7.2 Peer-to-Peer Engagement, Coaching & Overcoming Resistance
Key Takeaways
Investigate query quality, routing, responsibility, and workload before labeling a clinician resistant.
Use honest clinical-communication and workflow purposes; do not promise mortality-score, payment, litigation, or audit protection.
Patient-specific peer coaching that requests clarification is a query and follows the same substantive standards as other queries.
Combine defined response measures with query-appropriateness audits; agreement and speed alone do not establish accuracy.
Escalation timing and roles come from applicable local policy; preserve valid disagreement and avoid repeated queries intended to force a preferred answer.
Make the clinical question worth answering
Provider engagement begins with a usable process and a clear reason for the request. A query that is difficult to find, routed to the wrong person, or already answered in the record creates unnecessary work. Before labeling a clinician resistant, check the question, the delivery route, the available documentation, and the person's responsibility for the encounter.
Common concerns include interruptions, uncertainty about the purpose of a query, disagreement with its clinical premise, duplication, and difficulty recording a response. Ask which problem is occurring rather than assume one explanation applies to every clinician. A national time-use statistic from a different setting cannot establish the burden experienced by the local service.
Connect documentation to the clinician's actual work
Explain how an accurate account of the patient's condition supports clinical communication, care transitions, reliable reporting, and appropriate coding. Invite the clinician to identify where the current process obscures that account. This gives the conversation a practical purpose without promising a particular financial result, quality rank, or protection from litigation.
For example, “The covering team could not tell whether this condition remained active at discharge” identifies a communication gap. “This diagnosis will protect your mortality score” supplies an outcome-directed reason to prefer an answer. The latter is particularly inappropriate within a patient-specific clarification request.
General education may explain how documentation is used in classification and reporting. It should accurately describe the particular system and its limits. Not every inpatient encounter uses APR-DRGs, and a diagnosis does not automatically change a mortality model, GMLOS, or an individual clinician's public rating. Do not replace symptom language with a more severe condition unless the provider actually establishes that condition. Delirium, encephalopathy, hypoxia, and respiratory failure are not interchangeable labels to use for score improvement.
Distinguish education from a query
The 2026 ACDIS/AHIMA query brief distinguishes general documentation discussions and de-identified teaching from requests for clarification about a specific encounter. A peer-to-peer conversation becomes a query when it asks the provider to clarify that patient's documentation. Calling it coaching does not remove the query standards.
During general teaching, an advisor might explain the difference between describing an observation and documenting its clinical significance. During a specific case discussion, the advisor must present the relevant sourced record information neutrally, preserve the provider's independent judgment, and follow the organization's recording process. The request must not connect an answer to reimbursement, denial avoidance, or quality outcomes. Section 8.3 explains how to preserve a verbal query and obtain a record response.
Prepare a focused peer discussion
A useful local preparation sequence is:
- Check the need. Identify the unresolved point and whether the record already supplies a sufficient answer. Withdraw or correct a defective request through the approved process.
- Check the recipient. Confirm who directly assessed and cared for the patient and who is appropriate for the subject matter.
- Assemble the facts. Include relevant findings and conflicting information with their sources. Avoid turning a laboratory value or treatment into your own diagnostic conclusion.
- Choose the time and channel. Use the approved communication method and account for the clinician's workload. Immediate patient-care concerns follow clinical escalation rather than wait for a documentation discussion.
- Define the next action. The goal may be a response, a corrected query, a factual clarification, or a process fix. It is not agreement with the reviewer's preferred diagnosis.
This sequence is an operational example. It does not impose a national number of reminders, a response deadline, or a required hierarchy of job titles.
Use scripts that leave the clinical answer open
Concern: “I already documented this.”
An appropriate response is: “Let's review the entry together. If it answers the question, we can resolve the request through our process. The point we could not reconcile was the difference between these two dated entries.” If a genuine conflict remains, use a compliant clarification rather than dismiss the provider's concern or ask them to copy a favored phrase.
Concern: “The treatment makes the diagnosis obvious.”
A useful response is: “The treatment helps us understand the course, but we need your interpretation of the condition being evaluated or treated. Please clarify the relevant documentation through the approved response channel.” Do not affirm a diagnosis solely from ventilator settings, vasopressors, or medication use. Do not add an assumed cause such as aspiration pneumonia to make the answer more specific.
Concern: “These requests are filling my inbox.”
An appropriate response is: “Which requests are duplicative or reaching you when another clinician should answer? We can review a sample and the routing process. For the remaining necessary questions, let's confirm how you can find the indicators and record a response.” This treats workload and query quality as problems the program can investigate, rather than reasons to pressure the clinician.
These scripts introduce a discussion; they are not complete patient-specific query templates. Any actual request still needs its encounter-specific facts, appropriate format, and retained communication.
Learn from the workflow
A locally approved observation or shadowing exercise can show where requests interrupt care and how the query interface appears to the recipient. CDI staff might observe documentation workflow, while clinicians review a de-identified or appropriately authorized case with coding and CDI staff. Define the learning purpose, privacy safeguards, participants, and follow-up. An observation exercise is one option, not a proven universal best intervention or a requirement to watch bedside procedures.
Short case-based teaching can address a recurring verified gap. A service may benefit from a demonstration of where to record an answer, while another needs discussion of a clinical distinction with appropriate subject-matter expertise. Match the method to the demonstrated need. A lecture is unlikely to correct a message sent to the wrong covering team.
Give feedback that measures accuracy as well as activity
Use a small set of defined measures and examples. Response time can reveal routing delays; a sample audit can reveal unclear or unsupported queries. Do not rank clinicians solely by agreement or count every negative response as failure. A correct answer of no additional condition, another explanation, or inability to determine can improve record integrity.
For a hypothetical review, 18 of 20 queries receive a response, giving a 90% response rate under the stated denominator. A separate audit finds four of those 20 queries inappropriate. That is a 20% query-defect rate in the reviewed set. The response rate does not establish query quality, and the team should examine the defects rather than celebrate activity alone. Small samples and different patient populations also limit provider-to-provider comparisons.
Share individual feedback through appropriate local channels, check that the data are correct, and allow the recipient to explain a discrepancy. Use de-identified aggregate findings for broader education when suitable. The purpose is to identify a correctable process or documentation issue, not reward agreement with a predetermined diagnosis.
Escalate the unresolved issue, not a preferred answer
Follow the organization's approved escalation process after checking whether the query is appropriate and properly delivered. A local process may include CDI leadership, a physician advisor, a clinical department leader, or another designated role. The particular sequence and timing come from local policy and applicable governance.
Distinguish nonresponse from a clinical disagreement and from disruptive conduct. A supported disagreement is not insubordination merely because it does not change the DRG. Repeated clarification may be appropriate for new information or unresolved specificity, but sending the same question to another provider to circumvent an answer is improper pressure under the 2026 brief. Privilege or disciplinary decisions belong to authorized processes; the CDI reviewer does not impose an automatic suspension after a fixed number of days.
Close the loop by documenting the response or unresolved status according to policy, obtaining needed provider documentation, and addressing the workflow defect that caused unnecessary friction. The measure of engagement is a reliable, accurate clarification process that clinicians can use.
Primary references
- 2026 ACDIS/AHIMA query brief, verbal queries, appropriate recipients, multiple queries, and local escalation.
- AHIMA CDI ethical standards, collaboration and education.
- CMS risk-adjustment overview, model-specific interpretation rather than guaranteed score effects.
During peer coaching about a particular encounter, a clinician says, “The patient is ventilated, so the diagnosis is obvious.” Which response preserves diagnostic responsibility and query neutrality?
Agree that ventilation conclusively establishes the diagnosis and ask only for the preferred severity wording.
Explain which diagnosis would improve the unit’s mortality result before asking the clinician to select it.
Tell the clinician that no clarification can be sought because a treatment is already recorded.
Explain that treatment informs the review, then seek the provider’s supported interpretation through the appropriate documented clarification process.
In a hypothetical set of 20 queries, 18 receive a response. A separate appropriateness audit of those same 20 queries identifies four defective queries. What does this show?
The response rate is 90% and the observed query-defect rate is 20%; the response rate does not establish query quality.
The response rate is 90%, which establishes that 90% of the queries were clinically appropriate.
The defect rate is 2%, because four defects should be divided by a standard denominator of 200.
The figures establish that the physicians with the lowest agreement rates caused the query defects.
A provider has not answered a query after two reminders. The reviewer has not checked whether it reached the correct recipient or which escalation policy applies. What is the best next step?
Use a universal five-day physician-champion rule because all organizations must follow the same sequence.
Verify the need, routing, recipient, and applicable local escalation process before taking the designated next action.
Classify the silence as agreement and close the query as a confirmed diagnosis.
Refer the matter directly for privilege suspension because two reminders establish a national disciplinary threshold.
Sections you finish are checked off in the contents.