Optimized study design, precise site selection and faster patient search and identification for clinical research - in real time.
Q: What is Clinerion’s coverage? How many patients, how many countries, over how long?
A: We are in many countries in Europe, Asia, North America and South America, with a multitude of countries in the pipeline. Our current coverage is now visible on our homepage. [Updated November 28, 2018.]
The timeframe encompassed by our data differs depending on the availability of electronic health records and on their availability in the hospital’s current hospital information system. In general, our data goes back five years.
Q: How many hospitals and patients are included in the cloud?
A: We don’t actually include hospitals or patients in the cloud. We use the hospital infrastructure and our cloud is there to run queries, manage queries, create audit trails, etc.
Q: How do you handle countries where there are regulations preventing personal data from being exported overseas (including being hosted by overseas servers)?
A: The data we access is not stored in a centralized location, nor is it in the cloud. Patient data remains hosted locally at each of the hospitals that we work with.
Q: What can you do for a specific condition such as breast cancer?
A: We can identify eligible patients with due regard to detailed inclusion/exclusion constraints, e.g. diagnosis, medication, medical procedures, lab tests and demographics. Our system is able to incorporate time-sensitivity in the queries. All of these apply to any condition, and the results are in real-time.
Oncology patients, specifically, can be screened and filtered by attributes as cancer stage group, cancer T stage, cancer N stage, cancer M stage, etc.
Q: This seems to be geared at hospitals. How would this apply to local physician practices?
A: The system is not limited to hospitals. It also can be used for screening and analyzing patient data of an unlimited number of physician practices. We can include any health providing institution if it has a good level of EHR quality. In practice, we are more focused on hospitals as they can also function as eventual trial sites. It is also more efficient to connect hospitals with their high number of patients than physician practices which have lower numbers of patients. In certain countries that we cover, we are connected to both primary and secondary care facilities.
Q: What methods are available to connect the data from the hospital information system?
A: Clinerion uses an advanced, proprietary data extraction technology, which offers a number of advantages over existing technologies, including a detailed data model developed specifically for clinical research, ease of use and administration, built-in automatic data transfer to the Clinerion server, support from Clinerion, and regular updates to keep the model in line with future needs. In addition, the Clinerion Patient Network Explorer can extract patient records from any data warehouse through other protocols, including i2b2, OMOP, FHIR/HL7, etc., and proprietary APIs.
Q: What is the data lag?
A: The data refresh rate that we use routinely is 24 hours. However, we can work with a frequency of down to one minute. This is activated on a case-by-case basis, for instance when a clinical trial is run in an acute setting.
Q: Can your partner hospitals re-identify patient records with a time shift?
Q: Are you working with any sponsors, or CROs, yet?
A: As of September 2017, Clinerion has performed more than 20 recruitment studies and more than 70 feasibility studies for the top 10 pharmaceutical companies and largest CROs.
Q: How does ANID address the EU General Data Protection Regulation (GDPR)?
A: GDPR is a privacy regulation, covering personal data. Data which is de-identified and unlinked from identifiers no longer contains personal data and is therefore not covered by privacy legislation such as GDPR. Clinerion works with de-identified, unlinked data. A hospital would still need to conform to GDPR, regardless of their use of the Clinerion system.
Removing the need for a tracker (pseudonymization-ID), ANID enables hospitals to share fully de-identified, unlinked patient data for research reasons and it also allows patient identification for hospital internal clinical recruitment.
Q: Is anyone else doing ANID?
A: Our search for prior art during the submission of our patent application for ANID did not reveal any previous cases of the processes involved in ANID. ANID is new and Clinerion is unique in offering it.
Q: Is there a blanket method for removing PHI (Protected Health Information) that is the same across all countries, or do you remove fields depending on legal requirements on a country-by-country basis?
A: In general, we remove the standard set of PHI as defined by HIPAA. In addition, we also respect country-specific regulations and requirements. Local legal requirements can be accommodated by our model.
Q: Must a hospital get patient’s consent before de-identifying their data and using it for general research reason?
A: It is universally accepted that de-identified, unlinked data can be accessed without patient consent for research purposes. In the USA, consent is not required. In the EU, consent is not required if data is de-identified (but it is required if it is pseudonymized).
Q: Must a hospital get a patient’s consent before screening and identifying them as a potential candidate for a clinical trial?
A: In the USA, IRB (Internal Review Board, e.g. an Ethics Committee) approval is needed for screening and identifying potential clinical trial candidates without their consent. In the EU (and many other countries), no centralized rules are known, so far. Similar rules as in the USA may apply, locally. Which rules are valid must be clarified, in any case, by a hospital’s data privacy officer. Our advice is to get the patient’s consent upfront when first entering the hospital. With this in place, IRB approval would not be needed.
Q: Is this not just another procedure for pseudonymization rather than de-identification?
A: This is very emphatically not the case. The data patient data we receive and process are de-identified and unlinked from all identifiers; all identifiers are removed and no pseudonyms are used. The de-identification process is under the control of the hospital; we do not have the ability to re-identify the record. The fact that our technique allows authorized trial staff at hospitals to efficiently re-identify eligible patients is why our ANID technology is so powerful.