Managing Cdi Programs

---
name: managing-cdi-programs
description: Structures clinical documentation improvement queries with compliant physician engagement. Use when writing CDI queries, improving documentation specificity, or managing CDI programs.
tags:
  - management
  - medical-coding-and-billing
  - clinical
metadata:
  author: casemark
  practice_areas:
    - Medical Coding
    - Revenue Cycle
    - Health Information Management
  document_types:
    - Management Report
  skill_modes:
    - Management
    - Coordination
---

# Managing CDI Programs

Structures clinical documentation improvement (CDI) queries with compliant physician engagement, tracks query impact on coding accuracy and DRG assignment, and manages CDI program operations including concurrent review workflows, CDI specialist staffing, and performance metrics. Covers inpatient, outpatient, and risk adjustment CDI programs.

## Why This Skill Exists

Clinical documentation drives code assignment, which drives reimbursement, quality metrics, severity of illness reporting, and compliance risk. CDI programs bridge the gap between clinical knowledge and coded data — physicians often know more than they document. ACDIS data shows that effective CDI programs improve CMI by 0.05–0.15 points, recover $1,500–$3,000 per query that results in a DRG change, and reduce coding denials by 10–20%. However, non-compliant queries (leading, biased, or suggestive of diagnoses) create audit liability and can be flagged as upcoding schemes. Compliant CDI requires structured methodology.

---

## Checkpoint A — Intake

### Questions to Confirm Before Starting

1. What is the CDI program scope? (inpatient DRG, outpatient HCC/risk adjustment, quality metrics)
2. What are the target service lines or DRG families?
3. What is the current query volume, response rate, and agreement rate?
4. What query format is used? (verbal, written, electronic template, EMR-integrated)
5. What is the CDI-to-bed ratio and concurrent review coverage?
6. Are there known documentation gaps from prior audits or denial trends?
7. What physician engagement mechanisms exist? (education, feedback, peer comparison)

### Documents Required

- CDI program policy and procedure manual
- Query templates currently in use
- CDI metrics dashboard (query volume, response rate, agree rate, impact rate)
- CMI trend reports by service line
- Coding denial reports (documentation-related denials)
- CDI specialist staffing and assignment model
- Physician documentation education materials
- Payer audit findings related to documentation specificity

---

## Step 1 — Design Compliant CDI Queries

Follow AHIMA/ACDIS guidelines for non-leading, clinically grounded queries.

**Compliant query structure:**
1. **Clinical indicators**: Present the objective findings that triggered the query (lab values, vital signs, imaging results, treatments administered).
2. **Open-ended question**: Ask the physician to clarify the clinical significance of the indicators — NOT to add a specific diagnosis.
3. **Multiple-choice options** (when applicable): Provide clinically reasonable options including "unable to determine" and "clinically not applicable."
4. **Documentation instruction**: Request that the physician document their response in the medical record (progress note, discharge summary, or addendum).

**Compliant example:**
> The patient's WBC is 18,200, lactate is 3.8, temperature is 101.4°F, and blood cultures were drawn with IV antibiotics initiated. Based on your clinical judgment, please clarify the clinical significance of these findings. Possible considerations include: (a) Sepsis, (b) SIRS due to infection, (c) Bacteremia, (d) Other — please specify, (e) Unable to determine at this time.

**Non-compliant patterns to AVOID:**
- Leading: "Would you agree the patient has sepsis?"
- Yes/no: "Does the patient have acute respiratory failure?"
- Diagnosis-specific without clinical indicators: "Please add malnutrition to the problem list."
- Retrospective queries that ask physicians to document conditions not assessed during the encounter.

## Step 2 — Identify High-Impact Query Opportunities

Focus CDI review on cases with the highest documentation improvement potential.

**Inpatient DRG targets:**
- Cases where a single CC or MCC would shift the DRG to a higher severity level.
- Clinical indicators suggesting conditions not yet documented: acute kidney injury (creatinine elevation), malnutrition (BMI, albumin, prealbumin), respiratory failure (ABG values, ventilator settings), encephalopathy (mental status changes), sepsis (SIRS criteria + infection source).
- Cases with LOS exceeding the GMLOS for the current DRG — potential undercoded severity.
- Cases with high-cost treatments (ICU stay, ventilator, pressors) but low-severity DRG assignment.

**Common query opportunities by condition:**
| Clinical Indicators | Potential Query Target |
|---------------------|----------------------|
| Creatinine rise ≥0.3 mg/dL in 48h or ≥1.5× baseline | Acute kidney injury staging |
| BMI <18.5, albumin <3.0, weight loss >5% | Malnutrition type and severity |
| PaO2 <60, PaCO2 >50, on supplemental O2 or ventilator | Acute vs. chronic respiratory failure |
| WBC >12K or <4K, temp >38.3°C or <36°C, HR >90, RR >20 with infection | Sepsis vs. SIRS |
| Mental status changes, confusion, altered LOC | Encephalopathy type and etiology |
| EF <40% with fluid overload symptoms | Acute on chronic systolic/diastolic heart failure |

## Step 3 — Manage Concurrent Review Workflow

Structure the day-to-day CDI review process for maximum coverage and impact.

- **Admission review** (within 24–48 hours): Initial assessment of documentation completeness, principal diagnosis accuracy, and CC/MCC capture opportunities.
- **Continued stay review** (every 48–72 hours): Monitor for evolving conditions, new clinical indicators, and query opportunities as the patient's course develops.
- **Pre-discharge review**: Final documentation completeness check before the coder assigns codes. This is the last opportunity for prospective queries.
- **Post-discharge reconciliation**: Compare CDI working DRG to the final coded DRG. Investigate discrepancies — they indicate either missed query opportunities or coding issues.
- Prioritize high-CMI service lines: cardiac, pulmonary, sepsis/infection, surgical, oncology, neurology.
- Maintain a query tracking log with: query date, physician, condition queried, response date, response agreement, DRG impact.

## Step 4 — Track and Report CDI Program Metrics

Measure program effectiveness with industry-standard KPIs.

- **Query volume**: Total queries issued per month. Target varies by organization but typically 150–250 per CDI specialist per month.
- **Physician response rate**: Percentage of queries that receive a physician response. Target: ≥85%.
- **Query agreement rate**: Percentage of responses where the physician agrees with the query or documents a qualifying condition. Industry benchmark: 70–85%.
- **CDI impact rate**: Percentage of cases reviewed where CDI review resulted in a DRG change. Target: 15–25%.
- **CMI impact**: Monthly CMI with CDI vs. projected CMI without CDI. Track incremental CMI lift.
- **Revenue impact**: Dollar value of DRG upgrades attributable to CDI queries. Calculate as (upgraded DRG weight − original DRG weight) × blended rate.
- **Review coverage rate**: Percentage of eligible admissions reviewed by CDI. Target: ≥80% of medical and ≥60% of surgical admissions.

## Step 5 — Engage Physicians Effectively

Build sustainable physician collaboration for documentation improvement.

- **Peer comparison reports**: Show each physician their documentation specificity rates compared to department peers (anonymized). Competition drives improvement.
- **1:1 education sessions**: Target physicians with low query response rates or low agreement rates. Use specific case examples, not abstract guidelines.
- **Service line champions**: Identify physician leaders in each department to advocate for documentation improvement within their peer group.
- **Documentation templates**: Collaborate with physicians to build EMR templates that prompt for specificity (e.g., heart failure type, diabetes complications, malnutrition severity).
- **Feedback loop**: When a CDI query leads to a DRG change, inform the physician of the documentation improvement's impact. Positive reinforcement increases future compliance.
- **Avoid "query fatigue"**: Limit queries to clinically meaningful opportunities. Excessive queries for low-impact conditions erode physician trust.

---

## Checkpoint B — Review

- [ ] All queries follow AHIMA/ACDIS compliant format (non-leading, clinically grounded, open-ended)
- [ ] Clinical indicators are documented in the query to support the basis for the question
- [ ] Query opportunities target high-impact DRG changes, not marginal documentation tweaks
- [ ] Concurrent review coverage meets organizational targets
- [ ] Physician response tracking is current and non-responders are escalated
- [ ] CDI impact metrics are calculated accurately (DRG impact, CMI lift, revenue recovery)
- [ ] Post-discharge reconciliation identifies gaps between CDI working DRG and final coded DRG
- [ ] No query asks the physician to document a condition not supported by clinical indicators

---

## Quality Audit

- [ ] Query compliance is audited quarterly — random sample reviewed for leading or non-compliant language
- [ ] Agreement rates exceeding 95% are investigated — may indicate leading queries
- [ ] Agreement rates below 50% are investigated — may indicate poorly targeted queries
- [ ] CDI impact metrics are validated against actual DRG changes (not just projected changes)
- [ ] Physician response rates are tracked individually — non-responders receive escalation per policy
- [ ] CMI trends are analyzed for unusual spikes that might indicate overcoding risk
- [ ] CDI program ROI is calculated annually (program cost vs. incremental revenue captured)

---

## Guidelines

- Follow AHIMA/ACDIS "Guidelines for Achieving a Compliant Query Practice" (2019 update)
- Apply ACDIS CDI program standards for concurrent review workflow and staffing models
- Reference CMS ICD-10-CM Official Guidelines for conditions commonly queried (sepsis, respiratory failure, malnutrition)
- Use AHA Coding Clinic guidance for documentation specificity requirements by condition
- Never issue a query that directs the physician to a specific diagnosis — always present clinical indicators and ask for clinical judgment
- Never query for conditions where no clinical indicators exist in the record
- Mark with [VERIFY] any query where the clinical indicators are marginal or the condition is borderline
- Include disclaimer that CDI programs must operate within compliance boundaries and do not constitute guidance to upcode