PERSONAL PROTECTIVE EQUIPMENT DEMAND FORECASTING AND INVENTORY MANAGEMENT DURING COVID-19 (CASE STUDY: XYZ HOSPITAL)

• Main Author: Christie, Nadya
• Format: Theses
• Language: Indonesia
• Subjects: Manajemen umum
• Online Access: https://digilib.itb.ac.id/gdl/view/53409
• Institution: Institut Teknologi Bandung

Until January 2021, the number of new COVID-19 in Indonesia cases is increasing day by day, especially in West Java. The COVID-19 hospital's occupancy rate in West Java, particularly Bogor, Depok, Bekasi, and Greater Bandung, is around 80 percent, above the World Health Organization's suggested hospital occupancy, 60 percent. This predicts the hospital's personal protective equipment essential to accommodate the increasing demand, especially in XYZ hospital, a national referral for COVID-19, especially in West Java province. Lack of prediction methods leads to a shortage in Personal Protective Equipment (PPE), affecting the hospital's service to its patients. The challenge is to build an easy-touse system to predict the demand for personal protective equipment in the hospital during the COVID-19 pandemic. This study aimed to find the best forecasting method for COVID19 patients hospitalized in the hospital, determine the PPE demand, and find the hospital's best inventory models. This study compares three methods to forecast the COVID-19 patients hospitalized in the Negative Pressure Isolation room, Natural Air Flow Isolation room, ICUIntermediate room, and ICU-Isolation room. The methods are ARIMA (Auto-Regressive Integrated Moving Average), Single Exponential Smoothing, and Double Exponential Smoothing. Then the result of the COVID-19 forecast is inputted to the PPE Calculator to be translated into PPE demand. The best method is chosen based on the forecast accuracy measurement using MAPE (Mean Absolute Percentage Error), RMSE (Root Mean Squared Error), and MAD (Mean Absolut Deviance) / MAE (Mean Absolute Error). The method with the smallest forecast accuracy value is considered the best method. This study compares the EOQ (Economic Order Quantity) Model, Fixed-Time Period Model, and the Naïve Model for the inventory model. As a result, the best method to forecast the COVID-19 patients hospitalized is using ARIMA models because it has the least MAD, MAPE, and RMSE. Surprisingly, the EOQ model is still the best inventory control method even during the COVID-19 pandemic because EOQ has the best AIL among all methods.