Lessons from a model of effective eHealth implementation
Earlier this year, I took a three-month sabbatical at the Universidad Catolica (UC) and its extended healthcare institutions in Santiago, Chile, where I was able to make some observations on the implementation of an electronic medical record (EMR) and other eHealth developments that could provide some lessons for similar implementations in Australia.
The Chilean healthcare system is funded under a national health insurance scheme which is sourced by a seven per cent tax on incomes. The system has separated into two main streams of care delivery with public and private components. Care within the private system usually requires supplementary funding from the patient or insurance, and like in Australia, patients may have care in both systems.
Each component of the Chilean health system has significant differences in care delivery, with the public system in Chile experiencing similar care delivery issues as the public system in Australia. That is not to say that the private component of healthcare delivery is perfect, and it is definitely more expensive with significant overuse, underuse and inappropriate use of healthcare resources.
My working association during my sabbatical was predominantly within the UC private system, which covers both primary care clinics and hospital-based care, and I also taught some workshops in the department of biomedical informatics at the regional Duoc-UC San Joaquin, Valparaiso and Concepcion campuses, which provide TAFE-level education.
The Universidad Catolica group has a complex patient management system, with an EMR from Portuguese vendor Alert Lifesciences currently being rolled out. This is being accompanied by a powerful EMR support application called Epistemonikos and a bespoke temporal query system called ClinicalTime, which is being developed by UC physician-health informatician Daniel Capurro.
One of the strengths of the eHealth projects within UC is that the planning and constant evaluation is driven by the university faculty and the patient care departments, not government health departments. The EMR supports data collection and healthcare measures required by the government, with this data a direct byproduct of clinical care processes.
Because the EMR is clinically focused, doctors, nurses and other clinicians use the system for patient outcomes rather than for billing services, which have become the necessary adjuncts to the clinically directed project.
The EMR system permits real-time adjustment of applications and modules without significant disruption of day-to-day clinical use.
In my opinion, the successful direction of this project must go to the clinical leadership of Associate Professor Capurro, who has enlisted the active and enthusiastic support of his clinical staff and IT personnel. The teams leading this project meet up to twice weekly to discuss its evaluation and the overall adherence to the implementation time schedules.
Attending these meetings, it was obvious that the core enthusiasm and drive demonstrated by the participants revealed a strong belief in the value of this project. Development was not driven by an imposed external system, and this team and the clinicians “owned” the project.
The combination of frequent end-user driven goal definitions, evaluations and clinically focused applications with an adaptable system application infrastructure complies with many of the core success factors that were defined by researchers Paul Biondich and Burke Mamlin from their work in low and middle income nations.
Clinical decision support
I was also present at an educational meeting at a UC regional hospital called Clínica UC San Carlos de Apoquindo, where the Alert EMR was about to be implemented. The hospital director addressed the meeting and all individuals involved in the system's implementation and uptake had a voice in the proceedings. This includes both the objectors and the supporters.
The EMR has an excellent user interface, with real-time patient triage to care transitions and e-directed (guided) care management, such as patient-specific IV infusion times and rates and medication administration compliance schedules.
The system provides direct feedback on waiting times for patients in the emergency department – there was great concern that one patient had been waiting for 48 minutes!
I observed clinicians actually using the system to enter care plan protocols that supported their current care plans.
However, because clinicians are able to enter care plans, there is some variation at this stage of development. Dr Capurro is currently undertaking a project to refine the creation and design of clinical order sets to improve clinical workflow compliance and reduce the variation in care delivery.
In the development and design of clinical order sets, access to up-to-date knowledge and research is critical. The UC is using a powerful EMR support application called Epistemonikos, which was developed by the Epistemonikos Foundation, a not-for-profit organisation led by internal medicine specialist Gabriel Rada and created as a spin-off from UC and funded through international research funds.
The application is a free, multilingual database of the best available health evidence, and the UC team intends to integrate this system with the order sets management.
From my perspective, the Epistemonikos system is an ideal adjunct to the order sets program because it provides a functionality that, as former CEO of the California HealthCare Foundation Mark Smith put it, can “remove ourselves from the ‘unscientific, non data driven personal recommendations’ for care".
And as Andy Kanter of the Millennium Villages Project says, “The ability to feed back immediately to the people at the point of care … is critical for measuring and improving the quality of care."
Dr Capurro and colleagues are also developing a conjoint system called ClinicalTime which measures patient trajectories using time-dependent clinical and laboratory data. This has demonstrated that coded clinical data provides more accurate and dependable data for measuring healthcare than standardised coding using ICD-9.
Dr Capurro describes ClinicalTime as “a temporal query system” that is a valid method to add to the currently available ones to identify patient phenotypes in patient databases and improve the ability to re-use routinely collected electronic clinical data for secondary purposes.
Another potential benefit is in the use of clinical data to provide the early detection of those patients in an ICU environment who are likely to develop acute kidney injury (AKI). This is a clinical state which is often overlooked until late in the disease evolution in the ICU clinical environment.
The EMR has become the core patient care system in UC's outpatient clinics and the emergency department, which became paperless just over three years ago. This change occurred with minimal disruption after going live overnight.
The outpatient system was impressive. As a patient you can make your clinic appointment on the phone or on the internet in advance of attending. Patients can pay their accounts online in advance while making the appointment.
When arriving at the clinic, the patient or their carer enters their unique Chilean ID number at the kiosk in the foyer or they can swipe their card with this number. This unique ID number and card have been a part of the Chilean health system for decades.
The kiosk can display educational materials but the main purpose is to direct the patient to their designated clinic and its location. If you have not paid in advance you go to the cashier and pay there.
At your clinic location the kiosk provides an ID number for that visit and the room number. These are clearly displayed on TV screens and this information is supported by voice-over instructions.
In the consultation the clinician has access to the patient’s record and associated reports. The information is in both coded and document formats. At all stages the clinician is functionally looking at the one system.
The clinicians must enter the details of the clinical encounter into the record. In the wards, a dual e-record and paper coexist but the intention of the project governance committees is to have complete electronic integration in the not too distant future.
The Chilean government introduced legislation three years ago that states that all patients must be given a copy of their own record. The actual mechanisms for this are evolving.
The waiting lists for these clinics can be up to one to two days for chronic disease. A ‘long’ waiting period is one to two months, which is considered not to be completely efficient!
The UC EMR is also used as a data resource for capturing administrative healthcare data and is able to provide this in real time. For example, DRG data is captured on a daily basis.
The system was set to go live at an allied regional hospital of around 100 beds, with support staff living in for a period of time and providing 24/7 support, as I was leaving Chile. On-site training will occur in the ward location immediately adjacent to the emergency care wards.
While the cost of the EMR is confidential, it is significantly less than comparable implementations in many Australian hospitals.
My observation is that a significant eHealth project based on an appropriately designed EMR can succeed on a moderate budget as long as the objectives and design are driven by the direct patient care process.
Dr Terry Hannan is a consultant physician at Launceston General Hospital, a clinical associate professor at the University of Tasmania's School of Human Health Sciences, a fellow of the Australasian College of Health Informatics and a visiting fellow at Macquarie University's Centre for Health Informatics.
Posted in Australian eHealth