Aim The aim of the current study was to estimate and to compare the diagnostic accuracy of the ISI, AIS, and PSQI in identifying insomnia.
Method We perform a systematic search in electronic databases including EMBASE, PubMed, PsycINFO, CINAHL, and Chinese Electronic Periodic Services from their inception to until May 20, 2015. Summary sensitivity, specificity, and diagnostic odds ratios (DOR) against a reference standard were calculated for each study. The revised Quality Assessment of Diagnostic Accuracy Studies was used to assess the quality of each included study. All analyses were conducted using Stata 14.0, with midas and metandi user-written commands, SAS 9.0.2,with Proc Mixed module, and Review Manager 5.3. Random effects bivariate model was used for analyses.
Results We included 19 studies with a total of 4693 participants in data analyses. The summary estimates for the ISI, AIS, and PSQI in identifying insomnia in the studies were as follows: 88% (95% confidence interval CI = 0.79–0.93), 91% (0.87–0.93) and 94% (0.86–0.98); specificity: 85% (0.68–0.94), 87% (0.68–0.95) and 76% (0.64–0.85); and DORs: 41.93 (8.77–200.33), 67.7 (23.4–196.1) and 53 (15.5–186.2), respectively. No significant difference was observed regarding the pooled sensitivity, specificity, and DORs among the ISI, AIS, and PSQI (all P > 0.05).
ConclusionsThe current evidence indicates that the ISI, AIS, and PSQI are comparable and useful instruments for identifying insomnia in terms of diagnostic properties. Moreover, after comparisons of the diagnostic properties, sleep domains and feasibility of the scales revealed that the AIS is the strongest and most appropriate among the three instruments.
See more of: Evidence-Based Practice Sessions: Oral Paper & Posters