There is an age old debate amongst Star Trek fans: who was the better captain, Kirk or Picard? Regardless of a fan’s preference, one thing that they all would agree on is that it’s never a good thing to be a nameless member of an away-team, because that unfortunate soul never makes the return trip home. There are similarities between that red-shirted Ensign and a patient arriving in a hospital emergency department. There is little known about either. The reason the Ensign’s anonymity is for dramatic effect. Their death shows the risks that the away-team are taking when journeying into the unknown. The reason for hospital care providers knowing little about a patient arriving in their ED is largely due to the primary care data gap. Although for different reasons, being unknown places both the unsuspecting Ensign and data deficient patients at risk. Like Star Trek, eHealth in Ontario has a final frontier: primary care data. And just like space, the territory is vast and uncertain.
The gap in eHealth
Currently, eHealth in Ontario is comprised of repositories and portals: OLIS for lab, DI-r for imaging, Panorama for public health immunization and CGTA or cSWO for hospital data. Some of this data captures interactions with primary care. As an example, if a patient’s family doctor requisitions lab work, it will likely end up in the Ontario Lab Information System (OLIS). The lab information will then be available to care settings outside of primary care either through a portal or through direct integration with Electronic Medical Record systems (EMRs). In addition, if a patient receives treatment with a CCAC, the information about the encounter will be available in either the cGTA or cSWO portal. These portals are available to primary care providers as well as acute settings. In addition, other services such as Hospital Report Manager, HRM, and eNotification (both OntarioMD solutions) send hospital reports and notifications of hospital visits directly to primary care EMRs. However, when looking at the current eHealth map, it’s important to highlight that all data is pointed in one direction: towards primary care. Outside of the above two examples the data does not flow the other way.
How did the gap occur?
Starting in 2005, OntarioMD, sponsored by the Ontario Medical Association (OMA), made significant inroads in moving primary care physicians off of paper and onto electronic systems. Now in 2016, the majority of Ontario physicians and specialists are using one of the certified OntarioMD EMR, Electronic Medical Record, products. The following shows adoption in primary care throughout Ontario as of the end of 2015.
The Canadian Medical Association Journal estimates that these numbers represent approximately 83 per cent of adoption based on a January 2015 survey.
The EMR adoption progress that has been made in Ontario is a significant achievement. What helped make OntarioMD’s strategy a success is that it did not depend on building a complex repository. Instead, OntarioMD encouraged private sector companies to create solutions that OntarioMD would certify. This enabled OntarioMD to focus on governing software development rather than developing technology solutions themselves. By leveraging the private sector for development and partnering with their sales forces, an expedited provincial schedule of adoption and delivery of EMRs into physician offices was achieved.
However, the EMR good news story is accompanied with a caveat. In spite of the significant progress made with EMR adoption, the solutions developed by private sector vendors were not done with data sharing in mind. Therefore EMRs do not follow a universal standard of coded data, or a consistent ability to send HL7 messages to share information. As a result, most EMRs are like isolated planets in the expanse of space, unable to communicate with each other or to contribute to the overall electronic patient record.
Why is primary care data important?
Working on various eHealth programs like cGTA and cSWO I have heard knowledgeable leaders often say that 80 per cent of patient data resides in primary care. Although this statistic is anecdotal it doesn’t raise a Spockian eyebrow. Why? Because it’s logical. Consider a patient’s typical interactions with health care providers: their family doctor and the specialists that their family doctor refers them to. A patient only goes to a hospital when something unexpected happens or a diagnosis occurs that requires greater specialization and equipment to isolate and treat. Consider the following information that a family physician would know:
- Immunization history
- Allergies
- Addictions
- Other mental health concerns
- Family history
- Drug prescriptions
- Chronic conditions
- Out of country medical history, or any medical history that pre-dates electronic storage
- General health concerns and any consults or observations made during the course of the relationship with the patient
To support the assertion that 80 per cent of patient data resides within primary care, the Centre for Disease Control (CDC), issued a report that shows more than half of a patient’s care visits are with primary care providers. Therefore, if the majority of patient visits are with primary care, it is logical that most data about the patient resides in primary care.
How does the primary care data gap affect patient care?
To illustrate how the primary care data gap can impact to patient care, let’s review some possible patient scenarios:
- An elderly woman arrives in a hospital’s emergency department. She is confused. She cannot remember her medications or her physician’s name. The hospital staff are unaware that she is suffering from early dementia, a problem captured in her care provider’s EMR, along with her list of medications. They attempt to treat her but are concerned about administering any medications because they may conflict with what she has been prescribed by her primary care physician.
- A child sits in recovery after a tonsillectomy. A measles outbreak has occurred in the hospital, in the same wing where the child is recovering. The child has not yet had their measles immunization for a variety of logistical reasons and hospital staff are unaware that the outbreak could affect the patient.
- A teenage boy is taken to the hospital to undergo an emergency appendectomy while away from home playing in a hockey tournament. As he was rushed into surgery the care team is unaware that the boy is allergic to penicillin which is recorded in his primary care provider’s EMR.
Some of these scenarios have been used to make the case for primary care data sharing in Ontario. However, significant traction has yet to be made. The above scenarios illustrate the importance of having primary care data available in emergency situations.
To support the above scenarios with quantitative data the following report commissioned by the Empowered Patient Coalition shows a breakdown of adverse events that had patient impact during a hospital visits.
Further breakdown of this data shows that lack of knowledge of a patient’s case was seen by patients as a major contributor to the incidents that occurred.
The data sharing challenge
The above highlights how a lack of primary care data can place a patient at risk. But the challenge requires a solution more complex and delicate than engineers emphatically exclaiming “WE NEED MORE DATA!”
Some of these challenges were alluded to earlier and include:
- EMR systems are disparate. They are not integrated and were not designed for integration. They do not use common coded data, do not enforce standard data entry and, in most cases, are not built to transmit data.
- Outside of Family Health Teams, primary care offices tend to be small with limited number of support or IT staff. As a result they do not easily scale to meet the people and technology demands associated with securely transmitting personal health information.
- A common concern I’ve heard in my interviews with primary care physicians is the notion of peer review and concerns of being audited which could lead to scrutiny, second guessing and risk for the primary care practitioner.
The next steps?
Until tricorders are invented that can instantly scan people for all of their medical problems and assimilate the data without the need to electronic storage systems, more investment is required to build infrastructure, standards and integration.
Understandably this isn’t an appealing concept for most Ontarians. The notion of more dollars spent on eHealth investments is a polarizing topic. However, unlike previous initiatives that were complex and expensive from both a development and infrastructure perspective, integrating primary care data should leverage existing investments as much as possible.
To put some of the concerns physicians have about data sharing into context, most EMRs and physician offices support a SOAP method of documentation. SOAP is an acronym for subjective observations, objective observations, assessments and treatment plans. It is in essence primary care patient charting and can contain personal assessments. Personal physician documentation is one of the core concerns highlighted earlier and applications that support SOAP can understandably generate data sharing apprehension.
The good news is that this isn’t what other settings require. In an emergency, care providers do not have time to review a patient chart, or lengthy assessments in free form text. It is the coded data that enables decision making in emergency settings. Therefore, the focus needs to be on acquiring the decision making subset of data housed in primary care:
- Immunization history
- Allergies
- Drug prescriptions
- Chronic conditions
- Out of country medical history, or any medical history that pre-dates electronic storage
By leveraging existing repositories and focusing on acquiring minimal coded data, primary data can be liberated. Liberating primary care data will enable efficient clinical decision support in emergency situations: when quick retrieval and the ability to quickly assimilate information is critical.
The engineers are right, we do need more data. Hopefully Ontario embarks on the final eHealth journey without the need to build another repository, or spend significant time and money to achieve this objective.
In the end, primary care data needs to support acute settings. This may be more important than the inverse. By doing so it will ensure that acute care providers are not going into darkness when a patient emergency situation arises. After all, regardless of the situation, we want all the Ontarians’ to return from their away mission safely, without incident or complication if and whenever they require hospital services.
