This is never as simple as it sounds, and it may even sound obvious. That said, it is one of the critically important steps to ensuring a recall is executed accurately and timely.
At the end of the day, bad data costs money. It costs to acquire, compile, store, and (mis)manage/
maintain. It also costs money when sales are lost as a result of bad data. When looking at the cost of
data across multiple industries, these studies found the following impacts of poor data quality.
- A Forbes Insights and KPMG “2016 Global CEO Outlook” study found that 84 percent of CEOs are concerned about the quality of the data on which they base their decisions.4
- A LeadJen study found that bad data costs companies $20,000 per sales rep per year.5
- An Experian study concluded that 75 percent of companies are wasting 15 percent of their revenue due to poor data quality.6
- A Gartner study estimates that poor-quality data is costing them on average $14.2 million annually.7
Novasyte has learned over the past 10 years that in the med-tech industry the vast majority of
companies clean their data only after learning of a necessary field action or recall event. This can delay
the recall initiation, as often times the data has multiple sources and formats. By managing the data on
an ongoing basis, the likelihood of sending a notification to the correct consignee increases exponentially. This results in a faster response rate and ultimately a shorter recall.
EXAMPLE
As an example, say a recall has 1,000 consignees; following the traditional, manual method, it
will take on average five minutes to validate each consignee resulting in an initial 83 hour time
investment at the beginning of a recall. By having an updated (or recently updated) data set,
the recall launch can be significantly expedited.
Understand the Data Guidelines
Per 21 CFR 7.46, the information below outlines the minimum data as well as the preferred data an organization should try to work towards having readily up-to-date.8
The MINIMUM data set to successfully execute a recall includes:
- Consignee Address
- Number of Affected Units
The NICE-TO-HAVE data set that expedites a recall even further includes:
- Department Name
- Consignee Name
- Consignee Title
- Consignee Email
- Consignee Phone Number
- Department
How Much Internal Resource Time Will it Take?
Many med-tech companies have a need for tighter controls around data clean up and sanitization.
A study conducted by Blue Hill Research found that 28 percent of an internal data analyst’s time is
dedicated to data preparation.9 The average salary of a data analyst in the US averages $62,000
a year; this equals about $22,000 worth of their annual salary.9 This applies to OEM customer/consignee data, as often times the data comes from multiple sources and requires cleaning, analyzing and preparing prior to FDA submission.
The Importance of Asset Tracking
Beyond having updated consignee data, it is important to track products through every step in the recall
process.
Per 21 CFR 821.25, the FDA advises manufacturers to track certain types of devices and clearly identify
product distribution information to facilitate notification and recall in the event that it presents a serious
risk.10
Section 519(e) states the agency may require tracking of certain Class || or Class ||| devices and may even officially order a manufacturer to adopt a tracking method for a new device as a part of the pre-market clearance process.11
As mandated by the regulations, manufacturers will have three working days to provide critical
information regarding the location of a tracked product for devices that have not yet been distributed to
a patient. Further, manufacturers have 10 working days for devices that are intended for use by a single
patient over the life of the device, or following distribution or implantation in a patient, if requested.10
KEY TAKE AWAYS:
Evaluate data sources and formats.
Then develop a consistent master template.
The Novasyte Way
By working with Novasyte to manage your recall on the proprietary S.M.A.R.T. Platform, the consignee data will be processed through a multi-touch verification system that can reduce consignee validation time by more than half.
This is a critical step, among several, to expedite the recall on the front end, as the format and data accuracy positively contribute to the FDA closing a recall quickly. The sophistication, rigor and process of the S.M.A.R.T. Platform has allowed Novasyte clients’ to receive formal FDA recall termination approval in as quickly as three days.
To learn how to increase the speed and accuracy of a recall, download our white paper to learn:
- Current recall landscape in the med-tech industry
- Strategies for maintaining accurate consignee data
- Strategies for developing a methodical approach for a field strategy
- The risks associated with a de-centralized database, and why centralizing it is a necessity
- Case study outlining the S.M.A.R.T. method for managing medical device recalls
Sources:
4. Forbes Corporate Communications. (2017, May 31). Poor-quality data imposes costs and risks on businesses, says new Forbes Insights report. Forbes. Retrieved from https://www.forbes.com/sites/forbespr/2017/05/31/poor-quality-data-imposes-costs-and-risks-on-businesses-says-new-forbes-insights-report/#4af5a067452b
5. LeadJen News. (2017). Bad data costs companies 20,000 annually, per inside sales rep. Retrieved from https://www.leadjen.com/blog/leadjen-news/bad-data-costs-companies-20000-annually-per-inside-sales-rep-leadjenstudy-shows/
6. Adamson, G. (2014, February 3). Data is the oil that keeps the business cogs turning. Retrieved from https://www.edq.com/uk/blog/the-well-oiled-data-machine/
7. Friedman, T. & Judah, S. (2013). The state of data quality: Current practices and evolving trends. Gartner. Retrieved from https://www.gartner.com/doc/2636315/state-data-quality-current-practices
8. FDA Department of Health and Human Services General; Enforcement Policy, 21 C.F.R. § 7 (2017). Retrieved from https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfCFR/CFRSearch.cfm?CFRPart=7&showFR=1
9. Online Analytical Processing (OLAP). 28% of a data analyst’s time is spent on data prep. Retrieved from http://olap.com/28-data-analysts-time-spent-data-preparation/
10. FDA Department of Health and Human Services Medical Devices; Medical Device Tracking Requirements, 21 C.F.R. § 821.25 (2017). Retrieved from https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?fr=821.25
11. U.S. Food & Drug Administration. (2014, March 27). Medical device tracking: Guidance for industry and food and drug administration staff. Retrieved from https://www.fda.gov/MedicalDevices/ucm071756.htm
