Vol. 18 Issue 1
Making Each Patient One: Coping With Duplicate Patient Records
MPI coordinators offer tips and strategies on how to keep your MPI clean.
A patient is rushed into the emergency room (ER) unconscious and bloody. He had been hit by a car while riding his bicycle and the only form of identification on him is a library card. It's dirty and torn, but you manage to make out the name: Joe Pollock. You search the master patient index (MPI) database. There are five Pollocks but not a single Joe or Joseph. What you don't know is that Mr. Pollock is registered under his birth name William J. Pollock. What you do know is that you need to get him into the system ASAP so physicians can begin their care.
You create a new entry for Joseph Pollock, and, unless your hospital has a process in place to follow up and merge the records later, this ER stay could remain lost in the system forever. The next time William J. Pollock comes in complaining of stomach pains, his physician won't be able to see that he suffered internal bleeding while in the ER and therefore, can't completely assess his care.
An accurate MPI, whether in paper, electronic or somewhere in between, may be considered the most important resource in a health care facility, according to AHIMA Practice Brief Maintenance of Master Patient Index (MPI).
An MPI is only a beneficial resource, however, if it remains clean and free from duplicate medical records. When it's clean, each patient's complete medical informationfrom allergies, medication, disease and treatmentis all in one spot. But without processes in place to find and merge these duplicates, a patient's information could remain scattered, unknowingly, in two or more records.
Unfortunately, many HIM directors can't get duplicate medical records on their administration's radar screen. After all, duplicate medical records aren't a problem because they're around; they're a problem because they're hidden.
According to a recent poll in the Dec. 3 issue of ADVANCE "Does your office or facility have a process in place to help prevent duplicate medical record numbers?" half of the respondents either said no (17 percent) or they weren't sure (33 percent).
This is alarming because, without processes in place, duplicates often don't come to attention unless they pose a problem, or worse, negatively impact patient care.
That's the unfortunate reality that happened to Jan Baucom, RHIA, director of medical records at Gaston Memorial Hosptial, Gastonia, NC. In August 2003, an incident brought the issue of duplicate medical records to the forefront.
A patient was scheduled for surgery at Gaston, and when the physician requested the patient's medical record be brought to the operating room, he didn't like what he saw. The physician remembered having seen the patient on another day, and with only one record in front of him, knew this wasn't the patient's full medical story. He refused to do the surgery until the HIM department merged the records. The surgery was cancelled–all because of a dirty MPI.
"We had been telling the administration for a year and half that we needed to clean up, but I don't think people fully understood–'Why can't you just clean it up? Why do you need money to clean it up?'" Baucom recalled.
Identifying the Problems
After a number of meetings with the administration, they got to the root of the problems. Baucom's department used to discover duplicates by word of mouth, from systems in radiology and laboratory. They found three problems with this:
1.This gave them no way of knowing how many duplicates lay undetected in the system. It turned out the 220 duplicates Baucom's department investigated per month was a small percentage of the 2,100-plus duplicates actually being generated and the 20,000 that lay hidden in the system.
2.The process did nothing to prevent more duplicates from being created, which meant in the time it took Baucom's HIM department to merge one duplicate, three more could have been created in the same day. They were chasing their tails.
3.Without a system to identify duplicates before a problem arose, they ran the risk of negatively impacting patient care.
All of these problems can be solved with three main steps in the quest for a clean MPIand all need to be followed in this order, or you'll run yourself in circles.
Step 1: MPI Clean-Up
An MPI clean-up to purge your system of all undetected duplicates is essential. Without one, you'll be wasting time and energy shuffling around the same duplicate records over and over. A clean-up, however, requires manpower and time devoted solely to the task, either in the form of an MPI coordinator or vendor. "If an HIM department tries to deal with duplicates between their normal day-to-day activities they'll be treading water," said Robin Altendorf, MA, RHIA, QuadraMed MPI project manager.
When INOVA Health System, Falls Church, VA, hired QuadraMed to take care of its 155,000 duplicates, Altendorf hired and trained a temporary staff of 50 employees to work on-site around the clock for 11 monthsnow that's manpower. A vendor will also be able to bring in tools to auto-merge duplicates at an average of 2 to 4 minutes per record, compared to the 10 minutes it would take without the software.
When two facilities located just 2 miles apart, Bon Secours-Holy Family Hospital and Altoona Hospital, merged in 2005 to form Altoona (PA) Regional Health System, Diane Harris, RHIT, director of HIM, was faced not only with merging two medical record schemes into one, but also cleaning up duplicates created by two separate systems.
Add on that she was working in a paper environment, and an MPI clean-up becomes a lot more than just deleting and re-typing numbers. Harris decided it would be far too cumbersome to manually merge all of the paper records. Instead, Shelly King, RHIA, Siemens MPI consulting expert, created fields within Siemens' INVISION® system that listed all of the medical records numbers a patient ever had. This included every number, Harris saidfrom the patient's Bon Secours number, Bon Secours duplicate, Altoona number and Altoona duplicate, to the master Altoona Regional number. The HIM staff now pulls every record listed to get the patient's full history.
"It would be much too laborious to toggle between multiple databaseswe needed all the numbers in one system not only to clean up duplicates, but for MPI integrity as a whole," Harris said.
Siemens also allowed for an MPI purge to occur so that any record with no activity in the last 10 years was pulled and placed in an archive database that could still be accessed. The point of this, and an MPI clean-up in general, is to rid the system of any medical records that muddy the water when a registration clerk is trying to locate the right patient. "If you have more than 25 years of data, registrars have to leaf through inactive patients and it's harder for them to find the right match," Shelly King explained.
Step 2: Stop Them Before They Start
Now that you've rid your system of duplicates and are able to start fresh with a clean MPI, you'll need to implement prevention polices to make sure it stays that way.
"You can't have an MPI clean-up, wash your hands and say you're done," said Donna Coomes, MBA, RHIA, CPHQ, CCS, corporate director of medical records/HIM at Johnson City (TN) Medical Center. "If you don't pay attention to it in an ongoing effort, you're going to be back where you started with a dirty MPI."
She knows because she saw it happen. When her hospital was planning to move forward with an electronic medical record (EMR), Coomes knew they needed to clean their MPI, but kept getting hit with an administration "tired of writing checks" for MPI clean-up. "The CFOs said, 'this is the third time we've done this, why do we have to keep paying to get it cleaned up?'" Coomes recalled. The CFOs wouldn't move forward until they were assured plans would be put in place to keep it clean.
Duplicate records can be created in a number of waysmisspellings, typos, transpositions, nicknamesbut all come from the same place: registration. "You can't be successful with an MPI without being in good relations with your registration staff; everything they do directly affects us," said Amanda King, BS, RHIA, EMPI coordinator at Jackson-Madison County General Hospital, Jackson, TN.
Set clear policies for registration that include:
The complete demographic data that must be collected for each patient.
Exact standards on how to enter information (proper use of suffixes, hyphenated names, etc.)
All of the information the clerk must search for before creating a new medical record. (Social Security number first, then date of birth, etc.)
A list of potential questions to ask, including change of address/insurance, a recent marriage or nicknames. (Does the patient go by J. Lee, James or Lee?)
The registration staff should also be trained to conduct follow-ups with trauma patients and newborns, for instance, who are named things like Baby Roberts and Trauma Purple until they are given an official name or identified. If the registration clerk doesn't update this new information, the record could get lost in the system.
Step 3: Keeping it Clean
With preventative polices to help slow the creation of duplicates, the last step is to establish back-end polices to catch and clean up any duplicates that slip through.
Every MPI coordinator who spoke with ADVANCE agreed that the only way they've been successfully able to cope with duplicate medical recordsin a timely, cost-effective and preventative mannerhas been with the help of MPI software tools.
Prior to installing QuadraMed's suite of MPI software tools, Baucom's hospital saw 2,100 duplicates created each month; an 8 percent duplicate rate, far above the industry standard of 3 percent. For Baucom, the manual process of identifying duplicates on a case-by-case basis was such a time consuming and costly process, at 40 minutes and $8.96 per record, it became an "intolerable situation," she said.
Since installing the software tools, Gaston's duplicate rate has dropped to 0.49 percent and 1,000 duplicate records have been prevented each month–a savings of more than $100,000 per year in correction costs, Baucom said.
Front-end MPI software tools help ease the search process for registration clerks by showing the probability that the Bill Smith it found is the right one. Back-end tools can scan the MPI and generate a list of potential duplicates daily. "You will inevitably not be able to eliminate duplicate records completely, but with the tools in place you can be alerted daily and merge the records in a timely fashion," Harris said.
Just remember, an MPI clean-up has to happen before these tools are installed or they'll start locating thousands of duplicates and put you in over your head. "We had to start with a clean database or we'd be perpetuating the problem," Baucom said.
Every Day Focus
To keep the MPI a top priority in your hospital, it's important to assign someone the responsibility so the process doesn't fall to the wayside. When Shelly King leaves an MPI project, she leaves behind her skills with MPI coordinators. "Our goal is not to have to go back in a short time to clean up again," Shelly King said, "it's to give customers the tools and knowledge to keep it clean."
Coomes made the MPI the responsibility of the entire hospital; an MPI integrity team meets every other month. "It's part of keeping it on the radar screen," she said.
Accountability is key. Many MPI software tools have the capability to show which employees created duplicates and what mistakes were made. Harris' department works a duplicate list daily and registration is accountable for following up with those responsible. "All it takes is one employee who's not following the right steps to result in a dirty MPI," Shelly King explained.
Maintaining a clean MPI may seem like a huge undertaking, but the benefits of one are great; not to mention, the cost of doing nothing is even greater.
"There's a cost, yes, but there's also a cost of not doing anything at all and leaving your database unreliable for patient care," Baucom relayed. "I don't know that anybody could put dollars and cents on that."
Ainsley Maloney is an editorial assistant with ADVANCE.
Coping With Duplicate Patient Records In a RHIO
The RHIO Effect
As regional health information organizations (RHIOs) become a reality, the integrity of master patient index (MPI) data must become a top priority for facilities joining in, said Donna Coomes, MBA, RHIA, CPHQ, CCS, corporate director of medical records/HIM at Johnson City (TN) Medical Center.
"We scratch our heads every day when we ask other facilities 'Who works on your duplicates on a routine basis' and the answer is no one," Coomes said.
Her hospital is part of Mountain States Health Alliance, a 13-facility system that will soon be feeding into CareSpark, the not-for-profit RHIO spanning 17 counties in southwest Virginia and northeast Tennessee.
Coomes is well aware that for the RHIO to workfor it to successfully share patient information and be trusted by the physicians joining inthe data has to be clean and accurate or faith in the RHIO will diminish and its true value will be lost. What's the point of exchanging information if it's clogged with unusable data?
"We want to make sure the info we're sharing in the RHIO helps promote continuity of care, not adversely affect it by giving the wrong information," Coomes said. "MPI coordinators need to become more prevalent, or organizations run the risk of negatively impacting patient care."
CareSpark, which recently implemented an EMPI tool, Initiate Identity HubT Version 7.5, to link and match patient records across its providers, is currently operating in test mode, according to Deb Ingram, technical project manager at CareSpark.
In the meantime, Susan Torzewski, RHIA, MCSE, CareSpark's MPI administrator, is gearing up for an enormous undertaking-managing duplicates in a RHIO expected to take feeds from 1,200 physicians (945 outpatient, 255 inpatient) and 18 hospitals delivering care to 705,000 people in the region.
"I have done projects at different hospitals, but nowhere near the size and complexity of this project," Torzewski said. "With active feeds going pretty much every day; it's going to be a full-time job for me."
Rather than merging patient records, Torzewski will be providing linkages so that one patient's information, although spread out over multiple records, will be linked together.
Although CareSpark is not requiring that each hospital ensure a clean MPI before joining the RHIO, it is requiring that each hospital provide a minimum data set of name, date of birth and gender for each patient. "Of course we would strongly encourage each hospital to do their own clean-ups as they go," Torzewski explained. "We can tell them, for example, that we think they're sending duplicates in excess of 12 percent, which we expect will motivate them to work on improving and cleaning up their MPI."
What CareSpark will do is ask each entity to provide a list of medical records they've investigated but determined should never be merged. "There's no sense in CareSpark spinning our wheels trying to resolve something that was intentionally left unmatched 6 months ago," Ingram explained.
Ainsley Maloney is an editorial assistant with ADVANCE.