A PDS user allocates NHS numbers for a family of 7, that has not had previous NHS contact. All family members:
- are registered with the same GP practice
- live at the same address
- have the same surname
One month later, the same PDS user again allocates NHS numbers for the same 7 family members. The family registered with a new GP practice but still lived at the same address as previously quoted, but under a surname that has one letter missing from the previous surname. One of the patient registrations also has one letter missing from their forename, and a date of birth error.
This created 7 duplicate registrations on the PDS.
In 6 out of the 7 duplicate NHS numbers, the NHS numbers were allocated using the:
- surname with a one letter discrepancy
- exact same forename
- exact same date of birth
- exact same address
In one out of the 7 duplicate NHS numbers, the NHS numbers were allocated using the:
- surname with a one letter discrepancy
- forename with a one letter discrepancy
- same day and year of birth, but one month out
- exact same address
Seven family members each had duplicate PDS records for a period of 5 months, until the duplicate registrations were identified (one by the PDS Potential Duplicate Matching Algorithm, and 6 by an NBO Demographic Administrator), and were subsequently resolved by NBO.
The effect of a patient having duplicate NHS numbers is that one or more of the patient's clinical records may be incomplete, on which a clinician may make a clinical judgement or may have made a clinical judgement in the past. Duplicate NHS numbers are therefore clinically dangerous, and put patient safety at risk.
The PDS user's failure to validate patient data demonstrates bad data quality practice.