Help & Troubleshooting

Troubleshooting & Best Practices

Follow these guidelines to get accurate extractions and ratings every time. Most extraction issues come from two sources: image-based PDFs and PII that confuses the AI.

Before every upload — two quick checks

Can you click and select text in the PDF? (It must be text-based, not a scan)
Have you replaced the applicant's name, SSN, and DOB with placeholders?

PDF Requirements

Text-Based PDF — Correct

The PDF contains an actual text layer. Characters are stored as data, not as pixels.

How to check: open the PDF and try to click-drag to highlight a word. If text highlights, you are good.

"The applicant demonstrates 10% WPI..." ← selectable
Image-Based PDF — Will Fail

The PDF is a photograph or scan of a printed page. There is no text — only pixels. The AI has nothing to read.

How to fix: run OCR (Optical Character Recognition) using Adobe Acrobat, Smallpdf, or ilovepdf.com before uploading.

[image of text] ← no selectable text layer

PII Replacement Reference

Recommended

Use this as a checklist when preparing a document. Replace each type of PII with a neutral placeholder before uploading. The AI does not need any of this information to extract medical findings.

Category
Example Original
Replace With
Name
John Smith
APPLICANT_NAME
SSN
123-45-6789
XXX-XX-XXXX
Date of Birth
06/12/1975
[DATE OF BIRTH]
Address
456 Oak Street, Los Angeles
[ADDRESS REDACTED]
Physician Name
Dr. Maria Torres
TREATING_PHYSICIAN
Insurance Carrier
State Farm Ins.
INSURANCE_CARRIER
Claim Number
ADJ1234567
[CLAIM NUMBER]

You still enter the real date of injury, date of birth, and occupation group manually into the case fields — the AI only needs the medical findings from the document itself.

Best Practices

Replace PII Before Uploading
HIGH

Swap personally identifiable information with neutral placeholders before uploading any report. The AI only needs medical findings — not the patient's identity.

1
Open the PDF in Adobe Acrobat, Preview, or any PDF editor
2
Use Find & Replace or the Redact tool to replace the applicant's full name with APPLICANT_NAME
3
Replace the Social Security Number with XXX-XX-XXXX
4
Replace the date of birth with [DATE OF BIRTH] — the system uses the date you enter manually, not what is in the document
5
Replace street addresses with [ADDRESS REDACTED]
6
Replace physician names with TREATING_PHYSICIAN or DR_1, DR_2 if multiple
7
Replace insurance carrier names and claim numbers with generic labels
8
Save the redacted PDF and upload that version

The rating engine only needs WPI values, impairment codes, dates of injury, and occupation group. None of that requires the patient's name or SSN.

Verify the PDF Is Text-Based, Not a Scan
HIGH

The AI reads text directly from the PDF. If the document is a scanned image inside a PDF wrapper, there is no text to read and extraction will fail or produce incorrect results.

1
Open the PDF in any viewer (Preview, Chrome, Adobe Reader)
2
Try to click and drag to select a word in the body of the report
3
If text highlights, the PDF is text-based — safe to upload
4
If nothing selects or the cursor behaves like you are clicking an image, the file is a scanned image
5
To fix a scanned PDF: open in Adobe Acrobat and run Enhance Scans → Recognize Text (OCR). This converts the image to selectable text
6
Free alternatives: use Adobe online OCR, Smallpdf, or ilovepdf.com to run OCR before uploading
7
Re-check after OCR: try selecting text again before uploading the converted file

Even if OCR is applied, check the quality. Poor scan quality (low DPI, handwriting, rotated pages) can cause mis-reads in WPI values. Always review the extracted data before running the rating.

Always Review Extracted Data Before Rating
MEDIUM

AI extraction is highly accurate but not infallible. Before running the rating, verify every extracted field matches what is in the report.

1
After extraction completes, open the extracted data panel
2
Compare each WPI value against the corresponding page in the report
3
Confirm the impairment code matches the body part described by the physician
4
Verify the date of injury — a wrong year changes the applicable schedule
5
Check the occupation group — the AI infers this from job title descriptions; confirm it is correct
6
Verify apportionment percentages if the physician addressed industrial vs. non-industrial causation
7
Correct any field that does not match before proceeding to rating

The math is deterministic — garbage in, garbage out. A correct extraction is what makes the final rating defensible.

Upload One Report per Case
MEDIUM

Each case should correspond to one QME or AME report. Mixing multiple reports in one case can cause the AI to conflate findings from different evaluations.

1
Create a separate case for each QME or AME report
2
If a physician issued a supplemental report, upload it to the same case as an additional file — the AI will read both together
3
Do not combine the applicant's QME and the defendant's QME into one case — create two separate cases and compare ratings independently
4
For panel QMEs with multiple body parts evaluated in one report, a single upload is correct — the AI will extract all impairments

One evaluating physician per case keeps the extraction clean and the rating traceable to a single document.

Use Consistent Occupation Group Numbers
LOW

The occupation group drives the occupation adjustment step. Using the wrong group is one of the most common rating errors.

1
Look up the applicant's occupation in the 2005 PDRS Occupational Variant table
2
Use the numeric group code (e.g., 490 for Light work, 360 for Heavy work) — not a free-text job title
3
If the occupation is not listed, use the closest equivalent and note it in your case title
4
When in doubt, use the occupation group the QME physician referenced in the report
5
The AI will attempt to infer the group from the job title in the document — always confirm this is correct before rating

A difference of one occupation group can shift the final PD rating by 1–3 percentage points.

Common Issues & Fixes

Extraction returns no impairments

Likely Causes

  • PDF is a scanned image with no OCR text layer
  • PDF is password protected
  • Report does not use standard WPI terminology — the physician may use descriptive language instead of a numeric WPI

Fix

Run OCR on the document first, remove any password protection, and check that the report explicitly states a whole person impairment percentage.

WPI values are wrong or off by a decimal

Likely Causes

  • Low-quality scan caused OCR to misread digits (e.g., 10% read as 1%)
  • The PDF has multiple WPI tables and the AI picked the preliminary rather than the final value

Fix

Manually correct the extracted WPI in the data panel before rating. Compare against the specific page and paragraph in the report.

Occupation group is incorrect

Likely Causes

  • The report describes the job by title without a PDRS group number
  • The job title maps ambiguously to multiple occupation groups

Fix

Override the occupation group in the extraction panel. Look up the correct group in the PDRS Occupational Variant table.

Date of injury pulls the wrong year

Likely Causes

  • The report mentions multiple dates (filing date, exam date, injury date) and the AI selected the wrong one
  • The date format in the document is ambiguous (e.g., 01/02/23)

Fix

Correct the date of injury in the extraction panel. This is critical — the wrong year can change which schedule applies (pre- vs. post-SB 863).

Rating seems too high or too low

Likely Causes

  • Apportionment was not captured — the AI defaulted to 100% industrial when the physician apportioned
  • Pain add-on was included or excluded incorrectly
  • Occupation group is wrong

Fix

Review the extracted apportionment and pain add-on values. Check the physician's apportionment language in the report and correct the extracted values before re-rating.