evaluation were the employees engaged in the departments of the Clinical Centre of Serbia related to infectious medical waste man- agement: the Department of Hospital Epidemiology, the Depart- ment of Technical Affairs, the Department for Safety and Occupational health, the Department of Medical Ecology, and the Department for the treatment of infectious waste (Tubic, 2014). Using brainstorming and individual interviews techniques, the probabilities shown in Table 2 were obtained.
Using the inclusion-exclusion technique (Lisnianski and Levitin,
2003), the probability of the top event is calculated. The probability of infection spreading caused by malfunctioning of the infectious medical waste management is 0.0088. Since the disposal of infec- tious waste is made on a daily basis, the obtained probability in- dicates that the expected time between two accidents is approximately 113 days i.e. the expected frequency of accident is
3.2 times per year. According to Tubic (2014), based on current records of the number of injuries in the Clinical Centre of Serbia, conducted by the Department of Hospital Epidemiology, from September 2013 to August 2014, the number of reported injuries is
32, out of which ten percent of injuries happened combined with the other risk factors. We can conclude that the obtained results from our study are confirmed by observed frequency and real data in the Clinical Centre of Serbia. Bearing in mind the severity of the defined top event, the obtained probability can be considered excessive and should be decreased.
In order to effectively decrease the probability of infection spreading, it is important to identify the basic events which contribute the most to the resulting probability. For that purpose, we use importance measures.
The values of importance measures defined by (2)e(5) for all eleven basic events are presented in Table 3. Since each importance measure is calculated in a different way, these values are not directly comparable. However, the basic events can be ranked by their importance based on the obtained values. For each impor- tance measure, the importance of events is determined in the same way: the higher the obtained values the higher the rank of the event is. Therefore, the ranks obtained in accordance with the importance measures values are summarized in the Table 4.
According to the ranks from the Table 4, two out of eleven basic events are found out to be the most important: injury at work (P5) and not using protective equipment (P6). As it can be seen in Table 2, the probability of the event P5 is among the largest prob- abilities while the event P6 has the second-lowest probability. This result emphasizes the role of importance measures which take into account the probability of events as well as their position in the system's structure which is expressed through these events ap- pearances in minimal cut sets i.e. through structure function (6). On the other hand, two out of eleven basic events have the lowest rank according to all importance measures: failure to respect the