The COVID-19 pandemic recently challenges worldwide health system. Many countries have been establishing several rules in order to fight against the expected exponentially growth of affected people. After some weeks hospital bed occupancy started to grow because of infected people and this demand can potentially overlap hospital bed capacity. In this scenario, doctors would have to decide who should be treated and patients without adequate treatment can increase number of deaths. Thus, in order to avoid that patient demand overlap hospital bed capacity, this research propose a forecast and simulation approach in order to provide the number of intensive care beds and health care professionals that could be available in a short-term of 21 days ahead taking into account 3 patient demand scenarios. This research is applied to Rio de Janeiro city (Brazil) and all records were retrieved from Rio de Janeiro public database. Hospital length of stay (LOS) records were obtained with Rio de Janeiro City Health Department (RJHD). We concluded that to open the number of new intensive care unit (UCI) beds proposed by the government is lower than expected admission on the worst scenario.