Shared learning database

PrimarycareIT Ltd
Published date:
January 2017

We have built into EMIS Web (a clinical system for delivering integrated care), a number of optimisations which ensure that hypertension, CKD, AKI, asthma, urinary tract infections and diabetes are managed in accordance with NICE guidelines. Our work looks for triggers in relevant patients (e.g. entry of a blood pressure reading) and then advises the user (dependent on their level of skill) as to how best to manage that patient according to guidelines.

Self-management pathways are fully embedded within our work which helps surgeries to optimise their working and cope with the ever increasing demands upon primary care resources. We have constructed the "one template" - which is exactly as it sounds - one template which is dynamic and shows everything that is needed to be recorded for the patient the clinician has in front of them. It also aids clinicians by letting them know what monitoring the conditions or medications the patient is on requires and highlights important information like "sick day rules".

Does the submission relate to a general implementation of all NICE guidance?
Does the submission relate to a specific implementation of a specific piece of NICE guidance?
Full title of NICE guidance:


Aims & Objectives

To improve the care of patients registered at our surgery with hypertension, whilst also implementing a self-care system that allowed for remote monitoring of patients own blood pressure by themselves at home.


Prior to this we operated a very traditional model of hypertension care, just checking blood pressure at the surgery. We also tended to have quite a disease centric approach, with patients attending specific clinics for their particular conditions. There were a number of triggers for change. As we were getting busier with larger number of elderly patients with multiple morbidities there was a general feeling that we needed to work smarter to achieve better patient care.

We identified through audit presented at our clinical governance meeting that we had a large number of patients who had high blood pressure but hadn't been diagnosed as hypertensive. We also identified that our prevalence for hypertension was lower than that of surrounding practices with very similar demographics. This served as our baseline assessment. From the discussions at that meeting we decided that we were not optimising self-care and self-management plans as fully as we could. If we were to cope with the increased prevalence of hypertensive patients we needed to have a solution to allow for streamlining of workload.

As part of an away day at the practice we also recognised that often patients were coming in for multiple reviews for multiple conditions. Often this was for just one or two items of care that needed completing, but that there wasn’t an easy way for clinicians to know what overall information was needed for that patient – the only way would be to look at their summary screen and then individually open each template for each condition. There also wasn’t an easy way to know what investigations were needed for individual patients.

By creating the OneTemplate, we have created a dynamic template that will show all relevant information to the patient the clinician is seeing as well as letting them know what monitoring is required. It therefore follows that patient care will be much more streamlined, needing fewer but higher quality attendances saving burden not only on clinicians but also the practice administrative and reception team.

We therefore developed a tool within EMIS web which allowed clinicians to be able to choose from a 3 tier pathway of management:

1) full self-monitoring

2) self-monitoring but still seen for medication dose adjustment

3) standard care with monitoring and dose adjustment at the surgery. We discussed our plans with our patient participation group who were supportive.


We used EMIS Web to develop a series of "concepts", "protocols" and "templates". Concepts are a way in EMIS Web to test the patient you have loaded into the system against a defined number of criteria to give yes or no answers, e.g. is this patient hypertensive. These concepts can then be used within a protocol.

A protocol is a series of actions which are triggered by a specific event in EMIS, such as entering a blood pressure reading. The trigger can thus set a chain of evaluations off and allow for on screen information to be triggered or documents to be printed. In this way we were able to set up not only the self-management pathways but also implement the NICE-recommended control levels for optimal blood pressure management.

Concepts can also be used to create “visibility rules” for certain sections of templates. So for our OneTemplate we were able to create a series of complex concepts which then decided what information should be displayed. An example of this is the newly diagnosed asthmatic – where a concept was created to test this against the patient’s record and if appropriate a section of the template is then shown which prompts to ask the patient about occupational exposure which could be implicated. See the enclosed files 1-5 which show the total available information on just one page of our OneTemplate and then the test 1-3 for how these display with different patients. There are many more examples of the areas we have done this in the associated information submitted. See supporting files concept 1 and 2 for the list of some of the concepts for the OneTemplate.

The challenge for us was the translation of guidelines into the actual protocols. It took reading the guidelines many times, making notes and highlighting where we needed to make branches within the protocols to achieve the desired effect. The logic diagrams for what we were trying to achieve were sizeable (see protocols for an idea of the number of protocols needed just for hypertension give you an idea of the complexity).

For our hypertension work, we often had to reference several different guidelines e.g. diabetes, CKD as well as hypertension to make sure that we had dealt with the condition appropriately including co-morbidities and their effect on management. We then had to extensively test the constructed protocols with test patients to make sure that they achieved what we wanted to. We had to construct a bank of test patients of different ages and sexes with different conditions to make sure that we could adequately test the protocols.

We also had to make sure that when we released our work into the live system there was a robust feedback mechanism to let us know quickly of any problems and clear communication to the teams of the changes being made.

Achieving a change in approach involved getting all team members engaged with the process right at the beginning. In this way they understood the need for change and could understand why we were making the changes that we were. We took constant feedback from the team about any areas that were difficult – e.g. we created a separate quick link for receptionists to be able to quickly print out a self-monitoring sheet if a patient needed one.

Results and evaluation

We monitored both our prevalence of hypertension and also the control levels of not only hypertensive patients, but also separated out those with diabetes and CKD. We also separated out the different tiers of monitoring to see how these approaches affected control levels.

Our prevalence of hypertension leapt from 1205 to 1667 meaning we had an additional 462 patients diagnosed with hypertension.

We identified 144 self-monitoring patients in the first year and not only did these patients attend on average 2.6 less times per year, but their blood pressure was also lower by 4mmHg/3mmHg Systolic/diastolic respectively.

This demonstrates that we have managed to improve care whilst containing workload hence becoming a lot more productive as a surgery.

Feedback from the practice team is that the process of reviews of patients has been more streamlined and that they are much clearer as to what needs to be done for which patients.

Key learning points

Learn how to use concepts, protocols and templates within EMIS Web. A good place to start would be attending the National Users Group conference (COI I’m a committee member) to hear about how others are using their clinical system, or accessing the excellent materials on the NUG website.

Consider a clinician within the practice team becoming the lead in IT development to make sure that they have an overview of how to make the IT work best for your practice. In our view this needs to be clinician led to gain the necessary traction and visibility within the practice. They need protected time if they are to achieve this effectively.

Brain storm with the whole practice team, perhaps at an away day as to “easy wins” you can implement quickly – a friend of mine has implemented a protocol that when a call from a clinician to a patient is missed, the system displays the information the GP wanted to communicate to that patient as soon as the record is loaded – saving receptionists valuable time when the patient calls back. With these early easy wins, begin to develop more complicated concepts and protocols based on audits which have identified areas of need.

Engage early with all team members and plan how the whole team is going to implement the change. Take feedback from team members and act upon it quickly to gain confidence in the work you are doing.

We have put together a commercial package offering our work to other surgeries for those that are interested, this is at a very competitive price of £500 per annum and we will use all revenue to further develop other tools that can achieve the same aims.

Contact details

Dustyn Saint
Medical Director (and GP)
PrimarycareIT Ltd
Street Address:
2nd Floor, 43 Broomfield Road,

Category(s) that most closely reflect the nature of the submission:
Primary care
Is the submission industry-sponsored in any way?