Health Information Management

Avoid the push and pull when reporting complications of surgery

JustCoding News: Inpatient, September 30, 2009

Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

Reporting complications of surgery is often likened to a tug of war between hospitals and physicians. Hospitals need to be fairly compensated for treating patients who suffer unexpected events during surgery because treating these patients requires more resources and longer hospitalizations. But physician profiles and quality scores on Web sites such as Healthgrades.com suffer from over-reporting of these diagnoses. The push and pull of these opposing forces makes life difficult for the average coder trying to perform his or her job.

“We teach hospitals to look at the documentation and the patient’s severity, but then when you get to the physician’s side, these 900 [complication] codes may have negative ramifications in reference to their profiling,” says Mario A. Perez, III, RHIA, CCS, CCS-P, director of coding and documentation compliance for J.A. Thomas & Associates in Atlanta, GA.

Coding vs. clinical conundrums
From a coding perspective, reporting complications is relatively straightforward. According to the ICD-9-CM Official Guidelines for Coding and Reporting, coders must report the complication first. For example, when a physician documents that a patient has a postoperative ileus following bowel surgery, coders should report postoperative complication code 997.4 followed by code 560.1—provided that the ileus did not exist pre-operatively and was not an integral part of the surgery, says Robert S. Gold, MD, CEO of DCBA, Inc., in Atlanta, GA. “If there’s any doubt, clarify with the physician,” he says.

When a physician documents that the ileus was a normal physiologic response to handling the intestine (i.e., the physician documents ‘expected ileus secondary to bowel surgery’)—and the ileus did not prolong the hospital stay—coders should only report code 560.1. On the other hand, if the patient developed abdominal distension and vomiting, and a nasogastric tune was reinserted, then the ‘postoperative ileus’ went beyond the normal physiologic expectations, and it would be appropriate to assign code 997.4.

However, from a clinical perspective, the issue of what defines a complication versus an expected outcome of surgery is often unclear. Following are some examples.

  • Hemorrhage complicating a procedure (code 998.11) and acute blood loss anemia (code 285.1). For a patient admitted with a long bone fracture, there will be considerable loss of blood into the thigh as a result of the fracture. This blood loss happens regardless of whether the patient goes to the operating room (OR). The hematocrit may also drop low enough to meet the definition of anemia, notes Gold.

    Clinical documentation improvement (CDI) specialists should work with physicians to determine whether the complication is a true postoperative complication or whether it is the result of an expected event of the disease rather than of the surgery. When the patient’s anemia is the result of a long bone fracture, and the patient is treated for it, it is appropriate to assign code 285.1.

    When the documentation is unclear, CDI specialists should ask physicians the following question: “Do you consider this anemia due to blood loss from the long bone fracture, or was it a complication from the surgery?” If the former is true, it is correct to report code 285.1. If the latter is true, it may also be appropriate to report code 998.11.

    In contrast, when a true hemorrhage occurs—an unexpected laceration of a major artery in the pelvis, for example— it is appropriate to assign code 998.11.

    “If the patient has a ruptured aortic aneurism or an operative procedure for a major trauma, 285.1 would be expected, and 998.11 is likely inappropriate,” Gold says.

    CDI specialists should consider queries for complications of surgery due to unexpected events, such as an unexpected drop in hemoglobin. “If you see an operation that’s not associated with blood loss, for a disease problem that’s not associated with blood loss, and suddenly you see the patient’s hemoglobin drop in the chart, there’s likely a problem,” Gold says.
  • Postoperative respiratory insufficiency (code 518.5). This is another diagnosis that presents coding problems, Gold says. Some intensivists identify conditions as acute, critical care conditions when the patient is simply being reversed from anesthesia, resulting in inappropriate assignment of code 518.5, Gold says. “It’s inappropriate to ask the [physician] to document these things, and a lot of documentation improvement consultants train their students to do this,” he says. It results in a [major complication/comorbidity] for the hospital but results in a massive ‘ding’ for the surgeon.”

    In addition, many physicians elect to document “acute respiratory failure” in the above circumstance, leading to an inappropriate assignment of code 518.81 (acute respiratory failure). “Either one—518.5 and 518.81—is inappropriate to report for mere reversal from anesthesia,” Gold says.

    Gold says a good rule of thumb for postoperative respiratory insufficiency is to avoid querying physicians for the condition in planned instances of recovery, or as components of the surgery. It would be inappropriate to report code 518.5 for the following scenarios:
    • A patient has his or her wound packed open and is placed on a ventilator. The patient is sent back to the OR the next day to complete a planned staged procedure
    • A patient who undergoes a coronary artery bypass graft (CABG) or valve replacement comes out of the OR, and it takes three to four hours to get the patient off a ventilator while he or she is being reversed from anesthesia

In addition, many secondary diagnoses are components of a larger disease, so coders should not report them using complication codes. For example, a patient with a ruptured bowel generally presents with a tight abdomen, absent bowel sounds, and an elevated white count. Frequently the patient has ileus and sepsis—and when he or she comes out of the surgery, he or she still has ileus and sepsis and remains on antibiotics. “The question is, was it the surgery that caused the ileus, or the ruptured bowel?” Gold says. If it’s the latter, then coders should not report the ileus as a complication. Instead, coders should report the ileus as, present on admission (POA) or clarify its POA status with the physician.

In addition, all surgeries require some degree of routine postoperative treatment. For example, patients who undergo abdominal surgery typically don’t resume normal functioning of their large intestine until three days after the surgery. “When the surgeon writes on day one, ‘postoperative ileus,’ a coder may assign it to get the [complication/comorbidity], and the surgeons will get inappropriately [penalized] for it,” Gold says. Why is a complication code inappropriate in this instance? Paralytic ileus occurs in all patients who have major bowel surgery, he says, and all physicians treat it the same way.

“There is no additional utilization of resources, or observation, or anything different from the usual major bowel case,” says Gold. “It should not be reported at all. It does not meet [Uniform Hospital Discharge Data Set] criteria as a valid secondary diagnosis.”

Another classic example is patients who, after surgery for a coronary artery bypass graft, receive Dopamine to maintain blood flow through the kidneys following the operation. “Some people will look for postoperative shock in order to get a major CC—but everyone gets the Dopamine,” Gold says. “Don’t look for stuff like that just to get a major CC—it hurts the surgeon, and it’s not a valid code.”

Non-reporting runs risk of noncompliance
To avoid conflicts with physicians or sidestep tricky clinical questions such as those referenced in this article, some hospitals elect to sidestep the issue by not reporting complication codes. But Perez says this decision is non-compliant with coding regulations. When hospitals elect not to report complication codes, it reflects poorly on the facilities’ length of stay because patients with complications from surgery often require more resources and additional days in the hospital.

“Besides, the hospital now stands out as the only one that reports no complications—and that’s enough to lead to an investigation,” Gold says.

For example, a physician performs a procedure in a patient’s abdomen and inadvertently nicks the patient’s bladder, requiring surgical repair. When there are documented severe adhesions and considerable difficulty and time spent by the physician in lysing the adhesions, the entry into bowel or bladder is unavoidable. However, when there is no such documented difficulty, coders should code the event, because it can be construed that the hospital is trying to conceal the error, Perez says. Also, if the patient requires a longer length of stay, coding the complication provides evidence for the use of extra hospital and physician resources.

Assigning complication codes for a foreign body (e.g., a sponge) inadvertently left in a patient’s abdomen is another difficult matter that coders should report only under appropriate circumstances. When the count is incorrect, and the surgeon reopens a few layers and retrieves the sponge before the conclusion of the procedure, there was no foreign body inadvertently left in, Gold says. Therefore, reporting a complication code is inappropriate. However, if the case required a return to the OR from the recovery room, re-prepping, draping, and re-anesthetization, it constitutes a reportable complication, he says.

“Physicians need to be made aware that the identification of conditions in the postoperative phase justifies the patient’s severity, length of stay, and the resources,” Gold says. “But they are not always complications, nor are they all counted as complications. They are conditions the physician treats.”

In addition, in conjunction with a complication, a coder must also report an appropriate E code (i.e., external causes or effects), which are used for statistical reporting for risk management. “Facilities have been fined for not reporting these codes,” Perez says. “If you’re not doing that, you’re not providing good data quality from a statistical perspective.”

How ICD-9-CM codes affect HealthGrades, AHRQ scores
Complication codes affect physician profiles and hospital report cards on publicly available Web sites such as HealthGrades and the Agency for Healthcare Research and Quality. You can find a complete list of the codes used to determine the HealthGrades complication lists on the HealthGrades Web site.

Because so many don’t know about this list, confusion and miscommunication in many CDI and HIM/coding departments are rampant, Perez says. “Facilities aren’t getting these codes reported for fear that the physicians are angry, but you’re doing a disservice to your data and particularly for severity aggregate data that provides justification for length of stay and the cost of utilization of resources,” he says.

For example, after some recent changes in HealthGrades’ algorithms, code 285.1 is not one of the codes the Web site uses to determine complication rates, whereas code 998.11 is—a common misconception. “People need to be aware of and do due diligence on the list, if they’re being told what is and isn’t a complication code,” Perez says.

Another problem with code 285.1 is that the index of ICD-9-CM Manual defines it as Anemia, blood loss acute. But the tabular list defines it as ‘post-hemorrhagic anemia’—terminology that puts physicians on the defensive.

“[Post-hemorrhagic anemia] is what gets reported in the coding summary sheets, and the descriptor implies to the physicians that it’s a serious complication of surgery,” Perez says.
In many instances, instead of reporting a diagnosis as a complication of surgery, CDI specialists should work with physicians to ensure that the condition is properly documented as present on admission when the evidence exists that it, indeed, was present on admission.


Editor’s note: This story was originally published on the Association for Clinical Documentation Improvement Specialists Web site.



Want to receive articles like this one in your inbox? Subscribe to JustCoding News: Inpatient!

Most Popular