Sapior receives PIAG commendation for data privacy technology

June 28th, 2010 by admin

30 March 2009 – De-identification specialist Sapior Ltd announces it has received a commendation from the Patient Information Advisory Group (PIAG) (now National Information Governance Board – NIGB).

Sapior is working with St. George’s University of London to develop a groundbreaking privacy-enhanced data collection process called SAPReL (Secure And Private Record Linkage).

St. George’s and Sapior made a joint application for support under section 251 to link routinely collected primary care data, data from the IAPT programme in the demonstration sites, and the secondary uses service. PIAG has approved this access to patient-identifiable data for the purpose of pseudonymising or de-identifying it for use in a cohort study.

In its provisional approval letter, the Advisory Group described the application to be “an exemplar”.

“Members noted the use of technology in the pseudonymisation process to be highly commendable and an example of best practice in order to achieve the pseudonymisation process.” (Source: PIAG Meeting Minutes, Monday 8th December 2008, pg. 10)

“Our flexible data harvesting solution removes identifiable information at the data source in a consistent way to allow the vital ability to link with other sources,” says Robert Navarro, Managing Director of Sapior Ltd. “This means researchers can develop a more comprehensive view of an individual’s record whilst still protecting the sensitive data during use.”

Primary care data users optimistic despite breaches: NHS Alliance Mind the Gap quiz results released

June 28th, 2010 by admin

27 October 2008 – At a time when reports of data breaches are regular occurrences and surveys on working practices point to an NHS data protection time bomb, NHS Alliance delegates ranked their organisations highly on upholding the Confidentiality policy with regard to using patient data for secondary purposes.

Sapior asked delegates at the recent NHS Alliance 2008 conference to score their organisations’ enforcement of the Confidentiality policy in terms of using de-identified patient data for commissioning, performance management, clinical audit and other secondary purposes.

Most respondents scored their organisations very well with over a quarter scoring 5 out of a possible 6. 36% of organisations scored in the middle of the spectrum (3 or 4 out of 6).

However, remarks from respondents indicated that many, including those in key management roles, were simply guessing on their organisations data privacy practices. “We must be here (5 of 6) because data privacy is important,” said one non-exec director.

Not surprisingly, there were also mixed responses from within organisations. One Chief Executive scored the PCT at an impressive 5 out of 6, whilst a Finance Manager from the same PCT scored it at 2, saying “Don’t tell my CE. We have a lot to do in this area.”

Whilst not an accurate measurement tool, the goal of the quiz was to encourage delegates to give mindspace to the high profile issue of data privacy and security. In particular, to the ongoing practice of using identifiable patient data for secondary purposes which conflicts with Confidentiality policy.

About the quiz

Respondents were asked to consider six areas where patient data is used for secondary purposes. (See below) Organisations were given credit for each secondary use area the respondent believed it was using de-identified patient data. For example, if an organisation was using any de-identified data for commissioning work and public health, the respondent could claim a score of 2 out of 6.

Clearly, scoring for the Mind the Gap quiz is very generous and not an accurate measurement. For example, an organisation using a single set of de-identified data for clinical audit would receive full credit for that area even if it was also using many identifiable data sets for other clinical audit work.

Notably, not a single delegate responded that it was “Not my responsibility”. Several were comfortable with guessing on their responses, although 15% said they didn’t know for sure whether de-identified data was being used or not.

Areas of Secondary use:

  • Checking quality of care (e.g. clinical audit)
  • Managing NHS spending (e.g. PbR, PBC, QMAS)
  • Managing health service (e.g. commissioning)
  • Investigating healthcare concerns/complaints
  • Protecting public health
  • Supporting research
Breakdown of Scoring
# of secondary use areas using de-identified patient data # of respondents Percentage*
Don’t know 6 15%
1 1 3%
2 3 8%
3 7 18%
4 7 18%
5 10 26%
6 5 13%
Mean score: 3.5
Median score: 4

*Total percentage greater than 100 due to rounding.

Pseudonymised data not subject to Data Protection says EC

June 28th, 2010 by admin

23 July 2007

Financial justification for pseudonymisation to mitigate privacy breaches

The European Commission on Data Protection (DP) has defined Pseudonymised data as non-personal data and not subject to the Data Protection Directive in certain instances. This new position effectively permits organisations to meet DP responsibilities more economically when sharing data with partners, as in the case of cross-border data flows which have previously required expensive, complex, and time-consuming contracts.

In its “Opinion 4/2007 on the concept of personal data”, the European Commission Article 29 Data Protection Working Party (WP) clarified the notion of “personal data” thus enhancing legal certainty through the uniform interpretation of the EC Directive 95/46/EC. Adopted on 20th June 2007, the document describes the following conditions necessary to consider Pseudonymised data as non-personal data and thus not subject to the Directive:

  • the Data Controller pseudonymises or key-codes Personally Identifiable Information (PII) to be given to a Data Processor that does not receive the key
  • the goal of the processing must not be to identify individuals and influence or treat them differently from others.

In addition, the WP clarified its position on “retraceably pseudonymised” data which may be considered indirectly identifiable and thus subject to the Directive. If the linking to the individual is done by the Data Controller only under predefined circumstances, the risks to the individuals are considered to be low. In these cases, the WP claims the application of the Directive will be more flexible than if information on directly identifiable individuals were processed.

The Opinions of the WP can be found at:

http://ec.europa.eu/justice_home/fsj/privacy/workinggroup/wpdocs/2007_en.htm

“This enhanced legal certainty finally provides an economic justification for those organisations who have delayed implementing pseudonymisation in their privacy breach mitigation strategies,” said Robert Navarro, Managing Director of Sapior Ltd.

About Pseudonymisation

Pseudonymisation, a Privacy Enhancing Technology (PET), is essentially the replacement of Personally Identifiable Information (PII) – such as name, address or account number – with pseudonyms. Key-coded data are a classical example of pseudonymisation. Personally Identifiable Information (PII) is earmarked by codes, while the link between the code and the PII (like name, date of birth, address, etc.) is kept separately. Pseudonymised data can be used for audits, research, analysis, and administrative tasks or other work that requires access to relationships and trends in the data but not to PII.



BT Selects Sapior to Secure NHS Patient Data

June 28th, 2010 by admin

1 September 2005 – Sapior Ltd. announces that BT has awarded Sapior a nine-year contract to enhance data security and privacy for the new NHS Care Records Service. The Sapior Redbridge ES pseudonymisation solution will be used to de-identify sensitive patient information for use in secondary medical purposes rather than direct patient care.

Sapior has been selected as one of a number of suppliers working with prime contractor BT on the NHS Care Records service. This national patient record database will be one of the largest in the world and will eventually contain a summary care record for every NHS patient in England. The use of the Sapior pseudonymisation solution will ensure the confidentiality of these data, whilst they are being used for secondary activities such as financial transfers, management information and medical research.

“Sapior offers a mature, forward-looking data privacy solution that integrates easily and already meets significant future requirements,” explains Rob Story, NHS Care Records Service programme director, BT. “Sapior has been extremely responsive to the demands of this ambitious project.”

“Our significant experience in Business Intelligence enables us to understand and provide for the unique security needs of sensitive data being analysed or shared over extended time periods, as well as to accommodate the performance required by such a massive database,” said Robert Navarro, Managing Director of Sapior Ltd.

Pseudonymisation, a Privacy Enhancing Technology (PET), has been suggested by the UK Information Commissioner as a way to permit necessary access to patient information whilst hiding patient identities and other sensitive information. Also known as “reversible anonymisation”, pseudonymisation is essentially the replacement of identifiers – such as name, address or NHS number – with pseudonyms. Pseudonymised data can be used for audits, research, and administrative tasks or other work that requires access to relationships and trends in the data but not necessarily to all of the sensitive patient information.

More information about the NHS Care Records Service (CRS) Secondary Usage Service (SUS) can be found at:

http://www.connectingforhealth.nhs.uk/delivery/programmes/sus

Latest From Blog

Oct 25

The Cameron government has re-opened the debate on how much of ou ... Read...

Aug 27

I've heard many times through many media the need for "balance" i ... Read...

Latest News

Safemerge v2 released

May 2013 - Building on Sapior's market lead in enabling ethical d ... Read...

Self service Pseudo service launched

March 2012 - Sapior has launched a self service way to de-identif ... Read...