Hi, anasmedic!
Your posted data is actually the stats you're looking for. So consider where you extract the data from, there you might have person names, companies, sectors, positions, dates, times, type of accident, other data from the incidence, treatment days, follow up, ... and so.
With that detailed data is from where and with how you'd build a stat model regarding your business or corporate KPIs (key process indicators, just in case), and that stat model is which will reflect your actual posted data. You'll have to consider time periods, job positions, sectors, and all other classifications that might help you to analyze where, when, how much and why are you closer or farther from your goal, the famous zero.
My advice is to re-think the process from the data collection (entry, registration), define the proper way to store it (one worksheet, chained worksheets starting from company/sector/employee and ending in accident date / result), and state clearly what information do you and the organization want to extract from that data. Once done that, measuring the deviation from the goal is the simplest part of the job.
Regards!