Expert review · Work product
Work Samples
Four fictional excerpts show what Dr. Gill's work product looks like: how events are dated, how an opinion is tested against adverse evidence, how an AI-drafted note is compared with its source, and how missing records are requested.
- What these samples are
- Sample 1 (fictional): source-linked chronology excerpt
- Sample 2 (fictional): opinion-and-evidence matrix
- Sample 3 (fictional): AI documentation comparison
- Sample 4 (fictional): missing-records request list
- How actual work product differs
What these samples are
Attorneys and organizations reasonably want to see the work before they commission it. The four excerpts below show the structure Dr. Gill uses to keep a review honest: events dated by more than one clock, each opinion tested against the evidence that cuts against it, AI-drafted text compared with its source by category of change, and missing records requested with a stated reason.
Read these as illustrations
Every sample on this page is fictional. Names, times, page references, results and products are invented. Each sample states what it illustrates and what it cannot establish. Actual work product is shaped by the agreed scope, and a report prepared for a proceeding follows the requirements of that engagement and proceeding.
Samples 1, 2 and 4 use the same fictional matter, so they can be read as one review in miniature. Sample 3 uses a separate fictional encounter.
Fictional illustration: the matter
A 71-year-old woman with heart failure and chronic kidney disease is seen in a primary-care clinic on a Monday. Spironolactone is added to her existing lisinopril, and blood tests are ordered for about one week later. Her potassium result, finalized late on a Friday eleven days later, is critically high. She arrives at a hospital early Sunday morning with weakness and a slow heart rate. Counsel asks what the clinic knew, and when.
Sample 1 (fictional): source-linked chronology excerpt
Fictional illustration
What it illustrates: separate columns for when an event occurred, when it was entered, when information became available, and when anyone acknowledged or acted on it; a source for every row; and unknowns written as unknown rather than filled in. What it cannot establish: a chronology organizes evidence. It does not decide whether any interval was clinically acceptable, and an interface timestamp is not proof that a person saw the result.
| Event | Occurred | Entered | Available | Acknowledged or acted on | Source | Status and unknowns |
|---|---|---|---|---|---|---|
| Clinic visit; spironolactone 25 mg daily added; potassium and creatinine ordered for one week | Day 1 (Mon) 10:20 | Order Day 1 10:41; note signed Day 2 19:45 | Not applicable | Not applicable | Clinic note p. 112; order p. 115 | Documented Note signed the following evening; the order was entered during the visit. |
| Blood drawn at outside laboratory | Day 11 (Thu) 15:30 | Requisition Day 11 15:30 | Not applicable | Not applicable | Lab requisition p. 201 | Documented Drawn later than the one-week interval ordered; reason not recorded. |
| Result finalized: potassium 6.8 mmol/L (critical); creatinine 2.1 mg/dL (prior 1.4) | Day 12 (Fri) 16:52 | Interfaced Day 12 16:53 | Day 12 16:53 in the EHR | Day 15 (Mon) 08:14 | Lab report p. 203; acknowledgement entry p. 204 | Inferred Routing to the ordering physician's inbox is inferred from the interface time; the routing destination is not confirmed. |
| Laboratory critical-value call | Day 12 17:20 | Lab call log Day 12 17:21 | To answering service Day 12 17:20; to a clinician: unknown | Unknown | Lab critical-call log p. 207 | Documented Read-back recorded by "operator J." Unknown Whether the message was relayed, and to whom. |
| Answering-service message to on-call clinician | Unknown | Not in records produced | Unknown | Unknown | None | Unknown See missing-records items 1 to 3. |
| Hospital arrival with weakness and bradycardia; potassium 7.1 mmol/L | Day 14 (Sun) 06:40 | Hospital record Day 14 | To clinic: not established | Not applicable | Hospital record p. 310 | Documented Emergency care is outside Dr. Gill's scope; included for timing only. |
| Clinic phone note: daughter reports patient hospitalized | Day 15 (Mon) 08:40 | Day 15 08:44 | Not applicable | Day 15 08:44 | Clinic phone encounter p. 118 | Documented First clinic contact with the family appearing in the record. |
Two things in this excerpt matter more than any single row. The result was available in the system for about sixty-three hours before the first acknowledgement entry, and the record does not show what happened to the critical-value call after 17:20 on Friday. The chronology states both without supplying a story to connect them.
Sample 2 (fictional): opinion-and-evidence matrix
Fictional illustration
What it illustrates: one proposed conclusion stated precisely, with supporting and adverse evidence side by side, the alternative explanations, the limits, and the specific evidence that would change the opinion. Uncertainty is described in words, not given an invented numerical score. What it cannot establish: the matrix tests a factual and clinical conclusion. It does not decide negligence or causation, and it deliberately leaves causation to the disciplines that would need to address it.
| Element | Entry |
|---|---|
| Question | Before the patient's hospital arrival on Day 14, did a clinician receive the critical potassium result? |
| Proposed conclusion | On the records produced, the critical result reached the clinic's after-hours answering service at 17:20 on Day 12. The records do not establish that it reached any clinician before 08:14 on Day 15. |
| Supporting evidence | Laboratory call log with read-back to the answering service (p. 207). No phone encounter, on-call note or order in the clinic record between Day 12 17:20 and Day 15 08:14. The first acknowledgement entry is Day 15 08:14 (p. 204). |
| Adverse evidence | The laboratory's read-back shows the communication chain was working at its first step. Some practices document after-hours calls outside the EHR, so the absence of a note does not show that no call occurred. The clinic's critical-result policy, if it required the service to page the on-call clinician, would make it more likely that a page was attempted. |
| Clinical sources and version | The clinic's critical-result policy in effect on Day 12 (requested; not produced). ONC SAFER guidance on test-results reporting as a process reference only, in the edition current on the event date. Neither is treated as the legal standard. |
| Reasoning | The only documented recipient after 17:20 is the answering service. Receipt by a clinician would ordinarily leave a trace in a call log, an after-hours note or an order. None has been produced, but several of the records that would contain such a trace have not been requested yet. |
| Alternative explanations | (a) The service paged the on-call clinician, who called the patient but did not document the call. (b) The service paged a clinician who was not on call. (c) The service held the message for Monday under its own protocol. (d) The message was lost. |
| Causation | Not addressed. Whether contact on Friday evening would have changed the outcome depends on the hospital course and on cardiology or nephrology analysis. Emergency-department care is outside Dr. Gill's scope. |
| Limits | Answering-service records, the on-call schedule and the clinician's call records were not produced. EHR routing and viewing have not been technically verified. |
| What would change this opinion | An answering-service log showing a page and a callback. Telephone records showing a call from the on-call clinician to the patient. A separate after-hours log. A validated audit extraction showing a covering clinician opened the result before Day 15. |
| Status | Supported on the records produced; provisional pending missing-records items 1 to 5. |
A conclusion like this can be useful even though it does not resolve the dispute. It tells counsel exactly where the open question sits and what evidence would close it, in either direction.
Sample 3 (fictional): AI documentation comparison
Fictional illustration
What it illustrates: a clinically relevant difference between source material and a generated note, classified by type, with its support, clinical significance, and whether it was corrected and where it persisted. What it cannot establish: one encounter says nothing about how often a product makes any of these errors. The transcript is itself an imperfect source, and a difference is not automatically an error when the clinician supplied information not captured in the audio.
Fictional case record. Case FIC-03. A 45-year-old man with cough, accompanied by his spouse. Source packet: fictional visit transcript, the clinician's separately dictated examination findings, and the prior medication list. Product: a fictional ambient documentation tool; version and configuration not applicable. Final destination: the signed visit note, with linked problem and medication lists.
| Source (transcript or clinician input) | Generated draft | Final signed note | Classification | Clinical significance | Correction and persistence |
|---|---|---|---|---|---|
| Smoking not discussed | "Former smoker, quit 10 years ago." | Unchanged | Added assertion | Affects the differential and future lung-cancer screening eligibility | Not corrected; copied to social history |
| "If the fever comes back or you get short of breath, go to the emergency room tonight." | "Return precautions discussed." | Unchanged | Omission | A later reader cannot tell what thresholds the patient was given | Not corrected |
| "This might be early pneumonia, but more likely a viral bronchitis." | "Assessment: Pneumonia." | "Acute bronchitis, likely viral; pneumonia considered." | Changed certainty | An established pneumonia diagnosis would shape treatment and follow-up | Corrected in the note; "pneumonia" remained on the problem list |
| Chest pain not discussed | "Denies chest pain." | Unchanged | Negation | Implies a risk was screened for when it was not | Not corrected |
| Spouse: "I've been on lisinopril, and it gave me a cough." | "Patient reports cough since starting lisinopril." | Unchanged | Attribution | Wrong medication history; could prompt an unnecessary medication change | Not corrected; lisinopril added to the patient's medication list |
| "If it's not better in a week, we'll get a chest X-ray." | "Chest X-ray ordered." | Unchanged | Plan change | A later clinician may assume imaging was done | Not corrected; no order exists |
The examination findings in the final note matched the clinician's dictation and were not classified as unsupported. In an actual evaluation, the report would also record the review date, the known product version and configuration, any unavailable details, and, if review time was measured, how timing was defined.
Sample 4 (fictional): missing-records request list
Fictional illustration
What it illustrates: a request list in which each record is tied to the question it bears on and the reason it matters, ranked by how much it could change the analysis. What it cannot establish: a list identifies what is needed. It does not show that a record exists, and the legal route for obtaining it is for counsel.
| # | Record | Likely holder | Question it bears on | Why it matters |
|---|---|---|---|---|
| 1 | Answering-service message log and any call recordings, Day 12 to Day 15 | Answering service, through the clinic | Did the critical value reach a clinician? | The only direct evidence of what happened after 17:20 on Friday |
| 2 | On-call schedule, Day 12 to Day 14 | Clinic | Who was responsible for after-hours results? | Identifies the clinician whose receipt is in question |
| 3 | On-call clinician's call records or after-hours log | Clinician or clinic | Was the patient called? | Absence from the EHR does not exclude a call |
| 4 | Critical-result policy in effect on Day 12 | Clinic | What process was expected? | Separates a design problem from an execution problem |
| 5 | Audit extraction of routing and viewing for the Day 12 result | Clinic, through an EHR forensic specialist | Did anyone open the result before Day 15? | An acknowledgement entry is not necessarily the first view; extraction and validation require a technical specialist |
| 6 | Potassium and creatinine results for the prior twelve months | Clinic and laboratory | Was the starting risk apparent on Day 1? | Baseline for judging the monitoring plan |
| 7 | Pharmacy fill history for all medications | Pharmacy | What was the patient taking? | A fill shows dispensing, not ingestion, but can reveal other sources of potassium |
| 8 | After-visit instructions from Day 1 | Clinic | What was the patient told about laboratory timing and warning symptoms? | Bears on communication and on why the draw occurred on Day 11 |
| 9 | Complete hospital record from Day 14 | Hospital | Condition and timing at arrival | Needed for the chronology and by causation experts; emergency care itself is outside Dr. Gill's scope |
How actual work product differs
These excerpts compress a review into a few tables. An actual work product identifies the assignment, the materials considered and those missing, the method and sources relied on, each opinion with its reasoning and adverse evidence, causal limits, and material assumptions. Factual observations and inferences are kept visibly separate. The form follows the engagement: a consulting memorandum for counsel's internal use looks different from a retained expert's report in federal court, which must meet Rule 26(a)(2)(B), or a California expert designation under Code of Civil Procedure §2034.260.
For the method behind these samples, see How a Review Is Organized. For the reasoning in essay form, see The Medical Record Has More Than One Clock and A Medical Note Can Change Meaning Without Changing Much Text. Services: medical record review, clinical documentation, clinical AI.
Sources and scope
- ONC SAFER Guides (2016 edition for events before 2025; 2025 edition thereafter), Test Results Reporting and Follow-Up: an EHR safety process reference, used in Sample 2 only as such.
- Federal Rule of Civil Procedure 26(a)(2)(B): contents of a retained expert's report in federal proceedings.
- Code of Civil Procedure §2034.260: California expert declaration contents.
Questions counsel often ask
Are these samples drawn from real cases?
No. Every patient, clinic, laboratory, document page number, time and product in these samples is invented for illustration. They describe no real patient, organization, product test or engagement. They were written to show the structure and reasoning of the work product, not to report findings.
Will a report for my matter look like this?
The reasoning will; the format may not. The form of any work product is agreed for the engagement, and a report prepared for a proceeding follows that proceeding's requirements, such as the federal retained-expert report rules or California's expert designation statutes. These excerpts are illustrations, not a universal litigation form.
Are these templates validated instruments?
No. They are organizational templates for structuring a review: a way of keeping occurrence, entry and acknowledgement times apart, keeping adverse evidence in view, and stating limits. They are not diagnostic tests, scoring systems or certification tools, and they do not substitute for clinical judgment.
Have a matter that may fit?
Send a short, non-clinical inquiry naming the parties, forum, dates and question. Conflicts are screened before any records are sent, and every party is offered the same method and terms.
This page describes professional consulting services for attorneys and healthcare organizations. It is not legal advice or medical advice. An inquiry does not create an engagement, an attorney–client relationship or a physician–patient relationship. Opinions are formed only after review of the evidence and may support, limit or contradict the position of the party that requests them. Fees are never contingent on the content of an opinion or the outcome of a matter.