was shown to detect sleep apnoea in patients referred to a sleep laboratory with a lower sensitivity and specificity (68% and 49% for AHI ≥ 5, 62% and 43% for AHI ≥ 10, and 57% and 43% for AHI ≥ 15)
, and in patients undergoing pulmonary rehabilitation (sensitivity and specificity 62.5% and 53.8% with a cut-off for AHI of 10 or greater)
In our study, BQ demonstrated moderate to high sensitivity and low to moderate specificity: however, primary care practitioners require a practical and sensitive screening tool to identify patients at high risk of having OSAS. Because of the high prevalence of OSAS in primary care patients and the growing awareness of OSAS, general practitioners (GPs) are dealing with increasing numbers of patients with OSAS. Many patients who visit primary care physicians report risk factors, such as obesity and hypertension, and symptoms, such as snoring, sleepiness, and tiredness, suggestive of sleep apnoea. However, primary care providers often do not investigate these symptoms, and sleep apnoea frequently goes undiagnosed. Using the BQ as a screening tool for sleep apnoea in primary care population seems acceptable, being more convenient and less costly for health care users.
On the other hand, with the increased awareness of OSAS comes the risk that primary care physicians will refer patients with complaints of poor sleep, daytime sleepiness, and fatigue to sleep laboratories, without considering other diagnoses, such as insomnia, depression, and hypothyroidism
[28, 29]. Furthermore, excessive daytime sleepiness is a common complaint both in sleep-disordered patients and in a number of other non-medical and medical conditions, such as a shortage of sleep due to social or family reasons, shift working
 or daytime non-shift working
, and neurological disorders
[32, 33]. Therefore, given the need for stratification to determine patients who urgently require sleep studies, and in the face of rising health care costs and increased pressure for the more stringent control of health care resources, it would clearly be beneficial if primary care physicians could rely upon screening questionnaires to identify patients who are likely to have sleep apnoea, recommending sleep studies only for those patients who are at high risk of having a sleep disorder. In addition, most sleep clinics have long waiting lists, whereas patients who may have severe sleep breathing disorders warrant quicker diagnosis and treatment.
Our study presents some limitations that deserve comment. Firstly, there is a possible bias associated with the self-selection of patients. We found a high prevalence of OSAS in our group (91.5%), which is higher than estimates from community based surveys and is similar to the estimates found in surveys in sleep laboratories. As already mentioned, there may have been self-selection by the patients because those who had sleep symptoms might have selectively consented to the overnight polysomnography. Secondly, the mean BMI of the population in this study was high, with many patients being overweight and obese (29.6% and 54.5% respectively). The higher than average BMI of the sleep clinic population was likely to have affected the sensitivity and specificity of the BQ, as the BMI also factors heavily into the scoring of the BQ.