{"id":509,"date":"2020-02-05T21:17:18","date_gmt":"2020-02-05T18:17:18","guid":{"rendered":"https:\/\/www.aciltipakademisi.org\/?p=509"},"modified":"2021-11-16T04:10:01","modified_gmt":"2021-11-16T01:10:01","slug":"odds-orani-rolatif-risk-nnt-ve-nnh","status":"publish","type":"post","link":"https:\/\/tatd.org.tr\/atak\/2020\/02\/05\/odds-orani-rolatif-risk-nnt-ve-nnh\/","title":{"rendered":"Odds Oran\u0131, R\u00f6latif Risk, NNT ve NNH"},"content":{"rendered":"\n

Randomize kontroll\u00fc bir \u00e7al\u0131\u015fmada genellikle \u201cevet-hay\u0131r\u201d \u015feklinde dikotom (iki se\u00e7enekli) sonlan\u0131m \u00f6l\u00e7\u00fctleri de\u011ferlendirilir (\u00f6ld\u00fc-ya\u015f\u0131yor, hasta-de\u011fil, tan\u0131 var-yok). Makalede bu hedef sonlan\u0131m\u0131n geli\u015fti\u011fi ve geli\u015fmedi\u011fi hastalar\u0131n hem tedavi hem de kontrol gruplar\u0131ndaki oranlar\u0131 kar\u015f\u0131la\u015ft\u0131r\u0131l\u0131r. Bu tip bir sonlan\u0131m\u0131 bildirmenin olmazsa olmaz \u015fekli 2 x 2 Tablo<\/strong> kullanmakt\u0131r.<\/p>\n\n\n\n

\u00d6rne\u011fin, tedavi grubunda hastalar\u0131n %15\u2019inin, kontrol grubunda ise %20\u2019sinin k\u00f6t\u00fc bir sonlan\u0131ma ula\u015ft\u0131\u011f\u0131 (mesela \u00f6ld\u00fc\u011f\u00fc ya da maj\u00f6r bir olayla kar\u015f\u0131la\u015ft\u0131\u011f\u0131) Tablo 1<\/strong>\u2019de 2×2 tablosu verilen bir \u00e7al\u0131\u015fmay\u0131 ele alal\u0131m:<\/p>\n\n\n\n

\"Tablo<\/a>
Tablo 1. Farazi bir randomize kontroll\u00fc \u00e7al\u0131\u015fman\u0131n hedef sonlan\u0131mdaki hasta say\u0131lar\u0131<\/figcaption><\/figure><\/div>\n\n\n\n

Bu noktada baz\u0131 temel terimlerin tan\u0131mlamas\u0131n\u0131 yapmam\u0131z ve baz\u0131 temel hesaplamalar i\u00e7in gerekli form\u00fclleri vermemiz gerekli: (Kan\u0131ta Dayal\u0131 T\u0131p dili \u0130ngilizce oldu\u011fundan ve dilimizde bu terimlerin oturmu\u015f bir kar\u015f\u0131l\u0131\u011f\u0131 bulunmad\u0131\u011f\u0131ndan terimlerin kendimize g\u00f6re T\u00fcrk\u00e7esinin yan\u0131 s\u0131ra \u0130ngilizce kar\u015f\u0131l\u0131klar\u0131 da parantez i\u00e7inde verilmi\u015ftir.) Bu tan\u0131mlamalar\u0131 yaparken Tablo 2\u2019deki<\/strong> kutular\u0131n isimlendirmelerinden faydalanaca\u011f\u0131z.<\/p>\n\n\n\n

\"Tablo<\/a>
Tablo 2. Farazi \u00e7al\u0131\u015fma sonu\u00e7lar\u0131n\u0131n form\u00fcllerle yeniden d\u00fczenlenmi\u015f hali<\/figcaption><\/figure><\/div>\n\n\n\n

Tedavi verilen ya da Deneysel grubun tedavi etkisini ara\u015ft\u0131rd\u0131\u011f\u0131m\u0131z sonlan\u0131ma (olay, yani burada \u00f6l\u00fcm) ula\u015fma oran\u0131na deneysel grup riski [Experimental Group Risk = EGR]<\/em>, ya da deneysel olay h\u0131z\u0131 [Experimental Event Rate = EER] <\/em><\/strong>ad\u0131 verilir.<\/p>\n\n\n\n

Kontrol grubunda ayn\u0131 oran ise kontrol grup riski [Control Group Risk = CGR]<\/em>, ya da kontrol olay h\u0131z\u0131 [Control Event Rate = CER]<\/em> <\/strong>olarak adland\u0131r\u0131l\u0131r ve Form\u00fcl 1 ve 2<\/strong>\u2019deki \u015fekilde hesaplan\u0131rlar.<\/p>\n\n\n\n

\"EER_CER_formul\"<\/a><\/figure><\/div>\n\n\n\n

Bu iki oran aras\u0131ndaki fark, verdi\u011fimiz deneysel tedavinin kontrol grubundaki \u00f6l\u00fcm oran\u0131n\u0131 ne kadar azaltt\u0131\u011f\u0131d\u0131r. \u00d6rnek \u00fczerinden gidersek, EER %15 (tedavi grubunda 100 hastada 15\u2019i \u00f6lm\u00fc\u015f) CER %20 (kontrol grubunda da 100 hastadan 20\u2019si \u00f6lm\u00fc\u015f) iken tedavinin %20 olan \u00f6l\u00fcm oran\u0131n\u0131 %5 azaltarak %15\u2019e indirdi\u011fini s\u00f6yleyebiliriz.<\/p>\n\n\n\n

\u0130ki oran aras\u0131ndaki net farka net risk azalmas\u0131 [Absolute Risk Reduction = ARR]<\/em>) <\/strong>ad\u0131 verilir (Form\u00fcl 3)<\/strong>.<\/p>\n\n\n\n

\"ARR\"<\/a><\/figure><\/div>\n\n\n\n

Tedavinin etkinli\u011fini bildirirken kulland\u0131\u011f\u0131m\u0131z bir di\u011fer terim R\u00f6latif Risk ya da Risk Oran\u0131<\/strong>\u2019d\u0131r. Deneysel tedavi verilen hastalar\u0131n \u00f6l\u00fcm riskinin (EER)<\/strong> kontrol grubundaki hastalar\u0131n \u00f6l\u00fcm riskine (CER)<\/strong> oran\u0131 olarak ifade edilir. \u00d6rne\u011fimize g\u00f6re,<\/p>\n\n\n\n

\"RR\"<\/a><\/figure><\/div>\n\n\n\n

Yani tedavi grubundaki hastalarda g\u00f6r\u00fclen \u00f6l\u00fcm riski (oran\u0131) kontrol grubunun %75\u2019i kadard\u0131r.<\/p>\n\n\n\n

\"RR2\"<\/a><\/figure><\/div>\n\n\n\n

Bu tip dikotom sonu\u00e7lar bildirilirken en \u00e7ok kullan\u0131lan yol ise r\u00f6latif riskin bir ba\u015fka \u015fekilde ifadesi olan R\u00f6latif Risk Azalmas\u0131<\/strong> de\u011ferinin belirtilmesidir (Relative Risk Reduction = RRR)<\/em>(Form\u00fcl 5).<\/strong> Kontrol grubunda olsayd\u0131 \u00f6lecek olan hastalar\u0131n %25\u2019i tedavi al\u0131rsa \u00f6lmeyecektir.<\/p>\n\n\n\n

\"RRR\"<\/a><\/figure><\/div>\n\n\n\n

E\u011fer ara\u015ft\u0131rmac\u0131lar R\u00f6latif Riski (RR)<\/strong> belirli bir zaman dilimi i\u00e7in hesaplarlarsa buna da Hazard Oran\u0131<\/strong> (Riziko Oran\u0131)<\/strong> ad\u0131 verilir.<\/p>\n\n\n\n

\u00c7al\u0131\u015fmada belirtilmese bile \u00e7o\u011fu zaman \u201cX ilac\u0131 \u00f6l\u00fcm riskini azaltmada %30 ba\u015far\u0131l\u0131yd\u0131\u201d<\/em> ya da \u201cY tedavisinin etkinli\u011fi %80 idi\u201d<\/em> gibi bir \u00f6nermeyle asl\u0131nda R\u00f6latif Risk Azalmas\u0131<\/strong>ndan (RRR)<\/strong> bahsedilir. E\u011fer deneysel tedavi fayda sa\u011fl\u0131yorsa RRR isim de\u011fi\u015ftirerek r\u00f6latif fayda art\u0131\u015f\u0131 (Relative Benefit Increase = RBI)<\/strong>, e\u011fer kontrol grubunda g\u00f6r\u00fclen zarar\u0131 artt\u0131r\u0131yorsa r\u00f6latif zarar art\u0131\u015f\u0131<\/strong> (Relative Risk Increase = RRI) <\/strong>olarak ifade edilir.<\/p>\n\n\n\n

R\u00f6latif Risk Azalmas\u0131 (RRR) <\/strong>kullanman\u0131n en \u00f6nemli dezavantaj\u0131 tedavi verilmeyen (kontrol grubundaki) hastalar\u0131n olay riski (CER) hakk\u0131nda hi\u00e7bir bilgi vermemesidir. Bu sebeple de \u00e7ok k\u00fc\u00e7\u00fck riskler olsa bile aralar\u0131ndaki oransal fark\u0131 ifade etmesi y\u00fcz\u00fcnden etki \u00e7ok b\u00fcy\u00fck gibi g\u00f6r\u00fcnebilir. \u00d6rne\u011fin, %0,000012 olan \u00f6l\u00fcm riski tedavi ile %0,000009\u2019a d\u00fc\u015f\u00fcyorsa asl\u0131nda tedavinin sa\u011flad\u0131\u011f\u0131 net risk azalmas\u0131 sadece %0,000003 iken r\u00f6latif risk azalmas\u0131 %25\u2019dir. Bu gibi durumlarda net risk azalmas\u0131n\u0131n (ARR) <\/strong>ifade edilmesi \u00e7ok daha do\u011frudur. Bu tip \u00f6nermeleri farmas\u00f6tik \u015firketlerinin hekimlere tan\u0131t\u0131mda kulland\u0131\u011f\u0131 materyallerde s\u0131kl\u0131kla g\u00f6rebilirsiniz.<\/p>\n\n\n\n

Bu ifadelerin elle tutulur bir ger\u00e7e\u011fe d\u00f6n\u00fc\u015ft\u00fcr\u00fclmesi ihtiyac\u0131yla tedavi edilmesi gereken say\u0131<\/strong> (Number Needed to Treat (NNT)) <\/strong>terimi tan\u0131mlanm\u0131\u015ft\u0131r (Form\u00fcl 6)<\/strong>. Net risk azalmas\u0131n\u0131n tersi olan bu tam say\u0131 g\u00fcndelik pratikte alg\u0131lanabilir bir \u00f6l\u00e7ektir. \u00c7al\u0131\u015fmadan belirtilen tedavi s\u00fcresi boyunca geli\u015febilecek k\u00f6t\u00fc sonlan\u0131m say\u0131s\u0131n\u0131 (olay, \u00f6l\u00fcm vb.) 1 tane azaltabilmek i\u00e7in ka\u00e7 hastaya deneysel tedavinin uygulanmas\u0131 gerekti\u011fini g\u00f6sterir. Bir \u00f6nceki paragrafta yer alan \u00f6rnekten devam edersek ARR %0,000003 oldu\u011funa g\u00f6re<\/p>\n\n\n\n

\"NNT1\"<\/a><\/figure><\/div>\n\n\n\n

olarak hesaplanabilir. Bu say\u0131, tedavi verilmedi\u011fi takdirde \u00f6lecek olan 1 hastan\u0131n ya\u015fayabilmesi i\u00e7in 33 milyon hastaya tedavi verilmelidir anlam\u0131na gelir! 1 hastada tedavi faydas\u0131 g\u00f6rebilmek i\u00e7in milyonda 1 g\u00f6r\u00fclecek bir yan etkiden 33 ki\u015finin zarar g\u00f6recek olmas\u0131n\u0131n mant\u0131kl\u0131 olmad\u0131\u011f\u0131n\u0131 rahatl\u0131kla s\u00f6yleyebiliriz. Halbuki, RRR (r\u00f6latif risk azalmas\u0131) %25 gibi gayet anlaml\u0131 bir tedaviden bahsetmekteydik. Genel olarak, NNT ne kadar k\u00fc\u00e7\u00fckse tedavinin etkisi o kadar g\u00fc\u00e7l\u00fc, ne kadar y\u00fcksekse o kadar d\u00fc\u015f\u00fckt\u00fcr.<\/p>\n\n\n\n

\"NNT2\"<\/a><\/figure><\/div>\n\n\n\n

Yan etki i\u00e7in de ayn\u0131 mant\u0131kla advers bir olay g\u00f6r\u00fclmesi i\u00e7in tedavi edilmesi gereken hasta say\u0131s\u0131 (Number Needed to Harm (NNH<\/strong>)) hesaplanabilir. Bu hesaplamada net risk art\u0131\u015f\u0131 (ARI)<\/strong> kullan\u0131l\u0131r. Be\u015f y\u0131l boyunca tedavi etkisinin izlendi\u011fi bir statin \u00e7al\u0131\u015fmas\u0131nda kontrol grubunda %0,03 rabdomiyoliz g\u00f6r\u00fcl\u00fcrken bu oran tedavi grubunda %0,05 olarak g\u00f6zlenmi\u015f olsun.<\/p>\n\n\n\n

\"ARI\"<\/a><\/figure><\/div>\n\n\n\n

Tedavi verilen hastalarda kontrol grubuna g\u00f6re fazladan 1 ki\u015fide daha rabdomiyoliz g\u00f6r\u00fclmesi i\u00e7in 5000 hastaya 5 y\u0131l boyunca statin verilmesi gereklidir.<\/p>\n\n\n\n

\u00c7al\u0131\u015fmalar\u0131n baz\u0131lar\u0131nda yazarlar riskleri vermeyi tercih ederken di\u011ferlerinde Odds Oran\u0131<\/strong>n\u0131n (OR = OO)<\/strong> kullan\u0131ld\u0131\u011f\u0131n\u0131 g\u00f6receksiniz. Bu tercih ki\u015fisel bir tercih olup asl\u0131nda ayn\u0131 \u015feyleri ifade eden say\u0131lar\u0131n farkl\u0131 g\u00f6sterim \u015fekillerinden ibarettir. Odds terimi \u00f6zellikle \u00fclkemizde herkesin kafas\u0131n\u0131 kar\u0131\u015ft\u0131ran bir terim. Asl\u0131nda bu terim \u0130ngiliz kumar ve bahis jargonundan gelir ve olas\u0131l\u0131k istatistiklerine de ayn\u0131 kelimeyle ge\u00e7mi\u015ftir.<\/p>\n\n\n\n

Rulet oyunu \u00fczerinden \u00f6rnek verirsek, k\u0131rm\u0131z\u0131 ya da siyah gibi 2\u2019li bahislerden birine oynay\u0131p kazand\u0131\u011f\u0131n\u0131z takdirde yat\u0131rd\u0131\u011f\u0131n\u0131z her 1 birim kar\u015f\u0131l\u0131\u011f\u0131nda 1 birim daha kazan\u0131rs\u0131n\u0131z. \u0130lk ba\u015fta koydu\u011funuz kendi bahsinizi kaybetmedi\u011finiz i\u00e7in paran\u0131z 2 kat\u0131na \u00e7\u0131kar. Bu bahsin oddsu 1\u2019e 1, 1:1 ya da 1\u2019dir. \u201cKaybedersen 1 birim kaybedersin ama kazan\u0131rsan 1 birim daha kazan\u0131rs\u0131n\u201d, <\/em>ya da her \u201c1 birim kazanc\u0131n i\u00e7in ben 1 birim kaybederim\u201d<\/em> \u015feklinde de ifade edilebilir. Rulette herhangi bir say\u0131ya oynad\u0131\u011f\u0131n\u0131z takdirde risk etti\u011finiz ve kaybedebilece\u011finiz her 1 birim i\u00e7in 35 birim kazan\u0131rs\u0131n\u0131z, yani 1\u2019e 35 kazan\u0131rs\u0131n\u0131z, ya da odds 1:35 veya 35\u2019dir. Kazand\u0131\u011f\u0131n\u0131z takdirde elinizdeki para yat\u0131rd\u0131\u011f\u0131n\u0131z da dahil olmak \u00fczere ilk bahsinizin 36 kat\u0131d\u0131r.<\/p>\n\n\n\n

Bir ba\u015fka \u00f6rnek ile tekrarlayal\u0131m: 6 y\u00fczl\u00fc bir zar att\u0131\u011f\u0131n\u0131zda, tuttu\u011funuz say\u0131n\u0131n gelme olas\u0131l\u0131\u011f\u0131 (ihtimali) 6\u2019da 1 yani 1\/6, gelmeme ihtimali de 6\u2019da 5 yani 5\/6\u2019d\u0131r. Tuttu\u011funuz say\u0131n\u0131n gelme olas\u0131l\u0131\u011f\u0131n\u0131n (ihtimalinin), gelmeme olas\u0131l\u0131\u011f\u0131na oran\u0131 olan 1:5 olup odds 1:5 ya da 5\u2019dir. Bunu, \u201ctuttu\u011funuz zar\u0131n gelme ihtimali 1\u2019e 5\u2019tir\u201d<\/em> olarak da s\u00f6yleyebiliriz (6\u2019da 1 ve 1\u2019e 5 ayn\u0131 \u015fans\u0131n farkl\u0131 say\u0131sal g\u00f6sterimleri: ilki olas\u0131l\u0131k, ikincisi odds).<\/p>\n\n\n\n

Peki, zar at\u0131m\u0131nda tek say\u0131 gelme oddsu nedir? 6 say\u0131dan 3\u2019\u00fc tek 3\u2019\u00fc \u00e7ift oldu\u011funa g\u00f6re 3:3 ya da 3\u2019e 3\u2019t\u00fcr. Bu da 1 oddsa kar\u015f\u0131l\u0131k gelir (3:3 = 1:1). Yani, tuttu\u011funuz say\u0131n\u0131n (bu \u00f6rnekte tek say\u0131lar) gelme olas\u0131l\u0131\u011f\u0131, gelmeme olas\u0131l\u0131\u011f\u0131na (\u00e7ift say\u0131lar) denk oldu\u011fundan odds 1\u2019dir. Bu s\u00f6ylem istatistik bilimine ge\u00e7erken biraz erozyona u\u011frar ve kaybedece\u011fin:kazanaca\u011f\u0131n oran\u0131 olan odds, olma olas\u0131l\u0131\u011f\u0131\/olmama olas\u0131l\u0131\u011f\u0131, \u00f6l\u00fcm\/sa\u011fkal\u0131m, sa\u011fl\u0131kl\u0131\/hasta haline d\u00f6n\u00fc\u015f\u00fcr.<\/p>\n\n\n\n

D\u00f6rt g\u00f6zl\u00fc tablodaki \u00f6rne\u011fimize d\u00f6nelim: Tedavi grubunda 15 ki\u015fi \u00f6l\u00fcrken 85 ki\u015fi ya\u015fam\u0131\u015ft\u0131r (Form\u00fcl 7)<\/strong>. Yani tedavi grubunda \u00f6l\u00fcm oddsu 15:85 ya da 3:17\u2019dir. \u00d6len her 3 ki\u015fiye kar\u015f\u0131l\u0131k 17 ki\u015fi ya\u015fam\u0131\u015f, 20 ki\u015fiden 3\u2019\u00fc \u00f6lm\u00fc\u015f veya 20 ki\u015fiden 17\u2019si ya\u015fam\u0131\u015ft\u0131r. Kontrol grubunda \u00f6l\u00fcm oddsu 20:80 veya 1:4\u2019t\u00fcr (Form\u00fcl 8)<\/strong>. \u00d6len her 1 ki\u015fiye kar\u015f\u0131l\u0131k 4 ki\u015fi ya\u015fam\u0131\u015f, 5 ki\u015fiden 1\u2019\u00fc \u00f6lm\u00fc\u015f veya 5 ki\u015fiden 4\u2019\u00fc ya\u015fam\u0131\u015ft\u0131r. Tedavi ve kontrol gruplar\u0131n\u0131n oddslar\u0131 aras\u0131ndaki orana ise Odds Oran\u0131<\/strong> ya da tam ifadesi olmamakla beraber Olas\u0131l\u0131klar Oranlar\u0131n\u0131n Oran\u0131 (Odds Oran\u0131, OO)<\/strong> ad\u0131 verilir (asl\u0131nda olas\u0131l\u0131klar oran\u0131 sizin de g\u00f6rd\u00fc\u011f\u00fcn\u00fcz gibi oddsun kendisi. Dolay\u0131s\u0131yla odds oran\u0131na da olas\u0131l\u0131klar oranlar\u0131n\u0131n oran\u0131 dememiz laz\u0131m. \u0130\u015fte odds oran\u0131nda kafa kar\u0131\u015f\u0131kl\u0131\u011f\u0131n\u0131n sebebi buradan geliyor). Tedavi grubunun \u00f6l\u00fcm oddsunun kontrol grubun \u00f6l\u00fcm oddsuna oran\u0131 olan odds oran\u0131,<\/p>\n\n\n\n

\"OO\"<\/a><\/figure><\/div>\n\n\n\n

olarak hesaplanabilir (Form\u00fcl 9)<\/strong>. E\u011fer odds oran\u0131 1 ise tedavinin kontrol grubuna g\u00f6re farkl\u0131 bir olas\u0131l\u0131\u011fa sahip olmad\u0131\u011f\u0131n\u0131 s\u00f6yleyebiliriz. E\u011fer odds oran\u0131 1\u2019in \u00fcst\u00fcndeyse odds oran\u0131na bak\u0131lan olay ya da sonlan\u0131m tedavi ile artmakta, alt\u0131ndaysa azalmaktad\u0131r.<\/p>\n\n\n\n

\"Odds\"<\/a><\/figure><\/div>\n\n\n\n

Hangi Odds\u2019un hangi Riske kar\u015f\u0131l\u0131k geldi\u011fi Tablo 3<\/strong>\u2019de g\u00f6sterilmi\u015ftir. Odds ve Riskin birbirine d\u00f6n\u00fc\u015f\u00fcm form\u00fclleri Form\u00fcl 10 ve 11<\/strong>\u2019de verilmi\u015ftir.<\/p>\n\n\n\n

Odds\u2019lar g\u00f6r\u00fcld\u00fc\u011f\u00fc gibi daima risklerden say\u0131sal olarak daha y\u00fcksektir. Risk \u00e7ok y\u00fcksekse odds ve risk aras\u0131ndaki fark daha da b\u00fcy\u00fcr, risk \u00e7ok d\u00fc\u015f\u00fckse arada neredeyse hi\u00e7 fark kalmaz. Klinisyenler olarak anlamas\u0131 ve alg\u0131lamas\u0131 \u00e7ok daha kolay olan RR\u2019nin OR\u2019yle ili\u015fkisini bilmek ve her ikisini birbirine d\u00f6n\u00fc\u015ft\u00fcrebilmek durumunday\u0131z. \u015eans\u0131m\u0131za, t\u0131bbi uygulamalarda hastalar\u0131m\u0131z\u0131n kar\u015f\u0131 kar\u015f\u0131ya kald\u0131\u011f\u0131 riskler genellikle d\u00fc\u015f\u00fck oldu\u011fundan RR ve OR birbirine yak\u0131nd\u0131r. OR kullan\u0131ld\u0131\u011f\u0131nda tedavinin sanki daha etkinmi\u015f gibi bir alg\u0131lamas\u0131 olu\u015fur.<\/p>\n\n\n\n

\u00d6rne\u011fin, bir ilac\u0131n etkinli\u011fi (istenen sonu\u00e7) tedavi grubunda %45 kontrol grubunda %30 ise, tedavi grubunun oddsu 0,82 (0,45 \/ 0,55) iken kontrol grubunun oddsu 0,43 (0,30 \/ 0,70) olup, RR 1,5 (0,45 \/ 0,30) OR 1,91\u2019dir (0,82 \/ 0,43).<\/p>\n\n\n\n

\"Risk_Odds\"<\/a>
Tablo 3. Risk ve Oddslar\u0131n birbiriyle kar\u015f\u0131la\u015ft\u0131r\u0131lmas\u0131<\/figcaption><\/figure><\/div>\n\n\n\n
\"Risk\"<\/a><\/figure><\/div>\n\n\n\n

Peki, madem OO i\u015fleri bu derecede kar\u0131\u015ft\u0131r\u0131yor ve anlamas\u0131 da zor, neden kullanmaya devam ediyoruz? OO yukar\u0131da bahsetti\u011fimiz \u00f6rneklerdeki randomize kontroll\u00fc \u00e7al\u0131\u015fmalardan ziyade g\u00f6zlemsel kohort \u00e7al\u0131\u015fmalarda daha faydal\u0131d\u0131r. Kohort \u00e7al\u0131\u015famalarda, ara\u015ft\u0131rmac\u0131lar belirli risklere sahip hastalar\u0131 belirli bir zaman izler ve olu\u015fan istenmeyen olaylar\u0131n frekanslar\u0131n\u0131n kayd\u0131n\u0131 tutar. Vaka-kontrol \u00e7al\u0131\u015fmalar\u0131nda ise kat\u0131l\u0131mc\u0131lar halihaz\u0131rda zaten istenmeyen sonuca ula\u015fm\u0131\u015ft\u0131r. Ara\u015ft\u0131rmac\u0131lar bu sonlan\u0131ma ula\u015fm\u0131\u015f hastalara denk bir kontrol grubu se\u00e7er (bu kontrol grubunda ka\u00e7 ki\u015fi olaca\u011f\u0131na ve vakalarla oran\u0131na \u00e7al\u0131\u015fmac\u0131lar karar verir) ve geriye d\u00f6n\u00fck olarak her iki grubun ge\u00e7mi\u015finde risk varl\u0131\u011f\u0131n\u0131n oran\u0131n\u0131 kar\u015f\u0131la\u015ft\u0131r\u0131r. Bu tip \u00e7al\u0131\u015fmalar hastal\u0131klar\u0131n prevalans\u0131 hakk\u0131nda bilgi vermez, \u00e7\u00fcnk\u00fc hasta ve kontrol say\u0131s\u0131 hastal\u0131\u011f\u0131n toplumdaki prevalans\u0131na denk \u015fekilde de\u011fildir, ara\u015ft\u0131rmac\u0131lar\u0131n tercihleriyle olu\u015fturulmu\u015ftur.<\/p>\n\n\n\n

\u00d6rne\u011fin, KOAH geli\u015fiminde en az 1 y\u0131l soba bulunan bir evde ya\u015faman\u0131n etkisi bir vaka-kontrol \u00e7al\u0131\u015fmas\u0131yla incelenmek istense, KOAH hastas\u0131 olan bir grup ile kendi se\u00e7ece\u011fimiz bu gruba denk bir kontrol grubunun \u00f6zge\u00e7mi\u015flerindeki sobal\u0131 evde ya\u015fama durumlar\u0131n\u0131n frekanslar\u0131n\u0131 kar\u015f\u0131la\u015ft\u0131rmam\u0131z gerekir. 277 KOAH hastam\u0131z oldu\u011funu ve bunlar\u0131n 67\u2019sinde soba duman\u0131 maruziyeti oldu\u011funu varsayal\u0131m. Kontrol grubu olarak da KOAH d\u0131\u015f\u0131ndaki de\u011fi\u015fkenler a\u00e7\u0131s\u0131ndan bu gruba denk, poliklini\u011fimize ba\u015fvuran 283 hasta se\u00e7mi\u015f olal\u0131m ve bunlar\u0131n da 41\u2019inde soba maruziyeti belirlemi\u015f olal\u0131m. Bu durumda \u00f6rne\u011fimizin d\u00f6rt g\u00f6zl\u00fc tablo yap\u0131s\u0131 Tablo 4<\/strong>\u2019deki gibi olacakt\u0131r.<\/p>\n\n\n\n

\"Tablo<\/a>
Tablo 4. Farazi Vaka-Kontrol \u00c7al\u0131\u015fmam\u0131z\u0131n D\u00f6rt G\u00f6zl\u00fc Tablosu<\/figcaption><\/figure><\/div>\n\n\n\n

Bu tablodan yola \u00e7\u0131karak;<\/p>\n\n\n\n

\"Ornek1\"<\/a><\/figure><\/div>\n\n\n\n

risk nispeten orta seviyede oldu\u011fundan RR ile OO aras\u0131nda fark oldu\u011funu, OO\u2019nun daha y\u00fcksek oldu\u011funu g\u00f6r\u00fcr\u00fcz. Ara\u015ft\u0131rmac\u0131lar tam 2 kat\u0131 b\u00fcy\u00fckl\u00fckte bir kontrol grubu se\u00e7seydi sonu\u00e7lar\u0131m\u0131z de\u011fi\u015fir miydi? (Tablo 5)<\/p>\n\n\n\n

\"Tablo<\/a>
Tablo 5. Farazi Vaka-Kontrol \u00c7al\u0131\u015fmam\u0131z\u0131n D\u00f6rt G\u00f6zl\u00fc Tablosu, kontrol grubu iki kat\u0131 olsayd\u0131<\/figcaption><\/figure><\/div>\n\n\n\n

Bu durumda,<\/p>\n\n\n\n

\"Ornek2\"<\/a><\/figure><\/div>\n\n\n\n

G\u00f6r\u00fcld\u00fc\u011f\u00fc gibi, vaka-kontrol \u00e7al\u0131\u015fmalar\u0131nda r\u00f6latif risk tamamen kontrol grubundaki hasta say\u0131s\u0131na ba\u011fl\u0131 iken, odds oran\u0131 kontrol grubunun hasta say\u0131s\u0131 ile de\u011fi\u015fmemekte.<\/p>\n\n\n\n

Bu sebeple, odds oran\u0131 vaka-kontrol \u00e7al\u0131\u015fmalar\u0131nda kullan\u0131lmas\u0131 ve raporlanmas\u0131 gereken \u00f6l\u00e7\u00fctt\u00fcr. <\/strong><\/p>\n\n\n\n

OR\u2019nin di\u011fer \u00f6nemli \u00f6zellikleri ise \u015funlard\u0131r: <\/strong><\/p>\n\n\n\n

  1. Meta-analizlerde birbirinden \u00e7ok farkl\u0131 olay oranlar\u0131 bir araya getirilmeye \u00e7al\u0131\u015f\u0131ld\u0131\u011f\u0131ndan ayn\u0131 mant\u0131kla RR yerine OO kullan\u0131lmas\u0131 gerekir.<\/li>
  2. OO, maruziyete ba\u011fl\u0131 birbirini tamamlayan sonlan\u0131m\u0131nlar\u0131n y\u00f6n\u00fcne g\u00f6re birbirinin tersi de\u011ferler al\u0131r. Mortalite i\u00e7in bak\u0131lan OO 4 ise, sa\u011fkal\u0131m i\u00e7in ayn\u0131 \u00e7al\u0131\u015fmada OO \u00bc\u2019t\u00fcr.<\/li>
  3. Risk ne kadar b\u00fcy\u00fcrse b\u00fcy\u00fcs\u00fcn RR tam anlam\u0131yla y\u00fckselmez. Halbuki OO temel risk oran\u0131n\u0131n b\u00fcy\u00fckl\u00fc\u011f\u00fcnden etkilenmez. OO, 0 ile sonsuz aras\u0131nda de\u011ferler alabilirken, RR alamaz.<\/li>
  4. OO, maruziyet ile sonlan\u0131m aras\u0131ndaki ili\u015fki ve etkinin b\u00fcy\u00fckl\u00fc\u011f\u00fcn\u00fcn \u00f6l\u00e7\u00fcs\u00fc oldu\u011fundan lojistik regresyon i\u00e7in kullan\u0131lan parametredir. Tarihsel a\u00e7\u0131dan OO her t\u00fcrl\u00fc ili\u015fki bildirimi i\u00e7in tercih edilen ve daha do\u011fru kabul edilen \u00f6l\u00e7\u00fct olsa da modern istatistiksel y\u00f6ntemler kullan\u0131ld\u0131k\u00e7a \u00f6nemlili\u011fi azalmaktad\u0131r.<\/li><\/ol>\n","protected":false},"excerpt":{"rendered":"

    Randomize kontroll\u00fc bir \u00e7al\u0131\u015fmada genellikle \u201cevet-hay\u0131r\u201d \u015feklinde dikotom (iki se\u00e7enekli) sonlan\u0131m \u00f6l\u00e7\u00fctleri de\u011ferlendirilir (\u00f6ld\u00fc-ya\u015f\u0131yor, hasta-de\u011fil, tan\u0131 var-yok). Makalede bu hedef sonlan\u0131m\u0131n geli\u015fti\u011fi…<\/p>\n","protected":false},"author":1561,"featured_media":510,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"_lmt_disableupdate":"","_lmt_disable":"","footnotes":""},"categories":[21,10014],"tags":[],"class_list":["post-509","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-istatistik-yazilari","category-akademik-blog-yazisi"],"acf":[],"_links":{"self":[{"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/posts\/509","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/users\/1561"}],"replies":[{"embeddable":true,"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/comments?post=509"}],"version-history":[{"count":0,"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/posts\/509\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/media\/510"}],"wp:attachment":[{"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/media?parent=509"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/categories?post=509"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/tatd.org.tr\/atak\/wp-json\/wp\/v2\/tags?post=509"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}