Tıbbi araştırmalarda seçilen istatistiksel yöntemlerin önemi: genç erkek popülasyonda yüzeysel variköz ven risk faktörlerinin kestirimi üzerine modelleme çalışması

Year 2020, , 231 - 239, 18.06.2020

Feray Aydın , Dilek Dülger , Özgür Albuz

https://doi.org/10.32322/jhsm.695341

Abstract

Amaç: Hangi koşullar altında tek değişkenli analizin yetersiz olduğunu ve tıbbi araştırmalarda multipl lojistik regresyonun önemini açıklığa kavuşturmak birincil amaçtır. Bu amaçla retrospektif olarak planlanan genç erkek popülasyonda alt ekstremite yüzeysel venöz risk faktörleri model olarak seçildi.
Gereç ve Yöntem: Bu çalışma için 30 Mayıs 2016-1 Ocak 2019 tarihleri arasında 29 Mayıs Devlet Hastanesi Genel Cerrahi Kliniği’ne kaşıntı, ödem, bacak ağrısı ve şişlik gibi semptomları olan tüm hastalar retrospektif olarak incelendi.
Bulgular: Tek değişkenli analiz sonuçları, variköz venlerin gelişiminin, değişkenler açısından kontrol grubundan anlamlı derecede yüksek olduğunu göstermiştir. Sırasıyla, aile öyküsü [34 (%61,8) vs. 21 (%38,2)], p<0,008), sigara içme [37 (%59,7) ve 25 (%40,3), p<0,04], uzun süreli [90 (%53,6) 78’e karşı (%46,4), p<0,01], kabızlık [64 (%55,2) - 52 (% 44,8), p<0,001] ve kentsel yaşam tarzı [49 (%56,7) - 22 (%43,3), p<0,04], varikoz venlerinin gelişimi üzerine etkili bulunmuştur. Oysa, önemli tek değişkenli sonuçlar için ikili lojistik regresyonunu yaptığımızda; Variköz venlerin risk faktörleri açısından farklı sonuçlar saptadık; Sonuçlar aile öyküsünün, sigara içmenin, uzun süre ayakta durmanın ve kabızlığın variköz venlerin gelişiminde katkıda bulunduğunu ortaya koydu [(%61,8’e karşı %38,2, AOR: 2,62 (1,35 ve 5,07), p<0,04))], [(%59,7’ye karşı %40,3, AOR: 2,08 (1,13 ve 3,8) p<0,02)], [(%53,6’a karşı %46,4, AOR: 1,83 (1,06 ve 3,16) p<0,03)], [(%55,2 ve %44,8) AOR: 1,74 (1,03 ve 2,92) p<0,04)]. Öte yandan, kentsel yaşam tarzı ise önemsiz bulundu [(%56,7 ve %43,3, AOR: 1,73 (0,96 ve 3,15) p<0,07)*].
Sonuç: Bağımlı değişkeni etkileyen birden fazla bağımsız değişken olduğu durumlarda; bağımlı değişkenin bağımsız değişkenlerden nasıl ve ne şekilde etkilendiğinin tespitinde; çoklu lojistik regresyon testleri, tek değişkenli analizden daha başarılıdır. Tıbbi araştırmalarda seçilen yanlış veya eksik istatistiksel çalışmalar çok önemli yanlış klinik yorumlara neden olabilir.

Keywords

İstatistiksel ve matematiksel yöntemler, tıbbi araştırma, kestirimde modelleme

Supporting Institution

Yoktur

Project Number

Yoktur

The importance of the chosen statistical methods in medical research: study over modelling in estimation superficial varicose vein risk factors in young male population

Abstract

Objective: To clarify the under which conditions univariate analysis is insufficient and the importance of multiple logistic regression in medical research is the primary objective. For this purpose, lower extremity superficial venous risk factors were selected as a model in a young male population retrospectively planned.
Material and Method: All patients who presented to the General Surgery Clinic of 29 Mayıs State Hospital with symptoms of pruritus, edema, leg pain and swelling between May 30, 2016 and January 1, 2019 were retrospectively analyzed for this study.
Results: The results of univariate analysis showed that the development of varicose veins was significantly higher in terms of variables than in the control group. Family history [34 (61.8%) vs. 21 (38.2%)], p<0.008), smoking [37 (59.7%) and 25 (40.3), p<0.04], long-term standing, respectively [90 (53.6%) vs. 78 (46.4%), p<0.01], constipation [64 (55.2%) vs. 52 (44.8%), p<0.001] and the urban lifestyle [49 (56.7%) - 22 ( 43.3), p<0.04] were found to be effective on the development of varicose veins. However, when we do binary logistic regression for important univariate results; We found different results in terms of risk factors of varicose veins. The results showed that family history, smoking, long standing and constipation contributed to the development of varicose veins [(61.8% vs. 38.2%, AOR: 2.62 (1.35, 5.07), p (0.04)], [(59.7% vs. 40.3%, AOR: 2.08 (1.13, 3.8) p<0.02)], [(53.6%) 46.4%, AOR: 1.83 (1.06, 3.16) p<0.03)], [(55.2% and 44.8%) AOR: 1.74 (1.03, 2.92) p<0.04) ]. On the other hand, urban lifestyle was found to be insignificant [(56.7% and 43.3%, AOR: 1.73 (0.96, 3.15) p<0.07)*].
Conclusion: In cases where there is more than one independent variable affecting the dependent variable; In determining how and in what way the dependent variable is affected by independent variables; multiple logistic regression tests are more successful than univariate analysis. False or incomplete statistical studies selected in medical research may lead to very important false clinical interpretations.

Keywords

Statistical and mathemical methods, medical research, modelling in estimation

Project Number

Yoktur

References

• 1. Köklü N, Büyüköztürk Ş, Çokluk –Bökeoğlu Ö. Sosyal bilimler için istatistik. 2. Baskı. Pegem A Yayıncılık, 2007.
• 2. Arıcan E. Moleküler biyolojide kullanılan biyoistatistiksel yöntemler. file:///C:/Users/user/ Downloads /biyoistatistik_1_1.pdf.2017. Erişim Tarihi: 08.05.2017.
• 3. Belle GV, Fisher LD, Heagerty PJ, Lumley T. Bioistatistics: A methodology for the health sciences. 2st ed. John Wiley-Sons; 2010; 552-61.
• 4. Taşdelen B, Kanık EA. Sağlık araştırmalarında biyoistatistiksel yöntemlerin doğru kullanımı ve sunumu. Mersin Univ. Sağlık Bilimleri Derg 2009; 1: 1-13.
• 5. Bircan H. Lojistik regresyon analizi: Tıp verileri üzerine bir uygulama. Kocaeli Üniversitesi Sosyal Bilimler Enstitüsü Derg 2001; 2: 185-208.
• 6. Callam MJ. Epidemiology of varicose veins. Br J Surg 1994; 81: 167-73.
• 7. Rabe E, Partsch H, Jünger M, et al. Guidelines for clinical studies with compression devices in patients with venous disorders of the lower limb. Eur J Vasc Endovasc Surg 2008; 35: 494-500.
• 8. Cannon J, Rajakaruna G, Dyer J, Carapetis J, Manning L. Severe lower limb cellulitis: defining the epidemiology and risk factors for primary episodes in a population-based case-control study. Clin Microbiol Infect 2018; 24: 1089-94.
• 9. Porter JM, Moneta GL. Reporting standards in venous disease: an update. International Consensus Committee on Chronic Venous Disease. J Vasc Surg.1995; 21: 635–45.
• 10. Robertson L, Evans C, Fowkes FG. Epidemiology of chronic venous disease. Phlebology 2008; 23: 103-11.
• 11. Carradice D. Superficial venous insufficiency from the infernal to the endothermal. Ann R Coll Surg Engl 2014; 96: 5–10.
• 12. Öncü Öner T, Can Ş. Sağlıkta biyoistatistiksel uygulamalar. İzmir Kâtip Çelebi Üniversitesi Sağlık Bilimleri Fak Derg 2018; 3: 39-45.
• 13. Vollmer RT. Multivariate statistical analysis for pathologists. Part I, The logistic model. Am J Clin Pathol 1996; 105: 115–26.
• 14. Lemeshow S, Hosmer DW. Logistic regression. In: Armitage P, Colton T, Eds. Encyclopedia of Biostatistics. New York: J. Wiley, 1998. p. 2316–27.
• 15. Glantz SA, Slinker BK. Primer of applied regression and analysis of variance. New York: McGraw-Hill, Inc., 1990. ISBN:0070234078.
• 16. Hosmer WD, Lemeshow S, Klar J. Goodness of fit testing for multiple logistic regression analysis when the estimated probabilities are small. Biometrical J 1988; 30: 911-24.
• 17. Robert G, Rao JNK, Kumar s. Logistic regression analysis of sample survey data. Biometrika 1987; 74: 1–12,
• 18. Şahin M, Efe E. Lojistik regresyon yöntemi ile doğum ağılığına etki eden faktörlerin belirlenmesi. Black Sea J Health Sci 2018; 1: 22-7.
• 19. Chatfield C, Collins A. Introduction to Multivariate Analysis. Chapman & Hall, London, 1992.
• 20. Gibbons RD, Hedeker D. Random effects probit and logistic regression models for three-level data. Biometrika 1997; 53: 1527-37.
• 21. Scott A, Wild CJ. Fitting logistic regression models in stratified case-control studies. Biometrics 1991; 47: 497-510.
• 22. Feinstein AR. Multivariable analysis: an introduction. New Haven, CT: Yale University Press, 1996.
• 23. Dietz K, Gail M, Krickeberg K, Tsiatis A, Samet J. Statistics for Biology and Health. Logistic Regression A Self-Learning Text. 2 st ed. With Contributions by Erica Rihl Pryor 2002; 102-24.
• 24. Imai K, Enamorado T. POL572 Quantitative Analysis II Spring 2016; 63-5.
• 25. Van Doornmalen JPCM, Kopinga K. Temperature dependence of F‐, D‐ and z‐values used in steam sterilization processes. J Applied Microbiol 2009; 107: 1054-60.
• 26. Cox DR. Snell ES. Analysis of Binary Data. 2st ed. Chapman and Hall, London. 1989.
• 27. Jerome C. Joint dependence of risk of coronary heart disease on serum cholesterol and systolic blood pressure: a discriminant function analysis. Fed Proc 1962; 21: 58-61.
• 28. Korkmaz M, Güney S, Yiğiter ŞY. The importance of logistıc regression implementations in the Turkish livestock sector and logistic regression implementations/fields. J Agric Fac HRU 2012; 16: 25-36.
• 29. Frasin BA. Coefficient bounds for certain classes of bi-univalent functions. Hacettepe J MathematicsStatistics 2014; 43: 383–9.
• 30. Bagley SC, White H, Golomb BA. Logistic regression in the medical literature: Standards for use and reporting, with particular attention to one medical domain. J Clin Epidemiol 2001; 54: 979–85.
• 31. Trendelenburg F. “Über die Unterbindung der Vena saphena magna bei Unterschenkelvaricen”. [Brun’s] Beiträge zur Klinischen Chirurgie 1891; 7: 195-210.
• 32. Scott TE, Lamorte WW, Gorin DR, et al. Risk factors for chronic venous insufficiency: A dual casecontrol study. J Vasc Surg 1995; 22: 622-8.
• 33. Bradbury A, Evans C, Allan P, Lee A, Ruckley CV, Fowkes FGR. What are the symptoms of varicose veins? Edinburgh vein study cross sectional population survey. BMJ 1999; 318: 353–6.
• 34. Tolu İ, Durmaz MS. Frequency and significance of perforating venous insufficiency in patients with chronic venous ınsufficiency of lower extremity. Eurasian J Med 2018; 50: 99–104.
• 35. Pakdemirli A, Toksöz F, Karadağ A, Mısırlıoğlu HK, Başpınar Y, Ellidokuz H, Açıkgöz O.Role of mesenchymal stem cell-derived soluble factors and folic acid in wound healing. Turk J Med Sci 2019; 49: 914-21.
• 36. Athanerey A, Patra PK, Kumar A. J. Mesenchymal stem cell in venous leg ulcer: An intoxicating therapy. Tissue Viability 2017; 26: 216-23.
• 37. Burkitt DP. Varicose veins, deep vein thrombosis and hemorrhoids: Epidemiology and suggested aetiology. Br Med J 1972; 2: 556-61.
• 38. Mishra S, Ali I, Singh G. A study of epidemiological factors and clinical profile of primary varicose veins. Med J Dr. DY Patil University 2016; 9: 617-21.
• 39. Cornu-Thenard A, Boivin P, Baud JM, De Vincenzi I, Carpentier PH. Importance of the familial factor in varicose disease. Clinical study of 134 families. J Dermatol Surg Oncol 1994; 20: 318-26.
• 40. Musil D, Kaletova M, Herman J. Vasa. Risk factors for superficial vein thrombosis in patients with primary chronic venous disease. Vasa 2016; 45: 63-6.
• 41. Yun MJ, Kim YK, Kang DM, et al. A study on prevalence and risk factors for varicose veins in nurses at a university. Hospital Saf Health Work 2018; 9: 79–83.
• 42. Cleave TL. The neglect of natural principles in current medical practice. J R Nav Med Serv 1956; 42: 55-83.
• 43. Schultz-Ehrenburg U, Weindorf N, Von Uslar D, et al. Prospektive epidemiologische Studie &uuml; ber die Entstehungsweise der Krampfadern bei Kindern und Jungendlichen (Bochumer Studie I und II). Phlebol Proktol 1989; 18: 3-11.
• 44. Stocking K, Wilkinson J, Lensen S, Brison DR, Roberts SA, Vail A. Are interventions in reproductive medicine assessed for plausible and clinically relevant effects? A systematic review of power and precision in trials and meta-analyses. Hum Reprod 2019; 34: 659–65.
• 45. Baveja C P, Aggarwal P. Statistical analysis of microbiological diagnostic tests. Indian J Med Microbiol 2017; 35: 184-93.
• 46. Zhou J, Li Q, Gong G, Gong H, Hou Z. Analysis of prognostic factors and construction of a logistic regression model for patients with drug-induced liver failure. Zhong Nan Da Xue Xue Bao Yi Xue Ban=J Central South Univ Med Sci 2018; 43: 1337-44.
• 47. Sunderland KM, Beaton D, Fraser J, et al. The utility of multivariate outlier detection techniques for data quality evaluation in large studies: an application within the ONDRI Project. Manuel Montero-Odasso. BMC Medical Research Methodology 2019; 19: 1-16.