Browsing by Author "Ahmadian, Robab"
Now showing 1 - 6 of 6
- Results Per Page
- Sort Options
Publication Combining binary and continuous biomarkers by maximizing the area under the receiver operating characteristic curve(Taylor & Francis, 2020-03-08) Ahmadian, Robab; Ercan, İlker; Sığırlı, Deniz; Yıldız, Abdulmecit; Ahmadian, Robab; ERCAN, İLKER; SIĞIRLI, DENİZ; YILDIZ, ABDULMECİT; Bursa Uludağ Üniversitesi/Tıp Fakültesi/Biyoistatistik Anabilim Dalı.; Bursa Uludağ Üniversitesi/Tıp Fakültesi/Nefroloji Anabilim Dalı.; 0000-0003-1550-639X; 0000-0002-2382-290X; AAE-5602-2019; ABF-2367-2020; AAA-7472-2021; ABF-2368-2020In any clinical case, a decision is made with the maximum possible accuracy. To achieve such accuracy, in the presence of multiple diagnostic tests or biomarkers, biomarker combinations aim to achieve maximum accuracy. As existing biomarker combination methods combine only continuous biomarkers, therefore in this study biomarker combination for binary biomarkers was created by suggesting an approach using Youden's J statistic for combining binary biomarkers. The proposed approach will facilitate binary and continuous biomarker combinations. A simulation study was conducted to compare the performance of our proposed combination approach according to different sample sizes. Both in the analysis of real data and the simulation studies for different samples, the proposed approach has been shown to yield favorable results and higher area under the curve.Publication Comparison of reference intervals derived by direct and indirect methods based on compatible datasets obtained in Turkey(Elsevier, 2021-05-26) Özarda, Yeşim; Ichihara, Kiyoshi; Jones, Graham; Streichert, Thomas; Ahmadian, Robab; Ahmadian, Robab; Bursa Uludağ Üniversitesi/Tıp Fakültesi/İstatistik Anabilim Dalı.; 0000-0003-1550-639X; AAE-5602-2019Background: Indirect derivation of reference intervals (RIs) from the laboratory information system (LIS) has been recently pursued. We aimed at evaluating the accuracy of indirectly predicted RIs compared to the RIs established directly from healthy subjects in the nationwide RI study in Turkey, targeting 25 major chemistry analytes.Methods: LIS data were retrieved from the laboratory that performed measurements for the direct study. They were cleaned by limiting to outpatients with age 18-65 years, and by allowing only one record per year per patient. Evaluated were four indirect methods of univariate approach: Hoffmann, Bhattacharya, Arzideh, and Wosniok methods. Power transformation of the LIS dataset was performed either using the power (lambda) reported by the IFCC global RI study (the first two methods) or using a lambda predicted (the last two).Results: Compared to the direct study dataset, the LIS dataset showed a variable degree of alterations in peak location and shape. Consequently, lower-side peak-shifts observed in sodium, albumin, etc. led to lowered RI limits, whereas higher-side peak-shift observed in triglyceride, low-density lipoprotein cholesterol, etc. led to raised RI limits. Overall, 72% (62-81) of the RI limits predicted by indirect methods showed significant biases from direct RIs. However, the biases observed in total cholesterol, lactic dehydrogenase, etc. were attributed to a higher-side age-bias in LIS dataset. After excluding them, the overall proportion of biased RIs was reduced to 47% (38-54).Conclusion: To reduce prediction biases that remained after age adjustment, it is necessary to apply more rigorous data-cleaning before applying indirect methods.Publication Evaluating the sensitive question methods; Recommended Uludag adjustment for the crosswise model(Taylor, 2021-10-27) Ahmadian, Robab; Ercan, İlker; Ahmadian, Robab; ERCAN, İLKER; Bursa Uludağ Üniversitesi/Tıp Fakültesi/Temel Tıp Bilimleri/Biyoistatistik; 0000-0003-1550-639X; 0000-0002-2382-290X; AAE-5602-2019; ABF-2367-2020Sensitive questions are frequently asked in medical, psychological, and sociological research. In the social science literature, it is widely accepted that when respondents are asked such questions, social desirability bias affects survey research. Different methods in the literature aimed to reduce measurement errors such as social desirability bias and increase the reliability of participants' answers. These methods address the problem to increase the effectiveness of predictions by using indirect questioning techniques. The sensitive question methods evaluated in this study are randomized response technique, grouped answer method, unmatched-count technique, and crosswise model. These methods were evaluated with a comprehensive simulation study. The performances of the methods were evaluated according to the sample size and the sensitive issue prevalence in the hypothetical population. In the second part, in order to reduce the small sample size and low prevalence effect of the crosswise model, the Uludag correction of the crosswise model was proposed in the study. As a result, the crosswise model performs quite well compared to the other methods. In the use of the crosswise model, it is recommended to use the Uludag correction of the model proposed in this study in cases where low prevalence and small samples are studied.Item Hassas sorular sormaya yönelik yöntemlerin performansının karşılaştırması(Bursa Uludağ Üniversitesi, 2021-01-07) Ahmadian, Robab; Ercan, İlker; Bursa Uludağ Üniversitesi/Sağlık Bilimleri Enstitüsü/Biyoistatistik Anabilim Dalı.; 0000-0003-1550-639XTıbbi, psikolojik ve sosyolojik araştırmalarda yoğunlukla hassas konular veya son derece kişisel sorular sorulur. Hassas sorular üzerine anket araştırması, katılımcıların genellikle gerçeği ifade etmek istememesinden dolayı veya tamamen cevap vermeyi reddetmesinden dolayı zordur. Sosyal bilim literatüründe, katılımcılara kişisel ve hassas sorular sorulduğunda, sosyal tercih edilebilirlik yanlılığının anket araştırmasını etkilediği yaygın olarak kabul edilmektedir. Literatürde sosyal tercih edebilirlik yanlılığı gibi ölçüm hatalarını azaltmayı ve katılımcıların yanıtlarının güvenilirliğini artırmayı amaçlayan farklı yöntemler bulunmaktadır. Bu yöntemler dolaylı soru sorma teknikleri kullanımıyla, tahminlerin etkinliği yükseltmek için sosyal tercih edilebilirlik yanlılığı sorununu ele almaktadır. Tez çalışmamız ile değerlendirilen hassas soru yöntemleri, randomize cevap yöntemi, aday yöntemi, gruplandırılmış cevap yöntemi, eşsiz-sayım yöntemi ve çapraz modelidir. Tez çalışmasında bu yöntemler yapılan kapsamlı simülasyon planı ile karşılaştırılmıştır. Yöntemlerinin performansları örneklem büyüklüğüne ve popülasyonun mevcut hassas konu yaygınlık parametresine göre değerlendirilmiştir. İkinci bölümde, hassas konuların yaygınlık tahmini için kullanılan çapraz (Crosswise) modelin küçük örneklem büyüklüğü ve düşük prevelans etkisini azaltmak için tez çalışmamızda çapraz yöntemin Uludağ düzeltmesi önerilmiştir. Sonuç olarak, çapraz model diğer yöntemlere göre oldukça iyi performans göstermektedir. Çapraz modelin kullanımında, düşük prevelans ve küçük örneklemler ile çalışıldığı durumlarda bu tez çalışmasında önerilen çapraz modelin Uludağ düzeltmesinin kullanılması önerilir.Publication Knowledge and attitudes towards antibiotics use and an examination on patient's unrealistic health symptoms in Turkey(Modestum Ltd, 2022-04-01) Ahmadian, Robab; Uncu, Yeşim; Ercan, İlker; Toluk, Özlem; Ahmadian, Robab; UNCU, YEŞİM; ERCAN, İLKER; Toluk, Özlem; Bursa Uludağ Üniversitesi/Tıp Fakültesi/Biyoistatistik Bölümü Anabilim Dalı.; Bursa Uludağ Üniversitesi/Tıp Fakültesi/Aile Hekimliği Anabilim Dalı.; 0000-0003-1550-639X; 0000-0002-2382-290X; 0000-0001-6495-0839; AAL-6835-2021; D-9597-2016; ABF-2367-2020; AAE-5602-2019Background: This study aimed to assess the attitudes and knowledge of adult individuals about antibiotics. Since in Turkey, antibiotics are not sold without a prescription, it was also investigated whether Turkish people compelled physicians to prescribe antibiotics by showing unrealistic symptoms.Methods: A self-administered cross-sectional survey involving 1,057 respondents was conducted using a validated questionnaire. The questionnaire was prepared in four sections. The first section included attitude assessment and a five-point Likert scale has collected information. In the second section, knowledge questions examined participants' knowledge about antibiotics with Three-point Likert scale. The third section contained a sensitive question, and using the crosswise model was indirectly asked whether the participant has forced the physician to prescribe an antibiotic. The reliability test results for the attitude scale showed the a-value of Cronbach was 0.767, and KR-21 reliability for the knowledge scale was 0.713. Test-retest reliability coefficients for attitude scale was r=0.697, the coefficients for knowledge scale was r=0.887.Results: Turkish people's awareness about antibiotics is insufficient but at an acceptable level. The public is aware that antibiotics rapidly develop resistance when used widely and incorrectly. Nevertheless, %17 of them resort to showing unrealistic symptoms to force the physician to prescribe antibiotics.Conclusion: To conclude, Turkish people's knowledge and attitudes about antibiotics are insufficient but acceptable. However, a noticeable rate of patients resorts to showing unrealistic symptoms to force the physician to prescribe antibiotics.Item A nonnormal look at polychoric correlations: Modeling the change in correlations before and after discretization(Springer, 2016-03-08) Demirtaş, Hakan; Ahmadian, Robab; Atış, Sema; Can, Fatma Ezgi; Ercan, İlker; Uludağ Üniversitesi/Tıp Fakültesi/Biyoistatistik Anabilim Bölümü.; 0000-0003-1550-639X; 0000-0002-2382-290X; 0000-0002-1953-7735; AAE-5602-2019; 56689608500; 57185433800; 57185484200; 6603789069Two algorithms for establishing a connection between correlations before and after ordinalization under a wide spectrum of nonnormal underlying bivariate distributions are developed by extending the iteratively found normal-based results via the power polynomials. These algorithms are designed to compute the polychoric correlation when the ordinal correlation is specified, and vice versa, along with the distributional properties of latent, continuous variables that are subsequently ordinalized through thresholds dictated by the marginal proportions. The method has broad applicability in the simulation and random number generation world where modeling the relationships between these correlation types is of interest.