RECMGMT-L Archives

Records Management

RECMGMT-L@LISTSERV.IGGURU.US

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Frederic Grevin <[log in to unmask]>
Reply To:
Records Management Program <[log in to unmask]>
Date:
Sat, 15 Mar 2014 21:13:51 +0000
Content-Type:
text/plain
Parts/Attachments:
text/plain (57 lines)
In response to Ann Mangiaracina's posting:

"I'm meeting with our Office of the General Counsel to discuss revising our policy for when an original hard copy can be disposed of after being scanned."

This really covers more than the issue of legible, complete, and authentic reproduction of records that may be admissible in evidence during legal and/or regulatory proceedings, leaving aside specific statutory and/or regulatory requirements of the sort described by Richard King in his response.

An additional issue is the preservation of a legible, complete and authentic record over the very long term, i.e., more than twenty years (a rule of thumb).

Looking first at the quality control required for the production of legible, complete, and authentic images of paper records that may be admissible in evidence during legal and/or regulatory proceedings, the US National Standard ANSI/AIIM MS-44-1988 (R1993) "Recommended Practice for Quality Control of Image Scanners" applies (see http://www.techstreet.com/products/1275). 

With respect to the question of whether it is preferable to inspect 100% of the images produced or only a sample, see ANSI/AIIM TR34-1996 "Sampling Procedures for Inspection by Attributes of Images in Electronic Image Management (EIM) and Micrographics Systems" (see http://webstore.ansi.org/RecordDetail.aspx?sku=ANSI%2FAIIM+TR34-1996), which states that inspection of a sample is statistically valid, and offers a variety of sampling rates, depending on the level of confidence required. 

Being able, in legal or regulatory proceedings, to call in a statistician to testify to the validity of the sampling process is likely to be more effective and, it may even be possible to avoid this by showing that your image quality control process is based on the MS-44 Standard and TR-34 Technical Report, ALTHOUGH of course, you have to be able to demonstrate, to the satisfaction of the court and/or regulatory entity, that you have actually carried out the procedures of those documents. The latter calls for producing and maintaining adequate logs of the QC process, and possibly testimony by the staff performing it.

From my own personal experience with conversions, I can tell you that 100% inspection is valid only for PR purposes, as the process of viewing page-by-page results very quickly in fatigue and eye strain, and the effectiveness of the inspection suffers accordingly. 

Ann says "Our current policy is 'Scanned images need to be reviewed and deemed legible and stored in the trusted system prior to deleting the paper original. No page of any record, paper, or document shall be destroyed if any page cannot be reproduced on film with full legibility. Every unreproducible page shall be permanently preserved in a manner that will afford easy reference.' "

I have two issues with this policy statement. 

The first issue is that you may occasionally run across pages that are not reproducible in ANY imaging system (whether film-based or electronic) for the simple reason that the page has deteriorated (e.g., thermal fax paper). In such case, you may be better off inserting in the electronic image file a Comments sheet that states "illegible original" (or similar language that your GC can develop). This solution has precedents going back to the days of microfilm (see ANSI/AIIM MS-23-1998 "Standard Recommended Practice - Production, Inspection, and Quality Assurance of First-Generation, Silver Microforms of Documents" http://www.techstreet.com/products/1259).

My second issue is, if you find an "unreproducible page" (e.g., the fax paper), how will you "permanently preserve [it] in a manner that will afford easy reference" ? It is fading; thermal paper was never designed for long-term record-keeping!

Ann also said "Certain business areas have documents over 500 pages and we are finding it impossible to do a thorough review of every single page. So we are exploring adopting a less restrictive policy that uses due diligence to sample documents that provides a reasonable effort and will cover us in the cases when pages are missed or not legible. With that said, our paper record conversion to electronic process would include steps to help ensure they are complete and legible."

This would tie in well to using ANSI/AIIM MS-44 and TR-34. One of the simple steps, in addition to those described, is to scan both sides of every sheet. Good scanning software (for example, Kofax VRS--see http://www.kofax.com/software/vrs/features.php--this is NOT to endorse the product) can be set to automatically delete the blank pages.

Another way to check legibility is to convert the text image back to character code (i.e., letters, digits, etc.) by using Optical Character Recognition (OCR) software and setting the threshold of legibility fairly high. The software should then alert you to text that it cannot convert.

Ann also said "For example, we are considering letting the Document Control Group run a multipage paper record through our Ricoh MFD's ADF scanner or a production level Fujitsu scanner ADF ... " I would avoid at all costs using ANY multi-function copier for scanning important records when you want to destroy the paper. Those machines just aren't very good scanners (this will probably bring cries of outrage from vendors of such machines, but too bad). 

And also " ... which should represent a reasonable effort given that the reviewer verifies that the top page is legible and the job ran without jamming etc.   Is it fair to make an assumption that all of the pages of one document scanned within a system/process must be considered legible if the first page is legible?" As Abraham Miller pointed out, NO, this is not a fair assumption. Don't do it.

Following the preceding statement, you said "If not, then we would have the Document Control Group maintain the paper original for a period of time and then destroy.  In the event an error is identified by staff, the electronic version could be corrected." Well no; this does NOT substitute for a carefully-designed image QC program. If you're keeping the paper originals one (1) year, as mentioned by another respondent, you're ASSUMING that SOMEONE (who?) will by then detect most, if not all, of the errors in your process. A terrible assumption.

In addition, what about the indexing of your record images, the quality control of that indexing (an entirely separate process from image QC), and the storage of the images in a repository with appropriate access controls (oops, someone accidentally deleted all of the records for the critical contract!!), system backups, a reliable way to tie the index to the images, and a satisfactory (not too hard to use) search engine to find the specific record you're looking for?

Finally, what is the retention period for these records? If it's more than fifteen or twenty years, you're probably better off keeping the paper, as NO electronic system is viable over the very long term. Alternatively, you can output your document images and index to 16mm microfilm, which (correctly processed and stored) has a lifespan in excess of 500 years.

The bottom line is, good document management can be done, but it's not a trivial pursuit (pun intended).

Good luck,

Fred
---------------------------------
Frederic J. Grevin
[log in to unmask]
212-312-3903
Vice-President, Records Management
New York City Economic Development Corporation

List archives at http://lists.ufl.edu/archives/recmgmt-l.html
Contact [log in to unmask] for assistance
To unsubscribe from this list, click the below link. If not already present, place UNSUBSCRIBE RECMGMT-L or UNSUB RECMGMT-L in the body of the message.
mailto:[log in to unmask]

ATOM RSS1 RSS2