“Sully,” the Clint Eastwood-directed movie about US Airways Captain Chesley “Sully” Sullenberger, played by Tom Hanks, recounts the dramatic emergency landing Sullenberger made on the Hudson River after his airliner’s two engines were disabled when struck by a flock of geese.
While the newest commercial jets feature onboard flight-control computers, making them nearly autonomous, most passengers today would have second thoughts about boarding a plane without a human pilot. One could argue that an autopilot could not have landed US Airlines Flight 1549 as skillfully as did Sully on Jan. 15, 2009.
“When it comes to flying planes, humans are still better than computers at quickly assimilating unrelated facts and acting on them,” wrote Arnold Reiner in the March 2, 2016 edition of “The Atlantic.”
Could the same be true about computer-assisted coding (CAC)? Are certified coders and clinical documentation improvement specialists better than computers at quickly assimilating unrelated medical facts and acting on them?
In his article, “Towards the End of Pilots,” Reiner reports on how automation in the cockpit has contributed, in some cases, to a lack of attentiveness by the flight crew. Is it also possible, that the implementation of CAC is creating the same inattentiveness among coders?
“My observations point to decreased coder engagement with computer-assisted coding (CAC) as the root cause of increasing volume and the successes of third-party DRG overpayment audits,” wrote a senior appeals analyst for a major healthcare system. “Jane,” who is also a registered nurse, sent an email to ICD10monitor last week and spoke on the condition of anonymity because she was not authorized to speak on behalf of her facility.
“Computer-assisted coding was supposed to make coders more productive by scanning the desired parts of an encounter medical record for possible coding,” she said.
“Jane” told ICD10monitor that what she often sees is extra coding for a diagnosis that does not exist or is missing. She also said she sees incorrect procedure codes that do not clearly reflect the actual care and interventions ordered and completed by physicians. The missing procedure codes often would support instrumentation or implants chosen by the surgeon if present.
“While computer-assisted coding systems can be a valuable tool, HIM (heath information management) departments must find a balance between coder productivity requirements and CAC use,” said Sandra Brewton, a senior healthcare consultant. “The temptation for coders to just accept the suggested codes and move on to another claim can be too great, in the interests of meeting and exceeding productivity requirements.”
“What I find on a daily basis, is CAC, followed by a coding signature, with a minimum of two erroneous diagnosis or procedure codes on each and every claim I am asked to review,” Jane noted.
“If physicians really knew how their patients’ claims were being coded, I, for one, do not think they would be terribly pleased,” she added. “Physicians are being asked to learn more and more about ICD-10, CPT,® and HCPCS coding, yet I wonder how a physician/surgeon would react to the amount of coding done by computer algorithms?”
Jane noted that her job title might signal that her duties revolve around clinical appeals – but that’s not the case, she hastened to add. She said she spends more time reviewing claims for incorrect admission and discharge statuses and incorrect coding for both diagnosis and procedures.
Jane also said she sees diagnosis coding being scanned from informed consents, with possible complications of the planned procedure being coded on the claim.
“In one instance,” she wrote, “a patient claim was coded with a cardiac pacemaker in situ, with absolutely no such procedure in the patient’s history.”
Brewton said that this is a limitation of the CAC system in that that the software will “pick up” code-able terms from documents that are not appropriate to use for coding, thus reinforcing the need for certified coders to continue to carefully review documentation when coding claims.
“To some, the CAC system allows them to relax on documentation review, because the software has ‘reviewed the documentation for them,’’’ Brewton said. “In reality, especially with initial implementation of a CAC system, coders must review the documentation even more carefully due to the important responsibility they have to ‘teach’ that CAC system.”
Brewton, a certified coder who performs audits on behalf of Panacea Healthcare Solutions, also said that from what she has seen in her reviews, this is not what is happening currently.
“More often than not, CAC systems are implemented to help the bottom line, reduce the DNFB report, fill in gaps from coder shortages, and improve productivity,” Brewton said. “None of these goals address coding quality or accuracy.”
“Computer-assisted coding is only as good as where within the medical record, the scanning for diagnosis or procedure codes is implemented, and (it) should be based on how the electronic medical records (EMR) is constructed,” Jane wrote. “Coding quality review of CAC is essential.”
Jane also opined that patient satisfaction is heading south, noting that as more and more claims are being rejected by payers for coding errors, clean claims are being replaced by corrected claims at a much higher rate.
“I can see issues with both sides of the third-party overpayment auditing, overaggressive auditors, and careless, disengaged hospital coders,” she wrote. “But at the end of the day, these same overzealous auditors could also attest to the increasing coding errors caused by CAC that make those overpayment audits an easy slam dunk!”
Jane believes that outpatient and emergency room (ER) coding reviews are nonexistent, saying that the use of simple visit coding has come to be more mainstream.
In such cases, claims are not touched by a certified coder, but coded by algorithms – and then the bill is sent out the door.
“If facilities are going to practice this type of coding, regular audits must be performed by appropriately certified staff,” she wrote. “It would be best practice to have ER claims coded by certified coders.”
Jane also wrote that, having graduated from nursing school in 1971, she has seen what she described as the “metamorphosis of healthcare, both good and questionable.”
“The good news is that all the good came during my critical care clinical years, 32 to be exact, in the form of advancing standardization of care and the introduction of measurable clinical quality,” she said. “The bad news comes at the end of my nursing career. At the age of 63, I have but a few more years to absorb all the changes taking place on the managed care business side.”
Jane said she misses those days when a senior coder would spend the day in the surgical or cardiac intensive care, combing through charts and coding records a day or two at a time.
“Today it would take an army of experienced inpatient coders to code the number of medical records most hospital systems generate in one day – but when did coders become so indifferent to the value they bring to healthcare?” Jane asked.
Brewton believes that It is imperative that coders recognize their important role in HIM, an area she told ICD10monitor lies somewhere between clinical patient care and billing out the claims.
“To some coders, their responsibilities may be just a job,” Brewton said. “But in reality, it is a form of patient care, and incorrect coding can have many lasting effects – not only on facility revenue, but on the patients themselves.”
Jane said her comments about CAC should also be directed to healthcare leaders who make the decision to implement it as a method to increase coder productivity – but without the implementation of any monitoring of downstream effects.
“This is not the fault of CAC, nor does it diminish its value,” she wrote. “CAC is dependent on human engagement for what it needs to develop into what leaders think they are paying it to do: increasing coder productivity without lowering the quality of coding.”
Jane also expressed concern for physician and facility quality scores, noting that an overpayment auditor does not care whether the coding on the adjudicated claim is changed in the form of a corrected claim reflecting a correct DRG.
“They simply supply the A/R to the payers,” she said. “Those claims, with all sorts of diagnosis and procedure code errors, live on for other entities to use for data mining and for reporting. As a matter of fact, if the payer is a Medicare managed care plan, I would bet their preference is to leave the higher DRG just as it is.”
What also concerns Jane, she wrote, is the “slowly developing disengagement of coders to their actual value.”