An advantage of recording outcomes and other routine data electronically, is that it can be collated and analysed in near real time. This feature holds out the potential for surveillance use-cases such as, epidemiological studies and monitoring the safety of new treatments.
The US FDA has set up a nationwide electronic post-marketing product safety system called Mini-Sentinel (Platt 2015). The system has 18 data partners, encompassing data from almost 178 million patients. This allows the FDA to track potential safety issues associated with approved drugs and other medical products (Mini-Sentinel 2015).
The Mini-Sentinel project uses a distributed network (See Healthcare Data). The data partners, such as providers, maintain operational control over their data. They receive queries (data requests) from a coordinating centre and approve the release of the aggregated results. This protects patient privacy (Brown 2015). Because the data partners must approve each query and each data release, it takes a number of weeks to compile the results of a nationwide query. In the past, it has taken years for harmful adverse effects to be recognised and mitigated. This sort of system could reduce that time significantly (Mullard 2012).
Mini-Sentinel is a pilot project for a larger “Sentinel” project that is currently under development. These systems are particularly good at identifying adverse events that are relatively common in the general population, such as, heart attack, depression and suicide, which are otherwise difficult to associate with causal agents.
QSurveillance is a real time clinical surveillance system based on data from over 3,000 general practices throughout the UK. It was used to monitor the progress of the influenza pandemic in 2009 (Harcourt, Smith et al. 2012). Indicators such as, flu-like symptoms, pneumonia, antiviral use, deaths, gastroenteritis, heat stroke, vaccinations, etc. are extracted from the EMIS EHR, using a distributed network configuration. During the pandemic, daily extracts were provided to local and national authorities who where managing the response.
Surveillance systems require a methodology that is simpler than that required for comparative effectiveness research, therefore, the use of these systems is relatively more advanced. Access to routine data enables a shift from passive towards active surveillance. Passive surveillance approaches, such as the Yellow Card Scheme (https://yellowcard.mhra.gov.uk), operated by the UK Medicine and Healthcare Products Regulatory Agency (MHRA), rely on reporting of events or emerging case-studies. Active surveillance allows regulators to monitor for issues and to quickly mine distributed networks for answers to particular questions (Mullard 2012).
NHS Data Collections as a platform for a Learning Health System
What role for learning health systems in quality improvement within healthcare providers?
Dr Jeff Brown Interview
Professor Richard Platt Interview
Use of a large general practice syndromic surveillance system to monitor the progress of the influenza A(H1N1) pandemic 2009 in the UK