Evaluasi Terjadinya Infeksi Operasi Caesar di Rumah Sakit X Menggunakan Bayesian Bernoulli Model (BBM) dan Bayesian Bernoulli Mixture Model (BBMM)

Purwantoro, Nursetyo (2019) Evaluasi Terjadinya Infeksi Operasi Caesar di Rumah Sakit X Menggunakan Bayesian Bernoulli Model (BBM) dan Bayesian Bernoulli Mixture Model (BBMM). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211745000007-Undergraduate_Theses.pdf Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Infeksi dapat berasal dari komunitas (comunnity acquired infection) atau berasal dari lingkungan rumah sakit (hospital acquired infection) yang sebelumnya dikenal dengan istilah infeksi nosokomial. Rumah sakit X merupakan rumah sakit yang memiliki peralatan cukup lengkap untuk menangani kasus infeksi. Unit yang masih menemukan kasus infeksi hingga tahun 2018 adalah unit yang fokus untuk melakukan operasi ibu bersalin (operasi caesar). Terdapat variabel yang secara teoritis medis mempengaruhi risiko terjadinya infeksi an-tara lain faktor status imun atau disebut juga American Society of An-esthesiologists (ASA) dan jenis operasi. Dugaan sifat dari data adalah berdistribusi Bernoulli dengan adanya mixture. Pemodelan dilakukan menggunakan Bayesian Bernoulli Model (BBM) dan Bayesian Bernoul-li Mixture Model (BBMM) dilakukan untuk mengevaluasi terjadinya infeksi pada pasien operasi Caesar. Pemodelan Bayesian dilakukan karena data infeksi bersifat imbalance dan memiliki mixture. Hasil dari analisis dan pembahasan menyatakan bahwa terdapat 7 pasien yang terdapat infeksi dari 300 pasien yang diteliti. Identifikasi mixture menemukan 99 pasien masuk pada kategori mixture pertama yaitu tid-ak terantisipasi infeksi sedangkan 201 pasien terantisipasi infeksi. Perbandingan pemodelan Bayesian diperoleh model terbaik yaitu BBMM yang mempunyai nilai akurasi dan AUC lebih besar dibanding BMM. Terdapat 3 kovariat yang digunakan dalam pemodelan BBMM yang berpengaruh signifikan terhadap status infeksi pasien.
===================================================================================================================================
Infection can come from the community (comunnity acquired in-fection) or come from the hospital environment (hospital acquired in-fection) which was previously known as nosocomial infection. X Hospi-tal is a hospital that has enough equipment to handle cases of infec-tion. Units that still find cases of infection until 2018 are units that focus on performing maternal surgery (caesarean section). There are variables that theoretically medically influence the risk of infection among other immune status factors or also called the American Society of Anesthesiologists (ASA) and the type of surgery. The alleged nature of the data is distributed by Bernoulli with the mixture. Modeling was carried out using the Bayesian Bernoulli Model (BBM) and Bayesian Bernoul-li Mixture Model (BBMM) to evaluate the occurrence of infec-tions in Caesarean patients. Bayesian modeling is done because the infection data is imbalance and has a mixture. The results of the analy-sis and discussion stated that there were 7 patients with infections from the 300 patients studied. The mixture identification found 99 patients entered in the first mixture category ie infection was not anticipated while 201 patients were anticipated infection. The comparison of Bayesian modeling obtained the best model, namely BBMM, which has an accuracy value and AUC greater than BMM. There are 3 covariates used in BBMM modeling which have a significant effect on the pa-tient's infection status.

Item Type:	Thesis (Other)
Additional Information:	RSSt 519.542 Pur e-1 2019 3100019081116
Uncontrolled Keywords:	Bayesian, Bernoulli, Infeksi, Mixture
Subjects:	H Social Sciences > HA Statistics > HA31.3 Regression. Correlation. Logistic regression analysis. H Social Sciences > HA Statistics > HA31.7 Estimation
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Nursetyo Purwantoro
Date Deposited:	07 Nov 2025 03:40
Last Modified:	07 Nov 2025 03:40
URI:	http://repository.its.ac.id/id/eprint/64663

Actions (login required)

View Item