FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4200387958> ?p ?o ?g. }

• W4200387958 endingPage "1453" @default.
• W4200387958 startingPage "1441" @default.
• W4200387958 abstract "Free AccessReview ArticlesValidation of the STOP questionnaire as a screening tool for OSA among different populations: a systematic review and meta-regression analysis Darshit Patel, MSc, Jinny Tsang, BSc, Aparna Saripella, MSc, Mahesh Nagappa, MBBS, Sazzadul Islam, MSc, Marina Englesakis, MLIS, Frances Chung, MBBS Darshit Patel, MSc UCD School of Medicine, University College Dublin, Belfield, Dublin, Ireland Search for more papers by this author , Jinny Tsang, BSc Department of Anesthesia and Pain Management, Toronto Western Hospital, University Health Network, Toronto, Ontario, Canada Search for more papers by this author , Aparna Saripella, MSc Department of Anesthesia and Pain Management, Toronto Western Hospital, University Health Network, Toronto, Ontario, Canada Search for more papers by this author , Mahesh Nagappa, MBBS Department of Anesthesia & Perioperative Medicine, London Health Sciences Centre and St. Joseph Health Care, Schulich School of Medicine and Dentistry, Western University, London, Ontario, Canada Search for more papers by this author , Sazzadul Islam, MSc Department of Anesthesia and Pain Management, Toronto Western Hospital, University Health Network, Toronto, Ontario, Canada Search for more papers by this author , Marina Englesakis, MLIS Library and Information Services, University Health Network, Toronto, Ontario, Canada Search for more papers by this author , Frances Chung, MBBS Address correspondence to: Frances Chung, MBBS, Department of Anesthesia and Pain Management, Toronto Western Hospital, University Health Network, University of Toronto, 399 Bathurst St., Toronto, ON, Canada, M5T 2S8; Email: E-mail Address: [email protected] Department of Anesthesia and Pain Management, Toronto Western Hospital, University Health Network, Toronto, Ontario, Canada Department of Anesthesiology and Pain Medicine, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada Search for more papers by this author Published Online:May 1, 2022https://doi.org/10.5664/jcsm.9820Cited by:3SectionsAbstractEpubPDFSupplemental Material ShareShare onFacebookTwitterLinkedInRedditEmail ToolsAdd to favoritesDownload CitationsTrack Citations AboutABSTRACTStudy Objectives:Obstructive sleep apnea (OSA) is a sleep breathing disorder associated with adverse health outcomes, but it remains largely underdiagnosed. The STOP questionnaire is a simple tool for screening OSA and is widely used in various populations. The objective of this study was to determine the predictive parameters of the STOP questionnaire to detect OSA in sleep clinics, medical population, surgical population, commercial drivers, and the general population.Methods:Electronic databases were searched from January 2008 to April 2021. Pooled predictive parameters were recalculated using 2 × 2 contingency tables and random-effects meta-analyses were performed. The combined test characteristics at different OSA severities (any OSA [apnea-hypopnea index ≥ 5 events/h], moderate-to-severe OSA [apnea-hypopnea index ≥ 15 events/h], severe OSA [apnea-hypopnea index ≥ 30 events/h]) were used to compare the accuracy of the STOP questionnaire with polysomnography. The quality of the studies was evaluated using Cochrane Methods criteria.Results:Twenty-four studies met the inclusion criteria: 16 were in the sleep clinic population (n = 8,132), 4 in the medical population (n = 1,023), 2 in the surgical population (n = 258), and 1 study each on commercial drivers (n = 85) and the general population (n = 4,770). A STOP score ≥ 2 showed excellent sensitivity to the different OSA severities for the sleep clinic population (> 89%) and to severe OSA for the medical population (85.6%). In both populations, the STOP questionnaire also had excellent discriminative power to exclude severe OSA (negative predictive values > 84%). The pooled sensitivity and negative predictive values for the surgical population with moderate-to-severe OSA was 81% and 75%.Conclusions:This meta-analysis suggests that the STOP questionnaire is a valid and effective screening tool for OSA among these populations.Citation:Patel D, Tsang J, Saripella A, et al. Validation of the STOP questionnaire as a screening tool for OSA among different populations: a systematic review and meta-regression analysis. J Clin Sleep Med. 2022;18(5):1441–1453.BRIEF SUMMARYCurrent Knowledge/Study Rationale: Obstructive sleep apnea (OSA) is a common sleep breathing disorder associated with adverse health outcomes. The gold standard for diagnosis of OSA is in-laboratory polysomnography. However, it is expensive and time-consuming. The STOP (Snoring, Tiredness, Observed apnea, high blood Pressure) questionnaire is a simple tool widely used to screen OSA across different populations.Study Impact: In this study, through trends of high sensitivities and negative predictive values, we showed that the STOP questionnaire can help detect and rule out severe OSA in both medical and sleep clinic patients and moderate-to-severe OSA in surgical patients.INTRODUCTIONObstructive sleep apnea (OSA) is a highly prevalent health condition commonly associated with several medical comorbidities such as cardiovascular disease, obesity, and diabetes mellitus. 1 It is characterized by partial or complete recurrent obstruction of the upper airway. The prevalence of OSA has drastically increased, with over 400 million people worldwide having moderate or severe OSA. 2 The use of a polysomnography test (PSG) is considered the gold standard for the diagnosis of OSA to measure sleep variables such as sleep stages, respiratory efforts, oxygen saturation, and heart rate, but it is resource-intensive and time-consuming. 3 Requirements for a sleep specialist to attend laboratory PSG and long waitlists for appointments have prompted the use of portable devices for home sleep apnea testing (HSAT), but this still requires the expertise of sleep specialists for interpretation. With more than 80% of the world’s population living with untreated or undiagnosed OSA, 2 screening using questionnaires is easily accessible. It can encourage earlier diagnosis and treatment, relieve public health burden, 4 and address OSA-related mortality rates. 5The STOP questionnaire is an OSA screening tool containing four self-reportable items (STOP: Snoring, Tiredness, Observed apnea, high blood Pressure). 6 Each question is scored as yes or no with a total score of 4. The user is considered at high risk for OSA with a score of 2 or greater (STOP ≥ 2) and low risk for OSA with a score of zero or one (STOP < 2). 6The STOP questionnaire is used worldwide in many clinical settings. In particular, the STOP item measuring high blood pressure is highly influenced by comorbidity-specific characteristics that can be difficult to stratify, affecting the performance of the questionnaire for different medical conditions and region-specific lifestyles. Its translation to other languages to accommodate specific geographical areas may influence the accuracy of the STOP questionnaire. The objective of this systematic review and meta-analysis is to determine the predictive parameters of the STOP questionnaire for the screening of OSA in different populations such as patients from the sleep clinic, medical and surgical populations, commercial drivers, and the general population.METHODSLiterature search and data sourcesThe protocol for this systematic review and meta-analysis was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42021262764) and was carried out according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) reporting guidelines. 7 An information specialist (M.E.) designed search strategies including such terms as: “STOP questionnaire” or “STOP test” or “STOP instrument” or “STOP tool.” The following databases were systematically searched with no language restrictions from January 2008 to April 2021: MEDLINE, MEDLINE In-process, EmCare Nursing, Embase, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, PsycINFO, (all via the Ovid platform); the Web of Science (Clarivate Analytics), Scopus (Elsevier) and CINAHL (EbscoHost). Citation searching on Web of Science and Scopus was performed on the initial STOP validation study to capture any publications in which the article had been cited. Full-text searching was conducted via [email protected] and the University of Toronto Libraries PRIMO platform. A manual citation search was performed to obtain any related articles. The complete OVID search strategy is provided in Table S1 in the supplemental material.Inclusion criteria and selection of studiesOnce duplicates were resolved, three researchers (D.P., J.T., and S.I.) independently screened titles and abstracts. Two researchers (D.P. and J.T.) evaluated full-text articles for the following inclusion criteria: (1) STOP questionnaire was evaluated in adults, aged 18 and over, in the general population, commercial drivers, surgical population, patients with comorbidities, and different ethnic groups; (2) OSA diagnosis and STOP questionnaire were validated against laboratory PSG or HSAT; and (3) apnea-hypopnea index (AHI) or respiratory disturbance index (RDI) were used to diagnose OSA and grade its severity. The exclusion criteria involved a study population of children or adolescents under 18 years of age and studies not in the English language. Disagreements throughout the process were resolved through discussions between coauthors (D.P. and J.T.). Unresolved disagreements were subsequently discussed with other team members (A.S. and F.C.).Data extraction and managementData extraction from the included studies was carried out by two researchers (D.P. and J.T.) and recorded using a predesigned data collection form via an Excel worksheet. The STOP score ≥ 2 was accepted as the cut-off threshold. Any OSA was defined as AHI ≥ 5 events/h of sleep. Individuals with AHI ≥ 15 and AHI ≥ 30 events/h of sleep were identified as having moderate-to-severe and severe OSA. The predictive parameters at each AHI or RDI cut-off was used to generate a 2 × 2 contingency table for each study.Methodological quality of included studiesThe internal and external validity was independently assessed by two reviewers (D.P. and J.T.) using the Cochrane Method group’s guidelines on screening and diagnostic tests and disagreements were discussed with another researcher (A.S.). Internal validity was assessed using the following criteria: valid reference test, definition of disease, blind execution of the STOP questionnaire, independent interpretation of the results of the index test from clinical information, and study design. External validity was estimated using the following factors: disease spectrum, research setting, prescreening or referral, availability of demographic information, explicit threshold of STOP percentage of missing individuals, missing data management, and individuals’ selection for PSG.Statistical analysisMeta-analysis was performed using Review Manager Version 5.4 Copenhagen (The Nordic Cochrane Center, The Cochrane Collaboration, 2020) and MetaDisc Version 1.4 (Hospital Ramony Cajal, Madrid, Spain). The summary descriptive statistics were calculated for all variables. Mean and standard deviation were used for continuous variable and frequencies (n) and percentages (%) were reported for categorical variable.Using the random effects bivariate analysis model, the results of the paired and unpaired predictive parameters from each study were combined to obtain the following summary estimates with 95% confidence interval (CI): prevalence, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), diagnostic odds ratio, and area under the summary receiver operating characteristic curves. 8,9 The combined test characteristics were recalculated for each type of study population (sleep clinic patients, medical population with comorbidities, surgical patients, commercial drivers, and general population) at each AHI or RDI cut-off (AHI or RDI ≥ 5, ≥ 15, and ≥ 30 events/h) and forest plots were generated with the random effects model.We performed meta-regression and sensitivity analysis using Open Meta Analyst software 10 on various subgroups at each OSA severity for continuous variables (age, male sex, body mass index [BMI], prevalence, and sample size) and categorical variables (study type and validation tool).The aim was to measure the impact of these variables on the pooled estimates of sensitivity, specificity, and log scale diagnostic odds ratio. The reliability of the pooled estimates was further examined by leave-one-out meta-analysis. A value of P < .05 was considered statistically significant.RESULTSSearch results and study characteristicsThe initial search resulted in 5,726 studies, with 1 additional study identified through citations (Figure 1). After resolving duplicates and subsequently screening titles and abstracts, 4,001 articles were excluded, and 56 full-text articles were evaluated for eligibility. Of these, 32 studies were excluded for various reasons (Table S2). Twenty-four studies with 14,268 participants were included: 6,11–33 16 studies in the sleep clinic population (n = 8,132), 11–18,22,27–33 4 studies in the medical population (n = 1,023), 20,21,23,24 2 studies in the surgical population (n = 258), 6,19 and 1 study each in commercial drivers (n = 85) 25 and the general population (n = 4,770). 26 The included studies varied by country of origin: Croatia, 11 Egypt, 22 the United States, 14,17,20,24,26,27 Thailand, 28 China, 15,18,23,29 Greece, 30,33 Turkey, 25,31 Iran, 12,32 India, 13 South Korea, 16 Canada, 6 Brazil, 19 and Czech Republic. 21Figure 1: PRISMA flow diagram.AHI = apnea-hypopnea index, PRISMA = Preferred Reporting Items for Systematic reviews and Meta-Analyses, RDI = respiratory disturbance index.Download FigureFrom the sleep clinic population, 13 studies were included for meta-analysis at AHI ≥ 5 events/h, 11–14,16–18,22,27–30,32 14 at AHI ≥ 15 events/h, 12–14,16–18,22,27–33 and 11 at AHI ≥ 30 events/h. 12–14,18,22,27–30,32,33 Similarly, within the medical population, 2 studies 21,23 were included in the meta-analysis in AHI ≥ 5 events/h and ≥ 30 events/h and 3 in AHI ≥ 15 events/h. 20,21,23 Meta-analysis for the surgical population was possible in AHI ≥ 15 events/h, consisting of 3 studies. 6,19The study characteristics and demographic data are summarized in Table 1 and Table 2. The mean age (standard deviation) and BMI among all participants were 54 ± 14 years and 30 ± 7 kg/m2, and 65% of the patients were men. There were variations in the definition of OSA between studies, with 19 studies reporting as AHI ≥ 5 events/h or > 5 events/h, 6,11–14,16–18,21–24,27–30,32,33 1 study as AHI ≥ 10 events/h , 19 4 studies as either AHI ≥ 15 events/h, > 15 events/h, or RDI ≥ 15, 19,25,26,31 and 1 study as AHI ≥ 20 events/h. 15Table 1 Characteristics of included studies.First Author, Year (Country)Sample Size (N)Validation ToolOSA Definition (AHI events/h)OSA Prevalence n (%)No OSA (AHI < 5 events/h) n (%)Mild OSA (AHI ≥ 5 to AHI < 15 events/h) n (%)Moderate OSA (AHI ≥ 15 to AHI < 30 events/h or RDI ≥ 15 to RDI < 30) n (%)Severe OSA (AHI ≥ 30 events/h or RDI ≥ 30) n (%)Sleep clinic population Pecotic et al. 11 2012 (Croatia)217Lab PSGAHI ≥ 5216 (99.5)1 (0.5)66 (30.4)61 (28.1)89 (41.0) El-Sayed et al. 22 2012 (Egypt)234Lab PSGAHI ≥ 5204 (87.1)30 (12.9)27 (11.5)29 (12.4)148 (63.2) Boynton et al. 27 2013 (USA)219Lab PSGAHI > 5169 (77)50 (23)66 (30)41 (19)62 (28) Banhiran et al. 28 2014 (Thailand)303Lab PSGAHI ≥ 5222 (73.3)81 (26.7)73 (24.1)52 (17.2)97 (32.0) Luo et al. 29 2014 (China)212Lab PSGAHI ≥ 5196 (92.5)16 (7.5)26 (12.3)42 (19.8)128 (60.4) Pataka et al. 30 2014 (Greece)1,852Lab PSGAHI ≥ 51,498 (80.8)354 (19.2)251 (13.6)349 (18.8)898 (48.5) Yuceege et al. 31 2014 (Turkey)185Lab PSGAHI > 15166 (89.7)21 (11.34)53 (28.6)51 (27.6)60 (32.4) Sadeghniiat-Haghighi et al. 32 2015 (Iran)603Lab PSGAHI ≥ 5439 (72.8)164 (27.2)124 (20.6)114 (18.9)201 (33.3) Pataka et al. 33 2016 (Greece)204Lab PSGAHI ≥ 5165 (80.9)39 (19.1)29 (14.2)54 (26.5)82 (40.2) Kashaninasab et al. 12 2017 (Iran)250Lab PSGAHI > 5243 (97)7 (2.8)41 (16.4)43 (17.2)159 (63.6) Prasad et al. 13 2017 (India)210Lab PSGAHI ≥ 5164 (78.1)46 (21.9)28 (13.3)32 (15.2)104 (49.5) Sangkum et al. 14 2017 (USA)208Lab PSGAHI > 5162 (78)46 (22.1)62 (30)40 (19)60 (29) Hong et al. 15 2018 (China)2,208Lab PSGAHI ≥ 201,531 (69.3)677 (30.7)466 (21.1)352 (15.9)713 (32.3) Byun et al. 16 2020 (South Korea)778Lab PSGAHI ≥ 5569 (73.1)209 (26.9)145 (18.6)127 (16.3)297 (38.2) Medarov et al. 17 2020 (USA)344Lab PSG and HSAT (Embletta MPR)AHI > 5278 (80.8)66 (19.2)95 (27.6)183 (53.2) Wang et al. 18 2020 (China)105HSAT (Alice PDx)AHI ≥ 576 (72.4)29 (28)20 (19)21 (20)35 (33)Medical population Katzan et al. 20 2016 (USA)208 (Cerebrovascular patients)Lab PSGAHI ≥ 10127 (61.1)81 (28.9)127 (61.1)NANA Westlake et al. 21 2016 (Czech Republic)294 (Type2 diabetic patients)HSAT (ApneaLink, ResMed)AHI ≥ 5213 (72.4)81 (28)121 (41)61 (21)31 (10) Chen et al. 23 2019 (China)221 (Cerebrovascular patients)Lab PSGAHI ≥ 5165 (74.7)56 (25.3)49 (22.2)43 (19.5)73 (33.0) May et al. 24 2020 (USA)PAF 150Lab PSGAHI ≥ 5102 (68)48 (32)37 (24.7)65 (43.3)Control 150AHI ≥ 5104 (69.3)46 (30.7)35 (23.3)69 (46)Surgical population Chung et al. 6 2008 (Canada)177Lab PSGAHI > 5122 (68.9)55 (31.1)52 (29.4)31 (17.5)39 (22.0) Nunes et al. 19 2014 (Brazil)40 - CABGLab PSGAHI ≥ 1521 (52.5)19 (47.5)21 (52.5)41 -AbdominalLab PSGAHI ≥ 1517 (41.5)24 (58.5)17 (41.5)Commercial drivers Firat et al. 25 2012 (Turkey)85 (Bus drivers)Lab PSGAHI > 1546 (54.1)39 (46.9)46 (54.1)General population Silva et al. 26 2011 (USA)4,770 (CVD and respiratory disease patients)HSAT (Compumedics Portable PS-2)RDI ≥ 15948 (19.9)NARDI < 15 = 3,822 (80.2)RDI ≥ 15 603 (12.7)345 (7.2)AHI = apnea-hypopnea index, CABG = coronary artery bypass grafting, CVD = cardiovascular disease, HSAT = home sleep apnea testing, NA = not available, OSA = obstructive sleep apnea, PAF = paroxysmal atrial fibrillation, PSG = polysomnography, RDI = respiratory desaturation index, REI = respiratory event index (number of apneas and hypopneas per time in bed).Table 2 Demographic data of patients using STOP questionnaire.First Author, Year (Location)No. of patients (N)Study TypeAge, ySex,% MaleBMI, kg/m2STOP ScoreAHI/RDI (Mean)Minimum SpO2, %Sleep clinic population Pecotic et al. 11 2012 (Croatia)217Prospective54.3 ± 10.77730.1 ± 4.7NA31.4 ± 22.676.7 ± 13.1 El-Sayed et al. 22 2012 (Egypt)234Cross-sectional50.4 ± 11.385.537.8 ± 9.52.6 ± 1.045.6 ± 32.7NA Boynton et al. 27 2013 (USA)219Prospective46.3 ± 13.944.833.4 ± 8.8NANANA Banhiran et al. 28 2014 (Thailand)303ProspectiveNo OSA 49.4 ± 11.5 Mild OSA 50.2 ± 11.3 Moderate OSA 51.2 ± 11.5 Severe OSA 47.5 ± 12.561.4No OSA 24.4 ± 4.0 Mild OSA 26.1 ± 4.9 Moderate OSA 28.4 ± 4.9 Severe OSA 31.1 ± 6.3No OSA 1.7 ± 1.1 Mild OSA 2.1 ± 1.1 Moderate OSA 2.4 ± 1.1 Severe OSA 2.9 ± 0.9No OSA 1.9 ± 1.5 Mild OSA 9.6 ± 2.8 Moderate OSA 22.5 ± 4.4 Severe OSA 65.1 ± 14.9No OSA 90.0 ± 3.5 Mild OSA 83.6 ± 5.0 Moderate OSA 77.5 ± 8.6 Severe OSA 65.1 ± 14.9 Luo et al. 29 2014 (China)212Prospective44.8 ± 11.888.728.1 ± 3.72.5 ± 1.043.7 ± 28.274.3 ± 13.4 Pataka et al. 30 2014 (Greece)1852Retrospective52 ± 1474.432.8 ± 72.8 ± 0.933 ± 26.6NA Yuceege et al. 31 2014 (Turkey)185Prospective46.3 ± 10.365.9AHI ≤ 15: 29.2 ± 4.8 AHI > 15: 32.5 ± 6.1NANANA Sadeghniiat-Haghighi et al. 32 2015 (Iran)603Cross-sectional45.8 ± 12.774.829.2 ± 5.92.3 ± 1.0NA69.5 ± 28.1 Pataka et al. 33 2016 (Greece)204Prospective51.8 ± 13.877.532.8 ± 6.22.9 ± 0.929.7 ± 24.7NA Kashaninasab et al. 12 2017 (Iran)250Cross-sectional48.1 ± 1276NANA44 ± 31.2NA Prasad et al. 13 2017 (India)210Retrospective46.5 ± 13.772.931.9 ± 7.4NA39.2 ± 35.2NA Sangkum et al 2017 14 (USA)208Cross-sectional52.9 ± 0.936.136.9 ± 0.7NA18.6 ± 22.1NA Hong et al. 15 2018 (China)2,208Retrospective47.7 ± 13.977.926.5 ± 4.11.9 ± 1.124.4 ± 25.678.1 ± 13.8 Byun et al. 16 2020 (South Korea)778Cross-sectional49.3 ± 13.375.726.7 ± 4.1NA27.6 ± 39.080.2 ± 11.2 Medarov et al. 17 2020 (USA)344Prospective46.6 ± 144540.1 ± 9.92.3 ± 0.724 ± 24.9NA Wang et al. 18 2020 (China)105Retrospective46 ± 167526.6 ± 4.5NANA78.2 ± 10.6Medical population Katzan et al. 20 2016 (USA)208Retrospective55.5 ± 14.15131 ± 6.62.3 ± 0.75NANA Westlake et al. 21 2016 (Czech Republic)294Prospective63.9 ± 9.25931.1 ± 5.6NA13.6 ± 14.7NA Chen et al. 23 2019 (China)221Retrospective54 ± 12.779.627.2 ± 4.22.2 ± 1.125.5 ± 24.977.2 ± 13.1 May et al. 24 2020 (USA)300Case-control61.4 ± 11.963.331.4 ± 6.71.9 ± 113.4 ± 15.4NASurgical population Chung et al. 6 2008 (Canada)177Prospective55 ± 1349.730 ± 6NA20 ± 682 ± 11 Nunes et al. 19 2014 (Brazil)CABG - 40Prospective56 ± 77330 ± 4NA21 ± 1985.3 ± 8.5Abdominal - 4156 ± 86829 ± 5NA19 ± 1885.3 ± 4.6Commercial drivers Firat et al. 25 2012 (Turkey)85ProspectiveNA10029.4 ± 3.7NA21.1 ± 17.4NAGeneral population Silva et al. 26 2011 (USA)4,770Prospective62.4 ± 10.351.5NA2.9 ± 0.9NANAData are presented as mean ± SD where appropriate. AHI = apnea-hypopnea index, BMI = body mass index, CABG = coronary artery bypass grafting, NA = not available, OSA = obstructive sleep apnea, RDI = respiratory desaturation index.Among the sleep clinic population, the mean age was 49 ± 13 years and mean BMI was 30 ± 7 kg/m2. On the contrary, the surgical and medical population had mean age of 55 ± 12 years and 59 ± 13 years, respectively, with a mean BMI of 30 ± 6 kg/m2 in both populations. The 2 × 2 contingency tables and predictive parameters for each individual study are reported in Table S3.Quality assessment of included studiesThe internal validity results of the included studies are in Table S4. The types of study include 12 prospective studies, 6,11,17,19,21,25–29,31,33 6 retrospective studies, 13,15,18,20,23,30 5 cross-sectional studies, 12,14,16,22,32 and 1 case-control study. 24 All included studies used either laboratory PSG (20 studies, 6,11–16,18–20,22–25,27–32 84%), HSAT (2 studies, 21,26 8%) or a combination of both (2 studies, 17,33 8%) as a reference to verify the findings of the STOP questionnaire. No differences were observed in the capabilities of laboratory PSG compared to HSAT to detect OSA prevalence. Nine studies (38%) explicitly stated that physicians were blinded from interpreting the PSG and HSAT results and responses of the STOP questionnaire, 6,13,14,22,24,25,27,28,33 and 6 (25%) performed an independent interpretation of the questionnaire from clinical information. 6,13,22,27,28,33 These parameters suggest strong internal validity for 6 of the 24 included studies (25%). 6,13,22,27,28,33The evaluation of the included studies based on criteria for external validity is in Table S5. Twenty-one studies provided clear inclusion and exclusion criteria (88%), 6,12,14–33 while only 2 provided partial criteria (8%). 13,16 All studies met the criteria for providing enough information to identify the setting, no prescreening before the application of the STOP questionnaire, and inviting or randomly selecting patients to perform laboratory PSG/HSAT. All studies were able to provide age, sex, and BMI information for the studied population except for 1 that separated participants by the criterion of 45 years. 25 All studies used AHI/RDI ≥ 5 events/h as the definition of OSA with the exception of one study, which used AHI ≥ 15 events/h as the cut-off point for the diagnosis of OSA. 31 The percentage of participants missing or excluded from the initial sample pool was reported by 18 studies 6,13,15–19,21,23–27,29–33 (75%), and 8 (33%) provided clear explanations and basic characteristics for missing participants. 6,13,15,17,18,21,24,33Predictive characteristics of the STOP questionnaire in the sleep clinic populationThe pooled predictive parameters for STOP ≥ 2 are summarized (Table 3 and Figure 2). The prevalence of OSA (AHI ≥ 5 events/h), moderate-to-severe OSA (AHI ≥ 15 events/h), and severe OSA (AHI ≥ 30 events/h) was 80%, 60%, and 44%, respectively. A STOP score ≥ 2 showed excellent sensitivity at the different OSA severities (all OSA: 89.5% [95% CI, 88.6–90.4%], moderate-to-severe OSA: 90.5% [95% CI, 89.4–91.4%], severe OSA: 94.6% [95% CI, 93.5–95.6%]) and consistent diagnostic accuracy precision (all OSA and severe OSA: 0.74, moderate-to-severe OSA: 0.77). The PPV decreased with increasing OSA severity, which ranged from 50.5% (48.8–52.1%) to 84.3% (95% CI, 83.3–85.3%), while NPV increased and ranged from 43.8% (95% CI, 40.4–47.2%) to 84.8% (95% CI, 81.9–87.4%). The specificity is low across the different OSA severities, with values of 32.9% (95% CI, 30.1–35.8%), 30.9% (95% CI, 29.1–32.9%), and 24.4% (95% CI, 22.7–26.2%) for any OSA, moderate-to-severe OSA, and severe OSA, respectively.Table 3 The pooled predictive parameters of STOP ≥ 2 as the cutoff.Predictive Parameters (95% CI)All OSA (AHI ≥ 5 events/h)Moderate-to-severe OSA (AHI ≥ 15 events/h)Severe OSA (AHI ≥ 30 events/h)Sleep clinic population(13 studies; n = 5,536)(14 studies; n = 5,807)(11 studies; n = 4,400) Prevalence80.1 (79.0–81.2)60.4 (59.1–61.7)44.9 (43.4–46.3) Sensitivity89.5 (88.6–90.4)90.5 (89.4–91.4)94.6 (93.5–95.6) Specificity32.9 (30.1–35.8)30.9 (29.1–32.9)24.4 (22.7–26.2) PPV84.3 (83.3– 85.3)66.7 (65.3– 68.0)50.5 (48.8–52.1) NPV43.8 (40.4–47.2)68.0 (65.0–70.8)84.8 (81.9–87.4) Diagnostic odds ratio4.42 (3.17–6.15)5.03 (3.57–7.11)5.76 (3.84–8.64) SROC0.740.770.74Medical population(2 studies, n = 515)(3 studies, n = 665)(2 studies, n = 515) Prevalence73.4 (69.3–77.1)41.1 (37.3–44.9)20.2 (16.9–24.0) Sensitivity69.6 (64.7–74.2)76.6 (71.1–81.5)85.6 (77.3–91.7) Specificity51.8 (43.1–60.4)44.9 (39.9–50.0)41.1 (36.3–46.0) PPV79.9 (75.1– 84.0)49.2 (44.3–54.0)26.9 (22.2–32.0) NPV38.2 (31.2–45.6)73.3 (67.2–78.7)91.8 (86.7–95.2) Diagnostic odds ratio2.77 (0.95–8.10)2.48 (1.46–4.23)3.35 (1.84–6.08) SROC0.500.530.50Surgical population(1 study, n = 177)(3 studies, n = 258)(1 study, n = 177) Prevalence—41.9 (35.8–48.1)— Sensitivity—81.5 (72.9–88.3)— Specificity—40.7 (32.7–49.0)— PPV—49.7 (42.5–57.3)— NPV—75.3 (64.3–83.9)— Diagnostic odds ratio—2.94 (1.60–5.42)— SROC—0.71—Data expressed as percentage and 95% confidence interval (CI). AHI, apnea-hypopnea index, NPV = negative predictive value, OSA = obstructive sleep apnea, PPV = positive predictive value, SROC = area under summary of receiver operating characteristic curve.Figure 2: Forest plots: pooled sensitivity and specificity for OSA severity and population.Download FigureFigure 2: Forest plots: pooled sensitivity and specificity for OSA severity and population. (Continued)CI = confidence interval, FN = false negative, FP = false positive, OSA = obstructive sleep apnea, TN = true negative, TP = true positive.Download FigurePredictive characteristics of the STOP questionnaire in the medical populationThe prevalence of all OSA, moderate-to-severe OSA, and severe OSA in the medical population was 73%, 41%, and 20%, respectively. In this population, the STOP questionnaire showed a high sensitivity of 85.6% (95% CI, 77.3–91.7%) at severe OSA but a low to modest sensitivity at any OSA (69.6%; 95%CI, 64.7–74.2%) and moderate-to-severe OSA (76.6%; 95% CI, 71.1–81.5%). A STOP score ≥ 2 had excellent discriminative power to exclude severe OSA with 91.8% (95% CI, 86.7–95.2%) NPV. The PPV decreased from 79.9% (95% CI, 75.1–84.0%) to 26.9% (95% CI, 22.2–32.0%) with an increase in OSA severity.Predictive characteristics of the STOP questionnaire in the surgical populationThe prevalence of moderate-to-severe OSA in the surgical population was 41%. At this severity, a STOP score ≥ 2 showed a high pooled sensitivity (81.5%; 95% CI, 72.9–88.3%) but moderately low specificity (40.7%; 95% CI, 32.7–49.0%). The PPV and NPV were 49.7% (95% CI, 42.5–57.3%) and 75.3% (95% CI, 64.3–83.9%).Predictive characteristics of the STOP questionnaire in commercial driversThe STOP questionnaire was evaluated to detect moderate-to-severe OSA in commercial drivers by Firat et al. 25 The prevalence of moderate-to-severe OSA in this population was 54.1%. The sensitivity and specificity were 41.3% and 92.3%. The PPV and NPV were 86.4%% and 57.1%.Predictive characteristics of the STOP questionnaire in the general populationSilva et al evaluated the STOP questionnaire to detect moderate-to-severe and severe OSA in 4,770 participants in the Sleep Heart Health study. 26 The prevalence of moderate-to-severe OSA and severe OSA was 12.7% and 7.2%. The sensitivity for moderate-to-severe OSA and severe OSA was 62.0% and 68.8%, while the specificity was 56.3% and 59.5%, respectively.Meta-regression and sensitivity analysis of subgroupsThe meta-regression and sensitivity analysis of various confounders (continuous variables: age, sex, BMI, prevalence, and sample size) slightly changed the pooled estimates but did not impact the overall inference of our results (Table S6, Figure S1, Figure S2, and Figure S3). Similarly, the analysis of other confounders as categorical variables (study type and validation tool) did not influence the end inference of our results. To understand the impact/influence of a single study on the pooled estimates, we conducted the leave-one-out meta-analysis. There was no individual study that considerably affected the results as shown by the leave-one-out meta-analysis.DISCUSSIONThis systematic review and meta-analysis found that the STOP questionnaire was an effective tool to screen suspected OSA in the sleep clinic, medical, and surgical population. In the sleep clinic population, its high sensitivity ranging from 89.5 to 94.6% and consistent diagnostic accuracy (> 0.74 area under the summary receiver operating characteristic curves) across different severities of OSA helps identify those at high risk of OSA. For moderate-t" @default.
• W4200387958 created "2021-12-31" @default.
• W4200387958 creator A5004911176 @default.
• W4200387958 creator A5021767024 @default.
• W4200387958 creator A5035618784 @default.
• W4200387958 creator A5042138021 @default.
• W4200387958 creator A5080768708 @default.
• W4200387958 creator A5082293333 @default.
• W4200387958 creator A5088443346 @default.
• W4200387958 date "2022-05-01" @default.
• W4200387958 modified "2023-10-03" @default.
• W4200387958 title "Validation of the STOP questionnaire as a screening tool for OSA among different populations: a systematic review and meta-regression analysis" @default.
• W4200387958 cites W108384545 @default.
• W4200387958 cites W1490446417 @default.
• W4200387958 cites W1631437393 @default.
• W4200387958 cites W1726343988 @default.
• W4200387958 cites W1824523855 @default.
• W4200387958 cites W1989363300 @default.
• W4200387958 cites W1996293334 @default.
• W4200387958 cites W2011629460 @default.
• W4200387958 cites W2028774138 @default.
• W4200387958 cites W2042958005 @default.
• W4200387958 cites W2045667053 @default.
• W4200387958 cites W2060056139 @default.
• W4200387958 cites W2080427008 @default.
• W4200387958 cites W2082694529 @default.
• W4200387958 cites W2092218545 @default.
• W4200387958 cites W2093927492 @default.
• W4200387958 cites W2096763948 @default.
• W4200387958 cites W2120771911 @default.
• W4200387958 cites W2121013649 @default.
• W4200387958 cites W2121725012 @default.
• W4200387958 cites W2125397756 @default.
• W4200387958 cites W2126727011 @default.
• W4200387958 cites W2130864428 @default.
• W4200387958 cites W2139158349 @default.
• W4200387958 cites W2139168999 @default.
• W4200387958 cites W2152593325 @default.
• W4200387958 cites W2156098321 @default.
• W4200387958 cites W2156876629 @default.
• W4200387958 cites W2160762474 @default.
• W4200387958 cites W2285604059 @default.
• W4200387958 cites W2298714229 @default.
• W4200387958 cites W2401680069 @default.
• W4200387958 cites W241306484 @default.
• W4200387958 cites W246286872 @default.
• W4200387958 cites W2468554305 @default.
• W4200387958 cites W2483520518 @default.
• W4200387958 cites W2509358577 @default.
• W4200387958 cites W2515118094 @default.
• W4200387958 cites W2515710252 @default.
• W4200387958 cites W2549047171 @default.
• W4200387958 cites W2560289827 @default.
• W4200387958 cites W2570272552 @default.
• W4200387958 cites W2603912125 @default.
• W4200387958 cites W2623048247 @default.
• W4200387958 cites W2735970289 @default.
• W4200387958 cites W2787931652 @default.
• W4200387958 cites W2792504787 @default.
• W4200387958 cites W2792736409 @default.
• W4200387958 cites W2792981908 @default.
• W4200387958 cites W2793878697 @default.
• W4200387958 cites W2799314917 @default.
• W4200387958 cites W2804646038 @default.
• W4200387958 cites W282039873 @default.
• W4200387958 cites W2885734968 @default.
• W4200387958 cites W2938200700 @default.
• W4200387958 cites W2959442417 @default.
• W4200387958 cites W2982294814 @default.
• W4200387958 cites W2985225951 @default.
• W4200387958 cites W3030024133 @default.
• W4200387958 cites W3039767053 @default.
• W4200387958 cites W3073644491 @default.
• W4200387958 cites W3083107244 @default.
• W4200387958 cites W3085582654 @default.
• W4200387958 cites W3096943670 @default.
• W4200387958 cites W4240307074 @default.
• W4200387958 cites W4294215472 @default.
• W4200387958 doi "https://doi.org/10.5664/jcsm.9820" @default.
• W4200387958 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/34910625" @default.
• W4200387958 hasPublicationYear "2022" @default.
• W4200387958 type Work @default.
• W4200387958 citedByCount "6" @default.
• W4200387958 countsByYear W42003879582022 @default.
• W4200387958 countsByYear W42003879582023 @default.
• W4200387958 crossrefType "journal-article" @default.
• W4200387958 hasAuthorship W4200387958A5004911176 @default.
• W4200387958 hasAuthorship W4200387958A5021767024 @default.
• W4200387958 hasAuthorship W4200387958A5035618784 @default.
• W4200387958 hasAuthorship W4200387958A5042138021 @default.
• W4200387958 hasAuthorship W4200387958A5080768708 @default.
• W4200387958 hasAuthorship W4200387958A5082293333 @default.
• W4200387958 hasAuthorship W4200387958A5088443346 @default.
• W4200387958 hasBestOaLocation W42003879581 @default.
• W4200387958 hasConcept C105795698 @default.
• W4200387958 hasConcept C124936440 @default.
• W4200387958 hasConcept C126322002 @default.
• W4200387958 hasConcept C152877465 @default.

• next