Involve physicians and coders in measuring and improving clinical outcomes
Innovative program helped slash complication rates
A creative and comprehensive project to bring physicians and coders together to better measure clinical outcomes has helped one Texas hospital slash its complication rate by dramatically improving the quality of physician documentation. But the changes leading to the turnaround weren’t easy, and the fix wasn’t quick. At the heart of the solution was the need to create an atmosphere of trust and understanding between physicians and coders.
Five years ago, Covenant Medical Center in Lubbock, TX, began a project to benchmark clinical outcomes data with peer facilities in Texas and the Southern region. Results from that project indicated that the medical center could improve length of stay, mortality rate, and cost and complication rates for at least some of the 25 top diagnosis-related groups (DRGs) and procedures identified as being part of the hospital’s strategic priorities. The problem was, when the data were presented to the medical staff, physicians questioned their accuracy — particularly the accuracy of the complication rates, which were based on coded data entered into the hospital’s data repository.
"Basically, our physicians had a tendency to blame the coders for data quality," says Janice R. Noller, RRA, CCS, CPHQ, quality improvement specialist in the quality management department at Covenant. Back then, Noller says, the coders had been sending occasional notes to the physicians requesting clarification on certain coding issues based on the physicians’ documentation in the medical record. "They met with resistance from the physicians. It may just have been a communications problem," she says. "Coders and physicians have a problem communicating anyway, no matter where you go. They speak different languages, and they have to reach a common language somewhere."
Reaching common ground was a difficult proposition, however, because the physicians blamed the hospital’s coders for inflating complication rates by "just picking everything up as complications," Noller says. "Evidently, there was not a whole lot of communication explaining to physicians when and why things are coded as complications. As a result, they felt that there were too many issues with the coding, the documentation, and the whole communication process. We felt there was definitely an opportunity there to investigate and see what was going on."
The dispute was troubling to the quality management department, because it wasn’t yet clear whether the hospital’s high complication rates were driven by coding or by physician documentation. Quality managers wanted to ensure that data used in clinical outcomes monitoring were accurate and consistent. To investigate the complication rates, they first sought to define the term and the codes to be monitored on an ongoing basis. They reached a consensus with the director of medical records and the coding supervisor to use the ICD-9-M code range 996.00-999.9 in calculating the facility’s complication rate. In June 1996, a certified coding specialist was added to the quality management department to help with the project.
DRGs and complications studied
Noller used the hospital’s decision support system to list every DRG for a six-month period and the number of complications in the 996.00-999.9 code range for each. (See table for sample data, p. 19, top.) Then she looked at the total number of cases for each complication regardless of the DRG. (See table for sample data, p. 19, bottom.) After comparatively analyzing the two lists, she determined that her first priorities should be DRG 358 (uterine and adnexa procedures for nonmalignancy with CC, including hysterectomies), DRG 148 (major small and large bowel procedures with CC), and code 997.4 (digestive system complications).
Quality management performed an extensive medical record review on all the cases in DRGs 358 and 148 with the secondary diagnosis code of 997.4. As a result of that review, Noller identified two trends:
• Accurate code assignment had been made due to physician documentation of the term "postoperative ileus."
• A physician documentation pattern in discharge summaries was noted: "Patient’s postoperative course was complicated by ileus."
Quality management staff sought to determine if the clinical treatment of patients coded with this complication differed from those not coded with the complication. In most cases, those patients did not require more sources, additional length of stay, or additional monitoring. Given those facts, Noller questioned why the physicians documented the ileus complication in their discharge summaries. The physicians replied that the coding technicians weren’t familiar with the clinical aspects of ileus, as well as other conditions that commonly occur postoperatively but aren’t necessarily complications, such as atelectasis, hemorrhage, hematoma, and fever.
"The physicians were very adamant that some of the things that the coders were coding as complications were actually clinically things that normally occur after an open abdominal procedure," Noller says. Meanwhile, "[the coders] wanted me to be able to go to the physicians on a regular basis and say, This is the data, this is why it appears this way, and this is what we’d like you to do to help the coders. Meanwhile, the coders will try to help you in understanding why they’re asking what they’re asking for.’ This was not an overnight process. It was probably anywhere from a year to a year and a half before I even got a few physicians to finally give me some positive responses."
One of the biggest obstacles to improving the communication process was the physicians’ idea that the coding of a complication meant they had done something wrong. Noller reports that it took almost a year to convince them otherwise.
To facilitate greater cooperation among the physicians and coders, Noller initiated a three-pronged action plan:
1. She started a coding newsletter in January 1997 to improve communication and educate the medical staff, their office staffs, and hospital staff on coding and documentation issues. Noller writes the newsletter, titled Quality Notes, which is edited by a physician champion. Currently, the newsletter has expanded its focus to include other health care issues, such as case management, fraud and abuse, and state and federal health care legislation. About 1,000 copies are distributed every other month.
2. Noller formed an ad hoc group of physicians to work with her and the coding supervisor to determine a set of basic clinical guidelines to assist the coding technicians in making decisions when it came to coding a condition as a complication. The group concentrated on the "complications" regarded as problematic by the physicians: ileus, hemorrhage, hematoma, atelectasis, and fever. The group was formed about the same time the newsletter was launched, six months after Noller came on board at Covenant. "I thought that a six-month period was pretty good, to get these docs willing to work with us instead of being antagonistic," she says.
3. A few months later, Noller helped to develop a coding subcommittee of the hospital’s resource steering committee. The resource steering committee handles information management for the entire facility, including medical records, coding, and data quality. "It was felt that because of all the coding and billing issues out there, a coding subcommittee would be useful in discussing these issues and working through them," Noller says.
Also, a process was formalized for referring coding discrepancies found on medical record reviews performed by the quality management department. (See flowchart illustrating the chart review process, p. 20.) Two databases — one to assist in communicating coding and documentation trends from quality management to medical records, the other to help quality management keep track of all individual record reviews and results — were also constructed.
The results so far have been dramatic. For example:
• Within six months, the incidence of code 997.4 in DRG 358 decreased from 12.6% to 3.03%. A further review discovered that physician practice patterns in documenting ileus as a complication had changed.
• As a result of the various education efforts directed toward physicians, doctors are now actually requesting presentations on coding guidelines.
• Noller reports that the quality management department now believes the data used in clinical outcomes monitoring and reporting are much more accurate and consistent.
"I think the physicians are starting to understand where the coders are coming from," Noller says. She credits the hospital’s medical staff leadership for helping to facilitate the change. "Our medical director of quality improvement has always been able to work with our other medical staff leaders to get the medical staff involved in quality improvement housewide."
Also helpful was a medical staff leadership group called the Clinical Outcomes Improvement Team, made up of physician section chiefs. "They were the first to hear some of this information," Noller says. "At the time, they were concerned from a monetary standpoint, because our hospital was undergoing talks of a merger with one of our main competitors. Administration was saying, Listen, guys, if documentation will help improve our financial picture, then that’s what we need to do.’ Like every other quality initiative, it has to start from the leadership on down. So, through that leadership role and our physician champions, and showing the physicians that it’s not just the hospital that’s affected but their office and business health as well, it finally sank in a little bit."
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content