An OMPF Mini Reconstruction File – A Source I Want to Understand Better

A few months ago, I opened a 191-page PDF about my maternal grandfather, George Dubinsky: payment rolls, morning reports, movement orders, a statistical hospital record, and a Veterans Administration form from 1987. I had paid Twisted Twigs Genealogy to research and copy the records directly at the National Archives and Records Administration (NARA), and its cover called the resulting collection a “WWII Army OMPF Mini Reconstruction File.”¹

I know how to read many of these documents. What I did not fully understand was what kind of source I was looking at when I read them.

The package is not one continuous file created and maintained during George’s service. Its first section contains what remains of an NPRC file bearing reconstruction registry number R 6 980 936. Twisted Twigs added morning reports and other records copied from separate NARA holdings.² Some pages are digital images of records the Army created in 1936 or 1944. Others transmit information through several generations of extraction and decoding. They cannot all be evaluated in the same way.

What an OMPF Is Supposed to Be

An Official Military Personnel File (OMPF) is the accumulated administrative record of a service member’s career. It may contain enlistment and separation documents, assignments, promotions, medical records, awards, and correspondence.³ An intact OMPF brings many documents together under one name, but “official” does not mean infallible. Even an intact file may contain gaps, later-created entries, or conflicting information. George’s OMPF should have existed. Most of it almost certainly no longer does.

The Fire That Took the File

On July 12, 1973, a fire at the National Personnel Records Center in St. Louis burned out of control for twenty-two hours and destroyed 16 to 18 million Official Military Personnel Files. No duplicate or microfilm copies had been maintained.⁴ The damage fell hardest on Army records for personnel discharged between November 1, 1912, and January 1, 1960, an estimated 80 percent loss.⁵ George’s 1936 and 1945 discharges both fall within that span.

I wrote about the fire itself in “The 1973 NPRC Fire – When Flames Destroyed Military History“, so I will not retrace it here. The short version is that most of George’s original file is likely gone. What I now hold combines the few pages in his NPRC reconstruction file with image copies of auxiliary and organizational records located by a Twisted Twigs researcher at NARA.

What “Reconstruction” Actually Means

Sorting the material by origin reveals at least three kinds of sources.

Image copies of contemporaneous original records. A 1936 Casual Detachment Final Payment Roll documents George’s discharge from the Panama Canal Zone on December 8, 1936.

A handwritten entry dated November 25 on the same record concerns his reduction from corporal to private, not his discharge. A separate Final Payment Roll records his discharge at Indiantown Gap, Pennsylvania, on August 11, 1945.

The package also includes a February 1941 morning report placing “Corp. Dubinsky” at Fort George G. Meade, Maryland,⁸ a June 1944 rotation roster from the Pacific,⁹ and an August 1944 order confirming his return to the United States.¹⁰ These are digital image copies, not the physical papers, but the pictured records were created close to the events they document. They are strong sources for the administrative facts recorded, without making every entry automatically correct.

A layered derivative record. The hospital material has a much more complicated history. One page is a 1944 Surgeon General’s Office hospitalization listing on microfiche. The next is a decoding prepared on April 18, 2025. The printed source statement explains that the underlying Hospital Admission Card data were created for statistical purposes, did not retain names, used sampling, and were made available to NPRC in 1988. George is identified by service number 06880854.¹¹

The decoded page reports a February 1944 admission in New Guinea for undulant fever, or brucellosis. It also reports “loss of foot, part of, one” and an August 1944 disability discharge. Those last two fields conflict sharply with George’s later service. I am 100% sure that this is not George’s record as I know my grandfather was not missing a part of one of his feet!

The hospital admission itself, however, is not an isolated or obviously mistaken match. A morning report dated February 18, 1944, places First Sergeant George Dubinsky, under the same service number, with the Detachment of Patients at the 360th Station Hospital on Goodenough Island, New Guinea. A June 1944 report records him returning from hospital to duty.¹³ The independent records therefore support the time, place, and fact of a hospital episode. They do not support a lost foot or an August disability discharge.

Sampling explains why an admission might be absent from the database; it does not explain contradictory information attached to a matching service number. The problem may lie in the historical coding, the conversion of punched-card data, the alignment or interpretation of fields, or the 2025 decoding. I am leaving that conflict open until I can examine the underlying data and codebook.

A later original record containing information about earlier events. The surviving NPRC section also contains a 1987 Veterans Administration Request for Information, created four months after George’s death to verify service for a burial claim.¹⁴ As a document, it is an original record of that 1987 transaction. Its statements about George’s wartime service, however, were copied or summarized from earlier records. That information is secondary or of undetermined origin. The form is useful confirmation, but its distance from the events and unclear information trail affect the weight I give it.

This is the genealogical thread running through all 191 pages: source, information, and evidence are not interchangeable terms. A source may be original, derivative, or authored; the information within it may be primary, secondary, or undetermined; and the evidence I draw from it may be direct, indirect, or negative depending on the question I am asking.¹⁶

Testing the Reconstruction Against What I Know

The payment rolls establish the endpoints of George’s two periods of service: December 8, 1936, and August 11, 1945. The morning reports add a promotion trail that the surviving personnel pages do not preserve. He appears as a corporal in February 1941, was appointed sergeant on April 14, 1941, was a technical sergeant by July 1942, and was appointed first sergeant effective December 1, 1942.¹⁷

The 1944 records are especially revealing. The February patient-detachment report and June return-to-duty entry support a real hospitalization in New Guinea. A June rotation roster and August movement records then trace his return from the Pacific to California and onward for reassignment. Those documents fit his continued service and eventual assignment to Percy Jones Hospital before his 1945 discharge. What does not fit is not the hospitalization record as a whole, but two decoded fields within it.

What the Reconstruction Can’t Do, and Where I Go Next

A reconstruction cannot restore the single, continuous administrative context that an intact OMPF might have supplied. My next steps are more focused now:

• Obtain the original data record or best available image behind the Surgeon General’s Office listing and the codebook used to interpret it.
• Determine whether “loss of foot” and the disability-discharge result were present in the historical data or introduced through field alignment or later decoding.
• Correlate additional 360th Station Hospital and related unit morning reports between February and June 1944.
• Look for a fuller VA claim file beyond the single 1987 form.
• Ask Twisted Twigs for the archival citation data not recorded in the PDF: series, record group, repository, and roll, box, folder, or file identifiers for each document.

The Takeaway

A reconstruction is not merely a diminished OMPF. It is a research product assembled from records with different creators, purposes, histories, and levels of reliability. In this case, even the most troubling page became more understandable when I separated the supported hospital episode from two unsupported decoded fields.

Understanding a source well enough to use it responsibly means understanding how it was made.

---

Your Turn

Have you worked with a mini OMPF reconstruction or other fire-affected NPRC records? How did you sort image copies of original records from later extracts, and did you ever find a coded entry that contradicted the rest of the record?

---

1. Twisted Twigs Genealogy, “WWII Army OMPF Mini Reconstruction File for George Dubinsky,” cover and contents list, 2025, PDF 1. The documents cited below were located and imaged at the National Archives and Records Administration by a Twisted Twigs researcher, 2025, and are privately held by Kirsten M. Max-Douglas, Blue Ash, Ohio, 2026; that 191-page compilation is cited hereafter as the Dubinsky reconstruction file. ↩︎
2. National Personnel Records Center, reconstruction file jacket for George Dubinsky, service no. 06880854, registry no. R 6 980 936; Dubinsky reconstruction file, PDF 2. ↩︎
3. National Archives and Records Administration, “Military Service Records,” National Archives (https://www.archives.gov/veterans/military-service-records/service-record : accessed 10 August 2026). ↩︎
4. National Archives and Records Administration, “The 1973 Fire, National Personnel Records Center,” National Archives (https://www.archives.gov/personnel-records-center/fire-1973 : accessed 10 August 2026). ↩︎
5. NARA, “The 1973 Fire, National Personnel Records Center.” ↩︎
6. U.S. Army, Overseas Discharge and Replacement Depot, “Casual Detachment, Final Payment Roll,” roll no. 4, Panama, voucher no. 2000, line 8, George Dubinsky, private, discharged 8 December 1936; Dubinsky reconstruction file, PDF 8–10. ↩︎
7. U.S. Army, Separation Center No. 45, Indiantown Gap Military Reservation, Pennsylvania, “Final Payment Roll,” roster no. 424, voucher no. 1323, George Dubinsky, first sergeant, service no. 06880854, discharged 11 August 1945; Dubinsky reconstruction file, PDF 4–7. ↩︎
8. Headquarters, 28th Surgical Hospital (Fort George G. Meade, Maryland), morning report, 26 February 1941, entry for Corp. Dubinsky; Dubinsky reconstruction file, PDF 17. Original at National Archives at St. Louis. ↩︎
9. 274th Replacement Company, 14th Replacement Battalion, U.S. Army Forces in the Far East, “Rotation Personnel on TD with Base Quartermaster,” 7 June 1944, George Dubinsky, first sergeant, service no. 06880854; Dubinsky reconstruction file, PDF 184. ↩︎
10. Army Service Forces, Third Service Command, Fort George G. Meade, Maryland, Special Orders No. 207, extract, 10 August 1944, par. 54, George Dubinsky, first sergeant, service no. 06880854; Dubinsky reconstruction file, PDF 188. ↩︎
11. Office of the Surgeon General, U.S. Army, “U.S. Army SGO Hospitalization File Listing 1944,” microfiche 023, p. 254, service no. 06880854; Dubinsky reconstruction file, PDF 11; with “Information from the Hospital Admission Cards,” decoding created 18 April 2025, PDF 12. The decoding identifies the underlying source as Hospital Admission Card data files created by the Office of the Surgeon General for 1942–1945 and 1950–1954. ↩︎
12. Office of the Surgeon General, U.S. Army, “U.S. Army SGO Hospitalization File Listing 1944,” microfiche 023, p. 254, service no. 06880854; Dubinsky reconstruction file, PDF 11; with “Information from the Hospital Admission Cards,” decoding created 18 April 2025, PDF 12. The decoding identifies the underlying source as Hospital Admission Card data files created by the Office of the Surgeon General for 1942–1945 and 1950–1954. ↩︎
13. Detachment of Patients, 360th Station Hospital, morning report, 18 February 1944, Wataluma, Goodenough Island, New Guinea, George Dubinsky, first sergeant, service no. 06880854; Dubinsky reconstruction file, PDF 183. Also 360th Station Hospital morning report, June 1944, recording Dubinsky’s return from hospital to duty, PDF 182; the exact day is not legible on the supplied image. Compare the SGO listing at note 11. ↩︎
14. U.S. Veterans Administration, “Request for Information,” VA Form 70-3101-4, VA file no. 05 148 001, George Dubinsky, 20 November 1987, filed in NPRC reconstruction file R 6 980 936; Dubinsky reconstruction file, PDF 3. ↩︎
15. U.S. Veterans Administration, “Request for Information,” VA Form 70-3101-4, VA file no. 05 148 001, George Dubinsky, 20 November 1987, filed in NPRC reconstruction file R 6 980 936; Dubinsky reconstruction file, PDF 3. ↩︎
16. Elizabeth Shown Mills, “QuickLesson 17: The Evidence Analysis Process Map,” Evidence Explained: Historical Analysis, Citation & Source Usage (https://www.evidenceexplained.com/content/quicklesson-17-evidence-analysis-process-map : accessed 10 August 2026). ↩︎
17. Headquarters, 28th Surgical Hospital, morning reports, 26 February and 14 April 1941, PDF 17 and 30; and Unit B, 28th Surgical Hospital, morning reports, 2 July and 2 December 1942, PDF 145 and 181; Dubinsky reconstruction file. The 2 December 1942 report states his appointment as first sergeant was effective 1 December 1942. Original at National Archives at St. Louis. ↩︎