Quality manager offers tips to eliminate OASIS errors

Your guide to improving data quality oversight

As time draws closer to the implementation of home care’s prospective payment system (PPS), agencies will need to pay more attention to their Outcome and Assessment Information Set (OASIS) data collection and data quality oversight.

Tiny errors will cost agencies money, and those mistakes add up. At the VNA Health at Home of Watertown, CT, quality managers noticed a recurring problem of one or two discrepancies with a number or letter over several months. As minor as the error was, it could have easily added up to a double-digit error rate on certain OASIS items. So the agency implemented an improved data quality oversight program.

"Data quality oversight is difficult but it’s something that’s worthwhile, and especially with PPS it’s very important," says Sara Szafranski, RN, quality improvement manager of the agency, which serves 18 towns in western Connecticut.

Conduct data entry audits

State Medicare reviewers will expect agencies to have a quality oversight program in place, she adds. Here’s how the Connecticut agency designed its OASIS data oversight program to capture all errors:

Pay close attention to data entry audits.

VNA Health at Home conducts one of these audits each month, comparing scanned data input with paper documentation.

"We feed five data sets into the scanner and compare what the scanner has read into the software with what it says on paper to check the accuracy of the scanning software and the scanner itself," Szafranski says.

Agencies that have data clerks type in the information would have to do a similar comparison. However, agencies that use hand-held computers in which staff put in the data will have to handle this audit a little differently. Instead of comparing the input with a printed version, quality managers will need to do spot audits to check for numbers that are out of place or similar errors.

90% accuracy wasn’t good enough

The scanner used by VNA Health at Home initially had a 90% accuracy rate, which caused too many errors, so the agency fine-tuned the software to bring the accuracy rate to 95%, Szafranski says. "This means it’s scrutinizing little lines that make up handwriting in a box, and it asks for more verification from the data entry person," she adds. "So in that respect, it slows down the process, but it will reduce your errors, and we haven’t had a problem with it since then."

Now the scanner software is programmed to stop and request confirmation before inputting the data. It won’t recognize data that is outside of OASIS parameters. This way if the scanner reads a zero as a six at the start of care, the software will reject the data.

Audit clinical records. Each month, the agency reviews a sample of records and discharges and compares them with the start-of-care OASIS. "We compare the start-of-care OASIS with information gathered on the intake with regards to patients’ admission functional status, checking this against any therapy evaluations that were done," Szafranski says.

Then the same process is followed for discharge documentation, comparing OASIS data with nursing discharge notes. "If a patient has a wound, and the nurse is documenting that the wound is healed, then on the OASIS form it should show up that the wound is healed," Szafranski says.

Medicare reviewers will want to make sure that agencies haven’t beefed up patients’ diagnoses to make them look sicker than they are for the purpose of increasing reimbursement, she adds.

Check staff’s OASIS assessment skills. Each quarter, the agency’s supervisors or peer reviewers make home visits with clinicians to determine the accuracy of the interview process in collecting data, Szafranski explains. "Both the peer reviewer and the clinician complete an OASIS tool, and I look at the two and compare them as to accuracy," she adds. "They don’t talk to each other about it, and they arrive at the same time and leave at the same time so they will get the same information."

Szafranski has found that one of the biggest problem areas is in identifying the primary caregiver. "I found a fair number of discrepancies there," she says.

Staff had trouble describing caregiver

For example, on the caregiver question, the OASIS tool wants to know whether the primary caregiver is taking the lead and responsibility in providing and managing a patient’s care.

"Our staff have had a little discrepancy in describing who this is," Szafranski says. "One might say this is paid help, and another might say this is the daughter, a friend, or a neighbor."

Patients often say that everyone helps out, so they could be of little help in determining the primary caregiver.

Staff sometimes have disparities in listing diagnoses. Nurses are instructed to discuss any primary care diagnoses that are questionable because there often are secondary diagnoses that are more important to the home care plan of care. If a patient has chronic lung disease and diabetes, the agency might have been referred the case to stabilize the patient after a lung disease exacerbation. But when the nurse visits the home she may discover that the patient hasn’t been taking diabetes medications and has difficulty managing that disease.

"Diabetes becomes the primary diagnosis in home care, whereas the lung disease was the primary diagnosis in the hospital," Szafranski says. "We’re justified in doing this as long as we can show that the person’s bulk of need for skilled home care services relates to the diabetes diagnosis."

Also, some nurses may list more secondary diagnoses than do others. This sometimes is due to different observational skills. For instance, a patient’s paperwork may indicate that the patient doesn’t have seizures, but when the nurse visits the home, she might see new seizure medications.

The joint clinical audits have highlighted problems in how staff interpret the OASIS questions. The agency held counseling sessions with clinicians on a one-to-one basis to ask them to explain certain answers. For example, one nurse was not correctly interpreting the question about intractable pain. The nurse was going by her mental knowledge of intractable pain rather than reading the OASIS question and going by what that said.

"To her it meant you had pain that could not be relieved, such as appendicitis before a person goes to the emergency room," Szafranski says. "But the definition on the OASIS form is that it’s pain that’s not easily relieved and limits the patient’s ability and desire to perform physical activity."

The two interpretations are very different and could result in different payments from Medicare under PPS, she adds.

Szafranski says the agency’s goal is to provide those joint supervisor/clinician visits at least once a year for each field nurse and therapist. The agency will continuously educate staff as OASIS documentation issues arise, she adds.

"We’ll go over the individual OASIS indicators, helping staff assess these and learn more about how OASIS is designed and how to interpret the question so that you get the most accurate information," Szafranski says.

Sara Szafranski, RN, Quality Improvement Manager, VNA Health at Home Inc., 27 Princeton Road, Suite 101, Watertown, CT 06795. Telephone: (860) 274-7531.