Provincial Government Western Cape and UCT Head of Division of General Medicine, Dr Peter Raubenheimer, talks about how a local eHealth innovation is improving patient care and why clinicians are key to eHealth success.
There’s been a lot of talk about the development of an electronic Continuity of Care Record (eCCR) within Groote Schuur Hospital’s Department of Medicine. What were some of the drivers that led to the development of this eHealth solution and how were you involved?
Once discharged from the hospital every patient leaves with a discharge letter, or ‘Continuity of Care Letter’, that they’ll need to continue their treatment or book a follow up at another healthcare facility. The letter is a vital piece of information that serves multiple purposes, most importantly the transfer of information (continuity of care) about diagnoses, future treatment plans, prescribed medicine, etc. It also captures information that is valuable in recording morbidity data, and can simplify the coding and thereby costing of hospital stay.
The eCCR was conceived when we started optimising the paper version of the discharge letter to determine what information it needed to provide. When we looked at all the different agendas that this letter needs to meet it became clear that the only way to really do that was electronically. We needed to produce a letter that was legible, and useful for future care, but that also captured the data in a way that enables analysis. In addition, in order to allow for efficient coding and prescribing we needed to integrate ICD browsers, prompts and prescriptions. From very early on the design we aimed to satisfy these different and sometimes conflicting requirements, particularly in a very busy clinical setting where efficiency is essential.
We think we now have a template for a high quality letter that not only transfers information but also captures it for analysis. This is an exciting step towards developing routine provincial morbidity data for hospitalised patients. This, together with the current mortality data, will provide essential information to enable us to more accurately plan future services.
Is the eCCR part of a move to go paperless?
No, going paperless was never the primary intention for us. The system is rooted in two things: firstly, improving the quality of care so when a patient is discharged from hospital they have a legible letter that contains pertinent information about the patient’s journey and secondly, to capture data and interrogate it.
It is essentially the same purpose of all electronic patient records, but the principle for the eCCR wasn’t based on a desire to go paperless. There is however a great need to make healthcare information – as in an episode that succinctly captures what happened during a hospital stay – available for future care givers outside of where the patient was seen and of course the patient themselves. In this context, electronic has clear advantages. The province has for some time been very interested in making this sort of information available remotely and has other projects underway along these lines.
What are your thoughts on the healthcare industry going paperless in general?
Although ultimately going paperless is probably inevitable, we and the systems we have available – from infrastructure to software – are very far from ready. It’s extraordinarily expensive to go paperless; so before the move you have to ask what the gains are and know how to go about it. The UK spent billions trying to standardise records and failed. SA currently doesn’t have the required budget or infrastructure in place so we have to be careful about what solutions we think will be the answer to all our problems.
Paper is good, effective and efficient. It was our main competition when developing the eCCR system because paper is still quick and easy to use. In my opinion, we are still a way off from having interoperable electronic systems across our healthcare landscape and so paper still serves a major purpose. Individual facilities have made attempts to go paperless but until all the building blocks are in place it is too risky both financially and operationally to go paperless. Having said that, there’s no reason why you can’t expand on an eHealth system you’ve built, but buying a pre-packaged eHealth product is often unaffordable, and without the right infrastructure it’s not necessarily going to work for us.
Currently, everything we create electronically with eCCR gets printed and stored as a piece of paper, which I think is the right model for the local context. Everyone wants data, but what we have here focuses on a single episode of very important data. Much of other routine data collected is often of poor quality and difficult to analyse. We also need to balance our desire for data with some of the lessons we can learn from other regions trying to go paperless. In the US for example, there is a lot of dialogue about how the drive to use electronic systems detracts the patient-clinician personal interaction, which is a critical part of the consultation and something we continue to value very highly in our context.
Talk us through the development process of the eCCR system. Who was involved and how was it funded?
The vision of creating an improved discharge letter was shared by all. Robin Dyers, who was a registrar in public health, started developing the eCCR initially to create a unifying discharge letter that could be used province-wide with support from the University of Stellenbosch and the WC DoH Health Impact Assessment Unit. Different people volunteered their time to help him, with technical programming support from Shane Du Plooy at the Centre of e-Innovation, so initially there was no cost involved or associated funding.
The first step in the process was to see what was already out there. Robin found that there are over 200 versions of the discharge letter across numerous departments of every hospital in the province. So the next part of the process involved looking at best practice, what the requirements were and what kind of patient was receiving the letter. The simplest requirement was to have a one pager, so the project was initially piloted in paper form and developed in rapid improvement cycles using junior staff in Groote Schuur. 900 patients are discharged from the division of General Medicine in the hospital every month, so it is a great testing ground and vehicle to gather feedback for improvements.
It was critical to get buy in from all to continue the project, especially to allow people the time outside of their usual work to dedicate to this project. Robin spent a lot of time ensuring all stake holders were involved and satisfied with the process, while keeping the users (clinicians) at the centre of the design process. A budget was later allocated to buy more hardware equipment for the wards because once there’s a queue to use a computer staff will resort back to paper.
Is the eCCR interoperable with other systems?
The eCCR prototype is not networked in its current form. The pilot version however has shown the potential and funding was secured to develop a networked version. The web based redesign, or Enterprise version, will theoretically be available from anywhere, although it will probably use intranet rather than internet initially for security reasons. It will allow discharge letters to be done from any workstation in the province (or via VPN) – a crucial development function to improve efficiency and multiple user input into a discharge letter, and will store copies of past records for viewing.
Groote Schuur has been using the eCCR for over almost a year in the Department of Medicine, what has the uptake been like?
The response has largely been positive due to the system’s intuitive design that requires no training; to illustrate, we get new interns every four months and they are just told to use it.
There has been some frustrations trying to meet two objectives of a good information transfer while also capturing ICD-10 codes for data collection with its built in ICD-10 browser. ICD-10 codes don’t always match clinical communication and can be complex for untrained staff to enter accurately. Most importantly searching and entering is time consuming. As a result of these issues, quality control is an issue that needs to be constantly addressed.
The biggest frustration for the clinician is that it takes longer than simply writing on a piece of paper, but there’s acknowledgment of the benefits of generating morbidity data that we can all work with.
Overall the system works; we’ve never had any serious problems with it, every single patient leaves with a single letter and all the data is accurately captured.
How will the data sets collected be used?
Analysing the data for routine morbidity information, for example how many people are admitted to the hospital for strokes, will be a large project that will require particular expertise. For routine provincial data this expertise lies within the province’s Health Impact Assessment Unit, which collaborates with the MRC Burden of Disease Unit, for this kind of analysis. We hope that in time the eCCR system data will be used to generate routine information that can be fed back to clinicians and public health planners for information and planning of care. The available data can, of course, also be made available for any other individual research questions.
The future NHI talks about funding hospital admissions using Diagnosis Related Groups (DRG). There is interest in whether the data gathered can provide the best information to allow such coding for billing purposes.
Are there any other hospitals using the eCCR?
The eCCR has been piloted at Somerset Hospital in the Paediatrics Department, and is now being piloted in the Surgical Department at Groote Schuur. A provincial project is currently underway at Tygerberg Hospital’s Medical Department using the eCCR.
There were talks to roll out the system to other hospitals as it stands almost a year ago but the question of readiness and appropriateness across the board halted the roll out. We hope to pilot the web based enterprise version of the eCCR in the 2nd half of this year, but the pilot version has stood the test of time and continues to be used every day.
Are there any plans to roll out the system in other provinces?
There are no discussions with other provinces at the moment, although there is a lot of interest in measuring morbidity across the country. There’s no reason why it can’t be rolled out across the rest of the county because the eCCR is a standalone product that can theoretically be adapted to tie into any other IT system. However, provinces do tend to work in silos and they all have their own IT and healthcare system plans.
What key messages do you have for others who have identified a need that they want to address with an eHealth solution?
From my experience, I don’t believe you can buy a product that was designed for another system or country and expect it to work effectively. The strength of the eCCR is that it was designed by the users; it’s a bottom up design rather than top down implementation that uses regular feedback from clinicians about how it can improve their experience to ultimately improve patient outcomes.
We have to sell the product based on its benefit for the patients and clinicians. One thing that is not negotiable is efficiency and time; if anything takes up more of a user’s time it will be incredibly difficult to sell it to anybody – clinicians are incredibly busy and unduelly burdened by “paper work”. Our hook is that we say “if you fill in the codes we will give you the data,” which everyone wants in order to understand what’s going on in our health service nationally.
Clinicians are at the core of any change in the health service. They are the drivers. Don’t underestimate the power of the clinicians in the success of a system, if they don’t buy into it, it will fail. It’s crucial to have a relationship with clinicians and other stake holders and to engage them from the beginning. The system has to work for them first. That’s sometimes difficult for decision makers to remember because IT is sexy; everyone wants the latest eHealth systems but focusing on the technology and not the user will almost always result in a costly mistake.