Healthcare resource management by using data mining-predicting length of stay : case study : (birthing centers and maternity hospitals-Gaza Strip)‎

العناوين الأخرى

تطبيق تنقيب البيانات في إدارة الموارد الصحية-التنبؤ بمدة الإقامة : دراسة حالة : مراكز و مستشفيات التوليد-قطاع غزة

مقدم أطروحة جامعية

al-Tatir, Nariman Sami Saqr

مشرف أطروحة جامعية

al-Dayih, Wail Hamdi

أعضاء اللجنة

Fatayir, Tamir Sad
Dahliz, Khalid Abd Abd al-Salam

الجامعة

الجامعة الإسلامية

الكلية

كلية التجارة

القسم الأكاديمي

قسم إدارة الأعمال

دولة الجامعة

فلسطين (قطاع غزة)

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2015

الملخص الإنجليزي

Government hospitals in the Gaza Strip, Palestine are forced to use their scarce resources as efficient as possible.

One of these resources is the hospital bed capacity.

An optimal admission planning uses hospital bed capacity as efficient as possible.

In order to achieve such a planning, early predictions of the expected discharge moment of patients are needed.

Expected discharge moments can be predicted if the expected duration of admissions is known.

In other words, predictions of the expected length of stay (LOS) at admission are required.

In this study, LOS is defined as the number of days a patient is admitted to the hospital during an admission.

Predicting the LOS of patients in a hospital is important in providing them with better services and higher satisfaction, furthermore accurate predictions of patient LOS in the hospital can effectively manage hospital resources and increase efficiency of patient care.

The aim of this study is to applying data mining techniques to support successful decisions that will improve success of healthcare management and build accurate model to predict the LOS of maternity hospital.

The Data were collected from childbirth database.

The patient records of 22,461 instances were included in the analysis.

This thesis applied on three government maternity hospitals in Gaza strip.

The dataset used information of pregnant women who delivered in a public hospital in Gaza Strip between 1 January 2013 and 31 December 2013.

The techniques used are classification with decision tree.

LOS is the target variable, and 12 input variables are used for prediction.

The LOS is categorized into three classes, LOS 1 less than 24 hours, LOS2 from 24 hours to 72 hours and LOS 3 greater III than 72 hours.

A confusion matrix was obtained to calculate sensitivity, specificity, and accuracy.

Results: The overall accuracy of decision tree was 79.99 % in the training set.

Most normal delivery (77.54%) had an LOS ≤24 hours, whereas 20.5% of cesarean section had an LOS >24 hours.

Moreover, the study shows that LOS class recall is 97.56%.

Therefore the tree algorithms are able to predict LOS with various degrees of accuracy.

التخصصات الرئيسية

إدارة الأعمال

عدد الصفحات

120

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-727307