Study guide
4.2 Acute and Chronic Respiratory Failure, ABGs, and Ventilation Support
Certified Clinical Documentation Specialist CCDS
Track your progress as you read
Finished sections get a check mark.
1 Introduction & Exam Overview
1.1 Current Exam Facts 1.2 Eligibility Pathways and the 2,000-Hour Year 1.3 Allowed Resources and Test-Day Rules 1.4 Recertification, Fees, and How This Guide Maps the Eight Domains
2 IPPS Foundations
2.1 How IPPS Works and How CMS Updates It 2.2 MS-DRG Assignment, Relative Weights, and Grouping 2.3 Provider and Clinical Staff Documentation Duties under IPPS
3 Severity, CMI, Medical Necessity, and Oversight
3.1 CC versus MCC Recognition and CC Exclusion Lists 3.2 Case Mix Index and Why CDI Tracks It 3.3 Two-Midnight Rule, Medical Necessity of Setting, and Principal Diagnosis 3.4 MAC, RAC, CERT, OIG, and Denial-Vulnerable Practices
4 Nervous and Respiratory System CDI
4.1 Nervous System Conditions for CDI 4.2 Acute and Chronic Respiratory Failure, ABGs, and Ventilation Support 4.3 Pneumonia, COPD, PE, Pleural Effusion, and Respiratory Infections
5 Circulatory, Digestive, Hepatobiliary, Musculoskeletal, and Skin CDI
5.1 Circulatory Disorders 5.2 Digestive and Hepatobiliary Disorders 5.3 Musculoskeletal and Skin
6 Endocrine, Nutrition, Kidney, and Hematologic CDI
6.1 Diabetes, Dehydration, and Obesity 6.2 Malnutrition Criteria, Severity, and Audit Risk 6.3 AKI, CKD Staging, UTI, and Urinary Stones 6.4 Anemia, Coagulation, Sickle Cell, Lymphoma, and Leukemia
7 Infectious Disease, Other MDCs, and Clinical Indicator Tools
7.1 Sepsis, Severe Sepsis, Postoperative and Other Infections 7.2 Mental, Substance Use, Injuries/Toxicology, and HIV-Related Conditions 7.3 Medications, Diagnostic Tests, and Abbreviations as Clinical Indicators
8 Health Record Elements and Query Triggers
8.1 Which Record Elements Support Diagnosis and Procedure Codes 8.2 Attending versus Radiologist, Pathologist, and Other Providers 8.3 When a Physician Query Is Warranted 8.4 When Queries Are Not Appropriate
9 Query Timing, Verbal Queries, Formats, and Clinical Indicators
9.1 Concurrent, Retrospective, Pre-Bill, and Prospective Timing 9.2 Verbal Queries and Required Documentation 9.3 Open-Ended, Multiple-Choice, and Yes/No 9.4 Clinical Indicators and How to Source Them
10 Compliant Query Practice and Clinical Validation
10.1 Compliant versus Leading and Other Noncompliant Queries 10.2 ACDIS/AHIMA 2026 Guidelines for Achieving a Compliant Query Practice 10.3 Clinical Validation and Diagnoses without Clinical Support
11 CDI Program Metrics
11.1 Productivity, Query Volume, and Provider Response Rates 11.2 CMI, CC/MCC Capture, and High-Frequency DRGs 11.3 SOI/ROM as CDI Program Performance Metrics
12 Physician Metrics, PEPPER, Forecasting, and Public Data
12.1 Measuring Physician Documentation Performance 12.2 PEPPER, Forecasting, Hospital Financial Data, and Public Reporting
13 CDI Roles, Coder Collaboration, and Physician Education
13.1 Roles and Responsibilities of the CDI Specialist 13.2 Coder Roles Working with a CDI Department 13.3 Physician Education Tools, Forms, and Presentations
14 Provider Communication and DRG Reconciliation
14.1 Non-Confrontational Communication and When Verbal Beats Written 14.2 Working DRG versus Final DRG Reconciliation
15 Official Guidelines, Principal Diagnosis, and Secondary Diagnosis
15.1 Official Guidelines, Cooperating Parties, Updates, and AHA Coding Clinic 15.2 Principal Diagnosis Selection 15.3 Secondary Diagnosis Reporting Criteria
16 Sequencing, Coding Conventions, and POA Indicators
16.1 Sequencing When Two Diagnoses Equally Meet PD; Working DRG with Multiple Diagnoses 16.2 Integral Signs/Symptoms, the With Convention, and Uncertain Diagnoses 16.3 Present-on-Admission Indicator Assignment and Payment Impact
17 Hospital-Acquired Conditions, Disposition, and ICD-10-PCS Essentials
17.1 CMS Hospital-Acquired Condition Payment Provision 17.2 Discharge Disposition, Transfer Payment, Principal Procedure, and ICD-10-PCS Essentials
18 Ethics, Privacy, and Leading Queries as Compliance Failures
18.1 ACDIS Code of Ethics and CDI Goals beyond Reimbursement 18.2 HIPAA, Minimum Necessary, and Record Confidentiality 18.3 Leading Queries as a Compliance Failure
19 DRG Creep, OIG Oversight, PEPPER, and the False Claims Act
19.1 DRG Compliance, DRG Creep, and the OIG Work Plan 19.2 PEPPER Compliance Risks and the False Claims Act
20 Mortality Index, Observed-to-Expected Ratios, and Quality Data Sources
20.1 Documentation Impact on Mortality Index (SOI/ROM) 20.2 Mortality Reviews and Observed-to-Expected Ratios 20.3 Abstracted versus Claims-Based Quality Data
21 CMS Quality Programs, PSIs, and Infection versus Complication Codes
21.1 HVBP, HAC Reduction Program, HRRP, 30-Day Mortality, and Hospital IQR 21.2 PSI 90 Components and PSIs beyond the Composite 21.3 Healthcare-Associated Infections versus Complication Codes
4.2 Acute and Chronic Respiratory Failure, ABGs, and Ventilation Support
Key Takeaways Explain This Topic AI
PaO2 under 60 mm Hg and PaCO2 over 50 mm Hg with acidosis are common clinical indicators of hypoxemic and hypercapnic respiratory failure; they are not ICD-10-CM codes and do not let CDI assign respiratory failure without a provider diagnosis.
Specify acute, chronic, or acute on chronic, and hypoxemic versus hypercapnic versus both. A COPD retainer with a new acidotic PaCO2 rise is the classic acute-on-chronic query.
Intubation for airway protection is not automatically acute respiratory failure. Gas-exchange failure, work of breathing, and the provider's diagnostic statement still have to be there.
ICD-10-PCS mechanical ventilation uses consecutive hours: less than 24, 24 to 96, and greater than 96. Count clock hours on the respiratory record, not midnights, and do not code routine intraoperative ventilation that ends in the OR.
Noninvasive support (BiPAP, high-flow nasal cannula) can support a respiratory-failure diagnosis as treatment but does not get dumped into the invasive ventilation duration values.
Last updated: September 2026
4.2 Acute and Chronic Respiratory Failure, ABGs, and Ventilation Support
Quick Answer: Acute respiratory failure is a provider diagnosis . Common indicators — not codes — include hypoxemic failure with PaO2 under 60 mm Hg (or an equivalent SpO2) and hypercapnic failure with PaCO2 over 50 mm Hg plus acidosis . Specify acute , chronic , or acute on chronic , and hypoxemic versus hypercapnic versus both. ICD-10-PCS mechanical ventilation uses consecutive hours : less than 24 , 24–96 , and greater than 96 .
Domain II will hand you an arterial blood gas ( ABG ), a BiPAP start time, and an attending note that says hypoxia. The CCDS skill is knowing what that packet can support and what it cannot. ExamFinch's independent teaching here follows CMS IPPS inpatient logic and the ICD-10-CM/PCS Official Guidelines. It is not an ACDIS handbook reprint and does not claim ACDIS or CMS partnership.
Hypoxemic versus hypercapnic failure
Hypoxemic respiratory failure (often called type 1) is failure to oxygenate. The common teaching threshold is PaO2 under 60 mm Hg on room air, or an equivalent that accounts for supplemental oxygen (a low PaO2/FiO2 ratio, or an SpO2 that the provider treats as failure-range hypoxemia). Hypercapnic respiratory failure (type 2) is failure to ventilate. The common teaching threshold is PaCO2 over 50 mm Hg with acidosis (pH under 7.35), showing that the carbon dioxide rise is not fully compensated.
Those numbers are indicators , not codes. They appear in textbooks, respiratory protocols, and many hospital definitions. They are not an ICD-10-CM inclusion term that lets a coder or CDI specialist assign J96.- from a gas alone. The provider must document respiratory failure (acute, chronic, or acute and chronic, with hypoxemia, hypercapnia, or both when the record supports that detail). Symptomatic failure — dyspnea, accessory-muscle use, tripoding, declining mental status from CO2 narcosis, or a need for ventilatory support — belongs in the story. A PaO2 of 59 in an asymptomatic patient being weaned on a quiet night is not a DIY diagnosis.
Finding Common indicator reading What it is not
PaO2 < 60 mm Hg (or equivalent SpO2) Hypoxemic failure indicator Not an ICD-10-CM code and not automatic acute respiratory failure
PaCO2 > 50 mm Hg with pH < 7.35 Acute hypercapnic failure indicator Not hypercapnic failure when pH is normal and bicarbonate is chronically high
PaCO2 chronically high, pH near normal, HCO3 high Chronic hypercapnic / chronic respiratory failure indicator Not acute failure from the PaCO2 number alone
New acidotic rise in PaCO2 above a known retainer baseline Acute-on-chronic indicator Not a license to skip the provider's diagnostic words
Intubated for airway protection, gases normal Airway protection Not automatically acute respiratory failure
Normal ABG ballast you should still know: pH about 7.35–7.45, PaCO2 about 35–45 mm Hg, PaO2 about 80–100 mm Hg on room air, bicarbonate about 22–26 mEq/L. SpO2 near 90 percent is a rough bedside cousin of PaO2 60 mm Hg, not a perfect conversion. The A-a gradient , FiO2, PEEP, high-flow nasal cannula settings, and BiPAP backup rates further describe support. None of them is a code.
Acute respiratory distress syndrome (ARDS) is a specific hypoxemic syndrome (Berlin-type timing, bilateral opacities, P/F ratio strata, non-cardiogenic context). It is not a synonym for every low PaO2. If the intensivist is treating ARDS, query for that diagnosis rather than translating a P/F ratio of 90 into ARDS yourself.
Acute, chronic, and acute on chronic
Acute respiratory failure is a new, life-threatening gas-exchange problem. Chronic respiratory failure is a baseline state: home oxygen, chronic hypercapnia with metabolic compensation, a long-term tracheostomy and ventilator, or provider language of chronic ventilatory failure. Acute on chronic is an acute worsening superimposed on that baseline — the COPD retainer whose PaCO2 lives at 58 mm Hg with pH 7.38 at home and who arrives at 74 mm Hg with pH 7.25.
Missing acute-on-chronic specificity is a Domain II classic. Coding only chronic understates the acute decompensation. Coding only acute in a known retainer discards the chronic disease the grouper and risk models also need. The query cites the baseline (old ABGs, home O2 liter flow, pulmonary clinic notes) and the new acidotic gas, work of breathing, and support (BiPAP, intubation), then offers acute hypoxemic, acute hypercapnic, acute on chronic (with hypoxemic, hypercapnic, or both), chronic only, other please specify, and unable to determine. Do not offer a type the gases contradict. Do not mention MCC, SOI, or mortality index.
Original scenario. Mr. P. has GOLD IV COPD, 2 L home oxygen, and a clinic ABG last month of pH 7.37, PaCO2 59, PaO2 62, HCO3 33. He presents with increased dyspnea. Admission ABG on 4 L is pH 7.26, PaCO2 73, PaO2 48, HCO3 32. Respiratory therapy starts BiPAP. The H&P says COPD flare, hypoxic. CDI should not assign acute hypercapnic respiratory failure from the gas, and should not settle for hypoxia as the complete neurologic-respiratory story. A nonleading query asks whether the provider is diagnosing acute, chronic, or acute-on-chronic respiratory failure and whether it is hypoxemic, hypercapnic, or both, citing the baseline clinic gas, the new acidosis, PaO2 48, and BiPAP.
ICD-10-CM Official Guidelines also address sequencing . Acute respiratory failure may be principal when, after study, it is the condition that occasioned the admission. It may be secondary when it is present but another condition occasioned the stay, or when it begins after admission. When respiratory failure and another acute condition (pneumonia, CVA, pulmonary edema) equally occasion the admission and no chapter-specific sequencing rule controls, the two-or-more-diagnoses guideline may apply. CDI helps the provider state the circumstances of admission. CDI does not pick the higher-weighted MS-DRG.
POA follows the inpatient-order timestamp. Respiratory failure that is already being treated in the ED when the inpatient order is written is POA = Y , even if intubation happens in the ED. Failure that clearly begins on hospital day 4 is N if the record supports that timing.
The provider must document the diagnosis
Respiratory therapists, ABG instruments, and rapid-response notes generate indicators . Credentialed providers generate diagnoses . Do not code acute respiratory failure from an RT protocol that says start BiPAP for hypercapnia. Do not code it from a nurse's comment of desatting. Do not code it from a PaO2 of 55 that nobody clinically treats as failure. If indicators are strong and the diagnostic statement is missing or vague (hypoxia, SOB, needs O2), query.
Intubation is not the diagnosis. A patient intubated for airway protection after overdose, angioedema, or a seizure may have normal gas exchange on modest FiO2. Assigning acute hypoxemic respiratory failure because a ventilator is in the room is a clinical-validation failure. Conversely, a patient on 100 percent high-flow nasal cannula with PaO2 49, accessory-muscle use, and an intensivist note of acute hypoxemic respiratory failure may never be intubated and still have a supported diagnosis. Support mode and diagnosis are related, not identical.
Yes/no queries may not introduce respiratory failure as a new diagnosis. A first-time capture is multiple-choice (or open-ended) with Other, please specify. A yes/no may be appropriate later to substantiate an already-documented diagnosis against an ABG, or to settle POA, if unable to determine is included where required.
ICD-10-PCS ventilation duration
Invasive mechanical ventilation is reported with ICD-10-PCS values that turn on duration of consecutive hours , not on calendar days and not on the Two-Midnight rule:
- Less than 24 consecutive hours
- 24–96 consecutive hours
- Greater than 96 consecutive hours
Count clock hours from documented initiation of mechanical ventilation until discontinuation or extubation, using the respiratory flowsheet as the usual source. Seventy-five consecutive hours is the 24–96 value, even if the calendar shows Monday through Thursday and three midnights elapsed. Ninety-seven hours is greater than 96, even if the team thinks of it as a four-day vent. Do not convert hospital days into hours by multiplying by 24 unless the record actually shows that many consecutive hours.
Routine intraoperative ventilation that is part of general anesthesia and stops when the case ends is not coded as a separate ventilation procedure. If the patient leaves the operating room still ventilated and ventilation continues in the ICU, postoperative mechanical ventilation is in scope; hospitals document a start time from the clinical record rather than from CDI arithmetic. Weaning trials while the patient remains intubated on the ventilator still count as time on mechanical ventilation. If the current fiscal-year ICD-10-PCS Official Guidelines address interruptions and restart, apply that published rule — do not invent a gap of hours from memory.
Noninvasive ventilation ( BiPAP , CPAP) and high-flow nasal cannula can be powerful indicators that acute respiratory failure is being treated. They are not the same PCS duration values as invasive mechanical ventilation. Dumping four days of BiPAP into the greater-than-96-hours invasive code is a procedure-coding error. A tracheostomy with prolonged mechanical ventilation can move the stay into a different MS-DRG family (including pre-MDC grouping). Teach the duration breakpoint; do not recite unpublished relative weights.
Original scenario. Intubation at 07:00 Monday for hypercapnic failure; extubation at 10:00 Thursday (75 consecutive hours). A laparoscopic cholecystectomy on Tuesday used general anesthesia while the patient was already on the ICU ventilator. CDI validates 24–96 consecutive hours from the RT record, does not add a second intraoperative vent procedure, and still needs a provider diagnosis of the respiratory failure that occasioned intubation. Duration and diagnosis are two different documentation problems.
Pharmacology as indicators, not codes: furosemide when flash pulmonary edema is driving hypoxemia; methylprednisolone and frequent albuterol/ipratropium in COPD; norepinephrine when shock complicates failure; propofol or fentanyl for vent synchrony (which may also depress the GCS and confuse a coma query); heparin when PE is the suspected driver. Each drug supports a story. None of them assigns J96.- by itself.