The complicated process of reporting complications

 by James S. Kennedy, MD, CCS

Colleagues, as the Patient Protection and Affordable Care Act implementation progresses, the quality and outcomes of care face increasing public scrutiny. Policymakers will determine our complication rates and publicize them on the Internet, influencing public perception of our competencies and quality. Don’t believe me? Read in the Atlanta Journal-Constitution's article about some Georgia hospitals’ “high” pneumonia, heart failure, and myocardial infarction mortality (http://tinyurl.com/GA-mortality). See your own cost efficiency and quality profile on United Healthcare’s website (www.uhc.com) under the “Find a Physician” tab. Where is information to make these determinations obtained? From the ICD-9-CM codes assigned by our hospitals based upon documentation. Consequently, we have a vested interest in ensuring that complication codes are submitted accurately.

As we all know, certain conditions routinely occur in the postoperative period. AHA Coding Clinic for ICD-9-CM, third quarter 2009, p. 5, instructs coders to notrepresent these conditions as complications unless they are more than a routinely expected condition or occurrence, have a cause-and-effect relationship between the care provided and the condition, and are documented as a complication. If uncertain, coders must query the surgeon for clarification. Unless the answer is documented in the medical record, coders may be forced to code conditions as complications. This doesn’t reflect well on the surgeon or hospital.

At-risk complications and conditions
Risk adjustment involves ratios of actual to expected outcomes, so the sicker our patients are on paper, the better the outcome. Ask your quality or coding staff what factors determine expected outcomes.
Patient Safety Indicators (PSI) are influenced by the present-on-admission (POA) indicator. If the POA indicator is “U” (uncertain), this is counted as an “N” (no) and indicates that the condition was acquired as an inpatient under our watch. If it occurs during the outpatient or observation portion of the inpatient admission (before the order is written), the POA indicator is “Y” (yes) and does not count against the facility.
 
Medicare will use the RHQDAPU program quality measures to determine hospital (and possibly physician bundled) payments in fiscal year 2014. These include:

• Inpatient mortality only (risk-adjusted)
• AAA resection
• Hip fracture repair
• Composite mortality for selected medical conditions
• Patient safety
• PSI 06, Iatrogenic pneumothorax
• PSI 11, Postop respiratory failure
• PSI 12, Postop pulmonary embolism or DVT
• PSI 14, Postop wound dehiscence
• PSI 15, Accidental laceration
• A composite of these PSIs and others

Complete definitions are available at www.qualityindicators.ahrq.gov. 
 
Postoperative respiratory failure is a problem
The Agency for Healthcare Research and Quality (AHRQ) defines hospital-acquired respiratory failure (PSI 11) as the assignment of the following codes with a POA indicator of N or U:

• 518.81 (acute respiratory failure)
• 518.84 (acute and chronic respiratory failure) 

Fortunately, Coding Clinic requires all postop respiratory failure to be coded as 518.5 (pulmonary insufficiency following trauma and surgery), unless it is expected or integral to the surgery. While this addresses the AHRQ PSI issue, other rating companies (e.g., HealthGrades) consider code 518.5 a major complication.

What makes this dicey is that some intensivists routinely document “postop respiratory failure” in the immediate postop period (e.g., post-CABG) to justify medical necessity for CPT codes 94002–94003 for ventilator management.

The bottom line? Coders must know whether respiratory failure was POA or an expected or integral part of the postop recovery. Words like “expected postop respiratory failure” help; however, coders may still need additional clarification.
 
Iatrogenic pneumothorax
Code 512.1 (iatrogenic or postop pneumothorax) is a PSI unless POA or not coded at all because the pneumothorax was documented as integral to or expected with a procedure (e.g., chest or thoracic spine surgery). Pneumothoraces occurring in the outpatient or observation setting are POA and aren’t counted. Differentiate spontaneous, traumatic, and tuberculous pneumothoraces from those caused by treatment. 

Incidental lacerations in surgery
AHRQ has designated all accidental cuts, punctures, perforations, or hemorrhage during inpatient surgery, dialysis, injections, endoscopies, cardiac catheterizations, administration of enemas, and other procedures as patient safety issues. Coding Clinic requires coders to query in certain circumstances. These include:

• When a tear is documented in an op report (e.g., a small serosal tear of the stomach), coders should query the surgeon on whether the small tear was an incidental occurrence inherent in the surgical  procedure or whether the tear should be considered a complication of the procedure. Advice states that coders should report a full-thickness enterotomy during a difficult lysis of adhesions or an accidental suture of the bladder as an accidental laceration.
• When a localized arterial dissection occurs during angioplasty, the coder may ask whether it is clinically significant and thus should be reported. 

AHRQ designates pressure sores, pulmonary emboli, deep vein thrombosis, hemorrhages and hematomas, diabetic hyperosmolarity, and sepsis that are not POA as patient safety “never events.” This should remind us to define and document these conditions when patients are admitted, and to promptly respond to clarifications as to whether they are POA.

Don’t forget that a Stage 1 pressure sore manifests as nonblanchable erythema, and if it is POA and then progresses, it remains POA. And remember that many toxic-appearing infected patients requiring surgery may meet the definition of sepsis at the time of their admission. Read the most recent list of clinical indicators for sepsis at http://tinyurl.com/2008Sepsis.

Finally, because risk-adjusted mortality for hip fracture and abdominal aortic aneurysm is listed, ask what indicators predict risk in these populations and include them in your history and physical. 

Editor’s note: Dr. Kennedy is a general internist and managing director with Atlanta-based FTI Healthcare, which specializes in medical management, case management, clinical documentation, and quality reporting. Contact him at 615/479-7021 or james.kennedy@ftihealthcare.com. This article was initially published in the December edition of Medical Records Briefing, and was discussed during the November 18 ACDIS Member Quarterly Conference call.

Order the downloadable audio on-demand presentation Surgical Complications: Clinical Documentation Improvement for Compliant Coding and Accurate Quality Measures or call our customer service department at 800-650-6787 or e-mail customerservice@hcpro.com.