Case study: mpi software cleans up and prevents duplicate medical

We're the ideal place for homework help. If you are looking for affordable, custom-written, high-quality and non-plagiarized papers, your student life just became easier with us. Click either of the buttons below to place your order.

Order a Similar Paper Order a Different Paper
  1. Compare a contrast a duplicate, overlap, and overlay.
  2. Why is it important to avoid MPI errors? Illustrate with an example. 
  3. With the manual MPI cleaning method, it took 1 year to identify 60,000 duplicates. With the MPI cleanup software, it took 4 months to identify and fix 78,000 duplicates, because of the advanced duplication identification methods: phonetic research, deterministic search, and probabilistic algorithms. Explain these methods in your own words and use examples if that makes it easier to explain. (DO NOT quote the definitions provided. If you do, you will earn no points for this question).
  4. If you were a member of the “registration team” that was formed, what would be your recommendations for preventing the new duplicate numbers up-front?
  5. What knowledge and skills would you need to work in the area of MPI cleanup without the specialized software and with specialized software? After reviewing the HIM curriculum Links to an external site.( and checking out some of the course descriptions available, in which courses do you believe you may acquire some of that knowledge and skills? What other resources would you use to equip yourself for that type of job fully?  

Box 5-3

C as e Stud y : MP I S o f tw ar e C le a ns Up and P re v ents Du pl ica te

M ed ica l Rec o rd Nu mb e rs

Box 5-3 Case Study: MPI Software Cleans Up and Prevents Duplicate Medical Record Numbers

This case study concerns how two hospitals with a combined total of nearly 40,000 inpatient

discharges and 185,000 outpatient visits a year with different information technology and different

data fields came together beginning with a corporate person index (CPI). The challenge was to

implement a common CPI for two hospitals while performing a data integrity improvement of each of

the existing master patient indexes (MPIs). The target date for completion was 2 months.

The health information department was assigned responsibility for identifying “duplicate” patients in

the CPI between the two facilities and updating the respective separate MPIs to reflect the most

current information. The process began with a traditional deterministic search method to identify

duplicates within each facility’s MPI.

The following elements were researched in an exact deterministic match on all elements:

• Last name, first name
• Social Security number
• Date of birth

The problem with this method is that it is not possible to detect duplicates caused by errors in name

spelling or number transposition. Also, there is no easy way to cross-reference merged duplicate


After a year of manual clean-up effort and more than $60,000 in manpower, time had run out. When

the two hospitals’ MPIs were loaded into the CPI, another 78,000 duplicates occurred within the CPI.

An immediate and cost-effective solution had to be found. At that point the director of patient

financial services recommended using a computerized process that searched the MPI files by using

algorithms. Because time was of the essence, a product and a vendor who had an immediate

solution had to be selected. A vendor was selected to install electronic clean-up software.

The vendor accomplished the installation and had the software operating in less than 3 months. The

software was able to identify duplicates by using the following:

• Phonetic search
• Deterministic search
• Probabilistic algorithms that rank the information from 100 to 1 as a probable duplicate
record number

In addition, the software used name normalization, address normalization, and a stoplight method of

highlighting potential duplicate data fields, thus expediting the process of duplicate number decisions

by the clerical clean-up team.

At the time the two hospitals decided to embark on this partnership, there was a multidisciplinary

conversion team formed that included all ancillary departments, patient financial services, the

internal audit manager, computer services, and medical records. The high volume of duplicates

affected the ancillary departments. In addition, as the clean-up was occurring, the health information

department became frustrated that duplicates continued to occur as a result of new duplicates being

generated every day during registration.

In addition to the conversion team, a registration team was formed. This team worked with the

director of health information management (HIM) and the health information support manager

responsible for making recommendations about decision rules for the clean-up process. The

decision rules regarding which medical record number would “survive” as the permanent number

was a collaborative effort among the vendor experts, computer services, and the health information

professionals. The registration team worked on data element standardization between the two

facilities so the data element sex, for example, was F for female in both sites and M for male versus

using b for boy and g for girl. The MPI team started with the highest probabilistic matches because

the team was simultaneously cleaning the MPI and the CPI. However, as the clean-up project

continued, it was apparent that new duplicates continued to be created every day. The director of the

health information department felt that the project value was mitigated without “up-front” (at the time

of registration) prevention of new duplicate numbers. This viewpoint was strongly supported by the

ancillary departments. The blood bank, for example, was very supportive of duplicate prevention

software. As a result of the team support, the hospital elected to add up-front duplicate prevention


Unfortunately, the problem of duplicate medical record and guarantor numbers in health care

provider settings is not unique. What makes this case study interesting is that the electronic solution

of the 78,000 CPI duplicates was completed in less than 4 months. Today the MPI remains very

clean with a computer “refresh” run every 6 months. The computerized refresh presents only

duplicate numbers that have been issued since the last refresh and does not present duplicates that

have been rejected as “not a duplicate.”


1. Corporate person index (CPI): A person index that contains person names and medical record

numbers for multiple facilities. Person names may be the patient or guarantor of the account

2. Cross-reference: Archive of numbers file that, when created electronically, allows all numbers that

pertain to the same patient to be filed together in the cross-reference

3. Deterministic: Able to establish an association between two files by searching for an exact match

against a set of given criteria (e.g., medical record number) (AHIMA Practice Brief: Fundamentals of

Building a MPI/Enterprise MPI, 2010 available at

4. False positive: If a record is reported as a duplicate but on review is determined not to be a

duplicate and is rejected, careful attention has to be given as to where in the 100 to 1 probabilistic

scoring the majority of these occur

5. Merge: Once the patient number that is to survive is selected, the computer electronically merges

data from nonsurvivor (retiree) encounters together with the survivor number, therefore providing

continuity of patient information

6. Name normalization: A process that links common names such as William, Billy, and Bill together

under one name

7. Pairing rule: A data element field or fields used as a key by the scoring algorithm for identifying

potential duplicates

8. Phonetic file: A system to file and find names without regard to vowels; used to locate names that

sound alike but are not spelled the same way

9. Probabilistic: Capability or linking of data elements that compensate for discrepancies by using all

the information recorded in two files and attach a weighted value for matching discrepancies (AHIMA

Practice Brief, November 1997)

10. Refresh: Computer capability to locate all new person or patient numbers issued since the last

update, excluding those designated as “not a duplicate” during the earlier number clean-up efforts; to

“refresh” the file is to bring out all legitimate duplicates created since the last time the process was


11. Retiree: The duplicate patient number that is determined to be sent to an archive cross-reference

list and not to be used for further patient care

12. Scoring: Weighted or probabilistic algorithm for quantifying the relative validity of a potential


13. Street normalization: A table that brings “Berkshire St.” and “Berkshire Street” together

14. Survivor: The duplicate patient number that is selected to remain in the active file and that will be

used in new encounters

(Courtesy MediBase and Mary Mike Pavoni, RHIA, FAHIMA.)

Do you need academic writing help? Our quality writers are here 24/7, every day of the year, ready to support you! Instantly chat with a customer support representative in the chat on the bottom right corner, send us a WhatsApp message or click either of the buttons below to submit your paper instructions to the writing team.

Order a Similar Paper Order a Different Paper