3 Ways Data Can Improve Your Food Safety and Quality Program.
Let’s start with three problems, I’ve seen data create.
1) I have been surprised at how much data is collected by companies only to be filed and never looked at again. What is the reason for the massive data collection? In the food safety and quality world we can point to regulatory, customer, or third-party audit requirements for documentation, after all if it’s not documented it didn’t happen. The data is stored on paper or electronically in five locations, over three or four programs, basically it is piecemealed all over leaving it hard or impossible to get your head or hands around. The problem here is that it cannot be used in a meaningful way because it is too hard to mine the data.
2) To be clear, I love data to help support the food safety and quality department as well as the overall company. I find it is sometimes a struggle to understand what, when, where, why, who and how those data are collected. What data should be collected and can it improve the company in some manner. What I have seen companies do is create a knee jerk corrective action for a problem. Part of the solution to that problem is a document that collects some data, that may already being collected somewhere else or worse not even related to the original problem but over the course of time people have “improved” the document with new needs. Sometimes, it may not even be from a knee jerk reaction but a manager’s preference or idea at one time that is now antiquated. The problem here is that this data creates static noise that makes it hard to focus on important data. Not to mention, paying for someone to collect the data, another to verify the data, and sometimes a third person (or at least a third touch on the document) to trend the data.
3) Of course, some companies use this data to put together pretty power point ready charts or develop KPIs that may or may not help the company improve their food safety plan. A lot of the data that is collected and stored for years (on average five years, but at least shelf life plus one year), and never looked at again. Now it is important to understand that with food safety and quality systems; you have to say what you do and do what you say and prove it with documentation, verification, and validations. So, collecting data is key and needed, but what happens to that data and the labor cost going into that data seems like a loss to me to file away without at least trying to monetize the data.
Above are basically three things where I’ve experienced data go wrong. Not knowing the five W’s of the data to collect, not collecting data in a format/manner that is minable, and finally not monetizing the data collected. Below are three suggestions to help you get started on improving the way you and your organization can use data in a more thoughtful way.
1) Use a deviation log or CAPA log that collects all findings to one document. Start with a simple excel sheet, with tabs along the top. This allows the data you are collecting from deviations and findings to be sorted, categorized, and trend findings. Example maybe that you find a trend of metal shavings every other Thursday in one of your pieces of production equipment before the
start of second shift. This now allows you to investigate on the potential next Thursday. Let’s say the investigation found that is the PM schedule for lubricating a widget and the SOP for putting the equipment back together misses a crucial step causing the metal shavings. This trend may have been missed because of the infrequency of the PM.
2) Every facility is unique and what data is required could vary drastically. Keep in mind before asking someone to collect a piece of data, make sure it is useable. I remember this one time, where we had a mis-formulation of product. I was shocked that no one had noticed the mis-formulation. Knee-jerk corrective action went into place to have an organoleptic test and review conducted on a timed frequency during the production run. This seemed at the time, for that problem to be a good solution, until the next time the product was mis-formulated but it had passed the organoleptic test and review that was conducted. I then realized that the previous effort to solve the problem, not only did not address the root cause (larger problem), but also was not an effective screen at catching mis-formulations. Point being we were collecting useless information that still allowed a mistake to happen that wasn’t caught until the next day. Resulting in thousands of dollars or product not being able to be released into commerce.
3) Finally, you can analyze the data in an effective manner. This is the process I like to call monetizing the data. This is possible because of the first two steps; you now have your data in a format that is mineable, and you understand the true reason you are collecting that specific data. With this knowledge and the capability of mining the data in various ways you can support other departments throughout your organization. I recommend understanding what others on your team would like to see from the data. Maybe your maintenance team wants the report that breaks down foreign objects that could have come from equipment. They can use this data to improve their Preventative Maintenance Program. Maybe sanitation would like to see EMP findings, Pre-Op checks, titration records, Master Sanitation Schedule, and work orders to solve a problem area.