Limitations of this การวิเคราะห์อภิมาน must be considered. Firstly,
most of the eligible studies were cross-sectional (75%) in design และ clinic-based (69%) in setting which may limit the interpretation of the results with respect to the general population. Secondly, the quality of the included studies was not always optimal, demonstrated by the lack of reporting of non-responders, though the magnitude of this problem may be too small to substantially alter our conclusion. Thirdly, there was significant ความไม่เหมือนกัน (heterogeneity) of prevalence estimate across studies which could be partially due to ความไม่เหมือนกัน (heterogeneity) in publication year, evaluation method, study set- ting, population origin และ disease duration. Fourthly, the possibility of publication bias could not be fully excluded by Egger’s test. Finally, though NPI scale is a validated และ widely used tool to assess อาการประสาทจิตเวช, it could not cover all the NPS. Nevertheless, when the trim และ fill analysis was conducted, the overall imputation did not change the general result, which indicated the results are robust to the possibility of unpublished negative studies.